Potential for nature-based mitigation of coastal flood risks

From regional to global scale assessments

Potenties voor natuur-gebaseerde mitigatie van overstromingsrisico’s in kustgebieden

Een regionale tot globale studie

Dissertation for the degree of Doctor in Science at the University of Antwerp to be defended by

Rebecca Van Coppenolle

Promotor: Prof. Dr. Stijn Temmerman

Faculty of Science Department of Biology Antwerp 2018

Cover design: Ann Roelant, Niewe Media Dienst, University of Antwerp Front cover: Sarawak Mangrove Reserve, © Tim Laman

Sous les pavés, la plage

I

Acknowledgements

A bit more than 4 years ago, I took my chance and started a PhD journey at the

University of Antwerp. During this period, I got to meet and get the support of

many people who each in his own way helped me to achieve my goal and finish my

thesis.

To start, I want to gratefully acknowledge the University of Antwerp that funded

my researches and my time over the last years.

My greatest gratitude goes to Stijn Temmerman, my promotor, for your support

and enthusiasm at every step of the researches. Thank you for reading and re-

reading my work. It has been a real pleasure and a great learning to work with you

over the last years. My gratitude also goes to Christian Schwarz for his help over

the first years of the PhD and for always kindly answering my questions with

helpful advices.

I would like to thank the members of my doctoral commission and of my jury; Ivan

Janssens, Patrick Meire, Ivan Nijs, Tom Spencer and Nassos Vafeidis who took the

time to read my thesis and provided insightful comments and suggestions at the

very last stage of my PhD.

Furthermore, a big thank to all my colleagues at ECOBE for their friendship and

the nice discussions and laughs during the numerous coffee breaks, lunches,

parties and dinner out. A special thanks to Kristine, the perfect office-mate and

many thanks to Lotte, Dácil, Lindsay, Jeroen, Alexandra, Veerle, Lennert, Martijn

(collifriend!), Steven, Cedric, Alanna, Annelies, Jan-Willem, Rosanne, Niels, Dante,

Sam, Rose, Naeem, Katrien, Ken, Olivier, Jean-Philippe, Willem-Jan, Leo and Stijn,

you made this PhD adventure much more fun and easy.

Pour finir, je voudrais remercier mes amis et ma famille pour m’avoir écoutée et

rassurée quand j’en avais besoin et pour tous les moments passés ensemble au

cours de cette longue aventure. Tous vos petits mots d’encouragement au cours de

ces derniers mois ont été d’une grande aide !

Merci à Magali et Alexandre pour vos encouragements et votre compréhension.

Papa et Maman, merci pour votre soutien inconditionnel.

II

Summary

Coastal zones are threatened by increasing flood and erosion risks, due on the one

hand to sea level-rise and increasing storm intensity induced by global climate

change, and on the other hand to the growing coastal population density and

associated anthropogenic impacts that aggravate flood and erosion risks. Hence

there is an increasing need to implement efficient and sustainable coastal

protection solutions. Nature-based strategies, relying on the capacity of coastal

ecosystems to reduce the risks of shoreline erosion, attenuate wind waves and

storm surges, and to sustain themselves with sea-level rise by biogenic sediment

accretion are increasingly proposed and implemented for flood risk mitigation,

either as stand-alone solutions or in combination with the standard engineering

solutions.

The purpose of this thesis was to define and analyse the spatial distribution of

hotspots for nature-based storm surge flood risks mitigation, at regional to global

scale.

The creation of a GIS based model assessing the coastal plain area and population

benefiting from storm surge mitigation by salt marshes and mangrove forests

highlighted that over the world, about 30 % of the flood-exposed coastal plain and

40 % of the flood-exposed population benefit form storm surge mitigation. In

general, deltas, estuaries, bays and lagoons present the largest coastal plain areas

benefiting from storm surge mitigation. While the highest magnitudes of

mitigation are found when large and continuous tidal wetlands are located along

the main channels. The amount of coastal population benefiting from storm surge

mitigation by tidal wetlands is influenced by the location of the population relative

to the tidal wetlands’ location or the population density.

The assessment of the current extent of salt marshes, mangrove forests, seagrass

meadows and coral reef known to contribute to flood risk mitigation, in front of

highly populated and flood-exposed coastal cities shows that 75 % of the 136

studied coastal cities are fronted by existing coastal ecosystems. Therefore those

cities can potentially benefit from nature-based coastal flood and erosion risks

mitigation. Nevertheless, the cities with the largest populations and assets exposed

to flooding are generally fronted by small areas of coastal ecosystems. The largest

potentially available areas for tidal wetland restoration or creation in front of

those cities are found in deltas, estuaries, bays and lagoons. The presence and

extent of the different coastal ecosystems as well as the potentially available area

III

to restore/create tidal wetlands is linked to, amongst others, historic,

geomorphologic and socio-economic factors.

Overall, the global analyses presented in this thesis demonstrated that despite the

various worldwide coastal environments, multiple coastal areas can already

benefit from nature-based storm surge mitigation and that this nature-based

mitigation could increase with the restoration or creation of coastal ecosystems.

IV

Samenvatting

Kustgebieden worden bedreigd door het toenemende risico op overstromingen en

erosie, hetgeen enerzijds wordt veroorzaakt door stijging van de zeespiegel en

toenemende intensiteit van stormen teweeggebracht door klimaatverandering, en

anderzijds door de toenemende bevolkingsdichtheid in kustgebieden en

gerelateerde antropogene effecten die het risico op overstromingen en erosie

vergroten. Daarom is het een steeds grotere noodzaak om met efficiente en

duurzame oplossingen voor kustverdediging te komen om de kustbevolking te

beschermen en economische schade te voorkomen bij overstromings- en

erosiegevaar. Hierbij worden er steeds vaker risicoverminderende strategieën

voorgesteld en toegepast die gebaseerd zijn op de natuur. De werking van de op de

natuur gebaseerde oplossingen hangt af van het vermogen van de

kustecosystemen om het risico op kustlijnerosie te verminderen, windgolven en

stormvloeden af te zwakken en zichzelf in stand te houden tijdens stijging van de

zeespiegel door biogene sediment-aanwas. Op deze manier hebben ze het

voordeel zichzelf te kunnen aanpassen aan veranderende

klimaatsomstandigheden, zeespiegel en stormactiviteit, en daarom

kostenefficienter te zijn dan alleen het gebruik van traditioneel ontworpen

kustverdedigingsstructuren.

Het doel van dit proefschrift was het definiëren en analyseren van de ruimtelijke

distributie van hotspots waar stormvloeden worden afgezwakt door

kustecosystemen, d.w.z. de op de natuur gebaseerde vermindering van

stormvloedrisico op regionale tot globale schaal.

Ten eerste is er een model gecreëerd op basis van GIS om een inschatting te maken

van de kustvlakte en kustbevolking die profiteren van vermindering van

stormvloeden door schorren en mangrovebossen op delta- en mondiale schaal.

Onze bevindingen laten zien dat wereldwijd circa 30 % van de kustvlaktes en 40 %

van de bevolking blootgesteld aan overstromingsrisico profiteren van

vermindering van stormvloeden door bestaande getijdengebieden. Over het

algemeen bestaat het grootste deel van de kustvlaktes die ervan profiteren uit

delta’s, estuaria, baaien en lagunes. In delta’s en estuaria wordt de sterkste

vermindering aangetroffen wanneer grote aaneengesloten getijdengebieden langs

de hoofdgeulen liggen. Het aantal mensen van de kustbevolking dat profiteert van

vermindering van stormvloeden door getijdengebieden wordt aan de andere kant

ook beïnvloed door factoren als de locatie van de bevolking ten opzichte van het

getijdengebied, of de bevolkingsdichtheid en vestigingspatroon.

V

Een inschatting gemaakt van de huidige omvang van schorren, mangrovebossen,

zeegrasvelden en koraalriffen waarvan het bekend is dat ze bijdragen aan

risicovermindering, die voor dichtbevolkte kuststeden die zijn blootgesteld aan

overstromingen liggen, evenals de mogelijkheden om getijdengebieden voor die

steden te herstellen of creëren. Voor 75 % van de 136 bestudeerde kuststeden een

bestaand kustecosysteem waardoor er in potentie geprofiteerd kan worden van op

de natuur gebaseerde risicovermindering van vloed en erosie. Desalniettemin ligt

er voor de steden met de grootste bevolkingsaantallen en bezit die zijn

blootgesteld aan overstromingen in het algemeen een kleine oppervlakte aan

kustecosysteem. Als er gekeken wordt naar de mogelijkheden voor het herstel of

het creëren van getijdengebieden, liggen de grootste potentieel beschikbare

gebieden in steden die in delta’s, estuaria, baaien en lagunes. De aanwezigheid en

omvang van de verschillende kustecosystemen, alsook van het gebied dat

potentieel beschikbaar is om getijdengebieden te herstellen of creëren is

gekoppeld aan, onder andere, historische factoren, geomorfologische

eigenschappen en sociaal-economische kenmerken.

Over het algemeen demonstreren onze bevindingen de grote waarde van

kustecosystemen voor het verminderen van overstromings- en erosierisico’s aan

de kust op vele plaatsen over de hele wereld. In de toekomst moet er meer kennis

worden vergaard over mechanismes achter het vermogen van kustecosystemen

om overstromings- en erosierisico’s te verminderen en deze kennis moet gedeeld

worden met lokale gemeenschappen en beleidsmakers om hen meer bewust te

maken van op de natuur gebaseerde overstroming- en erosierisicovermindering

aan de kust en de integratie ervan in kustbeschermingsbeleid.

VI

Résumé

Les zones côtières sont menacées par une augmentation des risques d’inondations

et d’érosion. D’une part, le changement climatique induit l’augmentation du niveau

des mers et l’intensification des tempêtes telles que les cyclones, ouragans ou

typhon, sources de submersions marines. D’autre part, la densification des

populations côtières et leur développement économique aggravent la vulnérabilité

des zones côtières aux risques d’inondations et d’érosion. Par conséquent, le

développement de stratégies de protection côtière efficaces et durables est

indispensable. Les stratégies basées sur la nature (‘nature-based strategies’) sont

de plus en plus souvent proposées comme une solution permettant de réduire les

risques d’inondations et d’érosion. Déjà implémentées le long de certaines côtes,

elles sont majoritairement présentes en complément des structures d’ingénierie

classiques. Ces stratégies basées sur la nature reposent sur la capacité des

écosystèmes côtiers tels que les mangroves, les prés salés, les herbiers sous-

marins, les récifs coralliens, les dunes de sables, etc. à réduire les risques d’érosion

des côtes et à atténuer les vagues et les ondes de tempêtes, tout en ayant la

capacité de s’adapter à l’augmentation du niveau des mers par l’accrétion de

sédiments.

L’objectif de cette thèse est de définir et d’analyser la distribution spatiale des

zones d’intérêts (‘hotspot’) pour l’atténuation des ondes de tempêtes grâce à des

stratégies de protection basées sur la nature, à l’échelle régionale et globale.

La création d’un modèle SIG quantifiant la surface de la plaine côtière ainsi que le

nombre de personnes bénéficiant d’une atténuation de l’onde de tempête grâce

aux prés salés et mangroves a permis de mettre en évidence qu’au niveau mondial,

30 % des plaines côtières et 40 % des populations exposées aux aléas de tempêtes

bénéficient d’une atténuation de l’onde de tempête grâce à la présence des prés

salés et mangroves. De manière générale, les deltas, estuaires, baies et lagons

présentent les plus larges plaines côtières pouvant bénéficier de l’atténuation des

ondes de tempêtes, tandis qu’au niveau des deltas et estuaires, l’atténuation de

l’onde de tempête est de plus grande ampleur lorsque les zones de prés salés et

mangroves sont larges, continues et localisées le long des principaux chenaux de

marée. La population qui bénéficie de l’atténuation des ondes de tempêtes par les

prés salés et les mangroves est quant à elle influencée par d’autres facteurs tels

que la localisation des écosystèmes côtiers par rapport à la population, la densité

de population ou sa distribution dans la plaine côtière.

VII

L’estimation de l’étendue des prés salés, mangroves, herbiers sous-marins et récifs

coralliens reconnu pour leur capacité à atténuer les ondes de tempêtes, pour 136

villes côtières de plus d’un million d’habitants montre que 75 % de ces villes sont

entourées par un ou des écosystèmes côtiers. Ces villes peuvent donc bénéficier

d’une atténuation des ondes de tempêtes grâce aux écosystèmes existants.

Néanmoins, les villes ayant les plus larges populations et biens économiques

exposés aux tempêtes n’ont de manière générale peu ou pas d’écosystèmes côtiers

existants. Les villes pour lesquelles la possibilité de restaurer ou créer des zones

de prés salés et de mangroves est la plus grande sont généralement localisées dans

les deltas et estuaires. La présence et l’étendue des écosystèmes côtiers face aux

villes côtières ainsi que le potentiel de restauration ou de création d’écosystèmes

sont entre-autres liés à des facteurs historiques, géomorphologiques et socio-

économiques.

Finalement, les analyses globales présentées dans cette thèse démontrent que

malgré la grande variété des environnements côtiers, de nombreuses zones

côtières peuvent dès à présent bénéficier de la nature pour l’atténuation des ondes

de tempêtes, et que cette atténuation basée sur la nature peut encore augmenter

avec la restauration ou création des écosystèmes côtiers.

IX

Table of Contents

CHAPTER 1 GENERAL INTRODUCTION 1

CHAPTER 2 CONTRIBUTION OF MANGROVES AND SALT MARSHES TO NATURE-

BASED MITIGATION OF COASTAL FLOOD RISKS IN MAJOR DELTAS OF

THE WORLD 29

CHAPTER 3 IDENTIFYING GLOBAL HOTSPOTS FOR NATURE-BASED

MITIGATION OF COASTAL FLOOD RISKS 61

CHAPTER 4 POTENTIAL FOR NATURE-BASED FLOOD RISK MITIGATION IN

COASTAL CITIES AROUND THE WORLD 103

CHAPTER 5 A GLOBAL EXPLORATION OF TIDAL WETLAND CREATION FOR

NATURE-BASED FLOOD RISK MITIGATION IN COASTAL CITIES 137

CHAPTER 6 SYNTHESIS 171

REFERENCES 185

General introduction

1

CHAPTER 1 General Introduction

Chapter 1

2

The coastal zone, corresponding to the areas at the interface between the marine

and terrestrial environments, is delimited as the area within 100 km of the coast

(Barbier, 2015a). More specifically, the Low Elevation Coastal Zone (LECZ), which

is more vulnerable to coastal flooding and erosion risks, is the land area adjacent

and hydrologically connected to the sea (up to 100 km inland), below 10 m

elevation and above the low water mark (Agardy et al., 2005; McGranahan et al.,

2007; Millennium Ecosystem Assessment, 2005; Small & Nicholls, 2003).

Nonetheless, depending on the scales considered for a study and the topic of the

analysis, the definition of the coastal zone may differ. The coastal zone and LECZ

represent respectively 4 % and 2 % of the Earth’s total land surface and accounts

for about 30 % and 10 % of the world’s population (Barbier, 2015a; Barbier et al.,

2008; Lichter et al., 2011; Mcgranahan et al., 2006; McGranahan et al., 2007;

Neumann et al., 2015; Nicholls & Small, 2002).

The coastal zone and in particular the LECZ are vulnerable to flood and erosion

risks caused by so-called storm surges – these are exceptionally high sea levels

occurring when a severe storm depression (e.g. a tropical cyclone, hurricane,

typhoon or extra-tropical storm) is formed above oceans and seas and propagates

towards the coastline. Recent hurricanes, storms and typhoons highlighted the

enormous casualties and damages they can generate on the coastal zones (see

Figure 1.1 for selected examples of recent coastal flood events caused by storm

surges). In the coming decades, the LECZ is expected to suffer from increasing

rates of mean sea level rise and from increasing intensities of storms, both caused

by global climate change and both resulting in more extreme events causing

coastal flooding and erosion risks (Hallegatte et al., 2013; Kron, 2013; de Sherbinin

et al., 2007; Vitousek et al., 2017). Additionally, population densities in the LECZ

are projected to increase this century (Guzmán et al., 2009; Mcgranahan et al.,

2006; Neumann et al., 2015), resulting in an augmentation of the LECZ’s exposure

to the dramatic consequences of coastal hazards. Therefore, solutions to

sustainably protect and develop our coasts are essential. This thesis presents case

studies, ranging from regional to global scales, on the benefits of accounting for

coastal ecosystems in coastal risk assessments and in mitigation of coastal flood

and erosion risks.

General introduction

3

Figure 1.1 Examples of recent coastal flood events caused by storm surges: (a) Port of Galway (Ireland), Hurricane Ophelia, 2017; (b) New York (USA), Hurricane Sandy, 2012; (c) New Orleans (USA), Hurricane Katrina, 2005; (d) Belgium, SinterKlaas Storm, 2013; (e) Philippines, Typhoon Haiyan, 2013; (f) Houston (USA), Hurricane Harvey, 2017; (g) Vanuatu, Cyclone Pam, 2015; (h) Aytré (France), Storm Xynthia, 2010; (i) Sint Martin, Hurricane Irma, 2017.

1.1 Flooding and Erosion Risks

Coastal zones, or more specifically the LECZ, are subject to a multitude of threats

that are expected to be increasing with time. We present here the major sources of

coastal flood and erosion risks (sea level rise, changes in storm surge activity,

growing populations in the LECZ) and the anthropogenic impacts aggravating the

flood risks, introduce the basic mechanisms behind them, describe trends

(changes) in these risks, and discuss the impact they currently have and will have

in the future on coastal flooding and erosion risks.

1.1.1 Sea Level Rise

Sea level rise comprises of two sets of mechanisms: (1) a variation of the volume of

the ocean, i.e. the eustatic sea level rise, and (2) a variation of the land elevation

relative to the sea surface, i.e. tectonic uplift or subsidence of the land. Eustatic sea

level change and tectonic land elevation change together result in relative (local or

regional) sea level rise (IPCC, 2013; Rovere et al., 2016).

Chapter 1

4

The eustatic sea level rise is predominantly driven by the global climate change.

The warming of the Earth’s climate system due to anthropogenic activities is

contributing to the melting of the ice sheets and continental glaciers, adding water

volume to the oceans (IPCC, 2013; Rovere et al., 2016). In parallel, the storage of a

large part of the climate system’s heat in the oceans and the climate warming

result in an increasing sea surface temperature (SST). Since 1971 the oceans have

warmed by 0.44°C, causing the thermal expansion of the oceans’ waters (Church et

al., 2011; IPCC, 2013). The melting of glaciers and the thermal expansion of the

oceans are the major contributors to the observed global mean sea level rise since

the 1970s, i.e. for about 75 % (IPCC, 2013). The projections for global mean sea

level rise over the 21st century show that 30 to 55 % will be due to the thermal

expansion and 15 to 35 % to the melting of glaciers (IPCC, 2013). The global mean

sea level or eustatic sea level rose by 19 cm between 1901 and 2010. By 2100, it is

expected to rise by an additional 0.26 to 0.98 m (IPCC, 2013).

The relative sea level rise on the other hand is a local to regional phenomenon that

- apart from eustatic sea level change - is also driven by land uplift or subsidence

resulting in a falling or rising sea level respectively (IPCC, 2013; Rovere et al.,

2016). The mechanisms involved in the relative sea level rise correspond to

tectonic movements of the Earth’s crust (e.g. seismic activity, mantle flow),

sediment compaction in the coastal area (e.g. mechanical sediment compaction,

biological degradation or human-driven compaction), isostatic adjustment of the

Earth’s crust (i.e. response of the lithosphere to its loading or unloading due to

erosion, deposition, ice accumulation, deglaciation or water loading and

unloading), or to the results of human activities (e.g. dredging, river diversion, soil

drainage...) (IPCC, 2013; Newman, 1982; Rovere et al., 2016).

Relative sea level rise implies the gradual submergence and increased risk of

flooding of the coastal zone, as well as an increase of the risk of coastal erosion

processes, resulting in the destruction of existing coastal infrastructures as roads

or buildings (Bernstein et al., 2007; Nicholls & Cazenave, 2010). On the other hand,

sea level rise increases the risk of salt water intrusion into the groundwater, which

is expected to reduce the freshwater availability, especially in coastal areas

(Bernstein et al., 2007; Nicholls & Cazenave, 2010).

1.1.2 Storm Surges

Storm surges are exceptional variations in the sea surface level caused by the

atmospheric disturbance of a tropical or extra-tropical storm (Flather, 2001; Resio

& Westerink, 2008) (Figure 1.2). They are considered as potentially one of the

General introduction

5

most devastating features of a storm event, as they can cause extensive flooding,

and they can lead to severe coastal erosion events, albeit that the coastal

landscape may slowly recover to its original stage after the erosion event (Flather,

2001; Masters, 2012; McIvor, Spencer, et al., 2012; Resio & Westerink, 2008).

Especially tropical storms (or otherwise called cyclones in the Indian Ocean,

hurricanes in the Atlantic and North-eastern Pacific oceans, or typhoons in the

North-western Pacific ocean, Figure 1.3) can generate the most threatening storm

surges on Earth.

There are three main atmospheric forcing mechanisms contributing to the

creation of a storm surge (Figure 1.2):

(1) The strongest forcing mechanism is the wind set-up; a storm depression

(tropical or extra-tropical) causes a circulation of air, leading near the

Earth’s surface to strong winds blowing towards the low atmospheric

pressure centre of the storm depression. These winds start to spiral (rotate)

due to the Coriolis force. The drag exerted by these rotating storm winds on

the sea surface is creating a water circulation. The water pushed to the

centre of the storm sinks and, when in deep waters, is carried away by deep

ocean currents. Yet, as the storm is moving towards the shore and reaches

the shallower coastal waters, the deep water currents are hindered due to

friction in the shallow water and hence they can no longer carry away the

piled up water at the centre of the storm. The water partially spreads then

horizontally but mainly piles up to create a mound of water, i.e. the surge,

that will propagate inland when the storm moves from the sea towards land

and makes landfall (Flather, 2001; Masters, 2012; Resio & Westerink, 2008).

(2) The inverse barometric effect; the low atmospheric pressure at the centre of

the storm is raising the sea surface level by about 1 cm for each drop of 1

millibar, causing the water level at the eye of the storm to bulge and as such

adding height to the storm surge. In comparison to the wind forcing, this

effect is relatively low, and it is the only effect visible in deep waters

(Flather, 2001; Masters, 2012).

(3) The wave set-up forcing; when the waves generated by the high winds break

onto the shore, their energy (wave momentum) is transferred to the water

column causing an increase of the surge height (Flather, 2001; Komar, 1998;

Masters, 2012; McIvor, Spencer, et al., 2012). The storm waves are also the

major cause of coastal erosion during storm surge events (Flather, 2001;

Masters, 2012; McIvor, Spencer, et al., 2012; Resio & Westerink, 2008).

Chapter 1

6

Figure 1.2 Mechanisms behind the formation of storm surges and their inland propagation. Adapted from http://www.nola.com/hurricane/images/scourgeofsurge.pdf

Additionally numerous factors affect the height and extent of storm surges. They

are firstly influenced by the storm characteristics, as the radius of the maximum

winds (storms of larger radius generate higher peak water levels and flood

volumes), the storm forward speed (faster forward speeds are expected to create

higher surges in combination with lower flood volumes and inversely), the storm

track (the angle and direction at which the storm approaches the coast is decisive

for its impact) and the storm wind intensity and atmospheric pressure (stronger

winds as well as lower pressures produce higher surge) (Flather, 2001; McIvor,

Spencer, et al., 2012; Rego & Li, 2009; Resio & Westerink, 2008). The height of

storm surges is also related to the timing of the storm surge relative to the tide

level, particularly in areas of large tidal variations (e.g. macro-tidal areas). As such,

a storm surge reaching the coast around high tide in a macro-tidal environment

can lead to devastating consequences while the same surge reaching the coast at

low tide may be unnoticed (Flather, 2001; McIvor, Spencer, et al., 2012; Paul,

2009). Secondly, the bio-geomorphology of the coast largely influences the impact

of the incoming storm surge, e.g. the near-shore bathymetry (surges are higher in

large shallow-water coastal areas than in narrow and steep off-shore sloping

coastal areas), the concave or convex shape of the coastline (surges become larger

in case of landward converging coastlines such as funnel-shaped embayments and

estuaries), the geometry of estuarine and deltaic channels and water bodies

(channels allow the surge to propagate inland more easily) and the friction

exerted by the land surface that is flooded by the surge (increased surface

roughness or friction will slow the surge’s inland propagation and lower its

height) (Flather, 2001; McIvor, Spencer, et al., 2012; Rego & Li, 2009; Resio &

Westerink, 2008). Moreover, storm surges are generally associated with heavy

rainfall, which may cause additional flood risks through rainfall driven runoff and

General introduction

7

the freshwater discharge in estuaries and river deltas (McIvor, Spencer, et al.,

2012; Resio & Westerink, 2008; Woodruff et al., 2013).

In the coming decades, as a response to the global climate change, the current

frequency, intensity and tracks of the tropical and extra-tropical storms (Figure

1.3) and of the associated storm surges is expected to change (Knutson et al., 2010;

Tessler et al., 2015; Vitousek et al., 2017; Webster et al., 2005; Woodruff et al.,

2013). Tropical cyclones are formed and intensified over warm sea surfaces

(temperatures higher than 26°C) (Knutson, 2014; Webster et al., 2005). As oceans

and tropical oceans become warmer (IPCC, 2013), the maximal intensity of

cyclones might increase, as the most intense cyclones seem to not develop over

cool sea surface, but over warmer seas (Knutson, 2014). This will imply a rise in

the maximal wind speed of tropical cyclones by 2 to 11 % by 2100 (Knutson, 2014;

Knutson et al., 2010) and a higher occurrence of high intensity cyclones and fewer

occurrence of low intensity cyclones (Knutson, 2014; Webster et al., 2005). The

tracks of tropical and extra-tropical cyclones are also expected to shift with for

example a strengthening of storm tracks North of the British Isles or in the eastern

Pacific (Bengtsson et al., 2006). Observations and comparisons of the last two

cyclone seasons (2016 and 2017) to the general trend highlight that warmer SST

(sea surface temperatures) are at play in the strong cyclone activity of 2017 (Lim

et al., 2018). While the cyclone activity in 2016 showed some shifts in storm

tracks, with few cyclones in the main development region and the development of

some extreme and unusual storms in the Atlantic basin (Collins & Roache, 2017).

Chapter 1

Figure 1.3 Representation of the storms’ tracks and intensities following the Saffir-Simpson scale for the period 1842 to 2017 over the world based on the International Best Track Archive for Climate Stewardship (IBTrACS) (Knapp et al., 2010) and of the different storm appellations over the world

General introduction

9

1.1.3 Coastal Population Increase and Anthropogenic Activities

The world’s coastlines are populated since the early development of the

civilizations (Maddison, 2001; McEvedy & Jones, 1978). Current trends are leading

to the densification of the coastal population and to the growing value of the

associated assets located along the coasts (Guzmán et al., 2009; Small & Nicholls,

2003). By 2060, the population density of the LECZ is expected to reach 405 to 534

inhabitants per square kilometre, which corresponds to two times the current

LECZ’s density and ten times the current world’s average (Green & Short, 2003;

Mcgranahan et al., 2006; Neumann et al., 2015; Nicholls et al., 2008). This

increasing population, mostly concentrated in highly populated cities, or

‘megacities’, is putting more and more people and assets at flooding and erosion

risks (Bernstein et al., 2007; Von Glasow et al., 2013; Hallegatte et al., 2013;

Hanson et al., 2011; de Sherbinin et al., 2007).

Additionally, the increased human pressure and anthropogenic activities are

responsible for the disturbance, degradation and loss of natural coastal processes

and ecosystems, which are associated with increasing flood and erosion risks

(Auerbach et al., 2015; Balke & Friess, 2016; Barbier, 2014; Hanson et al., 2011;

Syvitski, 2005; Syvitski et al., 2009). Some of those disturbances are related to soil

drainage for urban and agricultural development that leads to soil compaction and

lower soil permeability or to the mining and extraction of groundwater, oil and gas

from the coastal substrate that contributes to coastal land subsidence. It is also

related to the shortage of sediment supply to deltas and estuaries due to upstream

river dams and levees that contribute in some areas to an increased erosion as the

limited deposition of sediments cannot counterbalance the land subsidence

(Kirwan & Megonigal, 2013; A. Murray, 2017; Pethick & Orford, 2013; Rovere et

al., 2016; Syvitski & Saito, 2007; Tessler et al., 2015). The embankment of tidal

wetlands along deltaic or estuarine channels for agriculture, aquaculture or urban

development purposes on the other hand leads to the loss of flood water storage

areas and to a rising level of the water in channels (Pethick & Orford, 2013;

Smolders et al., 2015). And the deepening and widening of the estuarine and

deltaic channels (e.g. for industrial shipping facilities) may facilitate the inland

propagation of tides and storm surges (Loder et al., 2009; Temmerman et al.,

2012).

Chapter 1

10

1.2 Ecosystems in the coastal zone

Coastal zones are home to a large diversity of marine and coastal ecosystems

(Figure 1.4), such as mangrove forests, salt marshes, dunes, seagrass meadows,

coral reefs, oyster reefs, rocky shores... In this PhD thesis our interest is mainly

focused on tidal wetlands (i.e. mangrove forests and salt marshes) and on seagrass

meadows and coral reefs, because they are widely considered to contribute to

mitigation of the storm surges impacts (see below) and because global data are

available for these ecosystems (Figure 1.4).

1.2.1 Ecosystems Types

Salt marshes in temperate regions and mangrove forests in sub-tropical and

tropical regions, together called tidal wetlands throughout this thesis, are

predominately present in river deltas or estuaries, where the low-lying coastal

plain is large and where low wave energy allows the development of vegetation on

muddy shores (Alongi, 2009; Scott et al., 2014; Wolanski & Elliott, 2015). They are

highly influenced by regular tidal flooding and drainage; their salt tolerant

vegetation colonizes the shores between the mean sea level and the highest tides

or extreme water levels, by following a vertical plant zonation related to the

exposure time to marine water, the stress generated by waves and the salinity

(Alongi, 2009; Mcowen et al., 2017).

Salt marshes are the most widespread tidal wetlands, mostly occupying

temperate latitudes. They are also found in polar regions and at the landward side

of mangrove forests in tropical latitudes with specific plant species adapted to

extreme cold or warm temperature conditions, respectively. Salt marsh vegetation,

consisting of tall halophytic grasses, herbs and low shrubs (Mcowen et al., 2017;

Scott et al., 2014; Wolanski & Elliott, 2015), is covering an area of about 55 000

km² over the full world (Mcowen et al., 2017).

Mangrove forests are dominated by trees and shrubs and occur in

tropical and subtropical regions, where the temperatures stay relatively warm (>

16°) all year round (Giri et al., 2011; Scott et al., 2014; Spalding et al., 2010;

Wolanski & Elliott, 2015). Different types of mangroves are found based on the

hydrodynamic conditions of the environment (Giri et al., 2011; Scott et al., 2014;

Wolanski & Elliott, 2015); riverine mangroves (R-type) are found in sheltered tidal

estuaries with an upstream input of freshwaters and nutrients and tidal creeks

draining the wetlands. The tree height and density is the highest in this type of

mangroves (≈ 20 m in height). Fringing mangroves (F-type) are located along

General introduction

11

coastlines, they are protected from the sea by coral reefs or headlands, and have

lower tree height and density. Basin mangroves (B-type) are present in inland

depressions where the water flow is very limited (water stagnation), and where

the trees do not exceed 10 m in height. Lastly, dwarf mangroves, located in

isolated and very stressful environments, are often present in clumps of very small

trees (≈ 2 m) (Giri et al., 2011; Scott et al., 2014; Spalding et al., 2010; Wolanski &

Elliott, 2015). Mangrove forests are covering an area of 152 000 km² over the

world (Giri et al., 2011; Spalding et al., 2010).

Seagrass meadows or seagrass beds develop in all climate regions.

Seagrass species are permanently submerged marine and estuarine flowering

plants and show large variation in shape and size (Green & Short, 2003). They

principally develop on sandy or muddy substrates in shallow clear waters in the

form of extensive beds, isolated patches or as part of a habitat mosaic (i.e. in

proximity and with ecological links with other marine and coastal habitats

(Duarte, 1991; Green & Short, 2003; Koch et al., 2006). Changing environmental

conditions are relatively well managed by seagrasses that are highly dynamic and

can migrate to new areas in relatively short timeframes (Green & Short, 2003;

O’Brien et al., 2017; Short & Neckles, 1999). Seagrasses are observed over an area

of about 177 000 km² over the world (Green & Short, 2003).

Coral reefs are marine habitats defined by a physical structure or skeleton

that keeps growing during the coral’s life by accumulation of calcium carbonate

and that is colonized by a multitude of organisms. Present between the 30°N and

30°S latitudes, corals need a solid substrate to develop, in addition to warm (> 16-

18°C), salty and clear water (e.g. low amount of suspended sediments) (Bessell-

Browne et al., 2017; Duckworth et al., 2017; Spalding et al., 2002). Yet, once

settled, they can grow vertically to adapt to slow changes of their environmental

conditions (e.g. rising sea level, lower light availability...). Several types of coral

reefs exist based on their location from the shore (e.g. fringing reefs, barrier reefs,

atolls) and their structure (Spalding et al., 2002). Coral reefs are present over

about 285 000 km² along the world’s coastlines (Spalding et al., 2002).

Chapter 1

Figure 1.4 Worldwide distribution of the four coastal ecosystems considered in this thesis, i.e. mangrove forests (Giri et al., 2011), salt marshes (Mcowen et al., 2017), seagrass meadows (Green & Short, 2003) and coral reefs (Spalding et al., 2002)

General introduction

13

Figure 1.5 Illustration of the four above-described ecosystems; (a) salt marsh, (b) mangrove forest, (c) seagrass meadow and (d) coral reef

1.2.2 Coastal Ecosystem Functions

Each ecosystem, whether it are tidal wetlands, coral or oyster reefs, seagrass beds,

coastal dunes or rocky shores, provides a range of ecological functions and

ecosystem services. A set of those ecological functions and services, sometimes

common to multiple ecosystem types, are described here.

Nutrient cycling inside the different ecosystems generates ecological functions and

services such as nutrient and food supply, water purification (e.g. pollutant uptake,

removal of excess nutrients) or carbon sequestration by the burial of the carbon in

anoxic soil (Barbier et al., 2011; Duarte et al., 2013; Duckworth et al., 2017;

Fourqurean et al., 2012; McLeod et al., 2011; Mumby & Steneck, 2018; Tack et al.,

2007; Temmerman et al., 2004; Teuchies et al., 2013). In terms of monetary value

of those ecosystem services, the water purification function of salt marshes in the

USA is estimated to provide an equivalent to traditional waste water treatments

amounting up to 15 000 US$ per acre (0.004 km²) (Barbier et al., 2011; Breaux et

al., 1995). While the carbon sequestration capacity of the tidal wetlands at an

Chapter 1

14

overall average rate of 210 g of CO2 per square-meter per year converted to the

Carbon Emission Reduction price corresponds to a monetary valuation of about 30

US$ per hectare per year (Barbier & Barbier, 2014; Chmura, 2003; McLeod et al.,

2011).

Due to the high concentration of nutrients and food in coastal ecosystems, as well

as due to their sheltered environment, they are providing a suitable reproduction

and nursery habitat for fish, shellfish and crustaceans (Barbier et al., 2011; S. Y.

Lee et al., 2014; Millennium Ecosystem Assessment, 2005; Ondiviela et al., 2014).

Mangroves, salt marshes and seagrass meadows provide a protection against

larger predators that cannot penetrate the complex structures of vegetation, while

larger fish (e.g. tuna, sharks...) find a diverse food source in coral reefs (Barbier et

al., 2011). Coastal ecosystems play a major role in the maintenance of healthy

fisheries. For example, in Thailand the mangrove forests can contribute to up to 1

000 US$ per hectare of capitalized increased offshore fishery production (Barbier,

2007). Coral reefs in the Philippines provide fishes for local consumption and live-

fish export for up to 45 000 US$ per km² per year and 10 000 US$ per km² per

year respectively (White et al., 2000). The loss of 127 km² of seagrass in Australia

was estimated to result in a loss of fishery production of 235 000 AU$ (Barbier et

al., 2011; McArthur & Boland, 2006).

Coastal ecosystems provide a highly valuable recreational and touristic

environment due to their unique landscapes and diversity of fauna and flora.

Subsequently, they generate revenues for the local populations, as in the

Seychelles where 40 000 tourists per year visit the marine parks generating

88 000 US$ (Mathieu et al., 2003), or along the North Carolina’s beaches where

tourists spend on average 166 US$ per trip (Barbier et al., 2011; Landry & Liu,

2009).

Materials provided by coastal ecosystems (e.g. wood from mangrove forests, lime

of coral reefs, grasses and herbs from salt marshes...) can be also highly valuable

and can be cut, extracted or harvested for human use (Barbier et al., 2011; Bolund

& Hunhammar, 1999; Reddy et al., 2016). In the UK, livestock grazing in salt

marshes generates about 15 £ per hectare per year of net income (King & Lester,

1995) and the gathering of mangrove products in Thailand is estimated at some

600 US$ per hectare per year (Barbier, 2007; Gedan et al., 2011).

Moreover, coastal ecosystems are providing a protection against coastal flood and

erosion risks (see below). In the Indian Ocean, the loss of coral reefs due to the

bleaching event of 1998 is estimated to have diminished the property value by 174

US$ per hectare per year due to the loss of coastal protection offered by the former

General introduction

15

coral reefs (Barbier et al., 2011; Wilkinson et al., 1999), while the presence of salt

marshes along the U.S. Atlantic and Gulf Coasts reduced hurricane induced

damages by more than 8 000 US$ per hectare per year (Costanza et al., 2008).

However, the anthropogenic development and pressure along the coastlines

during the last centuries resulted in large losses and degradations of the different

coastal ecosystems (Almeida et al., 2014; Hoeksema, 2007; Lotze et al., 2006; Scott

et al., 2014; Valiela et al., 2009). The worldwide loss of mangrove forests, salt

marshes and seagrass meadows is estimated at 20 to 50 % (Barbier et al., 2008;

McLeod et al., 2011; Millennium Ecosystem Assessment, 2005; Spalding et al.,

1997; Valiela et al., 2001), while 75 % of the global coral reefs are under threats

from the changing environmental conditions and anthropogenic impacts (Spalding

et al., 2002; Spalding, Ruffo, et al., 2014). Although recognized as a dramatic trend,

the loss of coastal ecosystems and services for the benefit of anthropogenic

activities is still ongoing (Arkema et al., 2013; Sutton-Grier et al., 2018; Tian et al.,

2016; Valiela et al., 2009). Projections for the next 100 years estimate the future

losses at 30 to 40 % of the actual considered coastal ecosystems via land

reclamation, wetlands degradation (e.g. over-exploitation, solid waste disposal...)

and shifts in the bio-geochemical conditions of the coastal areas, such as ocean

acidification threatening coral reefs (Blankespoor et al., 2014; Duckworth et al.,

2017; Duke et al., 2007; Gilman et al., 2008; IPCC, 2013; Ma et al., 2014; Pendleton

et al., 2012; Valiela et al., 2001).

1.3 Flood Protection Function of Coastal Ecosystems

An increasing number of studies have highlighted that coastal ecosystems

contribute to the mitigation and attenuation of wave and storm surge reaching the

coastal zone, and to the reduction of shoreline erosion rates (Barbier et al., 2011;

Cheong et al., 2013; Duarte et al., 2013; Guannel et al., 2016; Shepard et al., 2011;

Temmerman et al., 2013). The retention of sediments by natural coastal vegetation

(including salt marshes, mangrove forests and seagrass beds) and the roots

structure that stabilizes the sediments are crucial for the reduction of coastal

erosion risks, but also to allow vertical sediment accretion and build-up of soil

elevation in balance with sea level rise (Alongi, 2008; Kirwan et al., 2010, 2016;

Kirwan & Megonigal, 2013; McIvor et al., 2013; Ondiviela et al., 2014; Storlazzi et

al., 2011). The attenuation of wave and storm surge heights on the other hand is

provided by the friction exerted by the vegetation and soil surface on the moving

water column (Barbier et al., 2008, 2011; Gedan et al., 2011; Guannel et al., 2016;

Mazda et al., 1997, 2006; Narayan et al., 2016; Temmerman et al., 2013).

Chapter 1

16

A multitude of parameters are affecting the specific rate at which wave heights and

storm surge levels are reduced by the ecosystem. This rate of wave or storm surge

attenuation is often expressed as the amount of vertical wave height or storm

surge height reduction per horizontal distance travelled by the wave or storm

surge through the ecosystem (i.e. typically in cm/km for storm surge attenuation

rates). Factors that influence the specific wave and storm surge attenuation rates

provided by a specific ecosystem, include the topographic conditions (e.g.

shoreline slope, bathymetry...), the ecosystem and vegetation characteristics (e.g.

geometry, density, continuity of the vegetation cover, plant stiffness...) or the

extreme water event type, e.g. wind wave, storm surge or tsunami and their

characteristics (e.g. duration, intensity, track, wave period, wave height...)

(Leonardi et al., 2018; Marsooli et al., 2016; Resio & Westerink, 2008). As such,

defining a unique quantitative value of the rate of storm surge attenuation by a

distance travelled through the coastal ecosystems is not possible (see discussion

below and Table 1.1) (Barbier et al., 2008; Gedan et al., 2011; Koch et al., 2009;

Resio & Westerink, 2008).

Nonetheless, there is increasing knowledge on the rates of storm surge mitigation

by coastal ecosystems, and the factors that influence these rates. They are based

on three kinds of studies.

Firstly, there are the indirect measurements of the reduction of damages and

deaths behind coastal ecosystems during storms or tsunamis. Costanza et al.

(2008) estimated that the loss of 1 ha (0.01 km²) of marsh land could increase the

averaged storm damages by 33 000 US$ in the USA. Das & Vincent (2009)

observed fewer deaths in villages fronted by a mangrove forest during the cyclone

that struck India in 1999, yet this is also related to the elevation of the area and

therefore to the flooding depth as highlighted by Vermaat & Thampanya (2006).

Adriana Gracia et al. (2018) as well as Danielsen et al. (2005) reported that

villages located behind coastal vegetation belts experienced fewer damages and

deaths during the 2004 tsunami in the Indian Ocean and studies demonstrate that

the loss of coral reefs makes the construction of artificial coastal defences

necessary (Brown & Dunne, 1988; Ferrario et al., 2014; Frihy et al., 1996).

Although coastal protection seems to be often enhanced by the presence of the

coastal ecosystems, in some situation, the implementation of nature-based

strategies can also adversely lead to a reduced protection, if it is not implemented

wisely. In India for example, sand dunes were flattened and replaced by mangrove

trees which decreased the amount of nature-based coastal protection (R. A. Feagin

et al., 2010).

General introduction

17

The second type of studies are direct in situ measurements of storm surge height

reduction by the coastal ecosystems (Adriana Gracia et al., 2018; Fonseca &

Cahalan, 1992; Ondiviela et al., 2014). A number of studies along the marsh and

mangrove systems of the US Gulf coast (Lovelace, 1994; McGee et al., 2006;

Wamsley et al., 2010) resulted in a rule of thumb of 6.9 cm of storm surge height

reduction per kilometre of marsh crossed, although large variations of this rate are

possible (Leonardi et al., 2018; Stark et al., 2015). The storm surge attenuation

rates for tidal wetlands found in literature are presented in Table 1.1. They range

from 1 cm per kilometre of tidal wetlands up to 25 cm/km in salt marshes and 50

cm/km in mangrove forests. In general, the higher vegetation canopy and denser

root system of the mangrove forests is expected to exert more friction and

therefore attenuate the storm surge at a higher rate than the lower vegetation of

the salt marshes.

Thirdly, insights on storm surge height reduction come from hydrodynamic

modelling studies that simulate storm surge propagation over idealized or realistic

landscapes (Haddad et al., 2016; Hu et al., 2015; Liu et al., 2013; Loder et al., 2009;

Marsooli et al., 2016; Resio & Westerink, 2008; Sheng et al., 2012; Smolders et al.,

2015; Stark et al., 2016; Temmerman et al., 2012; Zhang et al., 2012). Such

modelling studies highlight that the effectiveness of storm surge attenuation by

coastal ecosystems depends on the storm surge forcing (e.g. forward moving

speed, duration, storm track...), the properties of the ecosystem itself (e.g. bed

elevation, surface roughness, vegetation induced friction, upstream/downstream

location in the deltas and estuaries, ecosystem size...) and the bathymetry and

topography of the surrounding large-scale coastal landscape (e.g. continental shelf

slope, channels structure...) (Leonardi et al., 2018). Each of these factors can

enhance or reduce the storm surge attenuation provided by the tidal wetlands; a

summary of the existing insights from the literature is presented in Table 1.2.

In tidal wetlands, i.e. salt marshes and mangrove forests, there are two main

mechanisms behind the storm surge mitigation (Figure 1.6); (1) the so-called

within-wetland attenuation, where the surge’s energy is absorbed through friction

induced by the vegetation canopy and by the topographic variations of the bed

surface (Barbier et al., 2013; Costanza et al., 2008; Leonardi et al., 2018; Mazda et

al., 1997, 2006), and (2) the so-called along-channel attenuation, where the excess

of water brought by the surge can flow into the low-lying tidal wetlands areas

adjacent to the channels within deltas and estuaries, thereby reducing the height

of a storm surge while it is travelling inland along a channel (Kobashi & Mazda,

2005; Leonardi et al., 2018; Smolders et al., 2015; Stark et al., 2016).

Chapter 1

18

Figure 1.6 Scheme illustrating the within-wetland and along-channel attenuation potential of tidal wetlands. Adapted from Stark (2016)

Compared to tidal wetlands, there are less available studies on storm surge

mitigation by seagrass meadows and coral reefs.

Seagrass meadows significantly influence the hydrodynamic environment by

reducing the flow velocity, dissipating the wave’s energy and stabilizing the

sediments. Consequently, they diminish erosion risks, and reduce storm surges at

a magnitude expected to be comparable to salt marshes (Duarte et al., 2013;

Fonseca & Cahalan, 1992). The maximal storm surge reduction happens in shallow

water and low wave energy environments, where the canopy height accounts for

more than 15 % of the water column (Adriana Gracia et al., 2018; Ondiviela et al.,

2014).

The structural complexity of coral reefs results in hydraulic roughness and a great

frictional effect of the reef on the water column (Ferrario et al., 2014; Harris et al.,

2018; UNEP-WCMC, 2006). The storm surge reduction, by coral reefs can reach 97

% under hurricanes conditions over the whole reef, with most of the attenuation

happening over the reef flat (i.e. the shallow part of the reef that extends from the

reef crest, or the seaward edge of the reef, and the shore) (Ferrario et al., 2014;

Harris et al., 2018; Principe et al., 2012; UNEP-WCMC, 2006). Coral reefs show

comparable storm surge attenuation capacities than artificial coastal defences

(Ferrario et al., 2014; Principe et al., 2012; UNEP-WCMC, 2006).

General introduction

Table 1.1 Storm surge attenuation rates across salt marshes and mangrove forests based on in situ measurements and hydrodynamic models. Adapted and completed from McIvor et al. (2012) and Stark et al. (2015).

Location vegetation type Event Attenuation

rate (cm/km)

Reference

Southern Louisiana coastal marsh Compilation of 7 storms between 1909 and 1957

1.6 - 20 United States Army Corps of Engineers (2006)

Louisiana marsh & open water Hurricane Andrew (1992), cat. 5 4.4 - 4.9 Lovelace (1994)

Het Verdronken Land van Saeftinghe, Western Scheldt, Netherlands

marsh & channels Simulations validated with in situ measurements

0 - 25 Stark et al. 2016

Cameron Prairie, Louisiana marsh Hurricane Rita (2005), cat. 3 10.0 Wamsley et al. (2010) calculated with data from McGee et al. (2006)

Sabine, Louisiana marsh Hurricane Rita (2005), cat. 3 25.0 Wamsley et al. (2010) calculated with data from McGee et al. (2006)

Vermillion, Louisiana marsh Hurricane Rita (2005), cat. 3 4.0 Wamsley et al. (2010) calculated with data from McGee et al. (2006)

Vermillion, Louisiana marsh Hurricane Rita (2005), cat. 3 7.7 Wamsley et al. (2010) calculated with data from McGee et al. (2006)

Ten Thousand Island National Wildlife Refuge, Florida

mangrove & interior marsh

Hurricane Charley (2004), cat. 3 9.4 Krauss et al. (2009)

Shark River, Everglades National Park, Florida

riverine mangrove Hurricane Wilma (2005), cat. 3 4.3 - 6.9 Krauss et al. (2009)

Ten Thousand Island National Wildlife Refuge, Florida

mangrove Hurricane Charley (2004), cat. 3 15.8 Krauss et al. (2009)

Everglades National Park, Florida mangrove Simulations validated with Hurricane Wilma (2005)

20 - 50 Zhang et al. (2012)

Everglades National Park, Florida no vegetation Simulations validated with Hurricane Wilma (2005)

6 - 10 Zhang et al. (2012)

Chapter 1

20

Table 1.2 Summary of the existing insights from the literature on the characteristics of storm surges, wetlands and channels that influence the attenuation rate of a storm surge (i.e. that enhance or reduce the attenuation)

Enhanced Attenuation

Reduced Attenuation

Mechanism References

Storm Surge

Level Shallow to moderate

Extreme Friction decreases with an increased water depth

Lawler et al., 2016; Resio & Westerink, 2008; Sheng et al., 2012; Wamsley et al., 2010

Duration Short (hours)

Long (days)

Long storms have more time to propagate inland and to fully flood the wetlands area

Resio & Westerink, 2008; Wamsley et al., 2010

Forward Moving Speed

Fast (e.g. 10 m/s)

Slow (e.g. 5 m/s)

Slow moving storm surges have more time to impact the coastal waters

Hu et al., 2015; Liu et al., 2013; Rego & Li, 2009; Sheng et al., 2012; Zhang et al., 2012

Wetlands

Soil Elevation High Low

Filling of the wetlands by the surge is longer and potentially still possible at the maximum surge height in high elevated wetlands soil

Loder et al., 2009; Smolders et al., 2015; Stark et al., 2016

Width Large Narrow Large wetlands are providing more storm surge water storage and for a longer time

Loder et al., 2009; Smolders et al., 2015; Stark et al., 2016

Location Upstream Downstream Upstream wetlands can store higher percentages of flood volume relative to downstream wetlands

Smolders et al., 2015

Accretion High Low Accreting wetlands keep up with sea level rise and maintain their mitigation capacities

Temmerman et al., 2012; Wamsley et al., 2010

Vegetated Wetland/Open

Water High Low

A high level of continuity in wetland vegetation cover provides higher attenuation rates

Hu et al., 2015; Loder et al., 2009; Schepers et al., 2017; Sheng et al., 2012; Zhang et al., 2012

Channels

Width Narrow Wide Wide channels concentrate and facilitate the landward propagation of the surge

Stark et al., 2016; Temmerman et al., 2012

Depth Shallow Deep Shallow wetlands channels exert more friction on the storm surge propagation

Stark et al., 2016; Temmerman et al., 2012

General introduction

21

1.4 Coastal flood defence strategies

Engineered coastal defence structures commonly consist of dikes, levees or dams.

Those structures can be found all over the world, and are especially widespread in

rich developed countries, as in The Netherlands, were most of the coastline is

delimited by dikes and embankments to prevent coastal flooding of large low-lying

lands that have been historically gained from conversion and drainage of coastal,

estuarine and deltaic wetlands (Hoeksema, 2007; Pierik et al., 2017; Wolff, 1993);

or as in the Mississippi delta or East Asian deltas (Ma et al., 2014; Temmerman &

Kirwan, 2015). These structures sometimes combine protective and recreational

functions as the ‘Promenade des Anglais’ in Nice (France) (Pranzini et al., 2015).

Whilst the knowledge and expertise on the implementation and functioning of the

engineered structures are large and well-established, their creation, construction

and maintenance is expensive (Leonardi et al., 2018; R. L. Morris et al., 2018;

Temmerman et al., 2013). Consequently, the presence of coastal protection

structures is not solely linked to the richness of the country but on the willingness

of the policy makers to invest in such structures (Nicholls et al., 2008). Some

coastlines present then low safety standards, because of non-reliable or non-

existent flood defence structures (Dasgupta et al., 2009; Mcgranahan et al., 2006;

Nicholls et al., 2008; de Sherbinin et al., 2007), that might create a false feeling of

security to the coastal communities (Sutton-Grier et al., 2015). Engineered coastal

protection structures present other disadvantages, for instance, they disturb

natural coastal processes such as sediment supply to tidal wetlands, which is an

essential process through which tidal wetlands can build up land elevation with

sea level rise (Firth et al., 2014; Sutton-Grier et al., 2015; Temmerman & Kirwan,

2015). In addition, although they are able to protect coastal communities against

the impacts of storm events during a certain period of time, they have a certain

lifetime and their strength weakens with age; furthermore engineered flood

defence structures do not have a self-adaptive capacity in response to changing

environmental conditions such as sea level rise (Firth et al., 2014; Leonardi et al.,

2018; R. L. Morris et al., 2018; Sutton-Grier et al., 2015) (Figure 1.7).

As coastal protection is a necessity and will need an enhanced efficiency and

adaptation in regards to the expected socio-economic and climatic evolutions and

associated threats, nature-based solutions for coastal flood and erosion risks

reduction are increasingly studied as an alternative and add-on to standard

engineered flood defences. The nature-based solutions rely on the ability of the

coastal ecosystems to attenuate the flood and erosions risks, while providing

additional ecological functions, or ecosystem services (e.g. carbon sequestration,

Chapter 1

22

water purification, habitat and nursery for different animals...) (Barbier et al.,

2011; McLeod et al., 2011; Sutton-Grier et al., 2018). They have the advantage to

be sustainable and self-adaptive, meaning that the need for human management of

the ecosystems is limited. Furthermore, the natural ecosystems are, under certain

conditions, prone to adapt to gradual environmental changes (e.g. sea level rise,

sea surface warming, change in sediment supply...) and cost-effective; their limited

management needs make them financially less demanding than hard engineering

structures especially in the face of climate change (Barbier et al., 2013; Rao et al.,

2013; Reguero et al., 2018; Schueler, 2017; Spalding et al., 2013)(Figure 1.7).

Figure 1.7 Traditional engineering coastal protection structures (top) versus nature-based coastal protection (bottom) and their associated impact on the environment. Blue arrows indicate an increase or decrease in intensity of storm wave, storm surge and sea level. Adapted from Temmerman et al. (2013).

General introduction

23

Over the last decade, nature-based coastal protection and mostly hybrid

structures, i.e. combinations of nature-based and hard engineering structures,

were increasingly developed at several locations around the world. It involves the

incorporation of the existing coastal ecosystems in the coastal protection planning,

but also the restoration or creation of those ecosystems. However, still little is

known on the capacities of the restored or created ecosystems for storm surge

mitigation. Although they are expected to not recreate a pristine environment

(Elliott et al., 2016; Lawrence et al., 2018), several studies suggest that they could

be able to provide ecosystems services as water quality regulation or wind waves

and storm surge mitigation, yet probably not to the same extent as natural

ecosystems (Bullock et al., 2011; Hobbs et al., 2009; Rupprecht et al., 2017;

Spalding, McIvor, et al., 2014). One of the largest examples is the creation, in the

aftermaths of the landfall of hurricanes Katrina and Rita (August and September

2005) in Louisiana, of the Coastal Protection and Restoration Authority (CPRA).

The CPRA, which for the first time integrates coastal restoration and hurricane

protection under a single clear voice for the full state, has the mandate to develop,

implement and enforce hybrid strategies for safe and sustainable coasts that will

protect the local communities, the industrial infrastructures and the natural

resources. Between 2007 and 2017, 135 projects of restoration and risks

reduction were funded and realized. They involved the restoration of 146 km² of

coastal habitat, the improvement of 454 km of levees and the construction of

about 100 km of barrier islands and berms. Following this, the five-year strategic

plan started in 2017 comprises 124 projects of ecosystem restoration and flood

risks reduction (Figure 1.8) financed for a total amount of 50 billion US$. It will

lead to the restoration of about 2 000 km² of coastal habitat, and the reduction of

the expected coastal damages by 150 billion US$ over the next 50 years (Boesch et

al., 2006; Coastal Protection and Restoration Authority of Louisiana, 2017; Day et

al., 2007). Similarly, the San Francisco Bay Joint Venture works, since 2001,

towards the restoration of wildlife and wetlands in the San Francisco Bay with the

combined objective to gain benefits for wildlife and coastal protection. To date,

some 550 km² of tidal flats, marshes and lagoons and 260 km² of seasonal

wetlands were protected, restored or enhanced (San Francisco Bay Joint Venture,

2018).

In North-western Europe, natural environments are also increasingly incorporated

in coastal planning to (re)create coastal habitats and deliver sustainable coastal

flood and erosion risks reduction to the local communities (Esteves, 2014;

Gardiner et al., 2007; Meire et al., 2014; Rupp-Armstrong & Nicholls, 2007;

SigmaPlan, 2017). In England and Wales, the Making Space for Water policy

(2005), shifted the coastal protection focus from hard engineering to hybrid

Chapter 1

24

strategies integrating the natural environment. It resulted in multiple projects of

‘managed coastal realignment’ (Figure 1.8) that consists of the landward

relocation of coastal flood defence structures in order to provide space for tidal

marshes development. The realignment will concern about 660 km of coastline by

2030 and recreate about 62 km² of intertidal areas (Esteves, 2014; French, 2006;

Pendle, 2013). Similarly, in Belgium and in The Netherlands, several projects of

‘depoldering’ are implemented. In Belgium, the SigmaPlan has the goal to improve

the protection against the flooding of the Scheldt river and to develop the valuable

nature along the Scheldt (SigmaPlan, 2017). The plan consists on the

strengthening and heightening of hard defence (especially dikes) and the

construction of controlled flooding areas in former polder areas, and some of these

controlled flooding areas will be designed as such that they can develop in tidal

flats and marshes. One of the major projects is the Hedwige-Prosper polder project

(Figure 1.8) straddling Belgium and The Netherlands, in which the tidal flats and

marshes will be restored over and area of 4.65 km² of formerly embanked land

(SigmaPlan, 2017). As in the UK, this project consists of the landward relocation of

the dikes enabling the creation of tidal marshes on formerly embanked land.

After the 2004 tsunami in South East Asia and the typhoon Haiyan in the

Philippines in 2013, the registered damages in the coastal zones highlighted lower

damages in villages located behind mangrove forests (Balke & Friess, 2016;

Dahdouh-Guebas et al., 2005; Danielsen et al., 2005), and increased the interest in

ecosystems for coastal flood and erosion risks mitigation in those regions. Several

associations and countries (i.e. Indonesia, India, Sri Lanka, Thailand and Malaysia)

collaborated to develop mangroves restoration projects (FAO, 2007; Schmitt,

2012) that allowed amongst others the restoration of 20 km² of mangrove forest

in Indonesia and the plantation of 310 000 seedlings over the Sri Lanka’s coasts

(Schmitt, 2012). Nature-based coastal protection strategies are then not just a

theoretical concept for coastal protection, they are already implemented at small

to large scales in several countries over the world (Esteves, 2014; Marois & Mitsch,

2015; MFF Pakistan, 2016; Schmitt, 2012; Spalding, McIvor, et al., 2014).

General introduction

25

Figure 1.8 (A) Completed or funded projects of the Louisiana Coastal Master Plan for 2017. It includes 79 marsh restoration projects, 13 structural protection projects, and 32 non-structural risk reduction projects (Coastal Protection and Restoration Authority of Louisiana, 2017); (B) location and type of 54 managed realignment projects along the UK’s coasts (Esteves, 2014); (C) Map of the depoldering of the Hedwige-Prosper polder along the Scheldt (Belgium and The Netherlands) (SigmaPlan, 2017).

Chapter 1

26

1.5 Thesis Objectives and Outline

Although an increasing amount of studies accounts for the contribution of coastal

ecosystem to nature-based mitigation of coastal flood and erosion risks on local to

regional scales there are, to our knowledge, no studies that have explored the

potential of nature-based coastal risk mitigation on regional to global scales.

Global scale studies have the advantage to reach more easily the local communities

and policy-makers and are then needed to promote the integration of coastal

ecosystems in coastal flood defence strategies. As the construction and

maintenance of hard engineering structures for coastal protection is expensive

and needs extensive maintenance, large portions of the world’s coasts have a low

protection and are vulnerable to coastal flood and erosion risks. Furthermore, the

implementation of nature-based or hybrid strategies for coastal flood and erosion

risks mitigation are often made in the aftermaths of destructive storm surge

events (i.e. Caribbean coast in the USA, South East Asia or in the Philippines).

Therefore, highlighting the worldwide presence of coastal ecosystems and their

potential for coastal flood and erosion risks mitigation is a necessity to increase

their incorporation in coastal defence strategies before having to deal with the

consequences of devastating storm surge events. As such, global scale studies will

promote the need for new local scale studies and spread to the different

stakeholders and policy-makers the possibility to implement nature-based

strategies while including the local characteristics of the area. Furthermore, with

the knowledge of the benefits coastal ecosystems can provide, local communities

and policy-makers can act to terminate the practices of land reclamation for

human use that weaken the coastal zone.

Subsequently, throughout this thesis, we pursued the general aim to identify

hotspots for nature-based storm surge flood risks mitigation, at a regional and

worldwide scale. Those hotspots were determined on the one hand as the deltas

plains, worldwide coastal plains and coastal cities with high needs for storm surge

mitigation, i.e. highly populated low-lying areas, low-lying areas with valuable

assets or coastal areas frequently exposed to storm surges. And, on the other hand,

as the deltas, the worldwide coastal plains and the coastal cities that have a high

potential for nature-based strategies via the incorporation of the existing or

potentially re-created coastal ecosystems into the coastal protection planning,

mostly as add-ons to the hard engineering structures.

Firstly, we created a GIS model based on globally available datasets that defines

the storm surge mitigation function of tidal wetlands for the low-lying areas of

General introduction

27

eleven highly populated deltas (Chapter 2) and for the entire world’s coastline

(Chapter 3).

Secondly, in order to create a comprehensive global assessment of the hotspots for

nature-based flood risks mitigation, we quantified the current extent of the four

above-described coastal ecosystems (salt marshes, mangrove forests, seagrass

meadows and coral reefs) in front of highly populated and flood-exposed coastal

cities (Chapter 4). Subsequently, in line with existing projects of nature-based

coastal risk mitigation by tidal wetland restoration or creation that are executed at

several places around the world, we estimated the potentially available areas for

tidal wetlands restoration or creation (Chapter 5) in front of those highly

populated and flood exposed coastal cities.

We conclude this thesis by a synthesis (Chapter 6) including the main findings of

the different chapters as well as some discussions on the limitations of our

approaches, on future researches and on managerial implications.

We explored the following aspects more specifically.

In Chapter 2, we created a GIS model based on globally available data and

relatively simple assumptions, simulating how a storm surge will be routed from

the open sea towards the potential floodplain within a delta. The model allows the

identification of the surface areas and population numbers within the delta which

are flooded via a flood pathway crossing through tidal wetlands. As such the model

identifies the areas and population that receive a mitigating effect from storm

surge attenuation by the tidal wetlands, and the magnitude of this mitigating effect

is estimated by the length of the flood pathway crossing through tidal wetlands. To

take a first step towards a global assessment of the potential contribution of tidal

wetlands to storm surge risk mitigation at a quasi-global scale, we applied the

model on 11 highly populated and flood-exposed deltas having tidal wetlands and

scattered over the world.

Subsequently, in Chapter 3 we made an upscaling of our quasi-global model to a

fully global model assessing the coastal areas and population numbers benefiting

from coastal flood risks mitigation by tidal wetlands at a worldwide scale. The aim

was to identify the specific areas, or hotspots, where mangrove forests and salt

marshes can provide the highest storm surge flood risk mitigation.

The following chapters are focusing on 136 coastal cities of more than 1 million

inhabitants that are exposed to coastal flooding due to storm surges. In Chapter 4,

we developed a GIS procedure defining the most likely pathway a storm surge

Chapter 1

28

would take to propagate from the open sea towards the city centres and the area

around this likely pathway that is potentially influencing the storm surge

propagation. In the area influencing the storm surge propagation, we quantified

the extent of four coastal ecosystems that are known to mitigate storm surge

propagation, i.e. mangrove forests, salt marshes, seagrass meadows and coral

reefs. As such, we identified the hotspot cities that have large surface area of

coastal ecosystems along their storm surge pathways, and therefore have a high

potential for nature-based flood risks mitigation. Furthermore, we aimed at

identifying the social and physical parameters explaining the spatial variations in

ecosystem surface areas in front of the 136 studies cities.

In Chapter 5, we explored the possibility of tidal wetlands’ restoration or creation

in the area influencing the storm surge propagation from the open sea towards the

city centre for the 136 coastal cities as defined in Chapter 4. Based on the

topography, the tidal amplitude, the land use and the population density, we

estimated the potentially available area (km²) where tidal wetlands could be

restored or created to enhance the nature-based flood risk mitigation in front of

the cities. We identified the hotspot cities for tidal wetlands restoration or creation

and a set of social and physical parameters influencing the size of the area

potentially available for tidal wetlands creation.

Deltas

29

CHAPTER 2 Contribution of mangroves and salt marshes to nature-based mitigation of coastal flood risks in major deltas of the world

Rebecca Van Coppenolle, Christian Schwarz, Stijn Temmerman

This chapter is based on Van Coppenolle, R., Schwarz, C., & Temmerman, S. (2018). Contribution of Mangroves and Salt Marshes to Nature-Based Mitigation of Coastal Flood Risks in Major Deltas of the World. Estuaries and Coasts, 41(6), 1699–1711. https://doi.org/10.1007/s12237-018-0394-7

Chapter 2

30

Abstract

Nature-based solutions are rapidly gaining interest in the face of global change and

increasing flood risks. While assessments of flood risk mitigation by coastal

ecosystems are mainly restricted to local scales, our study assesses the

contribution of salt marshes and mangroves to nature-based storm surge

mitigation in 11 large deltas around the world. We present a relatively simple GIS

model that, based on globally available input data, provides an estimation of the

tidal wetland’s capacity of risk mitigation at a regional scale. It shows the high

potential of nature-based solutions, as tidal wetlands, to provide storm surge

mitigation to more than 80% of the flood-exposed land area for 4 of the 11 deltas

and to more than 70% of the flood-exposed population for 3 deltas. The magnitude

of the nature-based mitigation, estimated as the length of the storm surge pathway

crossing through tidal wetlands, was found to be significantly correlated to the

total wetland area within a delta. This highlights the importance of conserving

extensive continuous tidal wetlands as a nature-based approach to mitigate flood

risks. Our analysis further reveals that deltas with limited historical wetland

reclamation and therefore large remaining wetlands, such as the Mississippi, Niger

and part of the Ganges-Brahmaputra deltas, benefit from investing in the

conservation of their vast wetlands, while deltas with extensive historical wetland

reclamation, such as the Yangtze and Rhine deltas, may improve the sustainability

of flood protection programs by combining existing hard engineering with new

nature-based solutions through restoration of former wetlands.

Deltas

31

2.1 Introduction

Global climate change induces acceleration of sea level rise and is expected to

increase the intensity of storm surges, and as such is threatening coastal and

deltaic areas worldwide (Bengtsson et al., 2006; Hallegatte et al., 2013; Hinkel et

al., 2014; Woodruff et al., 2013). Storm surges originating from severe storms,

such as tropical cyclones, propagate from the sea towards the land with surge

heights that can reach several meters above mean sea level, causing densely

populated low-lying areas in river deltas to be particularly vulnerable to storm

surge flood risks; (Day et al., 2007; Tessler et al., 2015). Additionally, the globally

averaged population density in the Low Elevation Coastal Zone (LECZ, i.e. less than

10 meters above mean sea level) is expected to grow from 241 people/km² in

2015 (i.e. five times the world’s average) to 405 to 534 people/km² by 2060

(McGranahan et al., 2007; Neumann et al., 2015; Small & Nicholls, 2003).

This increase of both coastal population density and risk probability of coastal

flooding events calls for the development of sustainable coastal management

strategies. Apart from traditional hard engineered flood defence structures, such

as dams or dikes, nature-based or ecosystem-based coastal flood defence is

increasingly proposed as an alternative or addition to traditional hard

engineering, and relies on the conservation and in certain cases restoration of

coastal and deltaic ecosystems (Cheong et al., 2013; Duarte et al., 2013;

Temmerman et al., 2013). Here we focus on salt marshes and mangrove forests,

which we collectively call tidal wetlands throughout this paper. Among their

valuable ecosystem services, tidal wetlands have the capacity to attenuate waves,

reduce shoreline erosion and inland storm surge propagation, and to sustain

themselves with sea level rise by allochthonous sediment accretion (Gedan et al.,

2011; Kirwan et al., 2016; Shepard et al., 2011; Temmerman & Kirwan, 2015). As

such, nature-based flood risk mitigation is increasingly regarded as a polyvalent,

self-adaptive and sustainable strategy (Temmerman et al., 2013).

Observational and hydrodynamic modelling studies have demonstrated the value

of tidal wetlands for storm surge mitigation due to the resistance exerted by the

wetland vegetation and topography on incoming storm surges, implying a

landward attenuation in storm surge height further referred to as storm surge

attenuation or reduction (Barbier et al., 2013; Costanza et al., 2008; Krauss et al.,

2009; Stark et al., 2015; Wamsley et al., 2010; Zhang et al., 2012). This attenuation

is quantified as the rate of vertical reduction in storm surge height per horizontal

inland distance over the delta plain (expressed in cm/km). It depends on various

factors as the flow resistance provided by the coastal geomorphology and its

Chapter 2

32

vegetation or land cover type and on the specific properties of the storm surge,

such as its height and duration (Loder et al., 2009; McIvor, Spencer, et al., 2012;

Stark et al., 2015; Wamsley et al., 2010). Observed rates of storm surge attenuation

over tidal wetlands range from a couple of cm/km to 25 cm/km (Krauss et al.,

2009; Stark et al., 2015; Wamsley et al., 2010), with maximum rates of up to 50

cm/km reported from a hydrodynamic modelling study in Florida’s mangroves

(Zhang et al., 2012). Although such local to regional observational and modelling

studies play an important role in advancing our understanding of the role of tidal

wetlands in storm surge risk mitigation, there are no upscaling studies yet that

have explored the potential contribution of tidal wetlands to storm surge risk

mitigation on a quasi-global scale.

Our study aims to take a first step towards such a global assessment of the

contribution of tidal wetlands to nature-based storm surge risk mitigation, by

selecting 11 of the most populated deltas around the world. We present results

from a GIS model based on globally available data and on relatively simple

assumptions to define the storm surge mitigation function of tidal wetlands for

low-lying delta areas and populations.

2.2 Material and Method

2.2.1 Study areas

For the selection of the studied deltas, the world’s deltas were ranked according to

their total population size as reported by Ericson et al. (2006). Starting from this

list, the selection of the deltas was firstly based on the presence of tidal wetlands,

then on the highest population, and lastly on their global distribution, so that at

least one delta per continent (North America, South America, Europe, Africa, Asia

and Australia) was selected (Table 2.1).

Some deltas could not be included in the study due to the lack of data regarding

the distribution of salt marshes as it is the case for the Pearl River delta in China.

Deltas

33

Table 2.1 Main characteristics and geographical distribution of the deltas selected for the study. A more extensive description of the deltas can be found in the Supplementary Information.

Country Tidal

wetland type

Population (Ericson et al., 2006)

Delta area as delineated

in the study (km²)

Tidal wetlands within the

delta (km²) Ganges-

Brahmaputra delta India/

Bangladesh Mangrove 111,000,000 78,453.9 6,431.6

Yangtze delta China Salt Marsh 42,100,000 61,251.9 121.1

Mekong delta Vietnam Mangrove 20,200,000 50,135.2 1,540.4

Chao Phraya delta Thailand Mangrove 13,700,000 19,177.3 174.3

Irrawaddy delta Myanmar Mangrove 9,720,000 28,744.0 1,252.0

Niger delta Nigeria Mangrove 3,730,000 16,714.8 5,646.6

Amazon delta Brazil Mangrove 2,930,000 42,028.3 1,676.0

Rhine delta The

Netherlands Salt Marsh 1,940,000 9,139.9 40.0

Mississippi delta USA Salt Marsh 1,790,000 36,894.3 6,014.3

Mahakam delta Indonesia Mangrove 706,000 2,425.1 503.5

Burdekin delta Australia Mangrove 5,800 1,441.4 129.7

2.2.2 Datasets

The following datasets were used.

The land elevation is defined with the NASA Shuttle Radar Topography

Mission (SRTM) Global 3 arc second V003 dataset (NASA JPL, 2013). The SRTM

dataset is so far the best-known Digital Elevation Model (DEM) available at a

global scale (Rodriguez et al., 2006; Sun et al., 2003). It is found to be more

accurate in areas with small slopes, such as deltas, yet there can be errors due to

reflection of the radar signal on vegetation canopies with an absolute vertical

error up to 16 m (http://www2.jpl.nasa.gov/srtm/datafinaldescriptions.html)

(Rodriguez et al., 2006; Sun et al., 2003).

The tidal wetlands distribution is based on different types of datasets. The

representation of mangroves is based on the Global distribution of mangroves from

the United States Geologic Survey (USGS, www.unep-wcmc.org) (Giri et al., 2011).

The distribution of salt marshes is a compilation of different country wide or

continent wide datasets as the European Commission program Corine Land Cover

of 2006 for Europe, the Geohabitats for Australia (Heap et al., 2001), the

Classification of Wetlands and Deepwater habitats for the United States (Federal

Chapter 2

34

Geographic Data Committee, 2013) and the Chinese wetlands mapping from Niu et

al. (2009).

The storm surge height is taken from the DIVA database (Hinkel et al., 2014;

Vafeidis et al., 2005). It corresponds to the storm surge water level above mean

sea level and is calculated by model simulations based on tidal levels, barometric

pressures, wind speeds and sea bed slopes for return periods of 10, 100 and 1000

years. The DIVA database uses the coastline of the Digital Chart of the World

(DCW, Environmental Systems Research Institute, ESRI, 2002) divided in

segments based on administrative and environmental parameters (Vafeidis et al.,

2005). To avoid inconsistencies due to differences in scale between the datasets

(e.g. tidal wetlands on the seaward side of the DIVA coastline), the DIVA coastline

was not directly used as the source of the flooding. Alternatively a ‘flood source’,

i.e. the coastline along the seaward delta front, is interpolated from a convex hull

based on the land area (the latter is defined more below). The storm surge heights

of the different segments stored in the DIVA coastline are then transferred to the

flood source with a shortest Euclidean distance algorithm (for further information

see the Supplementary Information).

The population distribution originates from the LandScan 2013 Global

Population Database (Bright et al., 2013). It represents the population over a 30

arc second grid resolution and integrates the diurnal movements and collective

travelling behaviour of the world population, i.e. the so-called “ambient

population”, averaged over 24 hours (Bright et al., 2013; Dobson et al., 2000). The

dataset of 30 arc second resolution was resampled to a resolution of 3 arc second

(to match the resolution of the SRTM land elevation dataset) based on the

guidelines of the LandScan documentation (Bright et al., 2013; UT BATTELLE

LLC.).

The extent of the world countries is the representation of the country boundaries

as they exist in January 2015 and is available through the ESRI platform (ESRI,

DeLorme Publishing Company, Inc., 2015). Due to the fact that the borders of the

different global datasets do not perfectly overlap (Lichter et al., 2011), the most

seaward extent of the emerging land was defined by merging the extent of the

world countries dataset and the tidal wetlands datasets, and this land extent is

further referred to as the land area.

The delta areas were delineated in accordance with spatial delineation of the

deltas in other studies (e.g. Coleman & Huh, 2004; Syvitski et al., 2009).

Deltas

35

The global datasets used present some limitations in regards to local data

accuracy and local data artefacts. Such limitations may include vegetation artefacts

in the SRTM dataset or the moderate resolution of the tidal wetlands datasets that

can omit some of the smaller wetlands areas for example. In addition, the use of a

global scale storm surge dataset based on the static calculation of the storm surge

height can introduce limitations related for example to the angle at which the

storm is making landfall that is not accounted for, or to the exact storm surge

height at one specific spot of the delta as the surge heights are defined for

segments that may not account for very local variations of the coastal area.

The resolution of all raster layers was converted to a 3 arc second grid based on

the World Geodetic System 1984 ellipsoid, which corresponds to the resolution of

the SRTM data grid.

2.2.3 Model description

The model simulates how a storm surge flood wave, entering a delta system from

the seaward delta front, would be routed in a landward direction through the

channels and over the potential floodplain of the delta, assuming that no artificial

flood protecting structures like dikes or dams would be present. The non-inclusion

of the protective structures as dikes or dams in the model refer to what would

happen if the protecting structure would fail. In this case of failure the model

highlights which coastal areas and coastal populations could still benefit from

storm surge mitigation thanks to the existing coastal ecosystems. As such, the salt

marshes and mangrove forests that existed before the construction of the

protecting structures are not accounted for in the study. It is based on a GIS

procedure developed in ArcGIS (10.3.1), and is similar to previously published

procedures that assess the coastal areas and number of people vulnerable to

storm surge flooding on regional to global scales (Arkema et al., 2013; Dasgupta et

al., 2011). The model does not simulate the full complexity of hydrodynamic

processes involved in flood propagation and therefore is not able to calculate

accurate flood depths and absolute values of reduction in flood depth behind tidal

wetlands. Nevertheless, it has the major advantage to be globally applicable to

make a relative comparison between delta systems around the world.

The input data of the model are globally available GIS data presented above in

combination with the storm surge attenuation rates derived from the range of

values found in the literature (Table 2.2) for the three land cover types considered

in this study (Table 2.3). The open water and channel areas are attributed a very

low attenuation value of 0.1 cm/km (Table 2.3) assuming that the flood height

Chapter 2

36

attenuation by friction over these areas is small. The remaining land area within

the delta is considered as a unique land cover type and is based on the literature

values of hydrodynamic modelling of storm surge propagation over coastal low

land areas (Table 2.2, (Zhang et al., 2012)), and on the assumption of a lower

attenuation rate over human-developed land (typically dominated by agricultural

land in deltas) than over natural wetlands. A conservative option was taken and

assumes that the remaining land areas have an average attenuation rate of 6

cm/km (Table 2.3). The attenuation rates for the tidal wetlands relate then to this

conservative rate over the remaining land. Additionally, the higher vegetation

canopy of the mangroves is expected to exert more friction and to result in higher

storm surge height attenuation (10 cm/km) than the lower vegetation of salt

marshes (8 cm/km) (Table 2.3). It is important to note, that there is much

uncertainty about which precise values to use for these attenuation rates as they

will vary with factors like the different land use. Currently little is known on how a

storm surge will be attenuated over an urban land use for example. For this reason

a sensitivity analysis was performed using a likely range of input values for the

attenuation rates for these land cover types (see Table 2.3), showing that the

model is relatively insensitive to it (see next section below).

Deltas

37

Table 2.2 Rates of storm surge height attenuation across various tidal wetland types based on observations during storm surges over tidal wetlands areas or on models calibrated by observations. Adapted and completed from McIvor et al. (2012) and Stark et al. (2015).

Location vegetation

type Event

Attenuation rate

(cm/km) Reference

Southern Louisiana coastal marsh Compilation of 7 storms between 1909 and 1957

1.6 - 20 United States Army Corps of Engineers (2006)

Louisiana marsh & open water

Hurricane Andrew (1992), cat. 5 4.4 - 4.9 Lovelace (1994)

Cameron Prairie, Louisiana marsh Hurricane Rita (2005), cat. 3 10.0 (Wamsley et al. 2010 calculated with data from McGee et al. 2006)

Sabine, Louisiana Marsh Hurricane Rita (2005), cat. 3 25.0 (Wamsley et al. 2010 calculated with data from McGee et al. 2006)

Vermillion, Louisiana Marsh Hurricane Rita (2005), cat. 3 4.0 (Wamsley et al. 2010 calculated with data from McGee et al. 2006)

Vermillion, Louisiana Marsh Hurricane Rita (2005), cat. 3 7.7 (Wamsley et al. 2010 calculated with data from McGee et al. 2006)

Ten Thousand Island National Wildlife Refuge, Florida

mangrove and interior marsh

Hurricane Charley (2004), cat. 3 9.4 (Krauss et al., 2009)

Shark River (Everglades National Park) Florida

riverine mangrove

Hurricane Wilma (2005), cat. 3 4.3 - 6.9 (Krauss et al., 2009)

Ten Thousand Island National Wildlife Refuge, Florida

mangrove Hurricane Charley (2004), cat. 3 15.8 (Krauss et al., 2009)

Everglades National Park, Florida Mangrove Simulations validated with Hurricane Wilma (2005)

20 - 50 (Zhang et al., 2012)

Everglades National Park, Florida no vegetation Simulations validated with Hurricane Wilma (2005)

6 - 10 (Zhang et al., 2012)

Chapter 2

38

Table 2.3 Storm surge attenuation rates attributed to the three land cover types considered in the study and used for the sensitivity analysis

Land cover type Attenuation rate

(cm/km) Attenuation rates for the

sensitivity analysis (cm/km)

Open water and Channels 0.1 Mangrove Salt Marsh

10.0 8.0

5, 10, 15, 20, 30, 40, 50

Remaining land 6.0 4, 6, 8, 10, 15, 20

The model works as follows. A cost distance algorithm is applied over the delta to

define the route that the storm surge follows during its landward propagation, i.e.

the storm surge flood pathway. The cost distance algorithm defines the flood

pathway between the flood source and every location (every pixel) of the delta

system as the route where the traveling cost of the storm surge is the lowest based

on the distance travelled and on the friction of the different land covers, or

attenuation rates (Table 2.3). It then allocates the cost of traveling, i.e. the

attenuation experienced by the storm surge, to every location. Subsequently,

pixels where the resulting storm surge height is higher than the land elevation are

considered at risk of flooding. The different steps of the model are presented for

the example of the Ganges-Brahmaputra delta in India and Bangladesh (Figure

2.1).

The model produces two main outputs for storm surge return periods of 10, 100

and 1000 years:

(1) It identifies the areas within the delta, and the number of people living in

those areas, that would be flooded via flood pathways crossing tidal

wetlands. We assume that areas flooded via those routes would benefit

more from nature-based attenuation of the storm surge, as compared to

areas flooded via pathways that do not cross tidal wetlands. In order to

select the pixels that are flooded or not flooded via tidal wetlands, two

scenarios were compared. The first scenario represents the existing extent

of tidal wetlands while the second scenario represents a situation where

the tidal wetlands would be replaced by the remaining land cover type and

its corresponding average attenuation rate (Table 2.3). Finally, all pixels

with a higher storm surge attenuation for scenario 1 as compared to

scenario 2 are identified as pixels having a storm surge pathway crossing

tidal wetlands.

Deltas

39

(2) It identifies the length of the flood route crossing tidal wetlands as a proxy

for the magnitude of nature-based storm surge flood risk mitigation. We

assume that the longer the flood wave travels through tidal wetlands

before it reaches inhabited land, the more that flood wave will be

attenuated. This was done by dividing the difference in storm surge levels

between the two scenarios by the difference in attenuation rates of the

tidal wetlands and remaining land (Table 2.3).

Figure 2.1 Maps of the Ganges-Brahmaputra delta illustrating the steps of the model. (a)Input data: topography of the delta (meters above mean sea level), area of the mangrove forest and location of the source of the storm surge. (b) Estimated flood-prone areas for a 1 in 100 year storm surge accounting for different storm surge attenuation rates over mangroves, open water and land area (Table 2.3), i.e. scenario 1. (c) Flood-prone areas with a flood pathway passing through the mangrove forest. (d) Mangrove forest length along the flood pathway for every pixel classified into the five distance classes

Further, we introduced a measure for the relative magnitude of nature-based flood

risk mitigation, which gives one value for the entire delta based on the length of

the tidal wetlands along the flood pathways. This length is classified into distance

classes i, which are classes of length of tidal wetlands along the flood pathway

(Table 2.4). The relative magnitude is calculated then in terms of land area (Mland)

and population (Mpop) benefiting from nature-based flood risk mitigation from the

following formula:

Chapter 2

40

𝑀 = ∑𝑁𝑖

𝑁∗ 𝑊𝑖

(Equation 1)

Where M is the relative magnitude of the nature-based flood risk mitigation, Ni is

the number of pixels or inhabitants in the distance class i, N is the total amount of

pixels or inhabitants having a storm surge pathway crossing tidal wetlands and W

is the weight of the distance class i (Table 2.4). The weight values follow a linear

function based on the mean of each length class.

Table 2.4 Value of the weight (W) of each class of length of tidal wetlands along the flood pathway

Distance class (i)

Classes of tidal wetlands length along the flood pathway (m)

Mean length of the distance

class (m)

Weight of the distance class

(W)

1 0 – 500 250 1 2 500 – 1000 750 3 3 1000 – 2000 1500 6 4 2000 – 5000 3500 14 5 > 5000 7500 30

2.2.4 Sensitivity analysis

As the range of values for storm surge attenuation over land cover types underlie a

certain variability in the literature, a sensitivity analysis on the Ganges-

Brahmaputra delta (India and Bangladesh), i.e. the largest delta in our study, was

performed to assess the dependence of our results to the chosen attenuation rates.

The tested parameters were the storm surge attenuation rate of the remaining

land (4, 6, 8, 10, 15 and 20 cm/km) and of the tidal wetland area (in this case

mangrove forest) (5, 10, 15, 20, 30, 40 and 50 cm/km) (Table 2.3).

2.3 Results

The results of the sensitivity analysis shows that the variation of the attenuation

rates of the mangrove forest has a limited impact on the land area benefiting from

storm surge mitigation by tidal wetlands, with averaged results of 1 812.60 km² ±

1.20 %; while the variation in the attenuation rates of the remaining land has a

higher influence, with averaged results of 1 664.20 km² ± 12.80 % of land area

benefiting from storm surge mitigation by tidal wetlands.

Deltas

41

In the model, we assume one single value for the attenuation rate of the remaining

land area, which is a simplification of the reality, as it is expected that the use of

different attenuation rates for different land use types (urban area, fields,

forests...) would have an impact on the propagation of the storm surge.

Nevertheless, the results of the sensitivity analysis suggest that the model output

is robust in regards to the tested range of input values for the storm surge

attenuation rates over the remaining land and tidal wetlands.

The results presented here after are for storm surges with a return period of 1 in

100 years. The results for the other return periods can be found in the

Supplementary Information and show qualitatively similar results.

2.3.1 Areas and populations benefiting from nature-based flood risk

mitigation

Figure 2.2 shows for each delta the land area (i.e. excluding the wetland areas

themselves) and the population benefiting from flood risk mitigation by tidal

wetlands, both in absolute numbers and in percentages. The percentages are

expressed relative to the total land area or total population at risk of flooding

within each delta for the current situation (scenario 1), i.e. the flood-exposed land

area or population.

Considering the land area benefiting from flood risk mitigation by tidal wetlands,

the response to the presence of tidal wetlands varies among deltas. In absolute

numbers, the largest land areas buffered by tidal wetlands are located in the

Mekong (13 806 km²), Yangtze (9 229 km²) and Mississippi delta (4 407 km²),

while the Burdekin (71 km²) represents the smallest land areas benefitting from

nature-based flood risk mitigation (Figure 2.3). Interestingly, this variation

between deltas in total land area buffered by tidal wetlands is not correlated to the

variation of the total tidal wetland area between deltas (Figure 2.3, Pearson’s r = -

0.073, p = 0.83). For example, the largest wetland areas are found in the Ganges-

Brahmaputra (6 432 km²), Mississippi (6 015 km²) and Niger delta (5 647 km²),

but these deltas do not represent the largest land areas buffered by tidal wetlands

(Figure 2.3). In terms of relative percentages, the Ganges-Brahmaputra, the

Irrawaddy, the Amazon and the Rhine deltas, have about 15 to 20 % of their flood-

exposed land area buffered by tidal wetlands. This percentage rises up to about 40

to 60 % for the Mekong, Burdekin and Yangtze deltas, while the other deltas

present percentages of more than 80 % of the flood-exposed land area benefiting

from flood risk mitigation by tidal wetlands.

Chapter 2

Figure 2.2 Land surface area (left/blue) and number of people (right/red) benefiting from flood risk mitigation by tidal wetlands. The map represents the absolute land area (km²) or population (number of people) through the colour of the symbols, while the size represents the percentage of land area or population buffered by wetlands relative to flood-exposed land area or population

Deltas

43

Figure 2.3 (a) Relation between the land area benefiting from flood risk mitigation (km²) and the tidal wetlands surface area (km²) (Pearson’s r = - 0.073; p = 0.83), (b) Relation between the population benefiting from flood risk mitigation (inhabitants) and the tidal wetlands surface area (km²) (Pearson’s r = - 0.17, p = 0.61), for the 11 deltas studied

A similar pattern is found for the population buffered by tidal wetlands. In

absolute numbers, the deltas with the highest number of inhabitants buffered by

tidal wetlands are the Yangtze (5 922 009 inhabitants) and Mekong (4 602 641

inhabitants) delta, while the Burdekin delta is by far the delta with the lowest

number of people (169 people) buffered by tidal wetlands. Also for the total

population buffered by wetlands within a delta, there is no significant correlation

to the total tidal wetland area within that delta (Pearson’s r = -0.17, p = 0.61). In

relative percentages, the Ganges-Brahmaputra, Irrawaddy, Amazon and Rhine

deltas show the lowest percentages of flood-exposed population benefiting from

flood risk mitigation by tidal wetlands (less than 20 %). The percentages of the

Mekong, Burdekin, Yangtze and Mississippi deltas are of 25.4, 46.8, 49.1, and 58.1

% respectively, and the other deltas have more than 70 % of their flood-exposed

population buffered by tidal wetlands, rising up to 98 % for the Mahakam delta.

2.3.2 Relative magnitude of the nature-based storm surge flood risk

mitigation

The magnitude of the nature-based storm surge flood risk mitigation defined via

the length of the flood pathway passing through tidal wetlands shows a large

variability among deltas, with mean distances inside tidal wetlands ranging from

234 m to more than 2 km.

The comparison of the relative magnitude of nature-based flood risk mitigation is

based on the magnitude M ( (Equation 1) in terms of land surface area (Mland) and

Chapter 2

44

delta population (Mpop) (Table 2.5). The magnitude is higher when a larger

proportion of land area or population is buffered by a wider length of tidal

wetlands.

Table 2.5 Value of the relative magnitude of nature-based flood risk mitigation in term of land area (Mland) and population (Mpop) for every delta

Delta

Ranking Ranking

Mland Mland Mpop Mpop

Ganges-Brahmaputra delta 6.26 3 1.53 9

Yangtze delta 1.07 11 1.02 11

Mekong delta 2.70 8 2.41 5

Chao Phraya delta 3.10 7 1.89 6

Irrawaddy delta 3.62 6 1.62 8

Niger delta 8.09 2 5.31 1

Amazon delta 4.47 4 2.95 4

Rhine delta 1.37 10 1.12 10

Mississippi delta 9.26 1 1.86 7

Mahakam delta 2.53 9 3.07 3

Burdekin delta 4.37 5 4.42 2

The relative magnitudes for land area range from 1.07 to 9.26 and for the delta

population from 1.02 to 5.31. Except for the Mahakam and Burdekin deltas, all the

deltas present higher magnitudes in terms of land area than in terms of

population. However, the degree of deviation between Mland and Mpop varies among

the deltas (Table 2.5). The Mississippi, Ganges-Brahmaputra and Niger deltas

present the higher differences, while the Chao Phraya, Amazon and Irrawaddy

deltas have smaller differences and the Yangtze, Burdekin, Rhine, Mekong and

Mahakam present the lowest differences between Mland and Mpop (Figure 2.4).

Deltas

Figure 2.4 Relative magnitude of nature-based storm surge flood risk mitigation for every delta in terms of land area (left/blue) and of population (right/red)

Chapter 2

46

The comparison of the tidal wetlands area (km²) and the magnitude of flood risk

mitigation for the land area (Figure 2.5) shows a significant positive correlation

(Pearson’s r = 0.89, p = 0.0003). For an increasing tidal wetland area the

magnitude of flood risk mitigation for the land increases. The correlation between

the tidal wetlands area (km²) and the relative magnitude of flood risk mitigation

for the population is non-significant (Pearson’s r = 0.17, p = 0.61).

Figure 2.5 (a) Relation between the tidal wetlands surface area (km²) and the relative magnitude of nature-based flood risk mitigation for the terrestrial land area (Mland) (Pearson’s r = 0.88; p = 0.0003). (b) Relation between the tidal wetlands surface area (km²) and the relative magnitude of nature-based flood risk mitigation for the population (Mpop) (Pearson’s r = 0.17; p = 0.61)

2.4 Discussion

Although nature-based coastal risk mitigation is rapidly gaining interest in the face

of global change (Cheong et al., 2013; Giosan et al., 2014; Spalding, McIvor, et al.,

2014; Sutton-Grier et al., 2015; Temmerman et al., 2013), current insights into the

role of tidal wetlands for storm surge risk mitigation are mainly based on local or

regional scale assessments (Costanza et al., 2008; Das & Vincent, 2009; Krauss et

al., 2009; Stark et al., 2015; Wamsley et al., 2010; Zhang et al., 2012), and

methodologies for intermediate to global scale assessments are scarce. Our study

contributes to fill this gap by developing a GIS model assessing the impact of the

presence of tidal wetlands along the storm surge flood pathway in deltas around

the world. We applied the model on 11 deltas, selected based on their population

density, the presence of tidal wetlands and a worldwide distribution. Our results

indicate that tidal wetlands provide storm surge mitigation to large percentages of

the flood-exposed land area (> 80 %) for 4 of the 11 studied deltas, and to large

percentages of the flood-exposed population (> 70 %) for 3 of the deltas. The land

area and population buffered by tidal wetlands within a delta were not found to be

Deltas

47

correlated to the total wetland area in the delta, suggesting that more complex

factors are at play, such as the spatial distribution of population within the delta,

the delta geomorphology or the location of the tidal wetlands compared to the

location of the population. The magnitude of the nature-based storm surge risk

mitigation, estimated as the length of the storm surge pathway crossing tidal

wetlands, was found to be significantly correlated to the tidal wetlands area within

a delta. The latter finding highlights the importance of the conservation and, where

possible, restoration of extensive continuous tidal wetland areas as a nature-based

approach to mitigate storm surge flood risks in deltas.

Our quasi-global modelling approach differs from the existing complex

hydrodynamic models applied in local to regional studies, which incorporate

physical mechanisms of storm surge generation on the open sea, landward surge

propagation and attenuation of peak surge levels by friction exerted by the

landforms and vegetation of tidal wetlands and other coastal land use types

(Haddad et al., 2016; Liu et al., 2013; Marsooli et al., 2016; Resio & Westerink,

2008; Smolders et al., 2015; Stark et al., 2016; Wamsley et al., 2010; Zhang et al.,

2012). Such hydrodynamic modelling approaches are data-demanding and

computationally expensive, enabling their application on local to regional scales,

but excluding their feasibility for application to many deltas worldwide. In

contrast, our modelling approach is simple, computationally much less intensive,

and based on input datasets that are globally available. As a first step, we

demonstrated its applicability for 11 deltas, yet, the same approach is applicable to

much more deltaic or non-deltaic areas around the world. The release of new

global datasets is highly interesting, and such updated datasets should, when

possible, be used in the future applications of our modelling approach. For

example, there is the recently published global distribution of saltmarshes by

McOwen et al. (2017) and the global dataset on storm surge levels by Muis et al.

(2016).

The results presented highlight that our relatively simple method based on

globally available data of generally lower resolution than local data can provide an

estimation of the tidal wetland’s capacity of risk mitigation at a regional scale.

Nevertheless, despite the advantage of its intermediate to global applicability, our

model, unlike hydrodynamic models, does not include parameters such as the

complex geomorphology of the delta, the storm surge characteristics (such as

storm intensity, duration, direction, forward moving speed) or the wetland

characteristics (such as vegetation and geomorphic properties), making it unable

to accurately predict the flooding extent, depth or duration as a result of a specific

storm surge event.

Chapter 2

48

Hence, a direct comparison between hydrodynamic models and our GIS model is

not very informative as the methods represent different conditions in their

simulations. Hydrodynamic models are setup for specific storm surge

characteristics, including specific wind velocity fields, storm track, duration etc.,

with typical output being among others the reduction of peak storm surge height

due to the presence of tidal wetlands (Wamsley et al., 2010; Zhang et al., 2012).

Whereas our GIS model is setup for ‘statistical’ storm surge height of a given

return period, neglecting the specific storm surge characteristics mentioned

above, and aims to identify the land area and population that is exposed to flood

risks via flood pathways crossing tidal wetlands and that will benefit from reduced

flood risks due to the presence of the tidal wetlands. Nevertheless, a number of

qualitative conclusions derived from both modelling approaches are comparable,

as discussed below.

The analysis of the sensitivity of the model output to the variation of the input

values for attenuation rates for the land area (1 664.20 km² ± 12.80 %) and for the

tidal wetlands (i.e. mangrove forests) (1 812.60 km² ± 1.2 %) reveals that the

model is robust in regards to those parameters. The reason why the model output

(i.e. the land area flooded via flood pathways crossing through tidal wetlands) is

relatively insensitive to the wetland attenuation rates that are applied can be

ascribed to two points. First, a main point explaining this low variation of the land

area benefiting from storm surge mitigation by tidal wetlands is the continuity or

connectivity of the tidal wetlands. A tidal wetland area dissected by channels,

embayment and land areas will introduce a non-linear parameter in the reduction

of the surge, as it will not only be related to the attenuation rate of the wetlands,

but also to the pathway the flood can follow. Because a higher attenuation rate is

applied for the wetlands and land areas compared to the water areas, the flood

will preferably propagate over areas of lower friction, such as the channels and

water bodies. As a result, the area benefiting from flood risk mitigation by tidal

wetlands is relatively insensitive to the range of attenuation rates that were

applied in our sensitivity analysis (5 to 50 cm/km for the wetlands and 4 to 20

cm/km for the land areas), as channels in the delta plain will remain the main

pathways of flood propagation.

The second point is related to the different land use classes. In the current design

of the model, the land area is divided into two classes, the tidal wetlands and the

other land area, whilst, the division of the land area into an increased number of

classes (e.g. forests, urban areas, agricultural fields...) is expected to modify and

refine the pathway of the storm surge in the delta area.

Deltas

49

Our analysis shows that there is no correlation between the surface area of the

delta’s tidal wetlands and the land area benefiting from storm surge mitigation.

This is corroborating the fact that even small wetlands can provide flood wave

attenuation to large areas (Gedan et al., 2011). The comparison of the Ganges-

Brahmaputra and the Chao Phraya deltas further illustrates this. The Ganges-

Brahmaputra delta has a total surface area of tidal wetlands of 6 432 km² that

provides flood risk mitigation to 1 821 km² of the delta area. In comparison, the

Chao Phraya delta has 174 km² of tidal wetlands and 1 666 km² of land area

benefiting from nature-based flood risk mitigation. The non-correlation between

the wetlands surface area and the land area benefiting from storm surge

mitigation can also be related to the effect of the tidal wetlands location in the

delta and along the channels that is known to influence their capacity to mitigate

storm surges (Smolders et al., 2015; Stark et al., 2015). Following the previous

example of the Ganges-Brahmaputra and Chao Phraya deltas, the effect of the

location of the tidal wetlands in the delta can be observed. Tidal wetlands of the

Chao Phraya delta are more scattered and present along the main channels of the

delta (see maps of the deltas in the Supplementary Information), which implies

that they influence the propagation of the flood wave for a large part of the delta (1

666 km²). In contrast, tidal wetlands of the Ganges-Brahmaputra delta are

clustered, leaving some of the main channels, such as the ones running to Kolkata

or Dhaka for example, exempt of tidal wetlands. Hence, the large tidal wetland

area within the Ganges-Brahmaputra delta (i.e. the Sundarbans mangrove forest)

provides nature-based flood risk mitigation to only a certain part (1 821 km²) of

this large deltaic area.

Nevertheless, the analysis shows a significant, positive correlation between the

surface area of tidal wetlands and the magnitude of nature-based storm surge risk

mitigation (Pearson’s r = 0.88, p-value = 0.0003). This implies that a delta with a

large and/or continuous surface area occupied by tidal wetlands will benefit from

a higher magnitude of flood risk mitigation due to a longer width of tidal wetlands

crossed by the flood pathway. This finding relates to several hydrodynamic studies

that have identified the importance of wetland continuity for effective attenuation

of storm surges (Loder et al., 2009; McIvor et al., 2012; Phan et al., 2015; Zhang et

al., 2012).

In addition, studies have pointed out that tidal wetlands dissected by deep and

wide channels provide less storm surge attenuation than wetlands with narrow

and shallow channels (Loder et al., 2009; Stark et al., 2016). On the large scale of a

delta area, our results show similar effects of the dissection of the tidal wetlands

by deltaic channels. A qualitative observation of the studied deltas shows that the

Chapter 2

50

deltas having the lower relative magnitude of nature-based risk mitigation are also

the deltas having large channels dissecting the wetlands, as for example the Rhine

or the Yangtze deltas.

When a delta has a high value of land area benefiting from flood risk mitigation,

like the 86 % of the Chao Phraya delta or the 95 % of the Mahakam delta, this does

not necessarily imply a high magnitude of storm surge flood risk mitigation. The

Mland of the Chao Phraya and Mahakam deltas are 3.10 and 2.53 respectively,

almost three times lower than the magnitude of the Mississippi or the Niger deltas.

This means that although the Chao Phraya and Mahakam deltas have a high land

area benefiting from flood risk mitigation by tidal wetlands, the magnitude of the

mitigation is rather small compared to the magnitude estimated for the Mississippi

or Niger delta for which the flood pathway is crossing in general a longer width of

tidal wetlands.

Neither the absolute number of people buffered by tidal wetlands nor the relative

magnitude of nature-based storm surge mitigation in terms of population is

correlated to the total tidal wetlands area in the delta, implying that other

parameters must be considered. This may be attributed to factors such as the

spatial population distribution and density, or the historical settlement of the

population and their spatial relation to tidal wetlands. The Niger, Burdekin and

Mahakam deltas have the population benefitting from the highest magnitude of

storm surge risk mitigation in terms of population, with values of 5.31, 4.42 and

3.07 respectively. The tidal wetlands in those deltas are located in between the

population and the sea and are present over the full extent of the coastline. They

differ in population distribution, as the population of the Niger delta is distributed

over the delta plain, while the population of the Mahakam and Burdekin deltas is

concentrated in cities. However, the presence of tidal wetlands along the coastline

and along either side of the channels is influencing the propagation of the flood

pathway that crosses the tidal wetlands before reaching the population located

behind. Those deltas are in contrast to other deltas, such as the Yangtze and Rhine

deltas, where the channels are wider, with smaller tidal wetlands located only in

some regions of the delta; or to the Ganges-Brahmaputra delta, where the tidal

wetlands occupy a large portion of the delta, but a large part of the population is

not located directly behind the wetlands resulting in a lower magnitude in term of

population.

Ranking the deltas according to the relative magnitude of storm surge mitigation

in terms of land area, demonstrates that the deltas with the highest ranking are

deltas where large tidal wetlands exist and where historic wetland reclamation

Deltas

51

and conversion into human land use has been limited, such as in the Mississippi,

Niger and Ganges-Brahmaputra deltas (Table 2.5). In contrast, the deltas with the

lowest ranking have experienced large scale historical wetland reclamation, such

as in the Yangtze and Rhine delta. Hence, the difference in historical land use

management of the deltas induces differences in the future management. Deltas

with limited wetland reclamation and large remaining wetlands, such as the

Mississippi, Niger and Ganges-Brahmaputra delta would benefit from investing in

the conservation of their tidal wetlands as a nature-based strategy to mitigate

storm surge flood risks, while deltas with extensive historical wetland

reclamation, such as the Yangtze and Rhine deltas, should not only rely on hard

engineering of flood defence structures but also invest in restoration of formerly

reclaimed wetlands where possible.

2.5 Acknowledgements

The author would like to thank the different data providers and Dr. Chen Wang for

her help in the gathering of the Chinese wetlands data. This work was funded by

the University of Antwerp.

Chapter 2

52

Supplementary Information

Deltas

Ganges-Brahmaputra delta, India and Bangladesh

Figure SI 2.1 Left: Topography and location of the tidal wetlands in the Ganges-Brahmaputra delta in India and Bangladesh. Right: Ambient population in the delta

Yangtze delta, China

Figure SI 2.2 Left: Topography and location of the tidal wetlands in the Yangtze delta in China. Right: Ambient population in the delta

Deltas

53

Mekong delta, Vietnam

Figure SI 2.3 Left: Topography and location of the tidal wetlands in the Mekong delta in Vietnam. Right: Ambient population in the delta

Chao Phraya delta, Thailand

Figure SI 2.4 Left: Topography and location of the tidal wetlands in the Chao Praya delta in Thailand. Right: Ambient population in the delta

Chapter 2

54

Irrawaddy delta, Myanmar

Figure SI 2.5 Left: Topography and location of the tidal wetlands in the Irrawaddy delta in Myanmar. Right: Ambient population in the delta

Niger delta, Nigeria

Figure SI 2.6 Left: Topography and location of the tidal wetlands in the Niger delta in Nigeria. Right: Ambient population in the delta

Deltas

55

Amazon delta, Brazil

Figure SI 2.7 Left: Topography and location of the tidal wetlands in the Amazon delta in Brazil. Right: Ambient population in the delta

Rhine delta, The Netherlands

Figure SI 2.8 Left: Topography and location of the tidal wetlands in the Rhine delta in The Netherlands. Right: Ambient population in the delta

Chapter 2

56

Mississippi delta, United States of America

Figure SI 2.9 Left: Topography and location of the tidal wetlands in the Mississippi delta in the USA Right: Ambient population in the delta

Mahakam delta, Indonesia

Figure SI 2.10 Left: Topography and location of the tidal wetlands in the Mahakam delta in Indonesia. Right: Ambient population in the delta

Deltas

57

Burdekin delta, Australia

Figure SI 2.11 Left: Topography and location of the tidal wetlands in the Burdekin delta in Australia. Right: Ambient population in the delta

Chapter 2

58

DIVA dataset

The DIVA coastline is based on the coastline of the Digital Chart of the World

(DCW, Environmental Systems Research Institute, ESRI, 2002) and stores the

values of storm surge height for every segment of the coastline and for several

return periods.

In the study, the origin of the storm surge was defined as a flood source (red

line in Figure SI 2.12) and the value of the storm surge heights was transferred to

this flood source via a Euclidean distance algorithm. The Euclidean distance

algorithm delineated for every segment of the DIVA coastline its area of influence

(thin black lines in Figure SI 2.12). Then, the flood source was segmented and the

value of every segment corresponds to the value of the area of influence in which

the segment is located.

Figure SI 2.12 Representation of the DIVA coastline (blue), the flood source (red) and the areas of influence of every segment of the DIVA coastline (black lines).

Deltas

59

Results for the 1 in 10 and 1 in 1000 year storm surges

Table SI 2.1 Results of the analysis for the three storm surge return periods, 1 in 10, 100 and 1000 year. The population corresponds to the number of people benefiting from flood risk mitigation by tidal wetlands. The percentage is calculated relative to the total number of people flooded in the case of the scenario 1, i.e. with the presence of the tidal wetlands. The surface corresponds to the square kilometres of land area (excluding the tidal wetlands area itself) benefitting from flood risks mitigation by the tidal wetlands. The percentage of surface area is calculated relative to the land area (excluding the tidal wetlands area itself) flooded in the case of the scenario 1.

Delta Country Population (number of people) Population (relative %)

1 in 10 year

1 in 100 year

1 in 1000 year

1 in 10 year

1 in 100 year

1 in 1000 year

Ganges-Brahmaputra

India and Bangladesh

1,114,765 1,322,735 1,541,264 16.3 16.8 15.4

Yangtze China 4,438,443 5,922,009 7,322,102 49.1 49.1 49.0

Mekong Vietnam 4,172,316 4,602,641 4,927,996 26.3 25.4 24.7

Chao Phraya Thailand 1,079,150 1,300,156 1,644,581 74.7 76.2 74.4

Irrawaddy Myanmar 223,766 274,041 293,823 9.7 9.5 8.9

Niger Nigeria 913,558 927,175 951,743 89.7 89.4 89.4

Amazon Brazil 44,289 45,286 46,663 12.3 12.2 12.5

Rhine The Netherlands 339,604 383,869 418,280 6.6 6.9 7.2

Mississippi USA 148,637 235,452 248,607 59.0 58.1 59.2

Mahakam Indonesia 175,773 238,500 240,363 98.4 98.0 97.7

Burdekin Australia 164 169 188 49.7 46.8 46.5

Delta Country Surface (km²) Surface (relative %)

1 in 10 year

1 in 100 year

1 in 1000 year

1 in 10 year

1 in 100 year

1 in 1000 year

Ganges-Brahmaputra

India and Bangladesh

1,538.7 1,820.9 1,538.7 19.0 19.5 19.0

Yangtze China 7,177.2 9,126.8 11,176.1 58.8 59.4 59.0

Mekong Vietnam 12,578.2 13,775.4 14,650.4 41.9 40.7 39.6

Chao Phraya Thailand 1,414.9 1,665.6 2,002.2 85.3 86.2 86.0

Irrawaddy Myanmar 1,593.2 2,022.8 2,163.0 17.4 17.3 16.3

Niger Nigeria 924.5 942.4 981.7 99.3 99.4 99.4

Amazon Brazil 1,832.7 2,109.4 2,373.4 17.4 17.6 18.1

Rhine The Netherlands 986.5 1,043.6 1,093.2 12.7 12.8 12.9

Mississippi USA 3,467.6 4,334.5 5,051.6 84.7 82.8 80.2

Mahakam Indonesia 375.6 525.2 553.7 94.8 94.3 94.7

Burdekin Australia 65.9 70.7 89.0 55.7 53.4 49.9

Chapter 2

60

Sensitivity analysis

In order to compare the flood-exposed areas predicted by the GIS model and the

areas that can be indeed flooded during flood events, the flooding extend of two

cyclones were selected, the cyclone Sidr of November 2007 for the Ganges-

Brahmaputra and the cyclone Nargis of May 2008 for the Irrawaddy delta. The

result of this comparison shows that about 50 % of the areas that were flooded

during those cyclone events are flood-exposed areas predicted by the model

(Figure SI 2.13). The model is not designated to predict the areas that are flooded

during specific events, as those areas depends on factors such as the specific

cyclone properties like the direction of the cyclone or its speed of forward

propagation, the failure of flood protecting structures like dikes and dams present

in both deltas, or the rainfall driven fluvial floods. Nevertheless, it is able to

designate as flood-exposed areas those areas that experienced storm surge

flooding from specific cyclone events.

Figure SI 2.13 Comparison of the delta’s flood-exposed areas predicted by the GIS model and the areas flooded during previous cyclone events for (a) the Ganges-Brahmaputra delta with the cyclone Sidr (15/11/2007) and (b) the Irrawaddy delta with the cyclone Nargis (02/05/2008).

Global hotspots

61

CHAPTER 3 Identifying global hotspots for nature-based mitigation of coastal flood risks

Rebecca Van Coppenolle and Stijn Temmerman

Chapter 3

62

Abstract

Low-lying coastal zones are increasingly exposed to flood risks due to global

change including sea level rise, increasing storm intensity and growing coastal

population densities. Nature-based risk mitigation, by conservation or restoration

of ecosystems, such as tidal wetlands (salt marshes and mangroves) that have the

natural capacity to mitigate storm surge related flood risks, has been

demonstrated by local to regional-scale studies, but yet, we currently lack global-

scale assessments of where hotspots are located of large flood-exposed coastal

areas and populations that can receive nature-based risk mitigation from tidal

wetlands. Here we present the results of a global-scale GIS model assessing the

worldwide contribution of tidal wetlands to coastal flood risks mitigation. It

identifies the inland surface areas and population numbers receiving storm surge

mitigation by mangrove forests and salt marshes, and it quantifies the distance

travelled by a storm surge through tidal wetlands as a measure of the magnitude

of storm surge mitigation. Results show that on a worldwide scale, about 30 % of

the flood-exposed low-lying coastal plain benefit from nature-based storm surge

mitigation by tidal wetlands, with the largest areas located in deltas, estuaries and

lagoons (e.g. Mississippi delta, Elbe estuary, Venice Lagoon). About 40 % of the

global flood-exposed coastal population receives nature-based storm surge

mitigation. The majority of that population (80 %) is located in five countries, i.e.

China, Vietnam, the Netherlands, India and Indonesia. Areas more exposed to

extreme storm surges (Eastern America, Caribbean Sea, Eastern Asia) include

hotspot areas where storm surges are travelling through wider tidal wetlands

generating higher mitigation, as for example in the Mississippi delta, Chesapeake

bay, Ganges-Brahmaputra delta or Yangtze delta. Our global scale assessment aims

to increase general awareness on the high capacity of nature-based flood risk

mitigation, and to inspire further local scale analyses in support of the wider

application of nature-based risk mitigation as a sustainable strategy to mitigate

increasing coastal flood risks around the world.

Global hotspots

63

3.1 Introduction

Coastal areas are increasingly exposed to flood and erosion risks due to sea level

rise, increasing intensity of storms and cyclones (Hallegatte et al., 2013; Hinkel et

al., 2014; de Sherbinin et al., 2007; Vitousek et al., 2017), and subsidence by

human actions such as reduction of sediment supply by river dams or conversion

and drainage of coastal wetland ecosystems into human land use (Adam, 2002;

Auerbach et al., 2015; Balke & Friess, 2016; Gedan et al., 2011; Kirwan &

Megonigal, 2013; Pethick & Orford, 2013; Storlazzi et al., 2011; Syvitski, 2005,

2008; Syvitski et al., 2009; Tessler et al., 2015; Thampanya et al., 2006).

In parallel, the coastal population will continue to grow, reaching globally

averaged densities of 405 to 534 people/km² by 2060 (or ten times the current

world’s average) (Guzmán et al., 2009; Kron, 2013; Mcgranahan et al., 2006;

Neumann et al., 2015), with more and more people concentrated in large coastal

cities (Von Glasow et al., 2013; Sengupta et al., 2018; United Nations, 2012),

increasing the number of people and assets exposed to coastal flood risks (Hanson

et al., 2011; Small & Nicholls, 2003).

The standard strategy for coastal protection is the construction of hard

engineering structures such as dams or dikes that protect the low-lying coastal

areas from the coastal flood and erosion risks (Adriana Gracia et al., 2018;

Pranzini, 2018; Rangel-Buitrago et al., 2018). However, those structures are more

and more challenged for their negative consequences on the natural environment

(disturbance of natural habitats, disturbance of sediment supply, accelerated

erosion at the bottom...) and the practical and financial difficulties to maintain

them in the face of projected climate and socio-economic changes. While nature-

based solutions, or combined hybrid solutions, are more and more regarded as a

sustainable, self-sufficient and cost-effective strategy to mitigate coastal flood and

erosion hazards (Adriana Gracia et al., 2018; Griggs, 2005; Temmerman et al.,

2013). Nature-based solutions are based on the conservation, restoration or

creation of coastal ecosystems, in particular mangrove forests and salt marshes

(further referred to as tidal wetlands), for their capacity to reduce the inland

propagation of storm surges, to reduce wind waves and shoreline erosion, and to

adapt to sea level rise by sedimentation (Costanza et al., 2008; Krauss et al., 2014;

McIvor, Möller, et al., 2012; McIvor, Spencer, et al., 2012; Shepard et al., 2011). In

the last decades, projects of nature-based coastal protection were developed in

several coastal areas around the world, as along the Mississippi delta plain

(Boesch et al., 2006; Coastal Wetlands Planning Protection and Restoration Act

(CWPPRA), n.d.; Day et al., 2007) or along the UK, Belgian and Dutch coastal plains

Chapter 3

64

and estuaries (R. A. Garbutt et al., 2006; Gardiner et al., 2007; Rupp-Armstrong &

Nicholls, 2007; SigmaPlan, 2017).

Tidal wetlands are increasingly recognized as having the capacity to attenuate

wind waves and storm surges, acting as a buffer in between the sea and the low-

lying coastal areas (Barbier et al., 2013; Costanza et al., 2008; Gedan et al., 2011;

Krauss et al., 2009; Mazda et al., 2006; Narayan et al., 2016, 2017; Wamsley et al.,

2010; Zhang et al., 2012). The mechanisms of storm surge reduction rely on the

friction exerted by the tidal wetlands’ geomorphology and vegetation on the water

column during the landward propagation of the surge (Barbier et al., 2013;

Costanza et al., 2008; Leonardi et al., 2018; Smolders et al., 2015; Stark et al.,

2016). Often expressed as a rate of storm surge height reduction per unit of

distance travelled through the tidal wetlands, the attenuation rates derived from

observations range from a couple of centimetres to 25 cm/km for salt marshes

(Krauss et al., 2009; Stark et al., 2015; Wamsley et al., 2010; Zhang et al., 2012),

and up to 50 cm/km for mangrove forests as reported by the hydrodynamic

modelling study of Zhang et al. (2012).

Existing studies on storm surge risk mitigation by tidal wetlands, discussed above,

mostly focus on local to regional scales and are mostly concentrated on specific

locations in the USA and to a lesser extent in Europe (Arkema et al., 2013; Das &

Vincent, 2009; Krauss et al., 2009; McGee et al., 2006; Stark et al., 2015), while

studies elsewhere in the world are much scarcer. As a consequence, until now

there is poor insight in the global scale possibilities for nature-based storm surge

mitigation by mangroves and salt marshes. From the upscaling of the GIS model

presented in Van Coppenolle et al. (2018) and based on globally available data, we

aim to identify the global hotspots of large flood-exposed coastal areas and

populations that can receive nature-based flood risk mitigation, defined as any

reduction of the storm surge due to its travelling through the existing mangrove

and salt marsh ecosystems. Such a global assessment should increase awareness

on the possibilities for nature-based risk mitigation and should stimulate further

developments in nature-based mitigation policies as a strategy against increasing

coastal flood risks.

Global hotspots

65

3.2 Methods

3.2.1 Datasets

The following datasets were used.

The values for the topography and the bathymetry are coming from the

General Bathymetric Chart of the Oceans (GEBCO) (British Oceanographic Data

Center, 2017) that represents a gridded bathymetry of the oceans coupled with the

NASA Shuttle Radar Topography Mission (NASA SRTM, NASA JPL, 2013) digital

elevation model of the continents. Both datasets have a resolution of 30 arc-

second. The SRTM dataset is found to be the best known global digital elevation

model (Rodriguez et al., 2006; Sun et al., 2003).

The worldwide distribution of the tidal wetlands was determined by the

Global distribution of Mangroves (Giri et al., 2011) and the Global distribution of

Saltmarshes (Mcowen et al., 2017) (USGS, www.unep-wcmc.org). The coastlines

delimiting the land and sea environment were defined by the combination of the

country boundaries as they exist in January 2015 (ESRI, DeLorme Publishing

Company, Inc., 2015), and the mangrove forests and salt marshes extent, as the

different dataset do not perfectly overlap (Lichter et al., 2011).

The storm surge heights for a 1 in 100 year return period accounted for

are coming from two datasets. The first one is the DINAS-COAST Extreme Sea

Level dataset from the Dynamic Interactive Vulnerability Assessment (DIVA)

database (Hinkel et al., 2014; Vafeidis et al., 2005). It corresponds to the storm

surge water level above mean sea level and is calculated by model simulations

based on tidal levels, barometric pressures, wind speeds and sea bed slopes for

return periods of 10, 100 and 1000 years. These storm surge water levels are

given for coastline segments, that correspond to the coastline of the Digital Chart

of the World (DCW, Environmental Systems Research Institute, ESRI, 2002)

divided in segments based on administrative and environmental parameters

(Vafeidis et al., 2005). The second storm surge height dataset is the Global Tide

and Surge Reanalysis (GTSR) (Muis, Verlaan, Winsemius, et al., 2016) that is also

based on the coastline segments of the Digital Chart of the World, as the DIVA data.

The GTSR corresponds to the near-coast global reanalysis of storm surges over the

period 1979-2014, with large validation of the results against observations. The

main difference between the two datasets is the static (DIVA) or dynamic (GTSR)

way of calculating the sea level extremes (Muis, Verlaan, Nicholls, et al., 2016). It

results in differences in the extremes, with an overestimation of the extremes by

Chapter 3

66

the DIVA data (mean bias of -0.19 m and mean absolute error of 0.23 m) and an

underestimation of the extremes by the GTSR dataset (mean bias of 0.55 m and

mean absolute error of 0.64 m) (Muis et al. 2016a; Muis et al. 2016b).

The population distribution originates from the LandScan 2013 Global

Population Database (Bright et al. 2013). It represents the population over a 30 arc

second grid resolution and integrates the diurnal movements and collective

travelling behaviour of the world population, i.e. the so-called “ambient

population”, averaged over 24 hours (Bright et al., 2013; Dobson et al., 2000).

The distribution of the historical tracks of the cyclones is coming from the

Global Cyclone Hazard and Frequency Distribution that is a compilation of more

than 1 600 storm tracks over the period of January 1980 to December 2000; the

wind speeds around the tracks have been modelled using the Holland’s model

(1997). The value of each cell corresponds to a decile ranking, a higher ranking

implies a greater frequency of the hazard relative to the other cells (Center for

Hazards and Risk Research - CHRR - Columbia University, Center for International

Earth Science Information Network - CIESIN - Columbia University, International

Bank for Reconstruction and Development - The World Bank, 2005; Dilley et al.,

2005)

Because there are local differences in the exact position of the coastlines between

the different datasets, e.g. tidal wetlands and land areas can locally appear on the

seaward side of the storm surge coastline segments and that the goal of the

analysis is to account for all the tidal wetlands and land areas that can influence

the landward propagation of the storm surge, the coastline segments of the storm

surge datasets were not directly used as the source of the flooding. Alternatively a

‘flood source line’ was defined via the creation of a buffer of 15 km around the

original storm surge coastline segments. Only the offshore limit of this buffer was

kept and defined as the ‘flood source line’. As such, the flood source line

corresponds to a simplified coastline 15 km offshore of the original storm surge

datasets coastline segments, to assure that all tidal wetlands and land areas that

influence the propagation of the storm surge are located on the landward side of

this flood source line. The storm surge heights of the different segments stored in

the DIVA and GTSR datasets are transferred to the flood source line segments of

various lengths (average length of 59.04 ± 87.03 km) with a shortest Euclidean

distance algorithm (for further information see Chapter 2).

The model has a resolution of 30 arc second in accordance with the original

resolution of the bathymetry (GEBCO), topography (SRTM) and population

(LandScan) datasets. In such, the other datasets, i.e. the countries boundaries, the

Global hotspots

67

salt marshes and the mangrove forests areas, were transformed to raster datasets

of 30 arc second resolution. This manipulation generated the loss of the smallest

salt marshes and mangrove forests areas (< 1 km²), however, due to the global

character of the model, those losses are unavoidable. Hence our model accounts

only for storm surge mitigation by wetlands patches larger than 1 km². We argue

that this is acceptable, as storm surge attenuation rates are up to 25 cm/km in salt

marshes and 50 cm/km in mangroves (e.g. Zhang et al. 2012; Mcivor et al. 2012;

Krauss et al. 2009; Wamsley et al. 2010; Lovelace 1994), hence less than 1 km

wide wetlands provide a relatively low degree of storm surge height reduction and

are not considered here.

The global datasets used present some limitations in regards to local data accuracy

and local data artefacts. Such limitations may include vegetation artefacts in the

elevation dataset that over-estimates the land elevation by one to several meters

(Rodriguez et al., 2006; Sun et al., 2003), or the moderate resolution of the tidal

wetlands datasets that can locally result in over- or under-estimations of the

surface area of the tidal wetlands (Giri et al., 2011; Mcowen et al., 2017). The

variable lengths of the DIVA coastline segment involve that some very local

characteristics of the coastal plain or some possible local increase or decrease in

storm surge height due to the geomorphology of the coast may not be accounted

for in the model.

3.2.2 Model

The model corresponds to the GIS procedure described in Van Coppenolle et al.

(2018), but applied on a worldwide scale, while in Van Coppenolle et al. (2018) it

was tested for 11 large deltas around the world. The model was developed in

ArcGIS (10.3.1) and Python (2.7). The model is similar to previously published

procedures that assess the coastal areas and number of people vulnerable to

storm surge flooding on regional to global scales (Arkema et al., 2013; Dasgupta et

al., 2011).

The model simulates how a storm surge flood wave would be routed from the

above-described flood source to the coastal plain, i.e. the land area below 10 m of

elevation (corresponding to the Low Elevation Coastal Zone and to the maximal

storm surge height of both datasets). The model resolution did not allow

accounting for flood protecting structures like dikes or dams, and hence it

evaluates the flood risks in case existing flood protecting structures would fail.

Four land covers are considered, i.e. (1) open water and channel areas, (2) salt

marshes, (3) mangrove forests and (4) remaining land area. For each of them, a

Chapter 3

68

storm surge attenuation rate, i.e. reduction of the surge height with distance the

surge has travelled over these land cover types (in cm/km), was defined based on

literature (Table 3.1) (for details, we refer to Van Coppenolle et al. 2018). The

storm surge propagation pathways were defined by the cost distance algorithms

that account for both the distance travelled and the friction generated by the land

covers. As such, every pixel of the coastal plain is associated with a given travelling

cost that is subsequently used in the cost distance algorithm, and corresponds to

the attenuation rate (Table 3.1) the storm surge will undergo by its propagation

from the flood source towards the pixel. The model is run for two scenarios: the

first scenario considers the current extent of the tidal wetlands, while, for the

second scenario, the tidal wetlands are considered as remaining land area. Finally,

the comparison of the two scenarios is used to determine the coastal areas and

populations that will be flooded via a storm surge flood wave that travelled

through tidal wetlands, namely, the areas where the storm surge attenuation is

higher in the case of the first scenario including wetlands.

Table 3.1 Attenuation rates attributed to the land covers considered in this study, in accordance with the model approach presented in Van Coppenolle et al. (2018).

Substrate Attenuation rate

(cm/km)

Tidal wetlands Mangrove Salt Marsh

10.0 8.0

Remaining land area 6.0 Open water and Channels 0.1

For every pixel located on the coastal floodplains, the length of the flood pathway

crossing through tidal wetlands was defined as an indication of the magnitude of

the storm surge mitigation by the tidal wetlands, as propagation through longer

distances of tidal wetlands are expected to generate a larger reduction of the

storm surge height. The distance travelled through tidal wetlands was calculated

by dividing the difference in storm surge height reduction between the two

scenarios by the difference of attenuation rates between the tidal wetland type,

either mangrove or salt marsh, and the remaining land (i.e. 10-6 cm/km, or 4

cm/km for mangrove forests and 2 cm/km for salt marshes). In the situation were

the two types of tidal wetlands were present, the value used for the division

corresponds to the averaged value, i.e. 3 cm/km.

From a comparative perspective, the simulations were made with the two storm

surge height datasets, the DIVA data and the GTSR data. The main discussion is

Global hotspots

69

based on the results of the DIVA data (as it shows the smallest error on predicted

extreme sea levels; see above and Muis et al (2016a,b)), while a comparison of the

results of the two datasets is presented in the Results section.

The model does not simulate the full complexity of atmospheric and

hydrodynamic processes involved in flood propagation and therefore is not able to

calculate accurate flood depths and absolute values of reduction in flood depth

behind tidal wetlands during specific storm surge events. Instead it calculates the

surface area and population numbers flooded via flood pathways crossing through

tidal wetlands, and it calculates the distance or length of the flood pathway

crossing through tidal wetlands, which are variables that are not dependent on

complex atmospheric and hydrodynamic processes during specific storm surge

events. As such, it has the major advantage to be globally applicable to compare

coastlines and coastal plains around the world.

For representation purposes, the model output will be presented on the coastline

segments defined by the Digital Chart of the World. A section of the coastal plain is

associated to each coastline segment via Euclidean distance. For each segment the

model output exists of (1) a value for the total surface area within the associated

coastal plain that is flooded via pathways crossing through tidal wetlands – further

called “area benefiting from storm surge mitigation”; (2) a value for the total

population number within the associated coastal plain that is flooded via pathways

crossing through tidal wetlands – further called “population benefiting from storm

surge mitigation”; (3) mean distance travelled by a storm surge through tidal

wetlands.

3.3 Results

3.3.1 Coastal Plain Areas Benefiting from Storm Surge Mitigation

The results show that for a 1-in-100 year storm surge event, without accounting

for any flood protecting structures and without the existing tidal wetlands

(scenario 2), 439 525 km² of the world’s coastal plain is exposed to storm surge

flood risks. However, when accounting for the currently existing tidal wetlands

(scenario 1), 135 911 km² (i.e. 31 % of the previous number) of the world’s coastal

plain benefits from a reduction in storm surge height as the storm surge pathway

passes through tidal wetlands (see Supplementary Information Figures SI 3.1, 3.2

and 3.5).

The locations having the largest coastal plain area benefiting from storm surge

mitigation by tidal wetlands (> 1 000 km² of the coastal plain associated to one

Chapter 3

70

segment, see Method section, segments are highlighted in red in Figure 3.1), are

mainly found in or close to deltas, estuaries and lagoons. Those hotspots include

the Northern part of the Yangtze delta, in front of the city of Yancheng (China),with

3 112.3 km² benefiting from storm surge mitigation associated to one segment of

the delta; the Northern part of the Elbe estuary (Germany), with one segment

having 2 215.2 km² of coastal plain benefiting from storm surge mitigation; the

Mississippi delta (USA) that has a segment with 2 050.4 km² benefiting from storm

surge mitigation; and the Laguna de Terminos in the Bahia de Campeche (Mexico)

that presents a segment with a surface area of 1 999.4 km² benefiting from storm

surge mitigation by tidal wetlands.

Not all coastline segments have the same length, and therefore, in order to

standardize the results, we also plotted what we call the standardized surface area

benefiting from storm surge mitigation, i.e. the absolute surface area divided by

the length of the associated coastline segment (km²/km) (Figure 3.2). The results

show that hotspots for storm surge mitigation by tidal wetlands per unit of

shoreline length also correspond to bays, lagoons, deltas and estuaries, yet, the

locations show some divergences with the absolute surface area benefiting from

mitigation. The hotspots are mainly located along the Northern European coasts

(France, Belgium, Netherlands, Germany and United Kingdom) and along the East

Asian coasts.

Global hotspots

Figure 3.1 Absolute surface area benefiting from a storm surge pathway crossing through tidal wetlands represented on the associated coastal segment (km²), with circles highlighting the segments for which the coastal plain area benefiting from storm surge mitigation is greater than 1 000 km².

Chapter 3

Figure 3.2 Standardized area, i.e. surface area per unit of shoreline length (km²/km), benefiting from a storm surge pathway crossing through tidal wetlands represented on the associated coastal segment, with circles highlighting the segments for which the surface area benefiting from storm surge mitigation is greater than 15 km² per 1 km of shoreline.

Global hotspots

73

3.3.2 Magnitude of Storm Surge Mitigation

Not only the surface area benefiting from a storm surge pathway crossing through

tidal wetlands, but particularly the distance travelled by the storm surge through

the tidal wetlands is a very relevant parameter determining the magnitude to

which wetlands can contribute to nature-based mitigation of storm surge flood

risks. Therefore the distance travelled by the storm surge through tidal wetlands

was defined for every pixel of the coastal plain and averaged over the areas

associated to each segment (Figure 3.3). The locations having the longest distance

(> 5 km, highlighted in red in Figure 3.3) travelled by the storm surge through

tidal wetlands are considered as having the highest degree of storm surge

mitigation. They are mainly found in areas where large tidal wetlands exist, again

in large deltas, such as in the Guayas delta in Ecuador, where parts of the coastal

plain benefit of storm surge mitigation by more than 12 km travelled through the

mangrove forests. The Kolyma delta in north-eastern Siberia presents large areas

of tidal wetlands (> 1 000 km²) that influence the storm surge propagation, as

some parts of the coastal plain can benefit from distances of more than 10 km of

salt marshes crossed by a storm surge. The coastal plains in the Mississippi delta

(USA), Chesapeake bay (USA), Ganges-Brahmaputra delta (India and Bangladesh)

or Tidung estuary in the North Kalimantan region of Borneo (Indonesia) also

benefit from more than 5 km travelled by a storm surge through tidal wetlands.

The comparison of the areas with a long distance of tidal wetlands along the storm

surge flood pathway (> 5 km) with the areas that are exposed to cyclone

conditions is presented on Figure 3.3. The value of the global cyclone hazard

distribution represents the frequency of the hazard relative to the other areas. Six

hotspots having a coastal plain area benefiting from storm surge mitigation by a

long distance (> 5 km) of tidal wetlands are also areas where the likelihood of

being exposed to cyclone hazards is greater than elsewhere (indicated in red

colours in Figure 3.3). In such, those areas whilst exposed to higher frequency of

cyclones could benefit from higher storm surge mitigation by the tidal wetlands.

When up-scaled to the country level, the results show that on the 114 countries

having tidal wetlands in their coastal plain, 101 benefit from flood risks mitigation

by the tidal wetlands (See Supplementary Information). The thirteen countries

that have tidal wetlands but no coastal area buffered by tidal wetlands are mainly

countries where the coastal plain is rapidly gaining in altitude, as in Equatorial

Guinea, where the 142 km² of mangrove forests are bordered by land areas with

an altitude rapidly reaching 5 m, while the 1 in 100 years storm surge is less than

2 meters high.

Chapter 3

Figure 3.3 Mean distance (m) travelled through tidal wetlands by a 1-in-100 years storm surge during its landward propagation. The circles highlight the segments for which the mean distance travelled through tidal wetlands by a storm surge is longer than 5 km, while red colours indicate hotspots were the long distance of wetlands coincides with high exposure to cyclones.

Global hotspots

75

3.3.3 Coastal Population Benefiting from Storm Surge Mitigation

Globally around 73.5 million people are exposed to coastal flood risks from a 1-in-

100 year storm surge in the case of scenario 2 (no protecting structures and no

tidal wetlands). When considering the current tidal wetlands (scenario 1), around

29.4 million people (i.e. 40 % of the previous number) benefit from nature-based

storm surge mitigation (See Supplementary Information Figures SI 3.3, 3.4 and

3.6). Hotspot areas (i.e. coastline segments with > 10 000 people benefiting from

storm surge risk mitigation) are predominantly located in large deltaic and coastal

lowland areas in Asia and Europe, such as in the Ganges-Brahmaputra delta (India

and Bangladesh), the Mekong delta (Vietnam), the Pearl and Red river deltas

(China), the Rhine-Meuse-Scheldt delta (Belgium and Netherlands) and the

Humber estuary (UK) (Figure 3.4). At a country level, the highest number of

people benefiting from storm surge mitigation by tidal wetlands is found in China

with 9.5 million people, followed by Vietnam (9.3 million people), India (2.8

million people), The Netherlands (1.7 million people) and Indonesia (1.2 million

people) (See Supplementary Information Figure SI 3.3). Those five countries

together make up for 83 % of the global population benefiting from nature-based

flood risks mitigation.

As for the coastal plain area, the population benefiting from storm surge

mitigation was also standardized by calculation per unit of shoreline segment

length (Figure 3.4). The hotspots where the population benefiting from storm

surge mitigation per 1 km of shoreline is the largest partly correspond to the

hotspots of the absolute number of people benefiting from storm surge mitigation

associated to one coastal segment, i.e. the Humber estuary (UK), the Weser estuary

(Germany) and the Rhine-Meuse-Scheldt delta (The Netherlands and Belgium) in

Europe, and most of the deltas and bays highlighted in Asia, at the exception of the

Ganges-Brahmaputra, Irrawaddy and Yangtze delta.

Chapter 3

Figure 3.4 Absolute number of people benefiting from a storm surge pathway crossing tidal wetlands represented on the associated coastal segment, with circles highlighting the segments for which the population benefiting from storm surge mitigation is higher than 100 000 people.

Global hotspots

Figure 3.5 Standardized population, i.e. people per unit of shoreline length (number of people/km), benefiting from a storm surge pathway crossing tidal wetlands, with circles highlighting the segments for which the population benefiting from storm surge mitigation is greater than 10 000 people per 1 km of shoreline.

Chapter 3

78

3.3.4 Comparison of the Storm Surge Height Datasets

In general, the GTSR sea level extremes have been found to underestimate storm

surge heights in comparison to the DIVA storm surge heights (See Supplementary

Information Figure SI 3.5) (Muis, Verlaan, Nicholls, et al., 2016; Muis, Verlaan,

Winsemius, et al., 2016). Whilst the correlation coefficient between the modelled

and observed sea level extremes is 0.70 for the DIVA database and 0.84 for the

GTSR database, the tropical cyclones in the GTSR data are under-represented, with

a bias that is larger in the tropical regions (Muis, Verlaan, Nicholls, et al., 2016).

Therefore, for the purpose of our analysis, we plotted the results from the DIVA

values in the main text of this paper, the results for the GTSR values are presented

in Supplementary Information.

Results from the simulations based on the GTSR dataset (Supplementary

Information Figures SI 3.8 and 3.9) show that for the 281 750 km² of coastal plain

exposed to 1-in-100 year storm surge flood risks, 80 307 km² are benefiting from a

storm surge pathway crossing tidal wetlands. Compared to the DIVA results

(Figure 3.1; resulting in 439 525 km² exposed to 1-in-100 year storm surges), the

GTSR dataset results in a smaller area exposed to coastal flood risks,

corresponding to an under-representation of the exposed areas by 36 %, which is

concordant with the difference observed between both datasets by Muis et al.

(2016). In regards to the hotspots of coastal plain areas benefiting from storm

surge mitigation (i.e. areas > 1 000 km²), fewer locations are identified based on

the GTSR dataset, yet, all GTSR hotspots are also identified as hotspots based on

the DIVA data.

The comparison of the mean distance travelled by the storm surge through tidal

wetlands before reaching the coastal plain also shows a fewer number of hotspots

(distance of more than 5 km) resulting from the GTSR data as compared to the

DIVA data. Most of the GTSR hotspots (Supplementary Information Figure SI 3.10)

are corresponding to the DIVA hotspots, while the Everglades in Florida and the

low-lying areas close to the Dee River in the United Kingdom are GTSR specific. In

the Everglades, the difference is relatively small between the two results, 500 m,

while for the Dee River, the GTSR storm surge height of 5.86 m (3.61 m for the

DIVA) results in a larger area identified as benefitting from nature-based storm

surge mitigation.

In terms of population that benefits from storm surge mitigation by the tidal

wetlands, over the 38.3 million people that are globally exposed to coastal flood

risks, 13.5 million can benefit from storm surge mitigation by the tidal wetlands,

Global hotspots

79

based on the GTSR data (Supplementary Information Figures SI 3.11 and 3.12). As

for the DIVA results, the highest absolute numbers of people benefiting from storm

surge mitigation are located along the Belgian-Dutch-German and East-Asian

coastlines.

The comparison of the results shows that the differences in the incoming storm

surge height between the two datasets is influencing the coastal areas and

populations benefiting from storm surge mitigation by the tidal wetlands.

Nonetheless, the general trends observed throughout the results are similar.

3.4 Discussion and Conclusion

In the face of global climate change and the associated increasing risks of coastal

flooding from more severe storm surges and expected sea level rise (Bengtsson et

al., 2006; Hallegatte et al., 2013; Hinkel et al., 2014; IPCC, 2013; Knutson et al.,

2010; Webster et al., 2005; Woodruff et al., 2013), the conservation of tidal

wetlands can contribute to the mitigation of coastal flood risks by their ability to

attenuate storm surges, reduce the impact of waves and shoreline erosion, and

accumulate sediments in balance with sea level rise (Kirwan et al., 2016; Krauss et

al., 2014; Lovelock et al., 2015; Sandi et al., 2018). As such, nature-based risk

mitigation can reduce the threats to flood-exposed coastal areas and populations

(Cheong et al., 2013; Costanza et al., 2008; Duarte et al., 2013; Sutton-Grier et al.,

2015; Temmerman et al., 2013). Current assessments on the role of tidal wetlands

for coastal flood risk mitigation are based on in situ observations (Das & Vincent,

2009; Krauss et al., 2009; McGee et al., 2006; Stark et al., 2015) and/or on

modelling studies (Arkema et al., 2013; Stark et al., 2016; Zhang et al., 2012) at

local to regional scales. Such site-specific studies have substantially advanced our

understanding of the mechanisms determining the rate of storm surge mitigation

by salt marshes and mangroves (i.e. how much the peak water level is reduced per

distance that the storm surge has travelled through salt marshes or mangroves).

However, we currently lack a global scale assessment of the possibilities for

nature-based mitigation of coastal flood risks. That is, the location of global

hotspots of large flood-exposed coastal areas and populations that can receive

nature-based risk mitigation from existing salt marsh and mangrove ecosystems.

Our study is to our knowledge the first worldwide assessment identifying such

global hotspots and quantifying the coastal area and number of people that can

benefit from nature-based coastal flood risk mitigation.

The presented model allows a global-scale assessment of the locations where tidal

wetlands are expected to play a role in the mitigation of storm surge flood risks,

Chapter 3

80

but it does not accurately define the intensity of the storm surge mitigation nor the

flooding depth, extent or duration resulting from a specific storm surge event.

Furthermore, our model analysis does not account for the socio-economic

limitations that may hamper the applicability of nature-based strategies. To

precisely evaluate the attenuation of a specific storm surge due to the presence of

a specific tidal wetland area, a hydrodynamic modelling approach would be

needed based on high resolution input data regarding the storm characteristics

(duration, intensity, track, wind velocity field...), the wetlands’ vegetation and

geomorphology (vegetation type, density and continuity, soil surface

topography...) as well as the geomorphology of the surrounding coastal area (off-

shore bathymetry, shoreline shape, flood protection structures...) (Leonardi et al.,

2018; Marsooli et al., 2016; Resio & Westerink, 2008; Smolders et al., 2015; Stark

et al., 2016; Temmerman et al., 2012, 2013). However, the high computational

demand of such a hydrodynamic modelling approach does not enable a worldwide

assessment, while the more simple approach of our model and the use of globally

available datasets enable its worldwide applicability with a much lower

computational demand than hydrodynamic models (Van Coppenolle et al., 2018).

The results show that about one third of the global flood-exposed coastal plains

(31 % for the DIVA data and 28.5 % for the GTSR data) and almost 40 % of the

global flood-exposed population (40 % for the DIVA data and 34.5 % for the GTSR

data) experience nature-based storm surge mitigation by existing mangrove

forests or salt marshes. The coastal plains with the largest absolute surfaces

benefiting from storm surge mitigation are mainly found in deltas (e.g. Mississippi

delta, Mekong delta), estuaries (e.g. Bahia Blanca estuary, Elbe estuary) or lagoons

(e.g. Laguna de Terminos, Venice Lagoon) (Figure 3.1), where large low-lying lands

are favourable to both the establishment of tidal wetlands and to storm surge

mitigation over large areas (Leonardi et al., 2018). The areas with the highest

absolute number of people benefiting from nature-based coastal flood risks

mitigation are located in densely populated deltas and estuaries, mainly in North-

western-Europe and East-Asia (Figure 3.4). On the 40 % of the flood-exposed

population benefiting from nature-based storm surge mitigation, the majority (i.e.

83 % for the DIVA data and 80 % for the GTSR data) is living within five countries,

being China, Vietnam, The Netherlands, India and Indonesia. The trends are the

same for the standardized surface area and population (i.e. calculated per unit

length of the coastline) benefiting from storm surge mitigation by the tidal

wetlands. Whilst for both, the number of hotspots is smaller in the case of the

standardized value, suggesting that some areas with large absolute areas or

population benefiting from storm surge mitigation are associated with long

shoreline segments. In addition, the hotspots for the absolute values are scattered

Global hotspots

81

over the world, while the hotspots for the standardized values are mainly grouped

in Western Europe and Southeast Asia.

Furthermore, the results illustrate that the areas more likely exposed to cyclones

also include hotspots for which the nature-based storm surge mitigation can be

the highest in terms of mean distance of tidal wetlands crossed by the storm surge

(Figure 3.3) (Leonardi et al., 2018; Loder et al., 2009; McIvor, Spencer, et al., 2012;

Stark et al., 2016). It is the case in the Mississippi delta, along the Chesapeake Bay,

in the Ganges-Brahmaputra delta or in the Golfo Dulce. Whilst other areas exposed

to severe storms benefit from a limited protection by the tidal wetlands as for

example in the Zambesi delta and the coastal areas in Mozambique.

Our relatively simple but global scale model highlights that numerous hotspot

areas around the world benefit from the conservation of salt marshes or mangrove

forests as part of strategies to mitigate and adapt to increasing coastal flood risks

associated with climate warming, sea level rise and increasing storminess. For

some of those hotspots nature-based coastal protection strategies are actively

implemented, mostly in addition to classical engineered flood defence structures

like dikes and levees, such as in the Mississippi delta, the Chesapeake Bay, the

Rhine-Meuse-Scheldt delta or the Humber estuary (Boesch et al., 2006;

Chesapeake Bay Program, 2000; Day et al., 2007; Elliott et al., 2016; A. Garbutt et

al., 2017; Gardiner et al., 2007; Rupp-Armstrong & Nicholls, 2007; SigmaPlan,

2017). In other hotspots, policy and decision-makers are only beginning to

account for the coastal protection value of tidal wetlands and nature-based

strategies are starting to be implemented along with classical hard engineering, as

in the Ganges-Brahmaputra delta, the Yangtze delta or the Mekong delta (Käkönen,

2008; Seavitt, 2013; Ysebaert et al., 2017). Nevertheless, the application of nature-

based strategies, and then the conservation and maintenance of the coastal

ecosystems, is highly related to the human land claim in coastal zones. As such, in

certain areas over the world, the implementation of those nature-based strategies

would be less realistic, like along the China’s coasts, where despite the local

implementation of nature-based strategies, large natural coastal areas are still

expected to be turned into human land use (Ma et al., 2014; Meng et al., 2017;

Wang et al., 2014).

Creating sustainable and cost-effective coastal protection strategies to adapt to the

increasing coastal flood risks would require policy makers to account for the

presence of tidal wetlands, and coastal ecosystems in general, in the design and

development of coastal protection structures and renounce to the practice of

large-scale tidal wetlands reclamation (Duke et al., 2007; McLeod et al., 2011;

Chapter 3

82

Pendleton et al., 2012; Spalding et al., 1997; Valiela et al., 2001, 2009). Indeed, salt

marshes and mangroves have been converted on large scales to human land use

such as for agriculture, aquaculture, industry and urbanization, and this is still

actively going on in several places around the world, especially in fast developing

regions such as in Southeast Asia (An et al., 2007; Jiang et al., 2015; Tian et al.,

2016; Wang et al., 2014). Apart from the fact that such large-scale tidal wetland

reclamation implies the loss of biodiversity and valuable ecosystem services, we

argue that wetland reclamation should be planned carefully and avoided as much

as possible in order to maximize the flood risk mitigation function of remaining

tidal wetlands. Additionally, where possible restoration or creation of tidal

wetlands should be considered to enhance the nature-based adaptation capacity of

the coastal communities exposed to coastal flood risks. Our global-scale

assessment aims to increase the awareness on the value of nature-based coastal

protection strategies and stimulate further site-specific studies as a first step

towards a more worldwide implementation of nature-based mitigation policies as

a strategy against increasing coastal flood risks.

Global hotspots

Supplementary Information

Results at the country level and percentages

Figure SI 3.1 Surface area (km²) of the flood-exposed coastal plain that benefits from storm surge pathway passing through tidal wetlands at a country level

Chapter 3

Figure SI 3.2 Percentage (%) of the flood-exposed coastal plain benefiting from a storm surge pathway passing through tidal wetlands at a country level

Global hotspots

Figure SI 3.3 Number of people in the flood-exposed coastal plain that benefits from storm surge pathway passing through tidal wetlands at a country level

Chapter 3

Figure SI 3.4 Percentage (%) of the flood-exposed population benefiting from a storm surge pathway passing through tidal wetlands at a country level

Global hotspots

Figure SI 3.5 Percentage (%) of the flood-exposed coastal plain associated to each segment benefiting from storm surge mitigation by tidal wetlands

Chapter 3

Figure SI 3.6 Percentage (%) of the flood-exposed population associated to each segment benefiting from storm surge mitigation by tidal wetlands

Global hotspots

89

Results for the DIVA dataset

Table SI 3.1 Values of the surface area and type of tidal wetlands as well as of the different variables related to storm surge risk mitigation based on a 1-in-100 year storm surge levels from the DIVA dataset at a country level for the countries having tidal wetlands along their coastline. The percentages of the fifth and seventh columns correspond to the percentage of the coastal plain or population benefiting from flood risk mitigation by tidal wetlands relative to the total flood-exposed coastal plain area or population.

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

Albania 29.1 Marsh 206.7 63.5 10166 19.8 357

Angola 245.2 Mangrove 11.0 9.2 18 0.3 895

Antigua and Barbuda

3.3 Mangrove

Argentina 982.2 Marsh 2 406.5 24.4 44 481 37.4 1 658

Australia 16,824.0 Mangrove & Marsh

10 775.5 50.0 7 120 10.5 1 155

Bahamas 306.4 Mangrove 831.8 45.7 922 26.0 342

Bangladesh 4,458.2 Mangrove 1 201.1 63.7 294 765 37.6 3 276

Belgium 2.0 Marsh 670.4 25.5 393 970 38.9 2 227

Belize 408.3 Mangrove 145.0 49.5 671 13.5 266

Benin 9.4 Mangrove 88.4 28.7 3 087 4.6 500

Brazil 8,350.5 Mangrove & Marsh

4 797.1 41.5 127 231 32.0 859

Brunei Darussalam 84.3 Mangrove 0.9 13.4 0 0.0 707

Cambodia 309.2 Mangrove 175.7 76.6 18 752 81.3 2 150

Cameroon 2,163.7 Mangrove 10.2 92.3 23 100.0 331

Canada 103.0 Marsh 955.0 2.1 12 762 17.5 241

Cayman Islands 58.2 Mangrove 5.7 87.5 5 141 71.0 61

Chile 5.2 Marsh 2.2 0.1 11 0.01 1 523

China 4,956.5 Mangrove & Marsh

13 683.7 38.9 9 520 175 54.8 1 195

Colombia 1,451.6 Mangrove 423.4 41.0 18 432 27.0 757

Congo DRC 221.1 Mangrove 7.7 100.0 1 100.0 1 253

Costa Rica 317.1 Mangrove 20.3 58.6 3 951 98.5 99

Côte d'Ivoire 4.3 Mangrove 5.1 54.6 329 95.9 738

Croatia 2.9 Marsh 1.3 2.4 104 1.9 98

Cuba 3,321.1 Mangrove 1 106.1 58.8 3 849 5.7 373

Cyprus 19.1 Marsh

Denmark 187.3 Marsh 750.1 46.1 37 573 32.5 483

Djibouti 0.8 Mangrove

Chapter 3

90

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

Dominican Republic

122.0 Mangrove 13.8 22.9 82 6.5 122

Ecuador 961.2 Mangrove & Marsh

890.1 67.0 20 086 32.2 1 866

El Salvador 296.3 Mangrove 10.0 57.1 170 38.9 380

Equatorial Guinea 141.9 Mangrove

Eritrea 6.6 Mangrove 6.7 3.8 1 0.1 69

Estonia 1.3 Marsh

Fiji 898.4 Mangrove 40.1 73.2 567 89.6 1 186

Finland 79.3 Marsh 174.9 6.6 27 470 22.6 270

France 657.0 Marsh 2 518.3 41.7 192 747 19.0 952

French Guiana 809.1 Mangrove 133.7 67.7 2149 73.6 1 274

Gabon 1,325.7 Mangrove 10.3 36.4 1 0.4 402

Gambia 514.3 Mangrove 18.6 65.5 5 150 38.5 307

Germany 142.6 Marsh 5 946.4 61.0 612 565 50.7 447

Ghana 54.4 Mangrove 216.9 37.5 31 598 46.8 657

Guadeloupe 15.6 Mangrove

Guatemala 286.0 Mangrove 1.7 20.0 1 7.7 84

Guinea 2,078.7 Mangrove 661.4 93.7 24 510 76.8 1 658

Guinea-Bissau 1,962.3 Mangrove 440.3 79.1 3 882 70.7 933

Guyana 162.7 Mangrove 458.3 28.2 9 452 4.6 470

Haiti 100.1 Mangrove 2.4 37.4 10 24.4 1 213

Honduras 457.5 Mangrove 35.7 30.8 162 8.7 404

Iceland 15.5 Marsh 11.9 1.6 11 0.1 331

India 3,209.8 Mangrove 4 517.7 22.6 2 792 390 41.1 734

Indonesia 22,263.4 Mangrove 7 045.5 46.8 1 192 607 30.3 928

Iran 49.0 Mangrove 206.1 3.4 1 058 0.9 267

Ireland 45.7 Marsh 138.1 22.3 17 840 37.3 285

Italy 370.2 Marsh 2 309.8 59.5 472 387 55.2 1 584

Jamaica 46.4 Mangrove 83.1 79.1 8 612 87.6 474

Japan 1.6 Mangrove

Kenya 243.9 Mangrove 320.2 45.3 2 701 50.8 525

Latvia 0.5 Marsh 0.9 0.5 23 0.2 102

Liberia 23.0 Mangrove 1.7 11.1 27 8.0 22

Madagascar 1,686.3 Mangrove & Marsh

1 005.3 65.5 9 838 16.3 1 083

Malaysia 4,513.5 Mangrove 351.3 31.9 188 224 48.0 1 355

Global hotspots

91

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

Martinique 3.3 Mangrove

Mexico 7,932.1 Mangrove & Marsh

8 249.1 60.2 102 142 42.2 1 502

Micronesia 37.4 Mangrove

Montenegro 1.3 Marsh

Morocco 8.3 Mangrove 39.4 3.7 117 0.2 598

Mozambique 2,241.4 Mangrove 4 235.6 76.9 93 528 40.3 1 267

Myanmar 3,962.4 Mangrove 2 812.8 36.3 315 508 24.5 851

Netherlands 149.7 Marsh 5 119.8 28.8 1 744 097 19.6 729

New Caledonia 106.8 Mangrove 24.8 46.4 63 21.0 217

New Zealand 150.7 Mangrove & Marsh

538.7 50.9 14 054 40.6 345

Nicaragua 470.5 Mangrove 137.9 49.4 984 73.9 874

Nigeria 5,550.9 Mangrove 106.4 60.1 37 054 54.1 1 324

North Korea 0.5 Marsh 24.0 1.9 19 309 8.0 290

Pakistan 339.0 Mangrove 1 795.5 30.8 42 586 28.4 623

Palau 28.0 Mangrove

Panama 1,231.9 Mangrove 189.5 58.7 1 988 15.7 450

Papua New Guinea 4,056.8 Mangrove 440.9 60.8 3 987 34.9 1 194

Peru 283.2 Mangrove & Marsh

68.4 3.7 670 24.9 668

Philippines 1,037.2 Mangrove 425.6 23.8 65 612 9.9 317

Portugal 182.0 Marsh 320.8 74.2 24 690 53.6 747

Puerto Rico 59.3 Mangrove & Marsh

17.9 33.9 1 268 6.6 408

Qatar 0.8 Mangrove 2.3 0.5 1 0.0 108

Romania 49.1 Marsh 6.1 1.9 82 2.7 171

Russian Federation 7,069.8 Marsh 7 068.4 15.2 251 0.3 7 753

Saint Lucia 0.8 Mangrove

Saudi Arabia 15.3 Mangrove 217.9 6.4 311 0.2 293

Senegal 902.9 Mangrove 406.7 57.1 1 436 29.8 2 066

Seychelles 3.4 Mangrove 2.5 19.9 0 0.0 77

Sierra Leone 1,138.1 Mangrove 78.9 64.6 539 71.5 768

Singapore 0.9 Mangrove 4.3 6.0 568 4.4 63

Slovenia 0.8 Marsh 7.8 92.9 3 297 42.9 119

Solomon Islands 94.7 Mangrove 15.2 6.8 124 5.2 124

Chapter 3

92

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

Somalia 4.3 Mangrove 11.1 1.8 13 0.0 163

South Africa 29.3 Mangrove & Marsh

1.4 3.2 3 1.8 25

Spain 815.2 Marsh 501.1 42.3 42 619 18.0 2 475

Sri Lanka 55.9 Mangrove 9.3 9.2 35 2.9 182

Suriname 651.4 Mangrove 1 229.7 93.4 49 567 87.0 1 122

Sweden 0.9 Marsh 4.2 0.3 1082 1.6 93

Tanzania 641.8 Mangrove 146.2 43.8 4401 48.9 237

Thailand 1,746.7 Mangrove 692.8 31.3 197 279 31.0 382

Timor-Leste 0.9 Mangrove

Togo 0.9 Mangrove 0.9 1.2 20 0.3 18

Trinidad and Tobago

47.9 Mangrove 9.3 57.9 369 61.5 864

Turkey 230.6 Marsh 31.2 26.1 38 0.02 142

Turks and Caicos Islands

157.5 Mangrove 10.3 61.90 0 285

United Arab Emirates

38.1 Mangrove & Marsh

515.6 34.62 71 469 11.1 194

United Kingdom 469.8 Marsh 2 431.3 46.20 492 834 46.2 907

United States 15,910.9 Mangrove & Marsh

15 495.6 52.45 656 117 38.1 1 542

Uruguay 15.5 Marsh 1.4 0.22 944 9.7 36

Vanuatu 4.1 Mangrove 1.7 12.47 32 36.4 328

Venezuela 3,055.3 Mangrove 660.8 39.20 21 008 16.0 1 831

Vietnam 1,602.6 Mangrove 15 940.8 61.36 9 277 995 61.5 857

Yemen 0.8 Mangrove 9.1 1.38 180 0.4 157

Global hotspots

93

Comparison of the DIVA and GTSR datasets

Figure SI 3.7 Maps representing the differences between the DINAS-COAST Extreme Sea Levels (DCESL, called DIVA here) and the Global Tide and Surge Reanalysis (GTSR) extremes for a return period of 100 years: (a) the sea level height for the DCESL/DIVA data relative to mean sea level; (b) the sea level height for the GTSR data relative to mean sea level; (c) the difference between DCELS/DIVA and GTSR; (d) whether the DCELS/DIVA extremes are within the 5 and 95 % confidence bounds of the fitted Gumbel distribution of the GTSR extremes. From (Muis, Verlaan, Nicholls, et al., 2016)

Chapter 3

94

Results for the GTSR dataset

Table SI 3.2 Values of the surface area and type of tidal wetlands as well as of the different variables related to storm surge risk mitigation based on a 1-in-100 year storm surge levels from the GTSR dataset at a country level for the countries having tidal wetlands along their coastline. The percentages of the fifth and seventh columns correspond to the percentage of the coastal plain or population benefiting from flood risk mitigation by tidal wetlands relative to the flood-exposed coastal plain area or population.

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

Albania 29.1 Marsh 58.4 73.1 185 81.1 388.4

Angola 245.2 Mangrove 4.2 7.7 13 0.6 281.6

Antigua and Barbuda

3.3 Mangrove

Argentina 982.2 Marsh 1 069.3 24.5 4 238 24.9 1 188.5

Australia 16,824.0 Mangrove & Marsh

7 120.3 52.0 3 622 16.6 1 042.0

Bahamas 306.4 Mangrove 338.1 45.8 30 7.6 435.5

Bangladesh 4,458.2 Mangrove 323.4 26.9 59 061 14.8 491.1

Belgium 2.0 Marsh 324.8 21.5 77 223 23.0 1 602.6

Belize 408.3 Mangrove 78.2 66.2 5 62.5 207.0

Benin 9.4 Mangrove 43.3 27.2 1 233 4.2 692.3

Brazil 8,350.5 Mangrove & Marsh

2 963.1 48.0 75 886 34.3 1 081.5

Brunei Darussalam

84.3 Mangrove

Cambodia 309.2 Mangrove 21.0 52.1 1 665 79.6 552.4

Cameroon 2,163.7 Mangrove 10.2 92.3 23 100.0 2 149.9

Canada 103.0 Marsh 813.6 2.0 10 020 18.1 328.3

Cayman Islands 58.2 Mangrove 5.7 87.5 5 141 71.0 241.2

Chile 5.2 Marsh 0.7 0.03 3 0.003 76.4

China 4,956.5 Mangrove & Marsh

13 521.2 37.5 7 116 768 45.8 713.8

Colombia 1,451.6 Mangrove 189.9 27.1 13 061 22.0 876.8

Congo DRC 221.1 Mangrove 6.8 100.0 1 100.0 552.1

Costa Rica 317.1 Mangrove 6.8 42.2 3 003 100.0 722.5

Côte d'Ivoire 4.3 Mangrove 4.3 83.3 80 100.0 104.4

Croatia 2.9 Marsh

Cuba 3,321.1 Mangrove 680.0 45.0 906 1.4 266.8

Cyprus 19.1 Marsh

Denmark 187.3 Marsh 947.6 45.4 34 574 49.3 547.4

Djibouti 0.8 Mangrove

Global hotspots

95

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

Dominican Republic

122.0 Mangrove 13.0 22.8 68 5.6 128.1

Ecuador 961.22 Mangrove & Marsh

484.5 59.3 6 508 21.7 1 874.9

El Salvador 296.34 Mangrove 3.3 66.7 5 7.4 458.1

Equatorial Guinea

141.89 Mangrove

Eritrea 6.64 Mangrove 1.7 2.1 114.6

Estonia 1.34 Marsh

Fiji 898.35 Mangrove 27.8 69.5 328 83.9 655.0

Finland 79.33 Marsh 120.2 6.5 24 300 24.9 271.5

France 657.02 Marsh 1 887.1 39.6 118 586 15.0 906.1

French Guiana 809.11 Mangrove 40.8 44.5 1 550 67.0 560.7

Gabon 1,325.70 Mangrove 5.1 42.9 458.6

Gambia 514.32 Mangrove 9.8 74.6 5 117 38.5 279.5

Germany 142.58 Marsh 5 599.1 55.1 601 389 47.1 469.0

Ghana 54.44 Mangrove 174.3 42.6 25 284 60.2 711.3

Guadeloupe 15.60 Mangrove

Guatemala 286.04 Mangrove 0.8 11.1 1 7.7 41.5

Guinea 2,078.73 Mangrove 418.4 92.6 10 881 60.1 1 501.7

Guinea-Bissau 1,962.32 Mangrove 147.0 75.8 805 53.8 839.7

Guyana 162.74 Mangrove 196.1 31.0 5 387 8.5 514.0

Haiti 100.08 Mangrove 1.6 49.9 10 100.0 441.0

Honduras 457.48 Mangrove 5.8 63.7 32 91.4 440.3

Iceland 15.50 Marsh 16.0 2.2 15 0.1 495.2

India 3,209.84 Mangrove 1 962.2 29.9 821 599 58.6 795.1

Indonesia 22,263.42 Mangrove 2 571.6 39.8 392 471 25.9 658.0

Iran 48.96 Mangrove 83.7 2.5 298 1.4 220.4

Ireland 45.67 Marsh 109.2 20.7 11 159 42.8 286.9

Italy 370.18 Marsh 74.2 24.2 13 544 20.1 186.6

Jamaica 46.40 Mangrove 18.7 65.7 2 055 97.4 224.5

Japan 1.56 Mangrove

Kenya 243.88 Mangrove 70.7 27.2 2 102 78.3 341.5

Latvia 0.47 Marsh

Liberia 22.96 Mangrove

Madagascar 1,686.30 Mangrove & Marsh

437.3 73.2 3 457 24.4 1 651.3

Malaysia 4,513.54 Mangrove 63.0 29.7 24 642 54.1 781.0

Chapter 3

96

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

Martinique 3.31 Mangrove

Mexico 7,932.09 Mangrove & Marsh

4 230.0 58.5 32 097 28.4 1 117.2

Micronesia 37.35 Mangrove

Montenegro 1.27 Marsh

Morocco 8.33 Mangrove 39.4 4.3 117 0.4 598.1

Mozambique 2,241.42 Mangrove 1 081.4 77.0 7 828 59.8 1 499.4

Myanmar 3,962.4 Mangrove 1 731.4 51.3 162 455 42.6 788.7

Netherlands 149.7 Marsh 3 980.8 27.5 1 329 940 19.3 800.4

New Caledonia 106.8 Mangrove 6.4 44.6 14 35.0 237.6

New Zealand 150.7 Mangrove & Marsh

136.9 29.6 2 069 49.5 210.1

Nicaragua 470.5 Mangrove 49.9 44.5 215 64.6 635.2

Nigeria 5,550.9 Mangrove 44.3 46.4 9 629 30.3 1 505.8

North Korea 0.5 Marsh 16.0 2.6 2 204 3.4 328.8

Pakistan 339.0 Mangrove 883.8 30.4 18 533 30.0 579.4

Palau 28.0 Mangrove

Panama 1,231.9 Mangrove 94.7 45.7 727 9.0 371.0

Papua New Guinea

4,056.8 Mangrove 361.4 80.2 431 6.1 1 278.6

Peru 283.2 Mangrove & Marsh

16.5 1.4 30 3.1 1 225.7

Philippines 1,037.2 Mangrove 200.6 22.8 13 843 7.5 297.9

Portugal 182.0 Marsh 191.7 73.5 11 048 55.0 628.2

Puerto Rico 59.3 Mangrove & Marsh

17.1 33.3 1 268 6.6 325.4

Qatar 0.8 Mangrove 2.3 0.6 1 0.01 108.0

Romania 49.1 Marsh 6.1 3.4 82 3.2 170.6

Russian Federation

7,069.8 Marsh 140.1 0.8 13 0.02 739.2

Saint Lucia 0.8 Mangrove

Saudi Arabia 15.3 Mangrove 176.4 8.6 218 0.2 318.2

Senegal 902.9 Mangrove 99.4 48.3 331 25.0 2 077.6

Seychelles 3.4 Mangrove 0.8 14.2 54.0

Sierra Leone 1,138.1 Mangrove 77.2 70.1 539 75.3 712.3

Singapore 0.9 Mangrove 3.4 5.4 26 2.4 52.5

Slovenia 0.8 Marsh 0.6 100.0 18.0

Solomon Islands 94.7 Mangrove 9.3 4.5 108 5.0 147.7

Somalia 4.3 Mangrove 5.1 1.3 1 0.00 161.0

Global hotspots

97

Country Tidal

Wetlands Area

Tidal Wetlands

Type

Flood-exposed coastal plain

benefiting from mitigation

Flood-exposed population

benefiting from mitigation

Mean distance of wetlands crossed

(km²) (km²) (%) (# people) (%) (m)

South Africa 29.3 Mangrove & Marsh

Spain 815.2 Marsh 192.9 39.7 9 704 11.1 1 062.5

Sri Lanka 55.9 Mangrove

Suriname 651.4 Mangrove 816.4 95.1 17 325 77.6 1 157.4

Sweden 0.9 Marsh 4.2 0.6 1 082 3.7 102.5

Tanzania 641.8 Mangrove 52.1 33.4 3 159 66.6 177.2

Thailand 1,746.7 Mangrove 223.8 22.3 22 487 16.7 340.2

Timor-Leste 0.9 Mangrove

Togo 0.9 Mangrove 0.9 1.8 20 0.8 18.0

Trinidad and Tobago

47.9 Mangrove 3.4 100. 156 100.0 792.7

Turkey 230.6 Marsh 14.0 36.7 8 0.1 171.6

Turks and Caicos Islands

157.47 Mangrove 2.4 60.0 130.8

United Arab Emirates

38.13 Mangrove & Marsh

294.2 24.3 47 677 10.1 227.5

United Kingdom 469.77 Marsh 3 643.4 54.9 595 079 48.2 1 032.9

United States 15,910.85 Mangrove & Marsh

13 717.0 48.1 165 060 29.2 868.7

Uruguay 15.50 Marsh

Vanuatu 4.11 Mangrove 0.8 7.1 10 16.4 150.5

Venezuela 3,055.27 Mangrove 253.5 27.7 6 150 18.6 584.8

Vietnam 1,602.57 Mangrove 4 412.0 57.0 1 628 479 51.4 701.7

Yemen 0.83 Mangrove

Chapter 3

Figure SI 3.8 GTSR results of the absolute surface area benefiting from a storm surge pathway crossing through tidal wetlands represented on the associated coastal segment (km²), with circles highlighting the segments for which the coastal plain area benefiting from storm surge mitigation is greater than 1 000 km².

Global hotspots

Figure SI 3.9 GTSR results of the standardized area, i.e. surface area per unit of shoreline length (km²/km), benefiting from a storm surge pathway crossing through tidal wetlands represented on the associated coastal segment, with circles highlighting the segments for which the surface area benefiting from storm surge mitigation is greater than 15 km² per 1 km of shoreline.

Chapter 3

Figure SI 3.10 GTSR results of the mean distance (m) travelled through tidal wetlands by the storm surge during its landward propagation. The circles highlight the segments for which the mean distance travelled through tidal wetlands by the storm surge is longer than 5 km, in red are the hotspots were the long distance of wetlands coincides with high exposure to cyclones.

Global hotspots

Figure SI 3.11 GTSR results of the absolute number of people benefiting from a storm surge pathway crossing tidal wetlands represented on the associated coastal segment, with circles highlighting the segments for which the population benefiting from storm surge mitigation is higher than 100 000 people.

Chapter 3

Figure SI 3.12 GTSR results of the standardized population, i.e. people per unit of shoreline length (number of people/km), benefiting from a storm surge pathway crossing tidal wetlands, with circles highlighting the segments for which the population benefiting from storm surge mitigation is greater than 10 000 people per 1 km of shoreline.

Cities

103

CHAPTER 4 Potential for nature-based flood risk mitigation in coastal cities around the world

Rebecca Van Coppenolle and Stijn Temmerman

Based on the paper submitted to Earth’s Future in July 2018

Chapter 4

104

Abstract

Nature-based risk mitigation is increasingly proposed as a strategy to cope with

global changes that increase flood risks in coastal areas. However, loss of coastal

ecosystems reduces their mitigating effect on coastal flood risks in many places

around the world. Here we identify global urban hotspots exposed to storm surge

flood risks, where conservation of existing coastal ecosystems can contribute to

nature-based risk mitigation. We present a global procedure identifying the most

likely pathways followed by storm surges from the open sea towards 136 cities

around the world, and quantifying the extent of mangrove forests, salt marshes,

seagrass meadows and/or coral reefs along these storm surge pathways. Cities

that combine large flood-exposed populations (> 400 000 people exposed to 1-in-

100 years storm events) and large potential for nature-based risk mitigation (>

200 km² of coastal ecosystems) are located in large river deltas and estuaries, such

as Khulna (Ganges-Brahmaputra delta in Bangladesh), Guayaquil (Guayas delta in

Ecuador), Ho Chi Minh City (Mekong delta in Vietnam) and New Orleans

(Mississippi delta in USA). Here conservation of mangroves and salt marshes plays

a key role. Cities with large populations and/or assets at risk, but few ecosystems,

are either located directly adjacent to coastlines, or where former wetlands have

been reclaimed, especially in European and Asian cities. Overall, an encouraging

75 % of the studied cities benefit from present ecosystems. Hence our study calls

for conservation and (re-)creation of coastal ecosystems as a sustainable strategy

for nature-based mitigation of increasing coastal flood risks.

Cities

105

4.1 Introduction

Coastal populations are exposed to increasing natural hazards due to socio-

economic and climatic changes (Barbier, 2014; Syvitski et al., 2009). Global climate

change causes accelerating sea level rise and increasing frequency of high intensity

storms, and as such increases the risks of coastal flooding and erosion by storm

surges and wind waves (Bengtsson et al., 2006; Knutson et al., 2010; Webster et

al., 2005; Woodruff et al., 2013). In addition, socio-economic developments are

leading to an increasing coastal population density and increasing value of assets,

in particular in coastal cities (Green & Short, 2003; Hallegatte et al., 2013;

Mcgranahan et al., 2006; Neumann et al., 2015; Nicholls et al., 2008; de Sherbinin

et al., 2007). Human activities in the coastal zone, such as the conversion of natural

wetlands into agricultural, industrial or urban areas, may interfere with local flows

of water and sediments, may cause land subsidence, and as such further aggravate

coastal flood and erosion risks (Adam, 2002; Auerbach et al., 2015; Kirwan &

Megonigal, 2013; Pethick & Orford, 2013; Syvitski et al., 2009). All together these

socio-economic and climatic changes highlight the need for sustainable protection

of coastal societies, and especially densely populated coastal cities, against

increasing flood and erosion risks.

The building of protective structures, such as dikes, levees and dams, is commonly

considered the standard solution to protect against coastal flood and erosion risks.

Although such coastal protection structures are implemented in several coastal

zones around the world, a high share of coastal cities remain highly vulnerable to

coastal flooding due to flood defences with relatively low safety standards (e.g.

protection against a 1 in 100 year event instead of 1 in 1 000 year event), or non-

reliable or absent coastal protection structures (Dasgupta et al., 2009; Green &

Short, 2003; Mcgranahan et al., 2006; Nicholls et al., 2008; de Sherbinin et al.,

2007). While the implementation of coastal protection is necessary, it is not solely

linked to the richness of the country, but the decision to build effective defences

against coastal flooding also strongly depends on the agenda of policy makers

(Nicholls et al., 2008).

Nature-based solutions for coastal flood and erosion risk reduction rely on the

natural ability of coastal ecosystems to mitigate flood and erosion risks (Cheong et

al., 2013; Gedan et al., 2011; Sutton-Grier et al., 2015; Temmerman et al., 2013),

while at the same time providing additional valuable ecosystem services as carbon

sequestration, contribution to fisheries production and water quality regulation

(Barbier et al., 2011; McLeod et al., 2011). The flood and erosion risk mitigation

capacities of natural coastal ecosystems were demonstrated in several studies,

Chapter 4

106

mainly for mangrove forests, salt marshes, seagrass meadows and coral reefs

(Guannel et al., 2016; Koch et al., 2009; Narayan et al., 2016). These coastal

ecosystems provide a barrier against landward propagation of wind waves

(Fonseca & Cahalan, 1992; Gedan et al., 2011; McIvor et al., 2012) and storm

surges (Ferrario et al., 2014; Guannel et al., 2016; McIvor, Spencer, et al., 2012;

Möller et al., 2014), and therefore reduce the risks of shoreline erosion and

flooding. Additionally, they are able to adapt to the rising sea level by natural

processes of sediment accretion (Buddemeier & Smith, 1988; Lovelock et al., 2015;

J. T. Morris et al., 2002), although this adaptation ability depends on local

conditions such as the rate of relative sea level rise, the tidal range, the availability

of sediments and nutrients (Alongi, 2008; Kirwan et al., 2010, 2016; Simas et al.,

2001). The magnitude of the flood risk mitigation, measured as the amount of

wave or storm surge height reduction per distance travelled through the

ecosystem, is driven by a multitude of parameters such as the phenological and

morphological traits of the species present in the coastal ecosystems, the

geomorphology of the ecosystem and of the wider surrounding coastal area, and

the characteristics of the wind waves or storm surges (Gedan et al., 2011; Koch et

al., 2009; Mazda et al., 2006). As a consequence, a unique quantitative value of the

rate of wind wave or storm surge height reduction by the different ecosystems

cannot be defined. Nevertheless, ranges of values of wave height and storm surge

height reduction are known from local studies.

Tropical mangrove forests have been shown to mitigate wind waves, storm surges

and to a limited extent small tsunamis (Alongi, 2008; Krauss et al., 2009; Mazda et

al., 2006; McIvor, Möller, et al., 2012; McIvor, Spencer, et al., 2012; Zhang et al.,

2012). Recorded values show that 500 m of mangrove forest can reduce small

wind waves (< 70 cm high) by 50 to 99 % (McIvor, Möller, et al., 2012), while

storm surge height reduction ranges from several centimetres to 50 cm per

kilometre travelled by the storm surge through the mangrove forests (Gedan et al.,

2011; McIvor, Spencer, et al., 2012; Zhang et al., 2012). The lower vegetation of

salt marshes is generally providing less friction than mangrove trees.

Nevertheless, salt marshes can reduce wind waves up to 80 % over several tens of

meters (Moller et al., 1999; Ysebaert et al., 2011) and can lower storm wind waves

by 60 % over 40 m (Möller et al., 2014). Salt marshes can also attenuate the height

of storm surges at rates of 1.7 to 25 cm per kilometre travelled through the marsh

(Leonardi et al., 2018; Shepard et al., 2011; Stark et al., 2015; Wamsley et al.,

2010). Seagrass meadows are expected to reduce the flood wave energy with a

magnitude comparable to salt marshes under wind wave and storm conditions

(Duarte et al., 2013; Fonseca & Cahalan, 1992), with the maximum reduction in

shallow water and low wave energy environments (Fonseca & Cahalan, 1992;

Cities

107

Ondiviela et al., 2014). Coral reefs can provide wave energy attenuation similar to

artificial defences (Ferrario et al., 2014), by reducing in average the wave energy

by 97 %, with most of the reduction provided by the reef crest (Ferrario et al.,

2014; Principe et al., 2012; UNEP-WCMC, 2006). While most of the studies on

wave attenuation by coastal ecosystems are focusing on moderate wind waves, the

evidence of wave mitigation under storm conditions is increasing (Guannel et al.,

2016; McIvor, Spencer, et al., 2012; Narayan et al., 2016; Stark et al., 2015).

As the flood and erosion defence function of natural ecosystems becomes

increasingly demonstrated, nature-based coastal defence is starting to be

implemented in a growing number of coastal areas, often as an add-on to

traditional engineered defence structures (Barbier et al., 2008; Costanza et al.,

2008; Gedan et al., 2011; McGranahan et al., 2007; Spalding, Ruffo, et al., 2014;

Temmerman et al., 2013). The conservation and/or restoration or creation of

ecosystem buffers between the sea and populations at risk, is often a more cost-

efficient strategy than only classic engineering solutions, as coastal ecosystems are

self-adaptive to sea level rise by sediment accretion (Alongi, 2008; Kirwan et al.,

2016; Simas et al., 2001) and hence need less maintenance than engineered

defence structures. Some observations by the National Park Service in the US

based on real projects show that the installation of engineering infrastructures

(e.g. seawalls, dikes, breakwaters...) cost usually from 6 500 to 9 800 $ per linear

meter, while nature-based projects usually have a cost of installation ranging from

0 (ecosystems already present) to 6 600 $ per meter (Beavers et al., 2016; Sutton-

Grier et al., 2018). The nature-based projects are also cost-efficient in terms of

maintenance and reparation, with observations of 0 to 328 $ per meter, in

comparison to the maintenance and repair costs of the engineering infrastructures

ranging from 0 to 1 710 $ per meter (Beavers et al., 2016). Furthermore, natural

ecosystems provide additional societal benefits through ecosystem services. This

makes nature-based solutions particularly relevant for areas with lower financial

resources. Yet, existing coastal flood protection programs that include nature-

based approaches are still relatively scarce on a global scale. Examples include the

large-scale restoration of marshes in the Mississippi deltaic plain with the

objective to reduce landward propagation of storm surges, in combination with

engineered flood defences (Coastal Wetlands Planning Protection and Restoration

Act (CWPPRA), n.d.; Day et al., 2007), or the projects of ‘managed coastal

realignment’, which is the landward relocation of flood defence structures to

accommodate space for coastal ecosystem development, and which is applied for

example in the UK and elsewhere in Europe (R. A. Garbutt et al., 2006; Gardiner et

al., 2007; Huguet et al., 2017; Rupp-Armstrong & Nicholls, 2007; SigmaPlan, 2017).

Chapter 4

108

Although nature-based coastal protection is increasingly valued in several places

around the world, we lack so far a comprehensive global assessment of where the

major global hotspots are located that have a high potential for nature-based

mitigation of coastal flood and erosion risks. With this study, we aim to bridge this

gap by quantifying how much surface area of the four above-described coastal

ecosystem types is present along the likely storm surge pathways between the sea

and the 136 coastal cities with the highest flood-prone populations in the world, as

defined by Nicholls et al. (2008). In addition, the study aims to identify the social

and physical parameters that influence the distribution of the different coastal

ecosystems in between the studied cities and the open sea. Finally, a comparison

between the surface area of the coastal ecosystems, the populations at risk and the

assets at risk was made as an aim to identify the global urban hotspots with large

flood-prone populations and assets and at the same time large flood-protecting

ecosystems.

4.2 Method

The coastal cities considered in the analysis have a population of more than 1

million inhabitants in 2005 (UN, 2005) and correspond to the 136 cities studied by

Nicholls et al. (2007). The term “city” corresponds here to the urban

agglomeration which is the area comprising the city centre and the built-up areas

contiguous to the administrative city boundaries (United Nations Department of

Economic and Social Affairs Population Division, 2016). For the analysis, the city is

represented as a single point, corresponding to the city centre. As the cities have a

certain area and might not be circular, the full extent of the city might not be

accounted for in the likely area influencing the propagation of the surge (see

below) and therefore the surface area of coastal ecosystems that provide

protection to the city may be underestimated

The first part of the analysis, performed in ArcGIS (10.3.1), was to identify the area

through which storm surges are likely to propagate from the sea towards the city.

This was based on the following steps.

The first step was to define the offshore source locations from which the

path of a storm surge would start and then would propagate inland. Storm surges

are generated at the open sea and from there propagate as a long wave towards

the coast and over the coastal floodplain. The source location of storm surges was

defined as the limit between the open sea, on the one hand (i.e. where the surge is

generated), and the emerged land including small water areas such as

embayments, estuarine and deltaic water surfaces, on the other hand (i.e. where

Cities

109

the surge is propagating over as a long wave). To create this limit, the General

Bathymetric Chart of the Oceans (GEBCO 14) that combines the ocean bathymetry

with the Shuttle Radar Topography Mission dataset of land topography was used.

This dataset has a grid resolution of 30 arc-seconds. It was classified into values of

1 for the emerged land areas (elevation > 0 m) and a value of 0 for the water

bodies (elevation < 0 m). A focal statistic analysis was then performed to draw the

limit of the ‘open sea’, i.e. excluding small near-shore water areas such as

embayments, estuarine and deltaic water surfaces, with a width smaller than

about 3 km. The procedure makes use of a moving window of 10 by 10 pixels that

applies to the pixel in the centre the most represented value in the window (0 or

1). The limit between those two areas is considered as the open sea limit, and

corresponds to the line where the path of a storm surge will start to be considered

in our analysis.

The second step assumed that a storm surge will preferably propagate

over water bodies (i.e. the embayments, estuarine and deltaic channels). Cost

distance algorithms (http://desktop.arcgis.com/en/arcmap/10.3/tools/spatial-

analyst-toolbox/how-the-cost-distance-tools-work.htm) were used to determine

the most likely pathway followed by a storm surge from the open sea to the city

centre. For the cost distance algorithms, the land areas were given a cost of 1000,

while the water bodies were given a cost of 1. As such the probable flood pathways

between the open sea and the city centre were defined for every city. A few

examples are presented in the Supplementary Information.

The third step defined the likely area influencing the propagation of the

storm surge. It assumes that a buffer area of 20 km around the probable storm

surge pathway is including both the potential variations in the storm surge’s

pathways and the area in which the presence of ecosystems (mangroves, salt

marshes, seagrasses, and coral reefs) can have an effect on the reduction of wave

and surge heights, and of erosion risks. With this procedure the buffer also extends

20 km offshore on the ‘open sea’ (see examples in the Supplementary Information)

and as such can also include the offshore ecosystem types (seagrasses and coral

reefs). These offshore ecosystems also contribute to wave, surge and erosion risk

reduction for the urban populations at risk, while far offshore ecosystems located

outside this 20 km buffer are not considered relevant anymore for risk reduction

in the city. Different values of buffer areas were tested (10, 20 and 30 km) to

evaluate their impact on the surface area of coastal ecosystems per city. The

surface area of coastal ecosystems inside the three buffer sizes (10, 20 and 30 km)

is presented in the Supplementary Information.

Chapter 4

110

The comparison of the coastal ecosystems surface area in the three buffer

size shows that for the cities that have no coastal ecosystems in the 10 km buffer

(i.e. 39 cities) four cities, (i.e. Xiamen, Yantai, Rangoon and Rotterdam) have a

presence of coastal ecosystems in their 20 km buffer area, and 6 more cities have a

presence of coastal ecosystems in the 30 km buffer (i.e. Qingdao, Athens, Kolkata,

Nagoya, Tokyo and Benghazi). For the cities having coastal ecosystems in the 10

km buffer area (i.e. 97 cities), 60 % have an increase of coastal ecosystems surface

area smaller than 60 km² in the 30 km buffer area, the other 40 % of those cities

have a higher increase in surface area. However, the difference in surface areas of

coastal ecosystems for the different cities and for the different buffer size remains

relative. As such, the comparison of the coastal ecosystems area between the cities

remains similar in the case of the use of the three buffer sizes.

The fourth step of the analysis was to determine the surface area of the

four coastal ecosystems accounted for inside the area influencing the propagation

of the storm surge. The spatial distribution of the different coastal ecosystems is

based on global scale datasets: the Global distribution of Mangroves (Giri et al.,

2011), the Global distribution of Saltmarshes (Mcowen et al., 2017), the World Atlas

of Seagrasses (Green & Short, 2003) and the World Atlas of Coral Reefs (Spalding et

al., 2002) from the United Nations Environmental Program – World Conservation

Monitoring Centre (www.unep-wcmc.org).

The second part of the study defined the surface area inside the 20 km buffer

around the probable flood pathway that probably consisted of coastal wetlands in

historic time and that was reclaimed for human land use throughout history. This

estimation of historically reclaimed coastal wetlands was used then to test if it can

explain the current distribution of coastal wetlands in between cities and the open

sea. As such, this part of the analysis is focussing specifically on two of the

considered ecosystem types, mangroves and salt marshes, that were in many

places around the world converted into human land use such as agricultural fields,

aquaculture ponds, industrial or urban areas - called here generally ‘land

reclamation’. This estimation of the surface area of land reclamation was made by

selecting for every city the land area that sits below the mean high tide, and

therefore is expected to be intertidal, but is currently not covered by mangroves or

marshes. The absence of mangroves and marshes in those areas is interpreted as a

historical conversion of former wetlands (mangroves, marshes) into human land

use.

The mean high tide is obtained from mean sea level augmented by the tidal

amplitude. The datasets used in this step of the analysis are the NASA Shuttle

Cities

111

Radar Topography Mission v3.0 at a resolution of 1 arc-second and the Finite

Element Solution (2012) – Global Tide from AVISO that contains the information

on the tidal amplitude. The Principal Lunar semi-diurnal component (M2) was

used to define the average tidal amplitude in front of every city. Due to data

limitations, the city of Helsinki (Finland) could not be included in this step of the

analysis.

The third part of our analysis aims to explore which parameters can explain the

presence and the size of the coastal ecosystems in front of the studied cities. For

this, two statistical analyses (logistic and linear regressions) were performed on a

set of social and physical parameters (Table 4.1). Except for the country’s GDP per

capita (http://databank.worldbank.org/data/home.aspx), the parameters are

obtained through the above-mentioned datasets and analysis in ArcGIS (10.3.1).

Except where mentioned, all the values are extracted for the probable area

influencing the propagation of the storm surge (i.e. the 20 km buffer around the

probable flood pathway).

Table 4.1 Physical and Social parameters extracted from the data for each city

Physical Parameters Social Parameters Type Unit Type unit

Latitude of the city Degree Short-distance population density (within 20 km around the flood pathway)

Inhabitants/km²

Distance between the sea and city

km Intermediate-distance population density (within 50 km around the flood pathway)

Inhabitants/km²

Coastline length km Long-distance population density (within 100 km around the flood pathway)

Inhabitants/km²

Area below mean high tide km² Country’s GDP per capita Constant 2005 US$

Shallow water area (> -100 m)

km²

Deep water area (< -100 m)

km²

Through logistic regression including all the studied cities and focussing on one

ecosystem at the time, each of the social and physical parameters (i.e. candidate

explanatory variables) was tested for association with the presence of the

ecosystem type (dependent variable). To do so, the cities were split in two groups

according to the presence or absence of the considered coastal habitat. The results

of the logistic regression consists of odds ratios (and their corresponding p-value)

that quantify the increase in odds of finding a coastal ecosystem for a specific

increase in the value of the explanatory variable.

Chapter 4

112

The influence of the candidate explanatory variables (social and physical

parameters) on the surface area of the coastal ecosystems (dependent variable)

was tested through linear regression. In this analysis, only the cities in which the

coastal ecosystem under study is present were retained. One assumption of a

linear regression model is the normality of the residuals. As deviations from

normality were observed in the data, the model was refitted and the dependent

variable used is the natural logarithm of the surface area of the ecosystem. As

such, the residuals showed a normal distribution. The results obtained from the

linear regression correspond to the regression coefficients of the slope, the

standard error of the coefficient and the associated p-value.

Finally the last part of the study focussed on a comparison of the surface area of

the four coastal ecosystems, the population at risk and the assets at risks in order

to identify the cities with both large population and/or assets at risks of flooding

and large coastal ecosystem areas. The population and assets at risk of coastal

flooding in every city were taken from the results of Nicholls et al. (2007) and

correspond to the population and assets exposed to coastal flooding due to storm

surge and high wind damages for a 1 in 100 years storm surge, without any

consideration of coastal defences or adaptations.

4.3 Results and Discussion

4.3.1 Coastal Ecosystems Areas

From the 136 cities studied, 101 cities have at least one natural coastal ecosystem

along their likely flood pathway, while 35 cities have a complete absence of coastal

ecosystems. The majority of cities (52) have only one ecosystem type along their

likely flood pathway, 39 cities benefit from the presence of two coastal ecosystem

types, while seven and three cities have respectively three and four ecosystem

types along their likely flood pathway (see Figure 4.1). The more frequent

combinations of ecosystem types are salt marshes and seagrass meadows, in front

of North American, European and Australian cities, while the combination of

mangrove forests and coral reefs is present in front of tropical region’s cities.

In total, the most represented coastal ecosystem is seagrass meadows with 5 195

km² followed by 4 890 km² of mangrove forests and 1 974 km² of salt marshes.

Coral reefs are much less present with a total of 282 km², partially due to their

more offshore location in comparison to the three other ecosystems.

The total surface area of natural coastal ecosystems varies greatly between the

different cities, with the highest surfaces being the 2 006 km² of mangrove forest

Cities

113

in front of Khulna in the Ganges-Brahmaputra delta in Bangladesh (but note that

there are no coastal ecosystems in front of Kolkata or Dhaka in the same delta) and

the 1 138 km² of seagrass meadow (1 022 km²) and mangrove forest (116 km²) in

front of Conakry in Guinea.

Based on the number of cities and the total surface area of coastal ecosystems per

continents, Africa is the continent with the largest averaged surface area of coastal

ecosystems per city (149.0 km²), while Europe is the continent with the lowest

averaged surface area of coastal ecosystems per city (5.7 km²) (Figure 4.2). The

surface of coastal ecosystems in front of cities in Africa, Asia and to a lower extent

South America shows a large variation, with a combination of cities having no or

small coastal ecosystems and cities having a high surface area occupied by coastal

ecosystems along their likely flood pathway, as Khulna in Bangladesh (2 006 km²),

Conakry in Guinea (1 138 km²) or Guayaquil in Ecuador (741 km²). Oceania is the

only continent where all the cities benefit from the presence of coastal ecosystems

along their probable flood pathway.

Chapter 4

Figure 4.1 Representation of the surface area (km²) occupied by coastal ecosystems (size of the circle) and of the different ecosystem types present (colours) within the 20 km buffer zones along the likely storm surge flood pathways for each of the 136 cities with highest flood-exposed populations in 2005.

Cities

115

Figure 4.2 Comparison of the surface area of coastal ecosystems (km²) in front of the studied cities per continent. The width of the boxes corresponds to the square root of the number of cities per continent. The cities of Khulna in Asia and Conakry in Africa are not represented due to their exceptionally high values of 2 006 km² and 1 138 km² respectively

4.3.2 Historical Influence on the Tidal Wetlands Distribution

The surface area of tidal wetlands, as mangroves and salt marshes, in front of the

studied cities is influenced by several parameters. It was tested if an explanation

could be found in the historical societal development of the continents. The

earliest highly populated continents are Asia and Europe (Maddison, 2001;

McEvedy & Jones, 1978; World Population History, n.d.), which was accompanied

by an increasing demand for food production and agricultural land. This was

particularly the case in or near coastal zones, where demand for food production,

urban and industrial development resulted in reclamation of marshes and

mangroves into human land use (Almeida et al., 2014; Hoeksema, 2007; Lotze et

al., 2006; Scott et al., 2014; Tian et al., 2016; Valiela, 2006).

Figure 4.3 represents for each city the surface area within the 20 km buffer zone

around the likely flood pathways where the land elevation is below the mean high

tide but no tidal wetlands exist presently. This area is considered here as an

estimation of the surface area of land that was reclaimed over time by conversion

of formerly existing tidal wetlands into human land use. The observation of the

estimated reclaimed land area per continent shows that Oceania, Africa and South

Chapter 4

116

America are the continents where there was the least reclamation, in contrast to

Europe, Asia and Northern America. The city with the largest extent of estimated

reclaimed land area is Hamburg in Germany with 1 391 km² of reclaimed

intertidal land area in its likely area for storm surge propagation, i.e. the Elbe

estuary. This number is in accordance with previous assessments, showing that

the original intertidal flood plain area decreased enormously by intertidal wetland

reclamation over the past centuries (Hamburg Port Authority, 2006; Hansen,

2015; Reise, 2005) (Supplementary Information Figure SI 4.2 ). Hamburg is

followed by Guangzhou, along the Pearl River delta in China, with an estimated

reclaimed intertidal area of 1 119 km², corresponding with reported estimations

of original wetland loss reaching 50 to 60 % over the delta region (Li & Lee, 1997;

Tian et al., 2016). Ho Chi Minh City in Vietnam has an estimated reclaimed

intertidal area of 785 km². For Rotterdam in the Netherlands this number is of 782

km², resulting from historical marsh embankment and drainage into agricultural

land – so-called “polders” – since the Middle Ages, and more recent construction of

the harbour of Rotterdam (de Haas et al., 2018; Pierik et al., 2017; Ysebaert et al.,

2016). The observations made here are consistent with other studies assessing the

land reclamation over the world (Airoldi & Beck, 2007; Ganong, 1903; Hatvany,

2003; Hoeksema, 2007; Murray et al., 2014).

Our analysis indicates that the lower surface area of coastal ecosystems in front of

European cities can be partly attributed to the long history of conversion of coastal

wetlands into human land use in Europe (often so-called polders). Throughout

Europe, evidence of coastal wetland reclamation for agriculture can be traced back

to the Middle Ages (Hoeksema, 2007; Scott et al., 2014), while in China, the

thirteenth century was already a time of high human influence in the Yangtze and

Yellow river deltas (Scott et al., 2014). In Asia, most of the cities are still fronted by

large areas of coastal ecosystems, mainly mangroves and salt marshes,

nonetheless, the estimation of reclaimed land in front of Guangzhou, Ho Chi Minh

City, Tianjin, Kolkata or Khulna are among the highest in the world, with

estimations ranging from 245 km² for Khulna to 1 119 km² of reclaimed land for

Guangzhou. Indeed these Asian cities are located in the world’s largest deltas,

where on the one hand large surface areas of tidal wetlands have been reclaimed

into human land use, but on the other hand also still relatively large wetlands

remain (Auerbach et al., 2015; Wang et al., 2014). An extreme example in this

respect, is Khulna, located in the Ganges-Brahmaputra delta, with a large

estimated reclaimed land surface area (245 km²) and at the same time the largest

remaining mangrove surface area (2 006 km²; in the Sundarbans) along its likely

flood pathway (Auerbach et al., 2015).

Cities

117

The growth of the population in North and South America and Africa appeared

later (from the end of the 18th century) with the industrialization and the end of

both the colonization and the slave trade. Yet, recent observations of coastal

ecosystems over the world are highlighting the dramatic trend of current large

ecosystem losses (Ma et al., 2014; Scott et al., 2014; Spalding et al., 2010; World

Population History, n.d.). The rise in worldwide commercial relationships

concentrated the increasing population to settle in or near the coasts and major

river mouths, in order to benefit from the water way connections with the rest of

the world, which at the same time increased the human pressure on the coastal

ecosystems (de Sherbinin et al., 2007). During the nineteenth and twentieth

centuries, all the continents knew a peak in land reclamation for human activities,

resulting in a major loss of coastal wetlands. In addition, the pollution,

acidification, warming and rising of the oceans among others are threatening

seagrass meadows and coral reefs (Green & Short, 2003; Spalding, Ruffo, et al.,

2014). It is estimated that 75 % of the world’s coral reefs are under threat

(Spalding, Ruffo, et al., 2014), while about one third of mangrove forests and

seagrass meadows disappeared and half of the world’s salt marshes were

destroyed or degraded over the last 30 years (Barbier et al., 2008; Millennium

Ecosystem Assessment, 2005; Valiela et al., 2009).

Chapter 4

Figure 4.3 Estimation of the surface area (km²) of tidal wetlands (mangroves and marshes) reclaimed within the likely area for storm surge propagation for every city.

Cities

119

4.3.3 Parameters Influencing Coastal Ecosystems Presence and Area

The long term historical evolution of human settlement and activities on the

continents in addition to the recent history of land reclamation and environmental

disturbance is only a part of the explanation behind the variation of the surface

area of coastal ecosystems along the cities’ likely flood pathway. In order to

identify the social and physical environmental variables that may explain the

variations in the presence or absence, or the surface area of the different coastal

ecosystems, logistic (presence/absence) and linear (surface area) regressions

were performed, both as simple and multiple regressions. The results of the simple

logistic and simple linear regressions are presented in the Supplementary

Information, while the results of the multiple logistic and linear regressions are

presented and discussed here after.

The multiple logistic regressions test the influence of each explanatory variable on

the presence or absence of the coastal ecosystem (i.e. dependent variable)

assuming all the rest remains constant. The odd ratio gives the factor of change in

odds to find an ecosystem for an increase in one unit of the explanatory variable

assuming that all the other explanatory variables remain constant. The Table 4.2

present the results of the multiple logistic regressions, i.e. the explanatory

variables significantly (p-value < 0.05) influencing the presence or absence of each

coastal ecosystem.

Table 4.2 Odds Ratio resulting from the multiple logistic regressions testing the influence of the explanatory variables (Table 4.1) on the odds of finding coastal habitats in front of the coastal cities for a significance of 95 % (p-value < 0.05), the ‘X’ corresponds to non-significant relations

Dependent variables

Explanatory variables

Unit of increase

Odd Ratio

Mangrove Salt Marsh Seagrass Coral Reef

Physical Parameters

Latitude 1 ° 0.961 X X 0.971

Distance between the sea and the city 1 km

0.954 X X X

Coastline length 1.006 X X X

Area below mean high tide 1 km²

X X 0.985 X

Shallow Water area (depth > -100 m) X X 1.003 X

Deep Water area (depth < -100 m) X X 1.005 1.007

Social Parameters

GDP Per Capita of the Country 1 US$ 0.999 1.0001 1.0001 X

Short-distance population density 1 inhabitant

/km²

X 0.9997 X X

Intermediate-distance population density X 1.001 X X

Long-distance population density X X X X

Chapter 4

120

The results highlight that the presence or absence of the four coastal ecosystems is

related to a set of explanatory variables. For example, the presence of a mangrove

forest is positively correlated to an increase of the coastline length assuming that

all the other parameters remain constant, while the increase of the latitude or the

GDP per capita each decrease the odds to find a mangrove forest, assuming that

when one variable changes, all the others remain constant. The negative relation

between the latitude and the presence of a mangrove forest is related to the

intrinsic characteristics of mangroves, as they develop in tropical regions that are

in low latitudes.

The condition that all the other parameters remain constant is obviously not

realistic in practice. However, it allows highlighting the ‘pure’ effect of each

parameter on the presence or absence of a coastal ecosystem.

Secondly, the multiple linear regressions focus on the influence of the tested

explanatory variables on the surface area of the four considered coastal

ecosystems. As for the multiple logistic regressions, the influence of each

explanatory variable is considered assuming all the other variables remain

constant.

Table 4.3 Regression coefficients resulting from the multiple linear regressions testing the influence of the explanatory variables (Table 4.1) on the size of the coastal habitats in front of the coastal cities for a significance of 95 % (p-value < 0.05), the ‘X’ corresponds to non-significant relations

Dependent variables

Explanatory variables

Unit of increase

Multiple Regression coefficients

Mangrove Salt Marsh Seagrass Coral Reef

Physical Parameters

Latitude 1 ° X X X X

Distance between the sea and the city 1 km

0.156 X X X

Coastline length 1.992 X X X

Area below mean high tide

1 km²

X X X X

Shallow Water area (depth > -100 m) 0.887 1.040 1.112 X

Deep Water area (depth < -100 m) X X X 1.021

Social Parameters

GDP Per Capita of the Country 1 US$ 0.998 X X X

Short-distance population density 1 inhabitant

/km²

X X 1.018 X

Intermediate-distance population density X 0.974 0.960 X

Long-distance population density 0.935 1.083 X X

Cities

121

The results show, for example, that the surface area of seagrasses is increasing for

an increase of the shallow water area, assuming all the other variables remain

constant. This seems to be related to the fact that seagrasses are growing in

shallow water areas, and as such an increase of their favourable environment lead

to an increasing possibility to have larger seagrass bed areas.

For the multiple logistic and linear regressions, the results highlight statistically

significant relations between the explanatory variables and the presence/absence

or surface area of the four coastal ecosystems. Nonetheless, those results must be

interpreted carefully, as the natural mechanisms behind the presence and size of a

coastal ecosystem are not accounted for. For example, according to the results, the

surface area of salt marshes is increasing for an increase of the population density

in the long-distance environment of the city (100 km). While this is statistically

significant, it seems unlikely in regards to the pressure the populations puts on the

coastal ecosystems.

The different regression analyses performed can then statistically back-up some

expected relations, as the disappearance of the mangrove forests and coral reefs as

the latitude increases, or the increase of the seagrass bed areas for an increase in

shallow water area. Yet, some relations must be interpreted carefully and should

be explored further to confirm or reject the statistical significance. Parameters as

the bio-geomorphological characteristics of the region (e.g. soil constitution,

granulometry), the water chemistry (e.g. salinity, pH, and temperature), or the

waves and currents conditions (e.g. orientation, energy...) are also expected to play

an important role in the establishment and development of coastal ecosystems and

differ from one city to another. As a consequence, some locations are not suitable

for the development of the considered coastal ecosystem types, such as the

Western coast of South America due to its steep topography and bathymetry (but

large deltas also exist there such as the Guayas delta in Ecuador) (Scott et al.,

2014).

4.3.1 Hotspots of Populations and Assets at Risk and Large Ecosystems

Finally, we compared our results of the geographical repartition of the coastal

ecosystems in front of every city to the people and assets exposed to coastal

flooding as defined by Nicholls et al. (2007), in order to highlight the share of the

world’s urban population and assets at risk that can benefit from the presence of

coastal ecosystems. Figure 4.4 shows on the one hand that some of the cities with

the largest populations at risk (> 412 000 people) can benefit from the presence of

large surface areas of coastal ecosystems (> 322.09 km²), such as Khulna

Chapter 4

122

(Bangladesh), Guayaquil (Ecuador), Shanghai and Guangzhou (China), Ho Chi Minh

City (Vietnam) and New Orleans (USA). All of these cities are located in large river

delta plains, where large mangroves or marsh areas exist in portions of the delta

plain between the city and the open ocean. On the other hand, a high number of

cities with a large population at risk of flooding either benefit from a limited

surface area of coastal wetlands (e.g. Boston, Rotterdam or Lagos) or are deprived

from any coastal ecosystem (e.g. Tokyo and other Japanese cities, Dhaka,

Amsterdam or Montreal). These are mostly cities either directly located along the

open sea (e.g. most Japanese cities) or where coastal ecosystems between the city

and sea have been historically reclaimed and turned into human land use (e.g.

Amsterdam, Rotterdam, Boston).

By looking at the ranking of the cities based on their population at risk, assets at

risk and surface area of coastal ecosystems, the top twenty cities with the largest

populations exposed to coastal flooding contain 71 % of the population and 69 %

of the assets at risk against only 15 % of the surface area of coastal ecosystems

that could participate to risk mitigation (Figure 4.5; all percentages are relative to

the total values for all 136 cities considered in this study). A similar result is found

for the top twenty cities with the most assets exposed to coastal flooding. In

contrast, the top twenty cities with the highest surface area of coastal ecosystems

contain 77 % of the coastal ecosystems found in front of the 136 studied cities, but

only about 30 % of both the population and assets at risk are found in those

twenty cities (Figure 4.5).

Several cities combine high values of population and assets at risk of flooding (12

cities with > 696 000 people), among which eight benefit from the presence of

coastal ecosystems, but only four have large surface areas of coastal ecosystems,

Miami (237 km²), New Orleans (241 km²), Guangzhou (304 km²) and Shanghai

(397 km²). In addition, Ho Chi Minh City is characterized by a large population at

risk (1 931 000 people) and a large surface area of coastal ecosystems (373 km²),

while Hong Kong combines a high value of assets at risk (36 billion USD) and a

large surface area of coastal ecosystems (349 km²) along its likely flood pathway.

In total, 6 of the 28 world’s cities having the highest values of population and/or

assets at risk benefit from the presence of large coastal ecosystem areas. For more

information of those top 20 cities, see Supplementary Information.

Cities

Figure 4.4 Representation, for each of the 136 most populated world’s coastal cities in 2005, of the surface area (km²) occupied by coastal ecosystems (size of the circle) and of population at risk of coastal flood damages based on the study of Nicholls et al. (2007) (colours)

Chapter 4

124

Figure 4.5 Representation of the population and assets at risk as well as the coastal ecosystem area for the top 20 cities with (1) the largest population at risk, (2) the highest amount of assets at risk and (3) the largest surface area of coastal ecosystems

4.4 Conclusion

The results show that on the 136 studied cities that have the largest flood-prone

population in the world, 75 % can benefit from existing coastal ecosystems, while

35 cities have no coastal ecosystems and 73 cities have less than 100 km². The

results highlight then that several densely populated and industrialized cities, i.e.

Rotterdam, London, Shanghai or Boston, can benefit from the presence of different

coastal ecosystems and that those ecosystems could be included in sustainable and

efficient coastal protection strategies. Our analysis implies that the parameters

influencing the presence of the coastal ecosystems on the one hand and their

surface area on the other hand are a combination of historical, social, and physical

factors. Several expected relations between the presence and extent of the

different coastal ecosystems could be backed up by the logistic and linear

regressions, while further research would be needed to fully understand the

influence of the studied and other parameters in the occurrence and extent of the

four types of coastal ecosystems in front of the world’s coastal cities.

Subsequently, future coastal protection strategies should account for the presence

or absence, and extent of coastal ecosystems in front of the city. For example, a city

Cities

125

like Khulna, with 441 000 people exposed to flood risks, but located far inland

(more or less 100 km from the open sea) and with large deltaic mangrove forests

(2 006 km²) between the city and sea, can largely benefit from nature-based flood

risk mitigation. The conservation or even restoration of formerly reclaimed coastal

ecosystems for those cities located in large river delta plains, can contribute to the

mitigation of storm surge flood risks, while providing additional valuable

ecosystem services. In contrast, while in the same delta, the cities of Kolkata

(almost 2 million people at risk) and Dhaka (844 000 people at risk) do not benefit

from the coastal flood and erosion risks mitigation offered by the mangrove forest

of the Ganges-Brahmaputra delta (i.e. the Sundarbans). Or a city like New York,

with 1 540 000 flood-exposed people and an estimated 23 km² of salt marshes

along its likely storm surge flood pathway, will benefit only marginally from

nature-based flood risk mitigation, as the available buffer area between the city

and open sea is small or non-existent. Those cities must primarily rely on investing

in hard engineering structures to protect their populations, and where possible

can benefit from the creation or restoration of coastal ecosystems as add-on to

hard structures, in order to obtain sustainable and cost-effective flood protection

solutions.

4.5 Acknowledgments

The author would like to thank Chris McOwen for sharing the salt marsh dataset

and Erik Fransen for the help on statistical analysis. The data used are listed in the

method and references. The research was funded by the University of Antwerp.

Chapter 4

126

Supplementary Information

Probable flood pathway

The probable flood pathway, i.e. the path preferably followed by the storm surge

between the open sea and the city, accounting for the friction exerted by the land

area and the water bodies, is presented here after for a selection of cities.

Figure SI 4.1 Representation of the probable flood pathways as calculated in the analysis for seven cities located at the end of long and small river channels (Guangzhou, Hamburg and Shenzhen), in bays (New Orleans), or at the coast (Hong Kong, Los Angeles, San Diego).

The comparison of the coastal ecosystems surface area in the three buffer size

(Table SI 4.1) shows that for the cities that have no coastal ecosystems in the 10

km buffer (i.e. 39 cities) four cities, (i.e. Xiamen, Yantai, Rangoon and Rotterdam)

have a presence of coastal ecosystems in their 20 km buffer area, and 6 more cities

have a presence of coastal ecosystems in the 30 km buffer (i.e. Qingdao, Athens,

Kolkata, Nagoya, Tokyo and Benghazi). For the cities having coastal ecosystems in

the 10 km buffer area (i.e. 97 cities), 60 % have an increase of coastal ecosystems

surface area smaller than 60 km² in the 30 km buffer area, the other 40 % of those

cities have a higher increase in surface area. However, the difference in surface

areas of coastal ecosystems for the different cities and for the different buffer size

remains relative. As such, the comparison of the coastal ecosystems area between

the cities remains similar in the case of the use of the three buffer sizes.

Cities

127

Buffer sizes

Table SI 4.1 The different surface area of coastal ecosystems (km²) are presented for the three buffer size tested around the probable flood pathways for every city

City Country Continent

Coastal Ecosystems Surface Area (km²) for each buffer size

10 km 20 km 30 km

Algiers Algeria Africa 1.4 5.3 8.9 Luanda Angola Africa 99.8 260.7 464.5 Buenos Aires Argentina South America 0 0 0 Adelaide Australia Oceania 61.0 148.1 233.2 Brisbane Australia Oceania 44.1 82.0 268.3 Perth Australia Oceania 65.8 161.6 286.5 Melbourne Australia Oceania 0.2 22.0 44.6 Sydney Australia Oceania 12.7 22.7 26.3 Khulna Bangladesh Asia 963.2 2006.0 3005.0 Dhaka Bangladesh Asia 0 0 0 Chittagong Bangladesh Asia 1.0 1.3 1.6 Grande Vitoria Brazil South America 19.8 20.0 20.0 Santos Brazil South America 31.7 70.1 79.1 Maceio Brazil South America 9.8 16.6 22.4 Natal Brazil South America 17.9 22.9 25.4 Belem Brazil South America 9.6 36.4 49.9 Porto Alegre Brazil South America 0 0 0 Fortaleza Brazil South America 10.3 15.0 53.8 Salvador Brazil South America 0.9 8.7 37.9 Recife Brazil South America 4.5 15.2 28.7 Rio de Janeiro Brazil South America 0.04 13.3 48.7 Douala Cameroon Africa 238.9 482.6 838.4 Montreal Canada North America 0 0 0 Vancouver Canada North America 0.5 2.5 3.3 Xiamen China Asia 0 55.8 82.0 Yantai China Asia 0 3.0 3.0 Zhanjiang China Asia 60.9 208.1 521.5 Dalian China Asia 0 0 0 Qingdao China Asia 0 0 3.0 Wenzhou China Asia 92.1 102.4 123.0 Ningbo China Asia 0.0 0.3 7.8 Shenzhen China Asia 9.0 69.0 277.5 Fuzhou China Asia 96.3 99.2 110.2 Guangzhou China Asia 180.1 304.0 491.9

Chapter 4

128

City Country Continent

Coastal Ecosystems Surface Area (km²) for each buffer size

10 km 20 km 30 km

Tianjin China Asia 5.7 9.3 9.8 Hangzhou China Asia 177.1 231.9 249.8 Shanghai China Asia 116.2 396.6 682.0 Barranquilla Colombia South America 60.4 98.6 145.5 Havana Cuba North America 3.7 11.0 22.2 Copenhagen Denmark Europe 5.0 23.4 39.6 Santo Domingo Dominican Republic North America 24.4 79.5 143.4 Guayaquil Ecuador South America 377.4 741.2 866.7 Alexandria Egypt Africa 0 0 0 Helsinki Finland Europe 1.0 3.4 3.4 Marseille France Europe 7.7 12.6 22.9 Hamburg Germany Europe 4.1 6.0 12.3 Accra Ghana Africa 154.3 426.6 571.7 Athens Greece Europe 0 0 7.7 Conakry Guinea Africa 303.1 1138.2 2454.4 Port-au-Prince Haiti North America 5.2 9.9 30.7 Hong Kong Hong Kong S.A.R. Asia 69.5 348.8 838.7 Vishakhapatnam India Asia 0 0 0 Kochi India Asia 0 0 0 Chennai India Asia 0 0 0 Surat India Asia 3.0 13.7 35.5 Mumbai India Asia 8.1 47.1 96.1 Kolkata India Asia 0 0 0.2 Palembang Indonesia Asia 112.4 263.0 388.3 Ujung Pandang Indonesia Asia 10.8 33.7 94.6 Surabaya Indonesia Asia 5.4 17.6 33.9 Jakarta Indonesia Asia 0.8 3.2 6.3 Dublin Ireland Europe 2.9 4.2 5.5 Tel Aviv-Yafo Israel Asia 0.8 2.7 2.7 Naples Italy Europe 0.9 2.2 12.5 Abidjan Ivory Coast Africa 0 0 0 Fukuoka Japan Asia 0 0 0 Nagoya Japan Asia 0 0 0.7 Hiroshima Japan Asia 0.1 1.5 2.3 Sapporo Japan Asia 0 0 0 Osaka Japan Asia 0 0 0 Tokyo Japan Asia 0 0 1.0 Kuwait Kuwait Asia 186.8 621.7 885.4

Cities

129

City Country Continent

Coastal Ecosystems Surface Area (km²) for each buffer size

10 km 20 km 30 km

Beirut Lebanon Asia 0 0 0 Benghazi Libya Africa 0 0 0 Tripoli Libya Africa 0 0 6.6 Kuala Lumpur Malaysia Asia 124.5 188.3 191.7 Rabat Morocco Africa 0 0 0 Casablanca Morocco Africa 0 0 0 Maputo Mozambique Africa 0.7 13.0 25.3 Rangoon Myanmar Asia 0 0.6 0.9 Rotterdam Netherlands Europe 0 5.7 7.3 Amsterdam Netherlands Europe 0 0 0 Auckland New Zealand Oceania 4.5 17.9 24.5 Lagos Nigeria Africa 9.2 49.7 87.0 Nampo North Korea Asia 0 0 0 Karachi Pakistan Asia 10.8 13.4 28.6 Panama City Panama North America 1.2 12.6 19.6 Lima Peru South America 0.2 2.9 4.5 Davao Philippines Asia 85.2 325.2 624.7 Manila Philippines Asia 174.0 522.3 958.3 Porto Portugal Europe 0 0 0 Lisbon Portugal Europe 1.8 9.7 23.7 San Juan Puerto Rico South America 17.1 34.8 55.9 St. Petersburg Russia Europe 0 0 0 Jeddah Saudi Arabia Asia 37.8 131.1 286.2 Dakar Senegal Africa 68.1 213.7 351.6 Singapore Singapore Asia 11.2 138.3 502.3 Mogadishu Somalia Africa 4.5 10.5 17.3 Durban South Africa Africa 0 0 0 Cape Town South Africa Africa 0 0 0 Ulsan South Korea Asia 0 0 0 Incheon South Korea Asia 0 0 0 Busan South Korea Asia 0 0 0 Barcelona Spain Europe 0.4 5.5 8.0 Stockholm Sweden Europe 0.1 0.3 0.6 Taipei Taiwan Asia 24.1 25.6 26.5 Dar-es-Salaam Tanzania Africa 23.3 53.1 77.5 Bangkok Thailand Asia 18.4 32.8 50.0 Lome Togo Africa 108.8 177.7 178.3 Izmir Turkey Europe 3.1 14.8 43.9

Chapter 4

130

City Country Continent

Coastal Ecosystems Surface Area (km²) for each buffer size

10 km 20 km 30 km

Istanbul Turkey Europe 0 0 0 Odessa Ukraine Europe 0 0 0 Dubai United Arab Emirates Asia 0.5 6.0 10.4 Glasgow United Kingdom Europe 0 0 0 London United Kingdom Europe 11.2 21.0 34.1 Washington, D.C. United States of America North America 37.8 47.9 51.8 Providence United States of America North America 4.4 7.6 10.7 Virginia Beach United States of America North America 2.0 8.6 22.6 Baltimore United States of America North America 1.1 4.8 15.1 San Jose United States of America North America 141.3 183.7 189.0 Portland United States of America North America 12.4 19.9 21.1 Seattle United States of America North America 0.1 0.1 0.2 San Diego United States of America North America 1.2 4.4 11.7 New Orleans United States of America North America 121.9 241.4 546.7 Boston United States of America North America 3.6 12.4 16.2 Tampa United States of America North America 9.2 56.3 96.7 Philadelphia United States of America North America 28.4 54.4 109.8 San Francisco United States of America North America 126.4 148.2 154.6 Houston United States of America North America 5.8 6.7 24.5 Miami United States of America North America 77.0 236.9 458.6 Los Angeles United States of America North America 0.1 2.6 3.5 New York United States of America North America 13.6 23.0 36.7 Montevideo Uruguay South America 15.5 18.5 21.6 Maracaibo Venezuela South America 4.0 6.1 72.3 Haiphong Vietnam Asia 31.9 55.7 69.9 Ho Chi Minh City Vietnam Asia 180.9 372.7 486.7

Cities

131

Elbe estuary embankments

Comparison of the embankments of low-lying lands along the Elbe estuary over

the last centuries (Hansen, 2015) and of the areas below mean high tide as defined

by our study along the Elbe estuary.

Figure SI 4.2 (A) Overview of the extent and period of construction of the embanked areas along the Elbe estuary adapted from Hansen (2015) and (B) representation of the area located below mean high tide in the likely pathway of storm surge propagation towards Hamburg according to our analysis

Chapter 4

132

Simple logistic and linear regressions

Table SI 4.2 shows the significant odds ratio resulting from the simple logistic

regression, i.e. the measure of the association between the explanatory variable

(social or physical parameter) and the dependent variable (presence of coastal

ecosystem). An odd ratio (OR) greater than 1 means that for an increase of one

unit in the explanatory variable, the odds of finding the ecosystem is increasing by

the OR to the power of 1.

Table SI 4.2 Odds Ratio resulting from the logistic regression testing the influence of the explanatory variable (Table 4.1) on the odds of finding coastal habitats in front of the coastal cities for a significance of 95 % (p-value < 0.05), the ‘X’ corresponds to non-significant relations

Dependent variables

Explanatory variables

Unit of increase

Odds Ratio (p-value < 0.05)

Mangrove Salt Marsh Seagrass Coral Reef

Physical Parameters

Latitude 1 ° 0.854 1.079 X 0.918

Distance between the sea and the city 1 km

X X X 0.887

Coastline length X X X 0.996

Area below mean high tide 1 km²

X X 0.991 X

Shallow Water area (depth > -100 m) 1.002 X X X

Deep Water area (depth < -100 m) X X X 1.006

Social Parameters

GDP Per Capita of the Country 1 US$ 0.999 1.0001 1.00004 X

Short-distance population density 1

inhabitants/km²

X 0.999 X X

Intermediate-distance population density X X X X

Long-distance population density X X X X

Cities

133

The Table SI 4.3 represent the results of the simple linear regressions determining

the parameters influencing the surface area of the coastal ecosystems in front of

cities. The regression coefficients in Table SI 4.3 are significant (p-value < 0.05),

they correspond to the factor by which the surface area of the ecosystem will be

augmented for an increase of one unit of the explanatory variable, as presented in

the Table SI 4.3.

Table SI 4.3 Regression coefficients resulting from the linear regression testing the influence of the explanatory variable (Table 4.1) on the size of the coastal habitats in front of the coastal cities for a significance of 95 % (p-value < 0.05), the ‘X’ corresponds to non-significant relations

Dependent variables

Explanatory variables

Unit of increase

Regression coefficients (p-value < 0.05)

Mangrove Salt Marsh Seagrass Coral Reef

Physical Parameters

Latitude 1 ° X 0.941 0.903 X

Distance between the sea and the city 1 km

1.040 1.018 X X

Coastline length 1.003 1.004 X X

Area below mean high tide

1 km²

X X X 0.987

Shallow Water area (depth > -100 m) X X 1.003 X

Deep Water area (depth < -100 m) X X X X

Social Parameters

GDP Per Capita of the Country 1 US$ X 0.999 0.999 X

Short-distance population density 1

inhabitants/km²

X X 1.0003 0.999

Intermediate-distance population density X 1.001 X 0.999

Long-distance population density X 1.002 X X

Chapter 4

Top 20 cities

Table SI 4.4 Top 20 world port cities ranked by population exposure to flood and wind damages and the corresponding surface and type of natural coastal ecosystems along their probable flood pathway

Rank City Country Population

2005 (000)

Population at risk (000)

Assets at risk

(US$bil)

Coastal ecosystem

(km²)

Coastal Ecosystem (type)

1 Mumbai India 18,196 2,787 46 47.1 Mangrove forest, Coral reef 2 Guangzhou China 8,425 2,718 84 304.0 Mangrove forest, Salt marsh 3 Shanghai China 14,503 2,353 73 396.5 Salt marsh

4 Miami USA 5,434 2,003 416 236.9 Mangrove forest, Salt marsh, Seagrass meadow, Coral reef

5 Ho Chi Minh City Vietnam 5,065 1,931 27 372.7 Mangrove forest 6 Kolkata India 14,277 1,929 32 0 7 New York USA 18,718 1,540 320 23.0 Salt marsh 8 Osaka Japan 11,268 1,373 216 0 9 Alexandria Egypt 3,770 1,330 28 0

10 New Orleans USA 1,010 1,124 234 241.4 Salt marsh 11 Tokyo Japan 35,197 1,110 174 0 12 Tianjin China 7,040 956 30 9.3 Salt marsh 13 Bangkok Thailand 6,593 907 39 32.8 Mangrove forest 14 Dhaka Bangladesh 12,430 844 8 0 15 Amsterdam Netherlands 1,147 839 128 0 16 Haiphong Vietnam 1,873 794 11 55.7 Mangrove forest 17 Rotterdam Netherlands 1,101 752 115 5.7 Salt marsh 18 Shenzhen China 7,233 701 22 69.0 Mangrove forest, Salt marsh 19 Nagoya Japan 3,179 696 109 0 20 Abidjan Ivory Coast 3,577 519 4 0

Cities

Table SI 4.5 Top 20 world cities ranked by assets exposure to flood and wind damages and the corresponding surface and type of natural coastal ecosystems along their probable flood pathway

Rank City Country Population

2005 (000)

Population at risk (000)

Assets at risk

(US$bil)

Coastal ecosystem

(km²)

Coastal Ecosystem (type)

1 Miami USA 5,434 2,003 416 236.9 Mangrove forest, Salt marsh, Seagrass meadow, Coral reef

2 New York USA 18,718 1,540 320 23.0 Salt marsh

3 New Orleans USA 1,010 1,124 234 241.4 Salt marsh

4 Osaka Japan 11,268 1,373 216 0

5 Tokyo Japan 35,197 1,110 174 0

6 Amsterdam Netherlands 1,147 839 128 0

7 Rotterdam Netherlands 1,101 752 115 5.7 Salt marsh

8 Nagoya Japan 3,179 696 109 0

9 Tampa USA 2,252 415 86 56.3 Mangrove forest, Salt marsh, Seagrass meadow

10 Virginia Beach USA 1,460 407 85 8.6 Salt marsh, Seagrass meadow

11 Guangzhou China 8,425 2,718 84 304.0 Mangrove forest, Salt marsh

12 Boston USA 4,361 370 77 12.4 Salt marsh, Seagrass meadow

13 Shanghai China 14,503 2,353 73 396.6 Salt marsh

14 London UK 8,505 397 60 21.0 Salt marsh

15 Vancouver Canada 2,188 320 55 2.5 Salt marsh

16 Fukuoka Japan 2,800 307 48 0

17 Mumbai India 18,196 2,787 46 47.1 Mangrove forest, Coral reef

18 Hamburg Germany 1,740 261 39 6.0 Salt marsh

19 Bangkok Thailand 6,593 907 39 32.8 Mangrove forest

20 Hong Kong Hong Kong SAR 7,041 223 36 348.8 Mangrove forest, Seagrass meadow

Chapter 4

Table SI 4.6 Top 20 world cities ranked by their natural coastal ecosystem surface along their probable flood pathway and the corresponding population and assets exposed to flood and wind damages

Rank City Country Population

2005 (000)

Population at risk (000)

Assets at risk

(US$bil)

Coastal ecosystem

(km²)

Coastal Ecosystem (types)

1 Khulna Bangladesh 1,494 441 4.41 2006.0 Mangrove Forest 2 Conakry Guinea 1,425 41 0.40 1138.2 Mangrove Forest, Seagrass Meadow 3 Guayaquil Ecuador 2,387 412 8.86 741.2 Mangrove Forest 4 Kuwait Kuwait 1,810 14 1.16 621.7 Seagrass Meadow, Coral Reef 5 Manila Philippines 10,686 113 2.69 522.3 Mangrove Forest, Seagrass Meadow 6 Douala Cameroon 1,761 11 0.13 482.6 Mangrove Forest 7 Accra Ghana 1,981 14 0.18 426.6 Mangrove Forest, Seagrass Meadow 8 Shanghai China 14,503 2,353 72.86 396.6 Salt Marsh 9 Ho Chi Minh City Vietnam 5,065 1,931 26.86 372.7 Mangrove Forest

10 Hong Kong Hong Kong SAR 7,041 223 35.94 348.8 Mangrove Forest, Seagrass Meadow

11 Davao Philippines 1,327 3 0.06 325.2 Mangrove Forest, Seagrass Meadow 12 Guangzhou China 8,425 2,718 84.17 304.0 Mangrove Forest, Salt Marsh 13 Palembang Indonesia 1,733 127 2.50 263.0 Mangrove Forest 14 Luanda Angola 2,766 1 0.02 260.7 Mangrove Forest, Seagrass Meadow 15 New Orleans USA 1,010 1,124 233.69 241.4 Salt Marsh

16 Miami USA 5,434 2,003 416.29 236.9 Mangrove Forest, Salt Marsh, Seagrass Meadow, Coral Reef

17 Hangzhou China 2,831 17 0.53 231.9 Salt Marsh 18 Dakar Senegal 2,159 18 0.17 213.7 Seagrass Meadow

19 Zhanjiang China 1,514 230 7.13 208.1 Mangrove Forest, Salt Marsh, Seagrass Meadow

20 Kuala Lumpur Malaysia 1,405 270 15.06 188.3 Mangrove Forest

Wetlands creation

137

CHAPTER 5 A global exploration of tidal wetland creation for nature-based flood risk mitigation in coastal cities

Rebecca Van Coppenolle and Stijn Temmerman

Based on the paper submitted to Estuaries, Coastal and Shelf Science in October 2018

Chapter 5

138

Abstract

Coastal cities around the world are increasingly exposed to flood risks due to

climate change, resulting sea level rise and more intense storm surges as well as

due to growing coastal population densities. Nature-based risk mitigation,

consisting of conservation or creation of coastal ecosystems that have the natural

capacity to adapt to sea level rise and to mitigate storm surges, is increasingly

proposed, but real-live implementation is yet limited to specific local cases. Our

study presents a global scale analysis of the surface areas available for potential

creation or restoration of tidal wetlands (salt marshes and mangrove forests) in

front of 135 highly populated, flood-exposed coastal cities, as part of nature-based

or hybrid strategies to buffer against coastal flood risks. Our results reveal that 34

% (4 624 km²) of the total land area within the influence zone of storm surge

propagation between the sea and the cities is potentially available for tidal

wetlands creation. Those areas mainly correspond to rural areas with a low

population density such as croplands, paddy fields or vegetated areas and to water

bodies. The key factors influencing the area potentially available for tidal wetlands

creation are the size of the low-lying coastal zone and the population density in the

close vicinity of the city, as 60 % (8 332 km²) of the land area below mean high

tide in front of the studied cities is urbanized or densely populated. Cities located

along deltas or estuaries and in bays and lagoons (e.g. Hamburg, Guayaquil,

Tianjin, Portland or San Jose) have generally larger low-lying coastal zones and

consequently larger potentially available areas for salt marshes and mangrove

forests restoration or creation for coastal flood risks mitigation. Our results

contribute to increasing evidence and awareness of the possibilities of nature-

based mitigation of coastal flood risks by restoring and creating tidal wetlands in

front of flood-exposed coastal cities around the world.

Wetlands creation

139

5.1 Introduction

Global climate change and the related intensification of coastal hazards is

threatening the coastal zone (Hallegatte et al., 2013; Neumann et al., 2015; de

Sherbinin et al., 2007; Vitousek et al., 2017). This increasing risk of coastal hazard

is resulting, among others, from sea level rise affecting the coasts by higher flood

and erosion risk due to the action of wind and storm waves (Gedan et al., 2011;

Storlazzi et al., 2011; Thampanya et al., 2006), but also from an increase in

frequency of tropical cyclones and extra-tropical storms of high intensity,

generating destructive storm surges when they reach the coastal area (Webster et

al., 2005; Woodruff et al., 2013).

In addition to the increasing threats due to climate change, the coastal populations

are facing socio-economic changes (Barbier, 2014; Hanson et al., 2011; Kron,

2013; Syvitski et al., 2009). In the Low Elevation Coastal Zone (LECZ, i.e. lower

than 10 m above mean sea level), the population is expected to increase and reach

by 2060 a global average density of 400 to 500 inhabitants per square kilometre

(Hanson et al., 2011; Mcgranahan et al., 2006; Neumann et al., 2015), or two times

the current global population density in the LECZ. This augmentation of the

population pressure in the coastal zone also implies an increase of the assets

exposed to coastal hazards (Barbier, 2014; Hallegatte et al., 2013; Hanson et al.,

2011). Furthermore, the human influence on its natural environment, such as a

reduced sediment supply to coastal zones by the trapping of sediments in

upstream river dams (Auerbach et al., 2015; Syvitski, 2005) or the extraction of oil,

gas or water from the substrate beneath coastal zones, is leading to the

intensification of coastal land subsidence that further contributes to the increasing

vulnerability of the coasts to flood and erosion risks (Balke & Friess, 2016; Kirwan

& Megonigal, 2013; Syvitski, 2008).

The need to develop strategic plans to mitigate these risks over the short and long

term is increasing. In this respect, there is an increasing interest for so-called

nature-based risk mitigation, i.e. the conservation, restoration and creation of

natural habitats as a contribution to coastal protection against flood and erosion

risks (Cheong et al., 2013; Costanza et al., 2008; Gedan et al., 2011; Sutton-Grier et

al., 2015; Temmerman et al., 2013). Tidal wetlands, including salt marshes in

temperate climates and mangroves in (sub-)tropical areas, are widely considered

as coastal habitats that provide a mitigating effect on flood and erosion risks in

more landward located, human-occupied coastal plains (Barbier et al., 2011;

Cheong et al., 2013; Duarte et al., 2013; Gedan et al., 2011; Temmerman et al.,

2013). These ecosystems have a certain capacity to build up elevation with sea

Chapter 5

140

level rise through sediment accretion (Kirwan et al., 2016; Kirwan & Megonigal,

2013; Krauss et al., 2014; McIvor et al., 2013; Temmerman & Kirwan, 2015), to

reduce wind waves and shoreline erosion, and to attenuate the landward

propagation of storm surges due to friction provided by the wetland’s vegetation

and topography (Barbier et al., 2008, 2011; Gedan et al., 2011; Guannel et al.,

2016; Narayan et al., 2016; Temmerman et al., 2013; van Wesenbeeck et al., 2017).

However, the observed decline in the world’s tidal wetlands over the recent

decades and the projections for the next decades are worrying (Duke et al., 2007;

IPCC, 2007; McLeod et al., 2011; Pendleton et al., 2012). The disappearance of tidal

wetlands is predominately an effect of the historical and present anthropogenic

pressures on coastal areas, by conversion of mangroves and salt marshes into

agriculture, aquaculture, urban and industrial areas. As such the loss of mangrove

forests and salt marshes over the last century was estimated at 20 to 50 % of their

total area (Food and Agriculture Organization (FAO) of the United Nations, 2007;

McLeod et al., 2011; Spalding et al., 1997; Valiela et al., 2001). Furthermore, the

degradation of the remaining mangrove forests and salt marshes through over-

exploitation for timber, over-fishing, pollution or solid waste disposal is reducing

their valuable ecosystem services, including their natural capacities to act as a

barrier against wind and storm waves (Food and Agriculture Organization (FAO)

of the United Nations, 2007; Scott et al., 2014; Spalding et al., 2010). Projections

for the next 100 years are estimating the future losses at 30 - 40 % of the actual

salt marshes and mangrove forest areas through continued land reclamation,

wetlands degradation and relative sea level rise (Blankespoor et al., 2014; Duke et

al., 2007; IPCC, 2007; Ma et al., 2014; Pendleton et al., 2012; Schuerch et al., 2018;

Valiela et al., 2001). Important to notice in this respect, is that recent regional to

global scale studies indicate that tidal wetlands have a high capacity to maintain

surface area with future sea level rise, while the most important threats come

from direct anthropogenic pressures, such as wetland conversion into human land

use and human infrastructure that prevents inland wetland migration with sea

level rise (Kirwan et al., 2016; Schuerch et al., 2018).

In this context of globally decreasing tidal wetland areas, the restoration or

creation of tidal wetlands in combination with coastal engineering solutions like

dikes or dams, is considered essential to allow populated coastal areas to mitigate

and adapt to the increasing risks of coastal flooding and erosion (Barbier et al.,

2008; Ma et al., 2014; Scott et al., 2014; Spalding et al., 2010, 2013; Valiela et al.,

2009). Plans of such hybrid coastal protection schemes which combine tidal

wetlands creation and hard engineering are already applied in various coastal

areas over the world. The restoration of the Mississippi delta through the

Wetlands creation

141

Louisiana Coastal Master Plan is one of the largest known examples of such

combination of coastal marsh restoration or creation, and engineering of levees for

coastal protection, with the expected restoration or maintenance of more than 2

000 km² of marshlands over the next 50 years (Boesch et al., 2006; Coastal

Wetlands Planning Protection and Restoration Act (CWPPRA), n.d.; Day et al.,

2007; RESTORE, 2017). Similarly, the San Francisco Bay Joint Venture works on

the protection, restoration and enhancement of, among others, about 550 km² of

tidal flats, marshes and lagoons and some 260 km² of seasonal wetland areas over

the San Francisco Bay with the combined objective to gain benefits for wildlife and

coastal protection (San Francisco Bay Joint Venture, 2018). At smaller scales,

countries in Europe are applying managed realignment of their engineered flood

defences (dikes), through landward relocation of dikes enabling the creation of

tidal marshes on formerly embanked land. For example, in England and Wales,

numerous projects are leading to the total realignment of about 660 km of

coastline by 2030 with the aim to create 62 km² of intertidal areas (Esteves, 2014;

Pendle, 2013); in Belgium, by 2030 about 40 km² of flood control areas are being

realized on formerly embanked land (Meire et al., 2014; SigmaPlan, 2017). After

the devastating 2004 tsunami in South East Asia and typhoon Haiyan in the

Philippines in 2013, field observations indicated smaller damages in villages

sheltered behind mangrove forests (Balke & Friess, 2016; Dahdouh-Guebas et al.,

2005; Danielsen et al., 2005). In response to these flood disasters, associations and

countries (i.e. Indonesia, India, Sri Lanka, Thailand and Malaysia) together

instigated mangroves restoration projects (FAO, 2007; Schmitt, 2012). This

resulted in, for example, the restoration of 20 km² of mangrove forest in Indonesia

and the plantation of 310 000 seedlings over the Sri Lanka’s coasts, of which about

60 % survived (Schmitt, 2012). In Pakistan, about 80 km² of mangrove forests are

restored along the coasts, while development of restorations strategies in the

Indus delta in collaboration with the IUCN (International Union for Conservation

of Nature) is scoping to restore 100 km² of mangrove forest in the delta (Marois &

Mitsch, 2015; MFF Pakistan, 2016; Schmitt, 2012; Spalding, McIvor, et al., 2014).

These examples of medium to large scale tidal wetland creation programs for

mitigation of coastal flood and erosion risks exemplify the potential of nature-

based mitigation programs on local to regional scales. A global scale analysis of the

potential of tidal wetland creation for coastal risk mitigation is lacking so far, but

could contribute to promote the more widespread implementation of nature-

based flood risk mitigation programs into policy on coastal zone management at

several places around the world. Our study aims to provide a global estimation of

the land surface areas where tidal wetlands, i.e. salt marshes and mangrove

forests, could be restored or created in front of the world’s most flood-exposed

Chapter 5

142

coastal cities, and to explore which factors influence the geographical variation in

the area available for tidal wetlands creation.

5.2 Method

The coastal cities considered in the analysis correspond to 135 cities studied by

Nicholls et al. (2008) that have a population of more than 1 million people in 2005

(UN, 2005) and are exposed to coastal flood damages generated by storm surges

and high winds without any consideration of coastal defences or adaptations. The

city of Helsinki in Finland could not be included due to data availability.

5.2.1 Data

The bathymetry is coming from the General Bathymetric Chart of the Oceans

(GEBCO) (British Oceanographic Data Center, 2017) and represents a gridded

bathymetry of the oceans at a 30 arc second resolution combined with the land

topography defined by the NASA Shuttle Radar Topography Mission (SRTM).

The NASA Shuttle Radar Topography Mission (SRTM) Global 3 arc second

V003 dataset (NASA JPL., 2013) is used as digital elevation model (DEM), as it is

the best known DEM available at global scale (Rodriguez et al., 2006; Sun et al.,

2003).

Information on the tidal amplitude in front of every city is derived from the

Finite Element Solution (2012) – Global Tide from AVISO. The Principal Lunar

semi-diurnal component (M2) was used to define the averaged tidal amplitude in

front of every city.

The location of the urban areas was determined from the Global Land Cover

by National Mapping Organizations dataset that classifies the status of the world’s

land cover into 20 categories (see below) based on the Land Cover Classification

System (LCCS) developed by the Food and Agriculture Organization of the United

Nations (Tateishi et al., 2014).

The population distribution originates from the LandScan 2013 Global

Population Database (Bright et al., 2013). It represents the population over a 30

arc second grid resolution and integrates the diurnal movements and collective

travelling behaviour of the population, i.e. the so-called “ambient population”,

averaged over 24 hours (Bright et al., 2013; Dobson et al., 2000). The dataset was

adapted to deliver values of population density following the guidelines of the

LandScan documentation (Bright et al., 2013; UT BATTELLE LLC., n.d.).

Wetlands creation

143

The location of the tidal wetlands, as salt marsh and mangrove forest, was

determined based on the Global distribution of Mangroves (Giri et al., 2011) and

the Global distribution of Saltmarshes (Mcowen et al., 2017) from the United Nation

Environmental Program – World Conservation Monitoring Centre (www.unep-

wcmc.org).

The study aims to identify areas that meet a selection of conditions for mangrove

and salt marsh development within the likely area of storm surge propagation in

front of the world’s most populated coastal cities, in order to contribute to the

nature-based mitigation of coastal flood risks. The likely area of storm surge

propagation in front of the cities is delineated using the procedure presented in

Chapter 4. The four conditions retained to identify the suitable locations for tidal

wetland creation are (1) elevation below mean high tide, (2) absence of existing

tidal wetlands, (3) location outside the urbanized area and (4) a population

density lower than 50 inhabitants per square kilometre. The logical steps are

presented in Figure 5.1.

The value of 50 inhabitants per square kilometre is an arbitrary value based on the

literature (Mcgranahan et al., 2006; Neumann et al., 2015). Scenarios with

different population threshold values (20, 35 and 50 inhabitants/km²) were

explored and are presented in Supplementary Information. The comparison of the

results based on the three population density thresholds is showing a general

increase of the average surface area available for tidal wetlands development of

8.99 ± 34.04 km² between the lowest and highest thresholds, while keeping the

relative difference between the cities very similar. It is important to note that the

considered threshold remains high and represents a theoretical exploration of the

potential to restore tidal wetlands. In practice, the displacement of a small number

of inhabitants may already generate so much societal-political resistance that tidal

wetland creation or restoration may be impossible to realize.

Figure 5.1 Representation of the logical steps for the selection of the pixels suitable for tidal wetlands restoration or creation

Chapter 5

144

The areas available for tidal wetlands creation were compared to the current land

cover based on the Global Land Cover by National Mapping Organization dataset.

For each city, the different land covers were defined based on five categories,

namely (1) the vegetated areas grouping the land cover classes of broadleaf and

needle leaf evergreen and deciduous forest, mixed forests, tree open areas, shrub,

herbaceous, herbaceous with sparse trees or shrub, sparse vegetation and

wetlands areas; (3) the cropland and paddy fields areas corresponding to the

cropland, cropland with other vegetation mosaic and paddy fields areas; (4) the

bare land corresponding to the bare consolidated or unconsolidated land, and (5)

the water areas corresponding to the water bodies (i.e. pounds or lakes and areas

of coastal water considered inland following the delineation of the country’s

limits).

Due to the global scale of the different datasets, the vegetated areas include areas

defined as wetlands and mangroves in the Global Land Cover dataset. This is

because the use of different global-scale datasets implies limitations in local data

accuracy or artefacts and the overlap of features. In this analysis, the extent of salt

marshes and mangrove forests is based on the Global distribution of saltmarshes

(Mcowen et al., 2017) and the Global distribution of Mangroves (Giri et al., 2011)

that are considered as more accurate than the Global Land Cover dataset.

A linear regression was performed to test the influence of social and physical

parameters (Table 5.1) (explanatory variables) on the geographical variations of

the size of the suitable area for tidal wetlands restoration or creation (dependent

variable). The parameters were extracted through ArcGIS (10.3.1), while the value

of the country’s GDP per capita was recovered from the World Bank data base

(http://databank.worldbank.org/data/home.aspx). More information on the

method can be found in Chapter 4. The regression was performed on the logarithm

value of surface area available for tidal wetlands creation augmented by one to

account for both the normality of the residuals and for the cities having no

available area for tidal wetlands creation. The results obtained from the linear

regression are the regression coefficients (slope), their standard error and the

associated p-value.

Wetlands creation

145

Table 5.1 Physical and Social parameters extracted from the data for each city, for the method, see Chapter 4

Physical Parameters Social Parameters Type Unit Type unit

Latitude of the city Degree Short-distance population density (within 20 km around the flood pathway)

Inhabitants/km²

Distance between the sea and city

km Intermediate-distance population density (within 50 km around the flood pathway)

Inhabitants/km²

Coastline length km Long-distance population density (within 100 km around the flood pathway)

Inhabitants/km²

Area below mean high tide km² Country’s GDP per capita Constant 2005 US$

Shallow water area (> -100 m)

km²

Deep water area (< -100 m)

km²

5.3 Results

5.3.1 Area Below Mean High Tide for Tidal Wetlands Restoration

The zone that sits below mean high tide was defined for every studied city and

divided into three categories, (1) the existing tidal wetlands, (2) the area

potentially available for tidal wetland restoration or creation (i.e. non-urban area

with less than 50 inhabitants/km²) and (3) the area not available for tidal wetland

creation (i.e. urban areas or with more than 50 inhabitants/km²) (Figure 5.2).

The cities with the largest areas below mean high tide are Hamburg in Germany (1

396 km²), Guangzhou in China (1 128 km²), Ho Chi Minh City in Vietnam (860

km²), Rotterdam in The Netherlands (784 km²) and Guayaquil in Ecuador (562

km²). At continental scale, the largest zones below mean high tide are found in

European cities (174 ± 355 km²; i.e. average ± standard deviation) where the

existing tidal wetlands are scarce (maximum of 12 km² in front of London in the

UK). Asia and North America also present large areas below mean high tide with

averages and standard deviations of 128 ± 201 km² and 80 ± 96 km² respectively.

In general the areas located below mean high tide are smaller in South America

(67 ± 130 km²), Africa (44 ± 106 km²) and Oceania (21 ± 11 km²).

The cities that have the highest potentially available area for tidal wetlands

creation are Hamburg in Germany (881 km²), Guayaquil in Ecuador (399 km²) and

Tianjin in China (233 km²), while eleven cities have no available space for tidal

wetlands creation (Figure 5.3). Over the 135 studied cities, the available area for

tidal wetlands is 34 ± 90 km² (average ± standard deviation), with large variations

Chapter 5

146

between the continents (Figure 5.4). Oceania and Africa present the lowest

available surface area for tidal wetlands development with 6 ± 3 km² and 12 ± 22

km² respectively, followed by Asia with 28 ± 41 km². Values for North and South

America are slightly higher with 40 ± 56 km² and 41 ± 94 km², respectively, while

Europe has the largest averaged surface area available for tidal wetlands

development with 73 ± 200 km².

Wetlands creation

Figure 5.2 Surface area located below mean high tide in front of every city (size of the symbol), categorized to surface areas occupied by (1) the existing tidal wetlands, (2) the area potentially available for tidal wetlands creation (i.e. non-urban areas with less than 50 inhabitants/km²) and (3) the area not available for tidal wetlands creation (i.e. urban areas or with more than 50 inhabitants / km²) (colour of the symbol)

Chapter 5

Figure 5.3 Surface area potentially available for tidal wetlands creation (size of symbols) and the current land use in those areas (colours of symbols)

Wetlands creation

149

Figure 5.4 Comparison of the available areas for tidal wetlands creation (km²) in front of the 135 studied cities, categorized per continent. The width of the boxes corresponds to the square root of the number of cities per continent. The city of Hamburg in Europe is not represented due to its high value of 888.8 km².

When looking at the current land use types within the areas that are identified as

potentially suitable for tidal wetlands restoration or creation, the most dominant

land use types are cropland and paddy fields, mostly in European, Asian, South and

North American cities (Figure 5.5), in combination with vegetated areas (1 672

km² and 1 621 km² respectively, or 36 % and 35 % of the total potentially

available area). The cities having the largest cropland and paddy fields areas

below mean high tide are Hamburg in Germany (624 km² of cropland, 71 % of the

potentially available surface area for tidal wetlands restoration or creation),

Guayaquil in Ecuador (70 km² of croplands and 41 km² of paddy field, 28 %),

Tianjin in China (67 km² of cropland, 30 %) or Rotterdam in The Netherlands (66

km² of cropland, 52 %) (see Supplementary Information). Croplands are mainly

found in North America and Europe, while paddy fields are mainly found in Asia,

and occasionally on large surfaces in other continents such as in Guayaquil

(Ecuador). For most of the cities, part of the potentially available area for tidal

wetlands restoration or creation is currently defined as water bodies by the Land

Cover Dataset; the largest areas are found in Asian and North American cities

(Figure 5.5), as in Tianjin (China) with 126 km² and San Jose (USA) with 110 km².

Bare land represents a really small share of the potentially available area for tidal

wetlands restoration or creation for all the continents, with less than 1 km² on

average (Figure 5.5).

Chapter 5

150

Figure 5.5 Comparison of the land use types in the potentially available areas for tidal wetlands creation in front of the 135 studied cities, categorized per continent. The width of the boxes corresponds to the square root of the number of cities per continent. Note that Y-axes have different scales.

5.3.2 Social and Physical Parameters Influencing the Potentially Available

Area for Tidal Wetlands Creation

A multiple linear regression was performed on several social and physical factors

susceptible to influence the potentially available area for tidal wetlands creation.

The significant regression coefficients (p-value < 0.05) are presented in Table 5.2.

Wetlands creation

151

Table 5.2 Regression coefficients resulting from the mutliple linear regression testing the influence of the social and physical factors (Table 5.1) on the geographical variation of the size of the tidal wetlands in front of the coastal cities for a significance of 95 % (p-value < 0.05). The ‘X’ corresponds to a non-significant regression coefficient.

Dependent variable

Explanatory variables

Unit of increase

Regression coefficients (p-value < 0.05)

Physical Parameters

Latitude 1 ° X

Distance between the open sea and the city 1 km

X

Coastline Length 1.002

Area Below Mean High Tide

1 km²

1.004

Shallow Water Area (depth > -100 m) X

Deep Water Area (depth < -100 m) X

Social Parameters

GDP Per Capita of the Country 1 US$ X

Short-distance population density 1 inhabitants

/km²

0.999

Intermediate-distance population density X

Long-distance population density X

The results can be interpreted as follows. One unit of increase (see Table 5.2) of

the explanatory variable is generating a multiplication of the surface area

potentially available for tidal wetlands creation by a factor corresponding to the

regression coefficient, assuming all the other variables remain constant. The

available area for tidal wetlands creation is positively linked to two variables, as

the coastline length and the area below mean high tide, while it is negatively

linked to the population density in the close vicinity of the city. It is important to

note that the hypothesis that all variables remain constant is highly improbable in

nature, as such the results must be interpreted carefully, as they mostly highlight

the ‘pure’ effect of each explanatory variable on the potentially available are for

tidal wetlands restoration or creation.

5.4 Discussion

Nature-based mitigation of coastal flood risks, by conserving, restoring or creating

coastal ecosystems that are known to attenuate the impacts of sea level rise, storm

surges, wind waves and shoreline erosion, is increasingly proposed as a

sustainable, cost-efficient strategy to mitigate and adapt to increasing coastal flood

risks (R. L. Morris et al., 2018; Narayan et al., 2016; Vuik et al., 2016; van

Wesenbeeck et al., 2014). Although this concept is increasingly adopted in

scientific literature, so far there are no global-scale studies that explored the

Chapter 5

152

potential of tidal wetland restoration or creation in front of flood-exposed coastal

cities, while such an analysis may contribute to increase global interest in the

implementation of nature-based risk mitigation programs into policy on coastal

zone management. We presented here a methodology for such a global-scale

analysis, which identified the potentially suitable and available areas for tidal

wetlands creation within the likely area of storm surge propagation from the sea

towards the 135 studied cities. Large variations in this available area between the

cities mainly reflect the differences in geomorphological setting, population

settlement and land use history within the likely area of storm surge propagation.

Note that the areas identified as potentially available for tidal wetlands restoration

or creation are theoretical areas based on parameters that do not include the

socio-economic limitations of the implementation of nature-based strategies. In

practice, socio-economic and political-governance factors highly influence the

possibility and the success of tidal wetland restoration or creation (Darwiche-

Criado et al., 2017; Hartman & Cleveland, 2018; Perillo et al., 2009). Although, such

socio-economic and political-governance factors are relatively poorly studied, the

public support and acceptance of tidal wetland restoration or creation is at least as

important as the financial and ecological considerations (Hartman & Cleveland,

2018; Perillo et al., 2009; Suman, 2019). In countries where the government’s

regulation capacities are weak for example, the conservation and restoration of

natural areas is often highly difficult; the natural areas are freely accessible by the

public with little monitoring of the different activities taking place. In general, the

local communities are willing to support the development of such nature-based

strategies, but not at their expense (e.g. livelihood reduction, land loss...), which

can seriously hamper the possibilities of tidal wetland restoration or creation

(Perillo et al., 2009; Suman, 2019).

In the studied factors, the geomorphic setting is a first factor that may explain the

differences in potential areas for tidal wetland creation. As shown by the logistic

regression analysis, the potentially available area for tidal wetlands creation is

increasing if the area below mean high tide increases, all other things remaining

constant, or with an increasing coastline length, all other things remaining

constant. This relates to the fact that cities located along deltaic or estuarine

channels or adjacent to bays and lagoons (i.e. having longer coastlines within their

zone of likely storm surge propagation), have greater low-lying zones and

subsequently larger potentially suitable areas for salt marshes and mangrove

forests development (Grobicki et al., 2016; Mcowen et al., 2017; Pennings &

Bertness, 2000; Scott et al., 2014; Spalding et al., 1997; Wolanski & Elliott, 2015).

This can be illustrated by the city of Hamburg (Germany) located adjacent to the

Wetlands creation

153

Elbe estuary at 110 km from the estuary mouth, for which our analysis identified

that the zone influencing the propagation of a storm surge includes a record area

below mean high tide of 1 396 km² and a potentially available area for tidal

wetlands development of 881 km² (see Supplementary Information Figure SI 5.1).

Another, tropical example, is Guayaquil (Ecuador), located alongside the main

river channel within the large Guayas river delta at 60 km from the open sea, for

which our analysis indicates an area below mean high tide of 562 km² and an area

potentially suitable for tidal wetlands development of 399 km².

Besides geomorphology, the potentially available area for tidal wetlands

restoration or creation is influenced by the population density in the short-

distance environment around the city (Table 5.2). It highlights the fact that the

population settlement in the low-lying zone can hamper the creation of new tidal

wetlands. This can be illustrated by Guangzhou in China, for instance, where, on

the 1 128 km² located below mean high tide (a huge area due to the location of

Guangzhou in the large Pearl river delta), only 8 km² are at present occupied by

existing tidal wetlands and 97 km² could be potentially available for new tidal

wetlands creation, while 91 % of the area is occupied by densely populated urban

zones (Figure SI 5.2). Indeed the city of Guangzhou is part of a huge agglomeration

occupying most of the delta. The situation is similar for a number of cities as

Nagoya, Kolkata or Shanghai (Figure SI 5.3), located especially in Asia (China,

Japan, India), where several of such large deltaic agglomerations developed with

little open space left for tidal wetlands restoration or creation (Figure 5.3).

The geomorphic and population factors together suggest that in the future, the

reduction of area between the open sea and the city, by marine transgression

through sea level rise and shoreline erosion, combined with an increasing

population in the coastal zone, i.e. coastal squeeze, may limit the potential

development of new tidal wetlands (Pontee, 2013; Rupp-Armstrong & Nicholls,

2007).

Land use history also plays a role. Over the world, the human influence on the

coastal areas and particularly around coastal cities led to the degradation,

destruction and conversion of hundreds of square kilometres of tidal wetlands,

leaving cities with few or no remaining tidal wetlands (Airoldi & Beck, 2007;

Alongi, 2008; Duke et al., 2007; He & Zhang, 2001). This is the case for many

European cities in an estuarine or deltaic setting, which present the largest areas

of croplands located below mean high tide level (Figure 5.3 & Figure 5.5).

Examples include Hamburg in Germany, London in the UK, or Rotterdam and

Amsterdam in The Netherlands, where the embankment and drainage of coastal,

Chapter 5

154

estuarine and deltaic wetlands into so-called “polders”, mostly for agriculture

purposes, dates back to the Middle Ages and even earlier (Airoldi & Beck, 2007;

Hansen, 2015; Hatvany, 2003; Hoeksema, 2007; Reise, 2005). As such, for instance

almost 2 500 km² of salt marshes were reclaimed along the Elbe estuary between

Hamburg and the sea (Figure SI 5.1) (de Haas et al., 2018; Hamburg Port

Authority, 2006; Hansen, 2015; Pierik et al., 2017; Reise, 2005; Vos, 2015).

Identified as a currently large “polder” (881 km²), especially consisting of cropland

(624 km²), our analysis identified this area as potentially available for tidal

wetland creation. In China, from the 1960s, the embankment of mangrove areas

into rice fields, aquaculture ponds or areas for industrial and urban development,

resulted in the loss of nearly 60 % of the Chinese mangroves; at a local scale, a city

like Hong Kong lost 85 % of its original mangrove forests (Li & Lee, 1997; Meng et

al., 2017). In other (sub-)tropical areas, historical land use changes from

mangroves into human land use, often aquaculture ponds, is widespread (Chen et

al., 2017; Deb & Ferreira, 2015; Meng et al., 2017; Scott et al., 2014; Valiela et al.,

2001; Zhu et al., 2016). In Guayaquil (Ecuador), for instance, the potentially

available area for tidal wetlands restoration coincides with present-day

aquaculture ponds that were created over the past decades in former mangrove

areas in the Guayas river delta (Delgado, 2013; Parés-Ramos et al., 2013).

Over the last decades, restoration of salt marshes and mangrove forests is

observed at several places around the world, with among others the restoration of

marshland in the Mississippi deltaic plain (Coastal Wetlands Planning Protection

and Restoration Act (CWPPRA), n.d.; Day et al., 2007), the restoration of tidal

marshes in the Rhine-Meuse-Scheldt delta (Eertman et al., 2002; Oosterlee et al.,

2018; Ysebaert et al., 2016), the projects of tidal wetlands restoration in the

Yellow river and along the Chinese coasts (An et al., 2007; Cui et al., 2009; Jiang et

al., 2015), or the reforestation of mangrove forest in front of Ho Chi Minh City in

Vietnam that was destroyed during the war by chemical spraying (Hong, 2001;

Marchand, 2008). Between 1978 and 2000, the efforts of reforestation in the Can

Gio region (Ho Chi Minh City, Vietnam) resulted in the restoration of around 200

km² of healthy and diverse mangrove forest (Hong, 2001).

The effectiveness of the creation or restoration of tidal wetlands strongly depends

on the current land use both in terms of the success of the tidal wetlands

development and of the future increase in coastal protection (Lewis & Brown,

2014; Q. Zhao et al., 2016). Therefore, locations where we identified possible areas

for tidal wetland creation may necessitate different measures and may experience

different rates of success, depending on their present land use type (Figure 5.3).

Firstly, the restoration or creation of tidal wetlands necessitates the presence of

Wetlands creation

155

several hydro-geomorphic conditions. For example, establishment of tidal wetland

vegetation may be limited when tidal flooding is too excessive and soil drainage

during ebb is poor. Consequently tidal wetland vegetation is usually only able to

grow in the upper portion of the intertidal zone, where soil and topographic

conditions allow good drainage. As such, the establishment of tidal wetland

vegetation where land use currently consists of water bodies (Figure 5.3) can be

very difficult as the water bodies should be drained or elevated to create a tidal

regime allowing the development of the wetland’s vegetation (Haltiner et al.,

1997). Areas that are currently used as agricultural or paddy fields for food

production (Figure 5.3), and that are currently protected from tidal flooding by

structures such as dikes, dams or levees, may be converted to tidal wetlands by

introducing a tidal regime, but also here, care should be taken that the elevation,

tidal inundation regime and drainage is suitable to allow successful wetland

vegetation establishment (Beauchard et al., 2011; Maris et al., 2007). Additionally,

those agricultural areas may be polluted with fertilizer, leading to high

concentrations of nitrate and phosphorus for example. When these agricultural

areas are restored in tidal wetlands, the release of those nutrients during tidal

cycles can lead to severe problems, such as on the water quality of the estuarine or

coastal system (Ardón et al., 2017; Shoemaker et al., 2017). Similarly, the pollution

in industrial or urbanized soils also implies limitations to the restoration or

creation of tidal wetlands that have to be accounted for.

Secondly, the effectiveness of tidal wetland creation for nature-based storm surge

mitigation also depends on the present land use type. Tidal wetlands reduce the

height of storm surges due to their bed roughness and the friction exerted by their

vegetation on the water column; the latter is dependent on amongst others the

vegetation density, height and stiffness (Shepard et al., 2011; Sutton-Grier et al.,

2015; Wamsley et al., 2009). For tidal wetlands, salt marsh vegetation (consisting

of grasses, herbs, and low shrubs) exerts less friction than mangrove forests, yet

they generate more friction on propagating storm surges than agricultural fields

(i.e. croplands or paddy fields) or bare soil surfaces (Mattocks & Forbes, 2008;

Passeri et al., 2018; Wamsley et al., 2009). The conversion of agricultural fields and

bare soil surfaces to tidal wetlands will then increase the friction on landward

propagating storm surges and hence will increase the attenuation rate of storm

surges. On the other hand, forested areas have a friction comparable to mangrove

forests (Mattocks & Forbes, 2008), making their conversion less interesting in

terms of storm surge attenuation. Nonetheless, the propagation of storm surges

through freshwater plants implies a salinity intrusion that is not well managed by

a number of freshwater species (Carter et al., 2018; Middleton, 2016; Stanturf et

al., 2007). Thus, keeping freshwater vegetation in order to protect the coastal

Chapter 5

156

population and areas from storm surges may be inefficient, as the resilience of the

freshwater vegetation to salinity intrusion is uncertain. Subsequently, in areas

with a high intensity and frequency of storm surges, the conversion of freshwater

vegetation to saltwater vegetation (i.e. tidal wetlands) might be seen as a valuable

nature-based strategy to increase the resilience of the vegetation to storm surges

(Middleton, 2016).

Local knowledge on how and where to restore and recreate tidal wetlands is

growing and highlights this unique character of each local setting and the

importance of understanding amongst others the different hydrodynamic,

geomorphic and ecological characteristics of the specific area that influence the

success of wetland creation (Balke & Friess, 2016; Elliott et al., 2016; Oosterlee et

al., 2018; Simenstad et al., 2006). Depending on the situation, the restoration or

creation of tidal wetlands may necessitate active re-conversion of the area by

restoring the natural hydrodynamic and subsurface hydrological flow patterns,

reshaping the topography of the area, restoring the sediment supply or planting

the appropriate vegetation for example, as in old aquaculture ponds, agriculture

fields or in more urbanized areas (R. A. Garbutt et al., 2006; Lawrence et al., 2018;

S.-M. Lee et al., 2012; Lewis & Brown, 2014; Spalding, McIvor, et al., 2014), while in

other places, wetland vegetation may spontaneously re-colonize the area without

much intervention once the appropriate hydrodynamic and bio-geomorphic

conditions are set (Eertman et al., 2002; Pethick, 2002). Although restoration

projects can be successful, the restored area will often not recreate a pristine

environment in terms of plant diversity, topography or hydrology (Bullock et al.,

2011; Elliott et al., 2016; Hobbs et al., 2009; Lawrence et al., 2018; Spalding,

McIvor, et al., 2014; Yepsen et al., 2014). However, restoration or creation of tidal

wetlands can be successful in a large variety of environments, and are expected to

be able to deliver ecosystem services such as water quality regulation, carbon

sequestration and protection against wind waves and storm surges (Adam, 2019;

Bullock et al., 2011; Hobbs et al., 2009; Spalding, McIvor, et al., 2014).

5.5 Conclusion

There is a pressing need for adaptation of the coastal zone to increasing threats

due to climate change (increased frequency of storm surges, sea level rise...) and

due to socio-economic changes (increasing coastal population density, coastal

megacities...). The development of nature-based and hybrid protection structures

for mitigation of coastal flood and erosion risks is increasingly regarded as a

sustainable and cost-efficient strategy over the long term.

Wetlands creation

157

Our study reveals that on the 135 studied cities, 60 % (8 332 km²) of the area

below mean high tide is urbanized or densely populated and 34 % (4 624 km²,

distributed over 124 cities) is potentially available for tidal wetlands restoration

or creation. Key factors influencing this potentially available space are the size of

the low-lying zone in front of the city (distance between the open sea and the city,

area below mean high tide...) as well as the population density in the close

surrounding of the city. The land use in the potentially available area for tidal

wetlands restoration or creation is mainly composed of croplands, paddy fields,

water bodies and vegetated areas, and influences the effectiveness of tidal wetland

creation for nature-based flood risk mitigation.

The analysis, which is based on global-scale datasets, is providing first estimations

regarding the globally available areas for tidal wetlands restoration or creation.

The development of specific successful restoration projects should necessitate

further local- to regional-scale analyses including a combination of scientific,

socio-economic, policy and management approaches. Local studies based on more

high-resolution datasets are needed to identify the potentially available area for

tidal wetlands restoration or creation at specific locations. This chapter has to be

regarded as a starting point to promote global awareness of the possibility to

restore or create tidal wetlands as nature-based risk mitigation in front of flood-

exposed coastal cities around the world.

Chapter 5

158

Supplementary Information

Population thresholds

Table SI 5.1 Values of the surface area available for tidal wetlands creation or restoration based on the three population threshold tested (20, 35 and 50 inhabitants/ km²)

Cities Continent Area for tidal wetlands development (km²)

< 20 inhab/km² < 35 inhab/km² < 50 inhab/km²

Abidjan Africa 15.0 16.8 18.1

Accra Africa 6.9 6.9 6.9

Adelaide Oceania 10.0 10.0 10.3

Alexandria Africa 86.8 96.3 101.4

Algiers Africa 0.0 0.0 0.0

Amsterdam Europe 71.1 97.4 123.8

Athens Europe 2.7 2.7 3.0

Auckland Oceania 5.0 6.5 6.5

Baltimore North America 3.8 4.9 5.5

Banghazi Africa 19.6 21.1 21.1

Bangkok Asia 86.3 89.5 95.8

Barcelona Europe 2.9 3.1 4.3

Barranquilla South America 105.8 113.8 118.4

Beirut Asia 0.0 0.0 0.0

Belem South America 19.7 25.9 27.8

Boston North America 4.8 5.3 6.1

Brisbane Oceania 2.1 5.5 6.1

Buenos Aires South America 0.4 0.7 1.4

Busan Asia 22.6 25.5 27.3

Cape Town Africa 0.0 0.0 0.0

Casablanca Africa 0.0 0.0 0.0

Chennai Asia 4.9 4.9 4.9

Chittagong Asia 3.9 4.2 4.2

Conakry Africa 15.1 15.7 15.7

Dakar Africa 0.0 0.0 0.0

Dalian Asia 6.0 7.0 8.0

Dar-es-Salaam Africa 2.5 2.7 2.7

Davao Asia 6.2 6.4 7.1

Dhaka Asia 97.3 102.4 107.2

Douala Africa 4.3 4.4 4.4

Dubai Asia 22.9 29.1 31.2

Dublin Europe 1.2 1.2 1.9

Wetlands creation

159

Cities Continent Area for tidal wetlands development (km²)

< 20 inhab/km² < 35 inhab/km² < 50 inhab/km²

Durban Africa 0.0 0.0 0.0

Fortaleza South America 3.6 3.8 3.8

Fukuoka Asia 2.2 2.7 3.3

Fuzhou Asia 14.0 15.5 19.5

Glasgow Europe 3.4 4.5 5.1

Guangzhou Asia 46.9 70.9 97.4

Guayaquil South America 323.9 372.5 399.4

Haiphong Asia 49.5 65.0 74.7

Hamburg Europe 505.3 747.6 880.8

Hangzhou Asia 2.7 4.2 7.4

Havana North America 0.0 0.0 0.0

Hiroshima Asia 3.2 3.4 3.8

Ho Chi Minh City Asia 49.6 68.8 98.1

Hong Kong Asia 14.5 17.4 17.6

Houston North America 16.3 17.5 19.2

Incheon Asia 17.4 20.8 21.9

Istanbul Europe 1.0 1.3 2.8

Izmir Europe 8.4 10.4 11.1

Jakarta Asia 5.6 9.0 9.0

Jeddah Asia 3.3 3.7 4.5

Karachi Asia 4.0 4.0 4.0

Copenhagen Europe 23.0 29.3 29.8

Khulna Asia 50.6 63.3 73.2

Kochi Asia 4.7 6.8 8.8

Kolkata Asia 15.1 20.2 23.7

Kuala Lumpur Asia 12.3 15.9 17.5

Kuwait Asia 2.7 3.7 3.7

Lagos Africa 13.7 14.2 15.6

Lima South America 1.6 2.0 2.0

Lisbon Europe 9.8 10.3 11.9

Lomé Africa 10.5 12.2 13.9

London Europe 21.8 35.5 48.4

Los Angeles North America 3.9 4.6 6.3

Luanda Africa 4.1 4.2 4.5

Maceio South America 7.0 8.0 8.3

Manila Asia 5.6 5.8 8.6

Maputo Africa 10.1 10.6 10.6

Maracaibo South America 8.5 9.1 9.5

Chapter 5

160

Cities Continent Area for tidal wetlands development (km²)

< 20 inhab/km² < 35 inhab/km² < 50 inhab/km²

Marseille Europe 3.4 3.9 3.9

Melbourne Oceania 2.2 2.6 2.7

Miami North America 2.2 2.9 2.9

Mogadishu Africa 4.4 4.4 4.4

Montevideo South America 9.4 10.1 10.8

Montreal North America 132.7 140.1 144.0

Mumbai Asia 12.8 16.2 20.7

Nagoya Asia 6.5 7.6 9.2

Nampo Asia 11.7 14.9 16.5

Naples Europe 0.8 0.9 0.9

Natal South America 8.8 13.8 14.8

New Orleans North America 107.5 112.9 116.0

New York North America 13.3 17.1 18.2

Ningbo Asia 5.6 8.2 8.2

Odessa Europe 49.9 50.5 50.6

Osaka Asia 0.0 0.0 0.0

Palembang Asia 38.0 39.9 45.9

Panama City North America 6.9 6.9 6.9

Perth Oceania 2.9 3.2 3.2

Philadelphia North America 71.3 77.6 80.7

Port-au-Prince North America 2.3 2.3 2.3

Portland North America 165.6 177.8 184.2

Porto Europe 0.0 1.3 1.3

Porto Alegre South America 20.9 23.4 25.3

Providence North America 2.9 3.1 3.7

Qingdao Asia 11.9 12.3 13.2

Rabat Africa 0.0 0.0 3.1

Rangoon Asia 4.8 10.3 33.6

Recife South America 1.1 1.1 1.1

Rio de Janeiro South America 3.1 3.8 3.8

Rotterdam Europe 66.4 94.0 126.6

Salvador South America 0.6 1.0 1.0

San Diego North America 2.4 3.0 3.0

San Francisco North America 8.5 11.9 13.3

San Jose North America 142.4 152.6 159.0

San Juan South America 8.7 9.4 9.9

Santo Domingo North America 2.1 2.1 2.1

Santos South America 3.4 4.2 4.5

Wetlands creation

161

Cities Continent Area for tidal wetlands development (km²)

< 20 inhab/km² < 35 inhab/km² < 50 inhab/km²

Sapporo Asia 10.4 13.1 18.2

Seattle North America 49.4 51.5 53.5

Shanghai Asia 6.9 8.3 9.7

Shenzhen Asia 6.9 12.1 13.7

Singapore Asia 25.2 27.5 28.5

Stockholm Europe 10.7 11.9 12.6

St Petersburg Europe 0.2 0.4 0.9

Surabaya Asia 30.6 44.1 46.2

Surat Asia 50.4 53.3 53.3

Sydney Oceania 2.2 4.1 4.4

Taipei Asia 2.1 2.6 2.6

Tampa North America 7.7 8.2 8.2

Tel Aviv-Yafo Asia 0.0 0.0 0.0

Tianjin Asia 174.1 211.6 233.4

Tokyo Asia 2.1 2.2 2.3

Tripoli Africa 0.0 0.0 0.0

Ujung-Pandang Asia 9.7 13.6 15.1

Ulsan Asia 1.0 2.0 2.3

Vancouver North America 1.4 5.1 8.2

Virginia Beach North America 3.8 4.3 4.3

Vishakhapatnam Asia 0.0 0.2 0.8

Grande Vitoria South America 45.1 49.9 55.7

Washington D.C. North America 48.5 51.7 54.7

Wenzhou Asia 5.6 5.8 8.4

Xiamen Asia 9.3 9.4 11.0

Yantai Asia 0.0 0.0 0.0

Zhanjiang Asia 60.1 66.3 73.9

Chapter 5

162

Land uses

Table SI 5.2 Surface areas (km²) of the land use types within the potentially available area for tidal wetlands creation or restoration in front of the 135 coastal cities, with the distinction between croplands and paddy fields.

City Continent Vegetated

Areas (km²)

Croplands (km²)

Paddy Fields (km²)

Bare Areas (km²)

Water Bodies (km²)

Abidjan Africa 10.7 0.9 6.6

Accra Africa 4.4 2.6

Adelaide Oceania 4.1 0.7

5.5

Alexandria Africa 43.9 33.7 23.8

Algiers Africa

Amsterdam Europe 78.5 35.2 10.0

Athens Europe 1.2 0.6

1.2

Auckland Oceania 4.2 2.3

Baltimore North America 2.4 0.1

2.9

Banghazi Africa 13.6 2.2 5.3

Bangkok Asia 13.2 21.4 5.8

55.5

Barcelona Europe 2.1 1.1 1.1

Barranquilla South America 93.1 8.0

17.2

Beirut Asia

Belem South America 24.7

3.1

Boston North America 1.3 4.8

Brisbane Oceania 3.1 1.1

1.9

Buenos Aires South America 0.3 1.1

Busan Asia 2.4 14.1 7.6

3.2

Cape Town Africa

Casablanca Africa

Chennai Asia 0.8 0.7 3.3

Chittagong Asia 0.9 1.8 0.3

1.3

Conakry Africa 11.1 0.7 3.9

Dakar Africa

Dalian Asia 3.5 0.2 4.3

Dar-es-Salaam Africa 0.8

1.9

Davao Asia 2.3 0.2 4.6

Dhaka Asia 14.1 8.7 34.8 0.8 48.9

Douala Africa 3.2 1.2

Dubai Asia 12.5

18.3 0.4

Dublin Europe 0.4 0.5 0.9

Durban Africa

Wetlands creation

163

City Continent Vegetated

Areas (km²)

Croplands (km²)

Paddy Fields (km²)

Bare Areas (km²)

Water Bodies (km²)

Fukuoka Asia 0.4 0.3 0.5

2.0

Fuzhou Asia 3.9 2.3 1.9 11.5

Glasgow Europe 3.2 0.5

1.4

Grande Vitoria South America 25.2 30.4 0.1

Guangzhou Asia 24.3 28.0 16.1 1.5 27.5

Guayaquil South America 210.4 70.4 41.3 77.3

Haiphong Asia 15.7 9.2 13.0

36.8

Hamburg Europe 251.5 623.9 5.5

Hangzhou Asia 0.3 2.0 1.9

3.2

Havana North America

Hiroshima Asia 1.2

2.6

Ho Chi Minh City Asia 35.6 14.7 27.2 20.6

Hong Kong Asia 8.5

9.1

Houston North America 8.5 3.0 0.7 0.6 6.4

Incheon Asia 13.9 3.6 1.7 0.3 2.5

Istanbul Europe 0.7 1.3 0.8

Izmir Europe 2.6 7.5 0.7

0.4

Jakarta Asia 0.8 0.8 7.3

Jeddah Asia 0.3

1.6 2.7

Karachi Asia 0.9 0.8 1.0 0.7 0.7

Copenhagen Europe 8.4 19.9

1.5

Khulna Asia 18.7 6.4 7.7 40.4

Kochi Asia 1.5

7.3

Kolkata Asia 1.5 4.6 2.5 15.1

Kuala Lumpur Asia 7.1 6.1 0.4 0.2 3.8

Kuwait Asia 0.6 0.5 1.5 1.2

Lagos Africa 11.6 0.7

1.5 1.8

Lima South America 0.9 1.2

Lisbon Europe 5.3 1.9

4.7

Lome Africa 11.9 2.0 0.1

London Europe 12.8 26.4

9.2

Los Angeles North America 1.9 0.2 4.2

Luanda Africa 1.0

3.5

Maceio South America 4.4 2.1 1.8

Manila Asia 0.3 0.3

8.0

Maputo Africa 4.6 6.0

Maracaibo South America 4.1

0.8 0.8 3.8

Chapter 5

164

City Continent Vegetated

Areas (km²)

Croplands (km²)

Paddy Fields (km²)

Bare Areas (km²)

Water Bodies (km²)

Marseille Europe 1.9 0.5 1.6

Melbourne Oceania 0.1 0.2

2.4

Miami North America 1.3 1.6

Mogadishu Africa 3.1 0.8

0.5

Montevideo South America 9.1 1.4 0.2

Montreal North America 29.3 54.2

60.5

Mumbai Asia 8.4 5.4 1.4 5.5

Nagoya Asia 1.1 2.8 2.6

2.6

Nampo Asia 0.7 3.8 0.8 0.1 11.1

Naples Europe

0.7

0.3

Natal South America 5.9 8.2 0.4 0.4

New Orleans North America 72.2 10.2 1.3

32.3

New York North America 7.2 0.4 10.7

Ningbo Asia 4.4

0.7

3.0

Odessa Europe 5.9 3.4 41.3

Osaka Asia

Palembang Asia 24.3 16.5 2.0 3.0

Panama City North America 2.1 0.7 0.5

3.7

Perth Oceania 1.3 1.9

Philadelphia North America 33.1 23.3 7.5 0.2 16.6

Port-au-Prince North America 0.3 1.1 0.3 0.6

Portland North America 81.4 52.9 9.6 0.2 40.2

Porto Europe 0.3 1.0

Porto Alegre South America 15.3 0.7 3.8

5.4

Providence North America 1.7 2.0

Qingdao Asia 6.1

7.1

Rabat Africa 0.7 2.4

Rangoon Asia 0.5 23.8 5.0 0.5 3.9

Recife South America 0.2 0.9

Rio de Janeiro South America 1.9

1.9

Rotterdam Europe 52.2 66.4 8.1

Salvador South America 0.3

0.7

San Diego North America 2.4 0.1 0.5

San Francisco North America 5.1

0.5 7.7

San Jose North America 42.8 4.3 1.9 110.0

San Juan South America 6.2

3.8

Santo Domingo North America 0.6 0.3 1.2

Wetlands creation

165

City Continent Vegetated

Areas (km²)

Croplands (km²)

Paddy Fields (km²)

Bare Areas (km²)

Water Bodies (km²)

Santos South America 3.5

1.0

Sapporo Asia 4.4 7.7 3.2 2.8

Seattle North America 3.8

0.4 49.2

Shanghai Asia 2.5 2.9 1.2 3.1

Shenzhen Asia 8.7

2.0 3.1

Singapore Asia 4.8 11.5 12.3

St. Petersburg Europe 0.4

0.5

Stockholm Europe 5.5 7.0

Surabaya Asia 5.2 3.4 8.1

29.5

Surat Asia 6.3 40.1 4.5 0.4 2.1

Sydney Oceania 0.9

3.5

Taipei Asia 0.2 2.4

Tampa North America 2.9 0.8 0.8

3.9

Tel Aviv-Yafo Asia

Tianjin Asia 35.8 67.2 2.0 2.0 126.3

Tokyo Asia 0.7 1.6

Tripoli Africa

Ujung Pandang Asia 1.7 0.8 12.6

Ulsan Asia 0.6 0.6 0.2

0.8

Vancouver North America 5.9 1.1 1.2

Virginia Beach North America 2.6

1.8

Vishakhapatnam Asia 0.3 0.5

Washington D.C. North America 26.3 8.0 0.4

20.1

Wenzhou Asia 3.8 1.1 0.7 2.8

Xiamen Asia 3.0 1.1

0.4 6.5

Yantai Asia

Zhanjiang Asia 5.9 10.7 5.7 51.6

Chapter 5

166

Simple linear regression

A linear regression was performed on several social and physical factors

susceptible to influence the potentially available area for tidal wetlands creation.

The significant regression coefficients (p-value < 0.05) are presented in Table 5.2.

Table SI 5.3 Regression coefficients resulting from the linear regression testing the influence of the social and physical factors (Table 5.1) on the geographical variation of the size of the tidal wetlands in front of the coastal cities for a significance of 95 % (p-value < 0.05). The ‘X’ corresponds to a non-significant regression coefficient.

Dependent variable

Explanatory variables

Unit of increase

Regression coefficients (p-value < 0.05)

Physical Parameters

Latitude 1 ° X

Distance between the open sea and the city 1 km

1.018

Coastline Length 1.003

Area Below Mean High Tide

1 km²

1.005

Shallow Water Area (depth > -100 m) X

Deep Water Area (depth < -100 m) X

Social Parameters

GDP Per Capita of the Country 1 US$ X

Short-distance population density 1 inhabitants

/km²

0.999

Intermediate-distance population density X

Long-distance population density X

The results can be interpreted as follow; one unit of increase (see Table SI 5.3) of

the explanatory variable is generating a multiplication of the surface area

potentially available for tidal wetlands creation by a factor corresponding to the

regression coefficient.

Wetlands creation

167

Examples of land reclamation

Historical land reclamation from the mouth of the Elbe estuary to the city of

Hamburg in Germany and comparison of the estimated land reclamation from our

analysis.

Figure SI 5.1 (A) Overview of the extent and period of construction of the embanked areas along the Elbe estuary adapted from Hansen (2015) and (B) representation of the urban, below mean high tide and potentially available for tidal wetlands creation areas in the likely pathway of storm surge propagation towards Hamburg according to our analysis

Chapter 5

168

Figure SI 5.2 (A) Representation of the urban expansion in the Pearl River delta and the city of Guangzhou via infrared-enhanced satellite images for the years 1979 and 2013. The red areas correspond to the delta vegetation, the blue areas to the water bodies and the grey areas to the urbanized land. From (H. Zhao et al., 2010) (B) representation of the areas urbanized, below mean high tide and potentially available for tidal wetlands creation in the likely pathway of storm surge propagation of the cities of Guangzhou, Shenzhen and Hong Kong according to our analysis.

Wetlands creation

169

Figure SI 5.3 (A) Representation of the land uses in for the city of Shanghai and the adjacent Yangtze river delta adapted from Haas et al. (2015) (B) representation of the areas urbanized, below mean high tide and potentially available for tidal wetlands creation in the likely pathway of storm surge propagation of the city of Shanghai according to our analysis.

171

CHAPTER 6 Synthesis

Chapter 6

172

The growing coastal populations and assets are increasingly put under threats of

coastal flooding and erosion risks (Von Glasow et al., 2013; Hallegatte et al., 2013;

Neumann et al., 2015) due to the effects induced by global climate change on sea

level rise and increasing storm activity (Bengtsson et al., 2006; Bernstein et al.,

2007; Webster et al., 2005). It is therefore necessary to set up sustainable and

effective coastal protection programs to mitigate the devastating consequences

that such hazards can have (Sutton-Grier et al., 2015; Temmerman et al., 2013).

One strategy that is gaining wide interest is the development of nature-based

solutions, consisting on the conservation, restoration or creation of coastal

ecosystems that have the abilities to mitigate flood risks associated with storm

surges, attenuate erosion risks from waves, and adapt to sea level rise by sediment

accumulation, and in addition provide other valuable ecosystem functions and

services (e.g. water purification, nursery for fishes and crustaceous, carbon

sequestration...) (Barbier et al., 2011; Gedan et al., 2011; Leonardi et al., 2018).

Nature-based strategies are then considered as sustainable, self-adaptive and cost-

effective solutions that can be established alone or often in combination with hard

engineering solutions (ecosystems in front of dikes), i.e. so-called hybrid

approaches (Buhl-Mortensen et al., 2017; Duarte et al., 2013; Temmerman et al.,

2013).

So far, the assessment of the contribution of coastal ecosystems to nature-based

storm surge flood risk mitigation is mainly based on local to regional scale studies

(e.g. Arkema et al., 2013; Das & Vincent, 2009; Krauss et al., 2009; McGee et al.,

2006; Stark et al., 2015). In this thesis we pursued the aim to upscale those

assessments from regional to global scales. With, for objective, on the one hand to

highlight the widespread presence of coastal ecosystems and their capacity to

attenuate storm surges as well as the possibility, under certain conditions, to

restore or create them for coastal flood and erosion risks mitigation and on the

other hand to compare the presence of the coastal ecosystems with the coastal

areas densely populated and exposed to storm surge flood risks. Furthermore,

with such global scale studies, we aimed to increase the awareness of the local

communities and policy-makers in the possibility to implement, at global scale,

nature-based strategies alone or in combination with hard engineering structures

for coastal flood and erosion risks mitigation. In such, we developed procedures to

identify the coastal plain areas and populations exposed to coastal flood risks via

flood pathways crossing through tidal wetlands, and hence that benefit from storm

surge flood risks mitigation provided by the tidal wetlands (Chapters 2 & 3); we

quantified the existing extent of different coastal ecosystems (salt marshes,

mangroves, seagrasses and coral reefs) in front of highly populated and flood-

Synthesis

173

exposed coastal cities (Chapter 4); and we estimated the potentially available

space for tidal wetlands restoration or creation in front of those cities (Chapter 5).

6.1 Nature-based mitigation of coastal flood risks

The contribution of tidal wetlands to storm surge mitigation was defined at the

delta scale (Chapter 2) and at the worldwide scale (Chapter 3), while the

potential for nature-based coastal flood risks mitigation was estimated at the city

scale (Chapter 4 & 5).

At the delta and worldwide scales, the results showed high variability in the

amount of coastal protection offered by the tidal wetlands against storm surge

flood risks, and suggests, as explored by Gedan et al. (2011), that even the smallest

tidal wetlands can provide a certain level of storm surge mitigation. The analysis

of the tidal wetlands’ protection to the delta coastal plains (Chapter 2) showed

that the three deltas with the largest percentage of flood-exposed surface area (>

80 %) and population (> 70 %) benefiting from storm surge mitigation are the

Mahakam (Indonesia), Niger (Nigeria) and Chao Phraya (Thailand) deltas. While,

at the global scale (Chapter 3), we found that about 30 % of the flood-exposed

coastal plain and 40 % of the flood-exposed population benefit from storm surge

mitigation by tidal wetlands.

Based on Chapter 2, we could observe that, on the one hand, scattered tidal

wetlands located along the main channels within a delta (e.g. Chao Phraya delta)

are contributing to a larger land area benefiting from storm surge mitigation, than

clustered tidal wetlands (e.g. Ganges-Brahmaputra delta) and on the other hand,

that tidal wetlands dissected by numerous and wide channels provide storm surge

mitigation to a lesser coastal plain area (e.g. Yangtze or Rhine deltas) than tidal

wetlands that are not so densely dissected by wide channels (e.g. Mississippi delta

or Niger delta). Similarly, the magnitude of storm surge flood risks mitigation, as

estimated in Chapter 2 by the length of the storm surge pathway that is crossing

through tidal wetlands, is the highest when large and continuous tidal wetlands

are present. This finding corroborates previous studies highlighting that the

continuity of the wetlands is a key factor in the storm surge mitigation capacity of

tidal wetlands (Loder et al., 2009; Phan et al., 2015; Smolders et al., 2015; Stark et

al., 2016; Zhang et al., 2012).

Low-lying coastal plains found in deltas, estuaries, bays and lagoons are the most

favourable environment for the development of extensive and continuous tidal

wetlands (Leonardi et al., 2018). In such, at the worldwide scale (Chapter 3), the

largest relative coastal plains (> 15 km² per kilometre of shoreline) benefiting

Chapter 6

174

from storm surge mitigation by tidal wetlands are in or close to deltas, estuaries,

bays and lagoons, and mainly located along the Northern European and Eastern

Asian coasts, e.g. Aiguillon bay (France), Rhine-Meuse-Scheldt delta (Belgium and

Netherlands), Ems estuary (Germany), Chao Phraya delta (Thailand), Mekong

delta (Vietnam) or Pearl river delta (China). Whilst, at the city scale (Chapter 5),

the cities located in deltas, estuaries, bays or lagoons are the hotspots for tidal

wetlands restoration, as they have extensive low-lying and flat coastal plains (e.g.

Hamburg (Germany) or Guayaquil (Ecuador)) mainly occupied by freshwater

vegetation, fields (e.g. croplands and paddy fields) or water bodies. However, some

of those deltas present densely populated urban agglomerations and as such have

less available space for tidal wetlands restoration or creation (e.g. Guangzhou

(China)).

Similarly, the longest distances (> 10 km) travelled by storm surges through tidal

wetlands, i.e. indicative for the highest magnitude of storm surge mitigation, are

also found in deltaic and estuarine environments such as in the Rio Guayas

(Ecuador), the Kolyma delta (North-eastern Siberia), the Yangtze and Northern

Yangtze delta (China), the Ganges-Brahmaputra delta (India and Bangladesh) or

the Tidung estuary (Borneo, Indonesia).

Results on the amount of people benefiting from storm surge mitigation by tidal

wetlands are highly variable and do not relate to the size of the tidal wetlands, but

on other factors as the location of the tidal wetlands relative to the spatial

distribution of the population, the population density or the history of human

settlement. The results of Chapter 2 reveal that tidal wetlands located either in

between the coastline and the populated areas (e.g. Chao Phraya delta), or along

both sides of the main channels within deltas (e.g. Niger and Mahakam deltas)

result in a higher impact of the wetlands on the propagation of the storm surge

and a larger population benefiting from the storm surge mitigation. At the

worldwide scale (Chapter 3), the coastal zones with the largest populations

benefiting from storm surge mitigation by tidal wetlands are in deltas, estuaries or

bays that are prone to the establishment of the coastal populations, but also to the

development of extensive tidal wetlands (Leonardi et al., 2018). Yet, at the city

scale (Chapter 4) although salt marshes, mangrove forests, seagrass meadows

and/or coral reefs are present in front of 75 % of the studied cities, the cities with

the largest flood-exposed population (> 500 000 people) and assets (> 36 billion

US$) group only 15 % of the total coastal ecosystem area fronting the 135 studied

cities. Only four cities with large population and assets at risk are benefiting from

large coastal ecosystems (> 237 km²; i.e. Miami, New Orleans, Guangzhou and

Synthesis

175

Shanghai), while Ho Chi Minh City and Hong Kong combine large coastal

ecosystems with, respectively, large flood-exposed populations.

At the global scale (Chapter 3), the largest flood-exposed populations (> 10 000

inhabitants per km of shoreline) benefiting from storm surge mitigation are

amongst others in the Rhine-Meuse-Scheldt delta (Belgium and Netherlands), the

Weser estuary (Germany), the Mekong delta (Vietnam) or the Red River delta

(China), while the degree of the storm surge attenuation is highly variable

between those areas due to the size of the tidal wetlands. Furthermore, results of

Chapter 3 show that more than 80 % of the flood-exposed population benefiting

from storm surge mitigation by tidal wetlands is located in only five countries (i.e.

China, Vietnam, India, Netherlands and Indonesia) that have highly populated low-

lying coastal areas.

The anthropogenic pressure on the coastal areas, via land reclamation among

others, resulted in large losses of coastal ecosystems. In order to better

understand the present-day distribution of tidal wetlands, i.e. salt marshes and

mangrove forests, we estimated the wetland areas that were reclaimed over time

for human use (e.g. agriculture, aquaculture, urban and industrial development)

(Chapter 4 & 5). Results from our analyses and from the literature (e.g. Airoldi &

Beck, 2007; Hatvany, 2003; Hoeksema, 2007; Murray et al., 2014; Scott et al.,

2014) suggest that Asia and Europe, as the earliest populated continents,

experienced the earliest, and largest, land reclamation by the creation of so-called

‘polders’. By the nineteenth and twentieth centuries, the practice of land

reclamation occurred on all continents resulting in major losses of coastal

ecosystems (Barbier et al., 2008; IPCC, 2013; Millennium Ecosystem Assessment,

2005; Pendleton et al., 2012; Valiela et al., 2001). The cities with the largest

reclaimed areas, estimated as the areas located below mean high tide (Chapter 4

& 5), are then mainly located in front of cities in European and East Asian deltas

and estuaries, e.g. Hamburg (1 400 km² of estimated reclaimed wetlands),

Guangzhou (1 100 km²), Ho Chi Minh City (800 km²) or Rotterdam (780 km²). The

analysis of the land use in the areas below mean high tide (Chapter 5) showed

that about two third of it is highly populated and/or urbanized and the remaining

third vegetated or occupied by agriculture fields.

From the analysis of a set of parameters susceptible to influence the presence and

extent of the coastal ecosystems in front of cities (Chapter 4), we found that

coastal ecosystems (i.e. salt marshes, mangroves, seagrass meadows and coral

reefs), depending on their type are influenced either positively or negatively, by

geomorphological factors (water depth, size of the area below mean high tide,

Chapter 6

176

distance to the sea...) and human induced parameters as the countries’ GDP per

capita and the population density in the vicinity of the city. However, the

differences in surface area of coastal ecosystems in front of the studied cities are

not fully explained by the studied parameters, implying that other factors play a

role, such as the bio-geomorphology (e.g. soil composition...), the water chemistry

(e.g. salinity, pH, temperature...), the hydrodynamic (e.g. tidal range, waves and

currents energy and orientation...) or other parameters as the anthropogenic

pressures on the ecosystems.

6.2 Limitations and suggestions for future research

The main purpose of this thesis was to present procedures for a regional to global

analysis of the contribution and possible contribution of a set of coastal

ecosystems to nature-based coastal flood and erosion risks mitigation.

Therefore, the use of global scale datasets in the GIS model and procedures

implied both a lower overall resolution and accuracy than in local studies or

hydrodynamic models and the loss of the local characteristics and specifications of

the coastal zone. It further implies the use of ‘statistic’ values of storm surge

heights for a given return period that neglect local and event dependent

characteristics of a storm surge. Indeed, the two storm surge datasets used (DIVA

and GTSR) present storm surge heights over coastline segments for return periods

of 10, 100 and 1000 years based on model simulations including a limited set of

parameters (e.g. tidal levels and barometric pressures or sea bed slope) and on

validation against past storm surges. It means that the storm surge heights used in

this thesis are to be regarded as broad scale quantifications necessitating further

refinement at local scale. Furthermore, the storm surge attenuation rates that

were used, were estimated for tidal wetlands at local to regional scales for specific

storm conditions and landscape characteristics (e.g. Krauss et al., 2009; Ondiviela

et al., 2014; Stark et al., 2015; Wamsley et al., 2010; Zhang et al., 2012).

Consequently, the storm surge attenuation rates remain location dependent and

focused on rates of storm surge attenuation within tidal wetlands. Attenuation

rates are then rarely known for storm surge propagation over other land use types

of the coastal floodplains, such as agricultural or urban land use, and different

vegetation types. Their up-scaling to a regional to global scale (as done in this

thesis) assumes an averaged and generally applicable value for the storm surge

attenuation rate by tidal wetlands, while in reality specific rate values may vary

with the local landscape conditions and storm characteristics (Guannel et al.,

2016; Leonardi et al., 2018; McIvor, Spencer, et al., 2012; Wamsley et al., 2010).

Therefore, efforts should be done to increase the knowledge on storm surge

Synthesis

177

attenuation rates over various land use types, namely tidal wetlands (salt marshes

and mangroves) and other coastal ecosystems (e.g. coral reefs, shellfish reefs (as in

Brandon et al. (2016), seagrass beds...), but also bare and vegetated soils in the

coastal floodplain (e.g. shrubs, herbaceous vegetation, forests, mixed types...),

urban, agricultural, and industrial land use types, etc.

Additionally, the need of global (or quasi-global) scale datasets of coastal

ecosystems implied a limitation of the coastal ecosystems considered in our

analyses. Therefore we focused on salt marshes and mangrove forests in Chapters

2 and 3, as the rate of storm surge height attenuation is best studied in literature

for salt marshes and mangrove forests, while largely unquantified for other coastal

ecosystem types. In Chapters 4 and 5, we estimated the spatial extent of four

coastal ecosystem types in front of flood-exposed coastal cities, namely salt

marshes, mangroves forest, seagrass beds and coral reefs. Other ecosystems such

as dunes, kelp beds, oyster beds, and others that may contribute to coastal risk

mitigation, even though they are known to provide a certain level of nature-based

coastal protection, couldn’t be incorporated in the analysis due to the lack of global

data on the worldwide spatial distribution of these ecosystem types (Beck et al.,

2017; Hanley et al., 2014; Hoggart et al., 2015; Narayan et al., 2016; Reguero et al.,

2014). In such, accounting for those coastal ecosystems in the GIS model of

Chapters 2 and 3 would modify the results. By highlighting, in addition to the

areas benefiting from the influence of tidal wetlands, the areas benefiting from the

presence of the other coastal and marine ecosystems as well as the areas

benefiting from the presence of multiple ecosystems that together provide higher

nature-based coastal flood risks mitigation (Guannel et al., 2016). Including more

coastal ecosystems in the cities’ analyses in Chapters 4 and 5 could lead to an

increase of the hotspots for nature-based coastal flood and erosion risks

mitigation.

The combination of improved knowledge on the values of storm surge attenuation

rates with an increased number of coastal ecosystems and datasets of higher

resolution (e.g. better delineation of the channels and wetlands, increased

information on the ecosystems vegetation density, height or structure...) would, in

the case of our GIS model, allow a more accurate prediction of both the landward

propagation of storm surges and the flood-exposed coastal plains and populations

benefiting from storm surge mitigation by coastal ecosystems.

Nevertheless, our global GIS model assessing the contribution of tidal wetlands to

coastal flood risks mitigation (Chapters 2 and 3), despite being of lower

resolution, for a selected number of coastal ecosystems and not simulating the full

Chapter 6

178

complexity of the hydrodynamic processes involved in storm surge propagation,

provides insights on the location and magnitude of storm surge mitigation by tidal

wetlands from regional to global scale. Its major advantages are to be globally

applicable, based on global or regional datasets and computationally much less

demanding than hydrodynamic models. Such quasi-global and global scale

analyses, by opposition to more local studies, allow the comparison of the

worldwide coastlines and selected deltas. Moreover, it defines and highlights

hotspots for nature-based storm surge flood risks mitigation on a global scale,

even for countries or world areas where such projects are not considered.

The GIS based procedures of Chapters 4 and 5 are on the other hand giving

estimations of the potential for nature-based coastal flood and erosion risks

mitigation in front of coastal cities corroborating the literature (e.g. Airoldi & Beck,

2007; Haas et al., 2015; Hansen, 2015; Zhao et al., 2010). The insights allow the

comparison of the different cities over the world in terms of existing coastal

ecosystems for nature-based storm surge flood risk mitigation and in terms of

potentially available space for tidal wetlands development to enhance the nature-

based storm surge mitigation. However, the estimates for tidal wetlands

restoration or creation are theoretical, which implies to account for other factors

such as the current land use or the socio-economic situation to fully determine the

possibility to restore or create wetlands in the highlighted areas. The current land

use involves parameters such as the soil elevation and hydrodynamic regime, the

soil pollution due to human use and the consequence of the release of those

pollutants, or the soil properties like the soil compaction that influences the

groundwater flow for example. Whilst the socio-economic situation may hamper

the restoration or creation of ecosystems due to the non-support or the

disapproval of such nature-based projects by the local communities that may be

willing to implement nature-based strategies, but not at their expenses (e.g. loss of

land, livelihood...) (Goeldner-Gianella, 2008; Temmerman et al., 2013). In addition,

the knowledge on ecosystem restoration is growing, and mostly highlights the

unique character of each project (Balke & Friess, 2016; Elliott et al., 2016).

Consequently, each project of wetland restoration or creation needs to rely on site-

specific, in depth scientific studies on the capacity of the wetlands to develop in

the delineated areas, depending on the hydrodynamic features, the

geomorphology, the sediment types, the tidal regime..., but also on an inclusive

analysis of the socio-economic situation of the area.

Additionally, the effects of global climate change on the coastal ecosystems for the

next centuries are not accounted for in the analyses. It is complex to define the

ecological effects of climate change as they are linked to different climate drivers,

Synthesis

179

the biotic and environmental conditions as well as to the anthropogenic activities

(Nicholls et al., 2018; Saunders et al., 2014). However, several studies highlight

that wetlands could suffer from sea level rise, with estimates, including

geomorphological and socio-economic feedbacks, giving up to 30 % of wetlands

loss by 2100, other studies estimate the loss to up to 90 % (Schuerch et al., 2018).

With as major factors for the resilience of the wetlands the sediment accretion for

vertical and lateral expansion and the accommodation space, to avoid the coastal

squeeze (Schuerch et al., 2018). Seagrasses on the other hand if exposed to the

changing climate conditions will have higher rates of photosynthesis, carbon

sequestration and growth, the latter, could lead to a long-term vulnerability of the

plants to storm conditions as their growth can modify their biomechanical

properties (De los Santos et al., 2017; Nicholls et al., 2018; O’Brien et al., 2017).

The coral reefs are expected to suffer from the warming and acidification of the

ocean (decreasing pH), yet it is not clear what would be their resilience to those

changes (Hoegh-Guldberg et al., 2007; Van Hooidonk et al., 2016; Nicholls et al.,

2018). Overall, global change over the next centuries is then expected to influence

the different coastal ecosystems and the changes in the environmental conditions

should be taken into account when planning for nature-based or hybrid coastal

protection strategies.

Overall, our global scale analyses emphasize the presence of coastal ecosystems

over the world’s coastlines and the possibility to account for them in coastal

planning strategies. And, the results presented in this thesis are then to be seen as

a step towards a better consideration of the benefits and value of the coastal

ecosystems for coastal flood and erosion risks reduction that could reach the

public and policy-makers.

Future research on the contribution of coastal ecosystems for coastal flood and

erosion risks mitigation should have two main objectives. Firstly, in order to

widen our understanding of nature-based coastal flood and erosion risks

mitigation, we should learn more about the mechanisms and processes behind the

attenuation of storm surges by coastal ecosystems while accounting for the

influence of the ecosystem structure and local landscape characteristics. This can

be done via in-situ monitoring of storm surge propagation over the coastal plain,

as it was done for several studies (e.g. Krauss et al., 2009; Stark et al., 2015) or via

the development of hydrodynamic models based on the insights resulting from the

previously mentioned field observations (e.g. Stark et al., 2016; Zhang et al., 2012).

Although an increasing number of such local-scale observational and modelling

studies is being published, the case studies are to be broadened to include a

Chapter 6

180

maximum of coastal ecosystem types, coastal bio-geomorphic settings, and storm

surge conditions.

Secondly, the definition, based on our GIS models and procedures, of hotspots for

nature-based coastal flood and erosion risks mitigation at the delta, city or global

scale provides inputs on where nature-based strategies could be applied.

Subsequently, it determines where local to regional studies on the implementation

of nature-based or hybrid coastal protection structures are needed. Those studies

should rely on local to regional high resolution datasets and account for the

landscape settings (including topography, bathymetry, geomorphology, slope of

the continental shelf...), the local storm surge characteristics (e.g. forward moving

speed, intensity, duration, direction...), the coastal land uses and their

characteristics (e.g. coastal ecosystems, agricultural fields, urban and populated

areas...) and the socio-economic situation. From there, the actual benefits of

coastal ecosystems for coastal protection would be refined for each considered

location and, where possible, coastal protection strategies could be implemented

to minimize the vulnerability of the coastal areas and communities to coastal flood

and erosion risks.

6.3 Implications for coastal zone management

The increasing threats exerted on coastal zones and associated populations and

assets, makes it imperative to create sustainable, cost-effective and efficient long-

term strategies to mitigate coastal flood risks (Adriana Gracia et al., 2018; Ma et al.,

2014). Nature-based strategies are increasingly proposed and implemented as

part of coastal risk reduction programs, for example along Northern European

coasts (Gardiner et al., 2007; SigmaPlan, 2017; Ysebaert et al., 2016) and in the UK

(Esteves, 2014; Pendle, 2013; Pethick, 2002), or in major coastal areas in the USA

(Boesch et al., 2006; Coastal Wetlands Planning Protection and Restoration Act

(CWPPRA), 1990; Day et al., 2007; Esteves, 2014; RESTORE, 2017). Nonetheless,

the historical trend of coastal ecosystem degradation and loss, and the

construction of hard coastal engineering structures with little attention for

adverse effects on the surrounding natural environment, are still problematic

(Hoeksema, 2007; Lotze et al., 2006; Valiela et al., 2009). Throughout this thesis,

we pursued the general aim to highlight the importance of conservation and

restoration/creation of coastal ecosystems and tidal wetlands as integrated

elements within programs for coastal flood risk mitigation. Whilst our

assessments are at regional to global scales, and consequently do not include

specific local conditions, they highlight that conservation and restoration/creation

Synthesis

181

of coastal ecosystems is beneficial to enhance the mitigation of storm surge and

erosion risks for coastal communities at several locations around the globe.

Here, in the scope of enhancing the coastal protection from storm surge and

erosion risks by accounting for nature-based strategies relying on four types of

coastal ecosystems (i.e. salt marshes, mangrove forests, seagrass meadows and

coral reefs), the focus was put on the environments where those ecosystems either

are present or can be developed. That does not imply, however, that coastal zones

not suitable for the establishment of the considered coastal ecosystems should not

rely on other and better suited ecosystems (e.g. dunes, oyster reefs...) to enhance

their resilience against coastal flood and erosion risks (Barbier et al., 2011; Cheong

et al., 2013; Grabowski et al., 2012). Sand dunes for example can mitigate storm

surges as they form geomorphic barriers against waves and storm surges, while

their above and below-ground vegetation increases resistance to the storm surge.

The above-ground vegetation will exert friction on the water column, in a similar

way as in tidal wetlands, while the below-ground vegetation provides sediment

stability and erosion resistance (Sigren et al., 2018; Silva et al., 2016).

Consequently, sand dunes are important elements of nature-based coastal

protection strategies such as in The Netherlands, in France or along the coastline

in Texas (USA) (Hanley et al., 2014; Rozé & Lemauviel, 2004; Sigren et al., 2014).

In the locations suitable for the development of the considered coastal ecosystems

(i.e. salt marshes, mangrove forests, seagrass beds and coral reefs), management

strategies should account for their presence or absence, but also to the protected

status of the area, such as international nature conservation legislations, e.g.

Ramsar or Natura2000 or national legislations. Along coastlines that were heavily

altered by human use over past and recent history (e.g. North European coasts),

the existing coastal ecosystems are often scarce or inexistent, limiting the

immediate application of nature-based coastal flood and erosion risks mitigation

(Chapters 4 and 5). For those coasts, which often rely on hard engineering

structures for coastal protection (e.g. dikes, dams...), ecosystems restoration or

creation should be stimulated as add-on to the existing engineered flood defence

structures, because vegetated ecosystems in front of these structures can

contribute significantly to reduced wave and storm surge impacts on the flood

defence structures, and hence can reduce the risk of flood disasters by failure of

these structures (Voortman et al., 2003; Vuik et al., 2016, 2018; van Wesenbeeck

et al., 2014). In those coasts, it is in addition crucial to include and gain the support

of the local communities in the development of such nature-based strategies. On

the other hand, many coastlines have large remaining ecosystems (e.g. Asian or

North and South American coasts) valuable for nature-based risks mitigation

Chapter 6

182

(Chapters 4 and 5), but the current widespread practice of ecosystem conversion

for human land use (e.g. aquaculture ponds, industrial areas...) makes them highly

vulnerable. Management practices should then invest in reducing the conversion

of coastal ecosystems to human land use, and in careful spatial planning of where

wetland conversion should be especially avoided in order to sustain their value for

nature-based storm surge risk mitigation. Nature-based strategies, although able

to reduce the risks of flooding, may not be sufficient on their own to fully protect

the low-lying coastal populations and associated assets. As such, hybrid strategies,

or the combination of nature-based strategies with other coastal protection

strategies, e.g. engineering structures, will often be of highest interest for coastal

protection.

As local communities are often unaware of the benefits and functions ecosystems

can provide, they may be reluctant to the development of nature-based strategies

(Esteves, 2014; French, 2006; Ledoux et al., 2004; The World Bank, 2017). A first

reason to this unwillingness of nature-based strategies is their skepticism about

the effectiveness of ecosystems for coastal flood and erosion risks mitigation,

while hard engineering structures make people feel safe (Esteves, 2014). Secondly,

the conversion of reclaimed land areas by previous generations (such as the

polders in Belgium and the Netherlands) into natural ecosystems is seen as a loss

of hard work and economic inputs (Goeldner-Gianella, 2008; Temmerman et al.,

2013). The integration of nature-based strategies into coastal planning

necessitates then to spread the scientific knowledge on the mechanisms and

processes behind the coastal flood and erosion mitigation capacities of the

ecosystems (Spalding, McIvor, et al., 2014; Sutton-Grier et al., 2015) as well as to

highlight the economic value of the ecosystem services such as coastal protection,

water filtration, habitat for fishes and crustaceous... (Barbier, 2015b; Dewsbury et

al., 2016; Himes-Cornell et al., 2018; Pascal et al., 2016) towards the local

communities, stakeholders, decision- and policy-makers. Both will then serve as

arguments to increase the support of the stakeholders and policy-makers for the

implementation of nature-based or hybrid strategies for coastal planning (De

Groot et al., 2013; Hill, 2015; Ledoux et al., 2004; Sutton-Grier et al., 2015).

Yet, if the coastal protection and other services provided by coastal ecosystems

along with their cost-efficiency and self-maintenance characteristics are

convincing arguments to maintain or create ecosystems, pilot studies on the

feasibility of nature-based strategies are necessary. The knowledge on the creation

or conservation of coastal ecosystems is growing and highlights the need to

understand the unique character of each coastal zone in regards to the

hydrodynamic, geomorphologic, ecological and socio-economic processes (Balke &

Synthesis

183

Friess, 2016; Simenstad et al., 2006). Successful nature-based strategies will then

rely on the capacity of the coastal managers to address the challenges created by

the different environmental and socio-economic situations. For example by

accounting for the needed environmental factors for ecosystems development (e.g.

favourable drainage, sediment supply...), by creating efficient population

settlement planning (e.g. to avoid coastal squeeze) and preventing the degradation

or modification of the environment that would affect the growth of the coastal

ecosystems (Lewis & Brown, 2014; The World Bank, 2017). In the case of

restoration projects, an extra vigilance is to be observed in defining the suitable

area for the establishment of the considered ecosystem, regarding the type of land

use for example; the conversion of agriculture or aquaculture fields into

ecosystems is often more feasible than converting urban or industrial areas.

Via the global analyses presented in this thesis, we demonstrated that despite the

various worldwide coastal environments, multiple coastal areas can already

benefit from nature-based storm surge mitigation and that this nature-based

mitigation could increase with the restoration or creation of coastal ecosystems.

Yet, only the combination of inclusive policy, long term management practices and

in depth scientific studies will lead to efficient nature-based coastal flood risk

mitigation strategies.

References

185

References

Adam, P. (2002). Saltmarshes in a time of change. Environmental Conservation, 29(01), 39–61. https://doi.org/10.1017/S0376892902000048

Adam, P. (2019). Salt Marsh Restoration. In Coastal Wetlands (pp. 817–861). Elsevier. https://doi.org/10.1016/B978-0-444-63893-9.00023-X

Adriana Gracia, C., Rangel-Buitrago, N., Oakley, J. A., & Williams, A. T. (2018). Use of ecosystems in coastal erosion management. Ocean and Coastal Management, 156, 277–289. https://doi.org/10.1016/j.ocecoaman.2017.07.009

Agardy, T., Alder, J., Dayton, P., Curran, S., Kitchingman, A., Wilson, M., et al. (2005). Coastal Systems. In Ecosystems and Human Well-being: Current Status and Trends, Volume 1 (pp. 513–550). Retrieved from http://www.millenniumassessment.org/documents/document.288.aspx.pdf

Airoldi, L., & Beck, M. (2007). Loss, Status and Trends for Coastal Marine Habitats of Europe. In R. N. Gibson, R. J. A. Atkinson, & J. D. M. Gordon (Eds.), Oceanography and Marine Biology An Annual Review, Volume 45 (1st Editio, pp. 345–405). https://doi.org/10.1201/9781420050943.ch7

Almeida, D., Neto, C., Esteves, L. S., & Costa, J. C. (2014). The impacts of land-use changes on the recovery of saltmarshes in Portugal. Ocean and Coastal Management, 92, 40–49. https://doi.org/10.1016/j.ocecoaman.2014.02.008

Alongi, D. M. (2008). Mangrove forests: Resilience, protection from tsunamis, and responses to global climate change. Estuarine, Coastal and Shelf Science, 76(1), 1–13. https://doi.org/10.1016/j.ecss.2007.08.024

Alongi, D. M. (2009). The Energetics of Mangrove Forests. Dordrecht: Springer Netherlands. https://doi.org/10.1007/978-1-4020-4271-3

An, S., Li, H., Guan, B., Zhou, C., Wang, Z., Deng, Z., et al. (2007). China’s natural wetlands: Past problems, current status, and future challenges. Ambio, 36(4), 335–342. https://doi.org/10.1579/0044-7447(2007)36[335:CNWPPC]2.0.CO;2

Ardón, M., Helton, A. M., Scheuerell, M. D., & Bernhardt, E. S. (2017). Fertilizer legacies meet saltwater incursion: challenges and constraints for coastal plain wetland restoration. Elem Sci Anth, 5(0), 41. https://doi.org/10.1525/elementa.236

Arkema, K. K., Guannel, G., Verutes, G., Wood, S. a., Guerry, A., Ruckelshaus, M., et al. (2013). Coastal habitats shield people and property from sea-level rise and storms. Nature Climate Change, 3(10), 913–918. https://doi.org/10.1038/nclimate1944

Auerbach, L. W., Goodbred Jr, S. L., Mondal, D. R., Wilson, C. a., Ahmed, K. R., Roy, K., et al. (2015). Flood risk of natural and embanked landscapes on the Ganges–Brahmaputra tidal delta plain. Nature Climate Change, 5(2), 153–157. https://doi.org/10.1038/nclimate2472

Balke, T., & Friess, D. A. (2016). Geomorphic knowledge for mangrove restoration: A pan-tropical categorization. Earth Surface Processes and Landforms, 41(2), 231–239. https://doi.org/10.1002/esp.3841

References

186

Barbier, E. B. (2007). Valuing ecosystem services as productive inputs. In Economic Policy (Vol. 22, pp. 177–229). https://doi.org/10.1111/j.1468-0327.2007.00174.x

Barbier, E. B. (2014). A global strategy for protecting vulnerable coastal populations. Science, 345(6202), 1250–1251.

Barbier, E. B. (2015a). Climate change impacts on rural poverty in low-elevation coastal zones. Estuarine, Coastal and Shelf Science, 165, A1–A13. https://doi.org/10.1016/j.ecss.2015.05.035

Barbier, E. B. (2015b). Valuing the storm protection service of estuarine and coastal ecosystems. Ecosystem Services, 11, 32–38. https://doi.org/10.1016/j.ecoser.2014.06.010

Barbier, E. B., & Barbier, B. E. B. (2014). A global strategy for protecting vulnerable coastal populations. Science, 345(6202), 1250–1251.

Barbier, E. B., Koch, E. W., Silliman, B. R., Hacker, S. D., Wolanski, E., Primavera, J., et al. (2008). Coastal ecosystem-based management with nonlinear in ecological functions and values. Science, 319(5861), 321–3. https://doi.org/10.1126/science.1150349

Barbier, E. B., Hacker, S. D., Kennedy, C., Koch, E. W., Stier, A. C., & Silliman, B. R. (2011). The value of estuarine and coastal ecosystem services. Ecological Monographs. https://doi.org/10.1890/10-1510.1

Barbier, E. B., Georgiou, I. Y., Enchelmeyer, B., & Reed, D. J. (2013). The Value of Wetlands in Protecting Southeast Louisiana from Hurricane Storm Surges. PLoS ONE, 8(3), 1–6. https://doi.org/10.1371/journal.pone.0058715

Beauchard, O., Jacobs, S., Cox, T. J. S., Maris, T., Vrebos, D., Van Braeckel, A., & Meire, P. (2011). A new technique for tidal habitat restoration: Evaluation of its hydrological potentials. Ecological Engineering, 37(11), 1849–1858. https://doi.org/10.1016/j.ecoleng.2011.06.010

Beavers, R. L., Babson, A. L., & Schupp, C. A. (Eds.). (2016). Coastal Adaptation Strategies Handbook (NPS 999/13). Washington, DC: National Park Service.

Beck, M. W., Losada, I. J., Reguero, B. G., Mendendez, P., & Burke, L. (2017). The global flood protection savings provided by coral reefs. Nature Communications, (2018). https://doi.org/10.1038/s41467-018-04568-z

Bengtsson, L., Hodges, K. I., & Roeckner, E. (2006). Storm Tracks and Climate Change. Journal of Climate, 19(15), 3518–3543. https://doi.org/10.1175/JCLI3815.1

Bernstein, L., Bosch, P., Canziani, O., Chen, Z., Christ, R., Davidson, O., et al. (2007). Climate Change 2007 : An Assessment of the Intergovernmental Panel on Climate Change. Change, 446(November), 12–17. https://doi.org/10.1256/004316502320517344

Bessell-Browne, P., Negri, A. P., Fisher, R., Clode, P. L., Duckworth, A., & Jones, R. (2017). Impacts of turbidity on corals: The relative importance of light limitation and suspended sediments. Marine Pollution Bulletin, 117(1–2), 161–170. https://doi.org/10.1016/j.marpolbul.2017.01.050

Blankespoor, B., Dasgupta, S., & Laplante, B. (2014). Sea-Level Rise and Coastal Wetlands. Ambio, 996–1005. https://doi.org/10.1007/s13280-014-0500-4

Boesch, D. F., Shabman, L., Antle, L. G., John W. Day, J., Dean, R. G., Galloway, G. E., et al. (2006). A New Framework for Planning the Future of Coastal Louisiana after the Hurricanes of 2005. Smithsonian.

References

187

Bolund, P., & Hunhammar, S. (1999). Ecosystem services in urban areas. Ecological Economics, 29(2), 293–301. https://doi.org/10.1016/S0921-8009(99)00013-0

Brandon, C. M., Woodruff, J. D., Orton, P. M., & Donnelly, J. P. (2016). Evidence for elevated coastal vulnerability following large-scale historical oyster bed harvesting. Earth Surface Processes and Landforms, 41(8), 1136–1143. https://doi.org/10.1002/esp.3931

Breaux, A., Farber, S., & Day, J. (1995). Using natural coastal wetlands systems for wastewater treatment: An economic benefit analysis. Journal of Environmental Management, 44(3), 285–291. https://doi.org/10.1006/jema.1995.0046

Bright, E. A., Coleman, P. R., Rose, A. N., & Urban, M. L. (2013). LandScan 2013. Oak Ridge, TN: Oak Ridge National Laboratory SE - July 1, 2012. Retrieved from http://www.ornl.gov/landscan

British Oceanographic Data Center. (2017). GEBCO 30 arc-second grid. Retrieved December 6, 2017, from https://www.gebco.net/data_and_products/gridded_bathymetry_data/gebco_30_second_grid/

Brown, B. E., & Dunne, R. P. (1988). The Environmental Impact of Coral Mining on Coral Reefs in the Maldives. Environmental Conservation, 15(2), 159–165. https://doi.org/10.1017/S0376892900028976

Buddemeier, R. W., & Smith, S. V. (1988). Coral Reefs predictions and suggestions for long-term research. Coral Reefs, 7, 51–56.

Buhl-Mortensen, L., Galparsoro, I., Vega Fernández, T., Johnson, K., D’Anna, G., Badalamenti, F., et al. (2017). Maritime ecosystem-based management in practice: Lessons learned from the application of a generic spatial planning framework in Europe. Marine Policy, 75, 174–186. https://doi.org/10.1016/j.marpol.2016.01.024

Bullock, J. M., Aronson, J., Newton, A. C., Pywell, R. F., & Rey-benayas, J. M. (2011). Restoration of ecosystem services and biodiversity : conflicts and opportunities, 26(10), 541–549. https://doi.org/10.1016/j.tree.2011.06.011

Carter, G. A., Otvos, E. G., Anderson, C. P., Funderburk, W. R., & Lucas, K. L. (2018). Catastrophic storm impact and gradual recovery on the Mississippi-Alabama barrier islands, 2005–2010: Changes in vegetated and total land area, and relationships of post-storm ecological communities with surface elevation. Geomorphology. https://doi.org/10.1016/j.geomorph.2018.08.020

Center for Hazards and Risk Research - CHRR - Columbia University, Center for International Earth Science Information Network - CIESIN - Columbia University, International Bank for Reconstruction and Development - The World Bank, and U. N. E. P. G. R. I. D. G.-U.-G. (2005). Global Cyclone Hazard Frequency and Distribution. Retrieved June 22, 2018, from http://dx.doi.org/10.7927/H4CZ353K

Chen, W., Wang, D., Huang, Y., Chen, L., Zhang, L., Wei, X., et al. (2017). Monitoring and analysis of coastal reclamation from 1995–2015 in Tianjin Binhai New Area, China. Scientific Reports, 7(1), 3850. https://doi.org/10.1038/s41598-017-04155-0

References

188

Cheong, S.-M. M., Silliman, B., Wong, P. P., van Wesenbeeck, B., Kim, C.-K. K., & Guannel, G. (2013). Coastal adaptation with ecological engineering. Nature Climate Change, 3(9), 787–791. https://doi.org/10.1038/nclimate1854

Chesapeake Bay Program. (2000). Chesapeake 2000. Manage, 1–13.

Chmura, G. L. (2003). Global carbon sequestration in tidal, saline wetland soils. Global Biogeochemical Cycles, 17(4). https://doi.org/10.1029/2002GB001917

Church, J. a., White, N. J., Konikow, L. F., Domingues, C. M., Graham Cogley, J., Rignot, E., et al. (2011). Revisiting the Earth’s sea-level and energy budgets from 1961 to 2008. Geophysical Research Letters, 38, 8. https://doi.org/10.1002/grl.50752

Coastal Protection and Restoration Authority of Louisiana. (2017). Louisiana’s Comprehensive Master Plan for a Sustainable Coast.

Coastal Wetlands Planning Protection and Restoration Act (CWPPRA). (n.d.). The Mississippi River Delta Basin. Retrieved August 17, 2017, from https://www.lacoast.gov/new/About/Basin_data/mr/Default.aspx

Coleman, J. M., & Huh, O. K. (2004). Major world deltas: A perspective from space. Retrieved from http://128.138.136.5/science/groups/wessman/projects/wdn/papers/colemanManuscript.pdf

Collins, J. M., & Roache, D. R. (2017). The 2016 North Atlantic hurricane season: A season of extremes. Geophysical Research Letters, 44(10), 5071–5077. https://doi.org/10.1002/2017GL073390

Van Coppenolle, R., Schwarz, C., & Temmerman, S. (2018). Contribution of Mangroves and Salt Marshes to Nature-Based Mitigation of Coastal Flood Risks in Major Deltas of the World. Estuaries and Coasts, 41(6), 1699–1711. https://doi.org/10.1007/s12237-018-0394-7

Costanza, R., Pérez-Maqueo, O., Martinez, M. L., Sutton, P., Anderson, S. J., & Mulder, K. (2008). The value of coastal wetlands for hurricane protection. Ambio, 37(4), 241–248. https://doi.org/10.1579/0044-7447(2008)37[241:tvocwf]2.0.co;2

Cui, B., Yang, Q., Yang, Z., & Zhang, K. (2009). Evaluating the ecological performance of wetland restoration in the Yellow River Delta, China. Ecological Engineering, 35(7), 1090–1103. https://doi.org/10.1016/j.ecoleng.2009.03.022

Dahdouh-Guebas, F., Jayatissa, L. P., Di Nitto, D., Bosire, J. O., Lo Seen, D., & Koedam, N. (2005). How effective were mangroves as a defence against the recent tsunami? Current Biology, 15(14), 1337–1338. https://doi.org/10.1016/j.cub.2005.07.025

Danielsen, F., Sørensen, M. K., Olwig, M. F., Selvam, V., Parish, F., Burgess, N. D., et al. (2005). The Asian Tsunami : A Protective Role for Coastal Vegetation. Science, 310(October), 643.

Darwiche-Criado, N., Sorando, R., Eismann, S. G., & Comín, F. A. (2017). Comparing Two Multi-Criteria Methods for Prioritizing Wetland Restoration and Creation Sites Based on Ecological, Biophysical and Socio-Economic Factors. Water Resources Management, 31(4), 1227–1241. https://doi.org/10.1007/s11269-017-1572-2

Das, S., & Vincent, J. R. (2009). Mangroves protected villages and reduced death toll during Indian super cyclone. Proceedings of the National Academy of Sciences of the United States of America, 106(18), 7357–7360. https://doi.org/10.1073/pnas.0810440106

Dasgupta, S., Laplante, B., Meisner, C., Wheeler, D., & Yan, J. (2009). The impact of sea level rise on developing countries: A comparative analysis. Climatic Change, 93(3–4), 379–388. https://doi.org/10.1007/s10584-008-9499-5

References

189

Dasgupta, S., Laplante, B., Murray, S., & Wheeler, D. (2011). Exposure of developing countries to sea-level rise and storm surges. Climatic Change, 106(4), 567–579. https://doi.org/10.1007/s10584-010-9959-6

Day, J. W., Boesch, D. F., Clairain, E. J., Kemp, G. P., Laska, S. B., Mitsch, W. J., et al. (2007). Restoration of the Mississippi Delta: lessons from Hurricanes Katrina and Rita. Science (New York, N.Y.), 315(5819), 1679–1684. https://doi.org/10.1126/science.1137030

Deb, M., & Ferreira, C. M. (2015). Potential impacts of the Sunderban mangrove degradation on future coastal flooding in Bangladesh. Journal of Hydro-Environment Research, 17, 30–46. https://doi.org/10.1016/j.jher.2016.11.005

Delgado, A. (2013). Guayaquil. Cities, 31, 515–532. https://doi.org/10.1016/j.cities.2011.11.001

Dewsbury, B. M., Bhat, M., & Fourqurean, J. W. (2016). A review of seagrass economic valuations: Gaps and progress in valuation approaches. Ecosystem Services, 18, 68–77. https://doi.org/10.1016/j.ecoser.2016.02.010

Dilley, M., Chen, R. S., Deichmann, U., Lerner-Lam, A. L., Arnold, M., Agwe, J., et al. (2005). Natural Disaster Hotspots A Global Risk Analysis. Earth Science. https://doi.org/10.1080/01944360902967228

Dobson, J. E., Bright, E. A., Coleman, P. R., Durfee, R. C., & Worley, B. A. (2000). LandScan: A global population database for estimating populations at risk. Photogrammetric Engineering and Remote Sensing, 66(7), 849–857. Retrieved from http://apps.webofknowledge.com.libezp.lib.lsu.edu/full_record.do?product=UA&search_mode=GeneralSearch&qid=3&SID=1D8vegIjNxsjNKm4nMs&page=1&doc=3

Duarte, C. M. (1991). Seagrass depth limits. Aquatic Botany, 40(4), 363–377. https://doi.org/10.1016/0304-3770(91)90081-F

Duarte, C. M., Losada, I. J., Hendriks, I. E., Mazarrasa, I., & Marbà, N. (2013). The role of coastal plant communities for climate change mitigation and adaptation. Nature Climate Change, 3(11), 961–968. https://doi.org/10.1038/nclimate1970

Duckworth, A., Giofre, N., & Jones, R. (2017). Coral morphology and sedimentation. Marine Pollution Bulletin, 125(1–2), 289–300. https://doi.org/10.1016/j.marpolbul.2017.08.036

Duke, N. C., Meynecke, J.-O., Dittmann, S., Ellison, A. M., Anger, K., Berger, U., et al. (2007). A World Without Mangroves? Science, 317(July), 41–43. https://doi.org/10.1126/science.317.5834.41b

Eertman, R. H. M., Kornman, B. A., Stikvoort, E., & Verbeek, H. (2002). Restoration of the sieperda tidal marsh in the Scheldt estuary, The Netherlands. Restoration Ecology, 10(3), 438–449. https://doi.org/10.1046/j.1526-100X.2002.01034.x

Elliott, M., Mander, L., Mazik, K., Simenstad, C., Valesini, F., Whitfield, A., & Wolanski, E. (2016). Ecoengineering with Ecohydrology: Successes and failures in estuarine restoration. Estuarine, Coastal and Shelf Science, 176, 12–35. https://doi.org/10.1016/j.ecss.2016.04.003

Ericson, J., Vorosmarty, C., Dingman, S., Ward, L., & Meybeck, M. (2006). Effective sea-level rise and deltas: Causes of change and human dimension implications. Global and Planetary Change, 50(1–2), 63–82. https://doi.org/10.1016/j.gloplacha.2005.07.004

References

190

Esteves, L. S. (2014). Managed Realignment : A Viable Long-Term Coastal Management Strategy? SpringerBriefs in Environmental Science. Dordrecht: Springer Netherlands. https://doi.org/10.1007/978-94-017-9029-1

FAO. (2007). Coastal protection in the aftermath of the Indian Ocean Tsunami: What role for forest and trees?

Feagin, R., Martinez, M. L., Mendoza-Gonzalez, G., & Costanza, R. (2010). Salt marsh zonal migration and ecosystem service change in response to global sea level rise: A case study from an urban region. Ecology and Society, 15(4). https://doi.org/14

Feagin, R. A., Mukherjee, N., Shanker, K., Baird, A. H., Cinner, J., Kerr, A. M., et al. (2010). Shelter from the storm? Use and misuse of coastal vegetation bioshields for managing natural disasters. Conservation Letters, 3(1), 1–11. https://doi.org/10.1111/j.1755-263X.2009.00087.x

Federal Geographic Data Committee. (2013). Classification of wetlands and deepwater habitats of the United States. FGDC-STD-004-2013. Second Edition, (August), 79. https://doi.org/FWS/OBS-79/31

Ferrario, F., Beck, M. W., Storlazzi, C. D., Micheli, F., Shepard, C. C., & Airoldi, L. (2014). The effectiveness of coral reefs for coastal hazard risk reduction and adaptation. Nature Communications, 5(May), 3794. https://doi.org/10.1038/ncomms4794

Firth, L. B., Thompson, R. C., Bohn, K., Abbiati, M., Airoldi, L., Bouma, T. J., et al. (2014). Between a rock and a hard place: Environmental and engineering considerations when designing coastal defence structures. Coastal Engineering, 87, 122–135. https://doi.org/10.1016/j.coastaleng.2013.10.015

Flather, R. A. (2001). Storm surges. Encyclopedia of Ocean Sciences, 2882–2892. https://doi.org/10.1016/B978-012374473-9.00124-7

Fonseca, M. S., & Cahalan, J. A. (1992). A preliminary evaluation of wave attenuation by four species of seagrass. Estuarine, Coastal and Shelf Science, 35(6), 565–576. https://doi.org/10.1016/S0272-7714(05)80039-3

Food and Agriculture Organization (FAO) of the United Nations. (2007). The World’s Mangroves 1980-2005. (FAO, Ed.). Rome. Retrieved from https://books.google.be/books?id=tLdlpOiuSmEC&pg=PA84&lpg=PA84&dq=978-92-5-105856-5&source=bl&ots=0yQSkC15h5&sig=4b8DAqZkt7MdXGHZ-J3q_7wrTeA&hl=fr&sa=X&ved=0ahUKEwiizqiL24naAhVM1RQKHbcYCfMQ6AEIQTAE#v=onepage&q=978-92-5-105856-5&f=false

Fourqurean, J. W., Duarte, C. M., Kennedy, H., Marbà, N., Holmer, M., Mateo, M. A., et al. (2012). Seagrass ecosystems as a globally significant carbon stock. Nature Geoscience, 5(7), 505–509. https://doi.org/10.1038/ngeo1477

French, P. W. (2006). Managed realignment - The developing story of a comparatively new approach to soft engineering. Estuarine, Coastal and Shelf Science, 67(3), 409–423. https://doi.org/10.1016/j.ecss.2005.11.035

Frihy, O., Fanos, A., Khafagy, A., & Aesha, K. (1996). Human impacts on the coastal zone of Hurghada, northern Red Sea, Egypt. Geo-Marine Letters, 1993(December 1993), 324–329.

Ganong, W. F. (1903). The Vegetation of the Bay of Fundy Salt and Diked Marshes: An Ecological Study. Botanical Gazette, 36(3), 161–186. https://doi.org/10.1086/676943

References

191

Garbutt, A., de Groot, A., Smit, C., & Pétillon, J. (2017). European salt marshes: ecology and conservation in a changing world. Journal of Coastal Conservation, 21(3), 405–408. https://doi.org/10.1007/s11852-017-0524-6

Garbutt, R. A., Reading, C. J., Wolters, M., Gray, A. J., & Rothery, P. (2006). Monitoring the development of intertidal habitats on former agricultural land after the managed realignment of coastal defences at Tollesbury, Essex, UK. Marine Pollution Bulletin, 53(1), 155–164. https://doi.org/10.1016/j.marpolbul.2005.09.015

Gardiner, S., Hanson, S., Nicholls, R. J., Zhang, Z., Jude, S., Jones, A., et al. (2007). The Habitats Directive, coastal habitats and climate change - case studies from the south coast of the U.K. In Proceedings of the International Conference on Coastal Management (pp. 1–10). Retrieved from http://eprints.soton.ac.uk/53142/

Gedan, K. B., Kirwan, M. L., Wolanski, E., Barbier, E. B., & Silliman, B. R. (2011). The present and future role of coastal wetland vegetation in protecting shorelines: Answering recent challenges to the paradigm. Climatic Change, 106(1), 7–29. https://doi.org/10.1007/s10584-010-0003-7

Gilman, E. L., Ellison, J., Duke, N. C., & Field, C. (2008). Threats to mangroves from climate change and adaptation options: A review. Aquatic Botany, 89(2), 237–250. https://doi.org/10.1016/j.aquabot.2007.12.009

Giosan, L., Syvitski, J. P. M., Constantinescu, S., & Day, J. (2014). Protect the world’s deltas. Nature, 516, 31–33.

Giri, C., Ochieng, E., Tieszen, L. L. L., Zhu, Z., Singh, A., Loveland, T., et al. (2011). Status and distribution of mangrove forests of the world using earth observation satellite data. Global Ecology and Biogeography, 20(2011), 154–159. https://doi.org/10.1111/j.1466-8238.2010.00584.x

Von Glasow, R., Jickells, T. D., Baklanov, A., Carmichael, G. R., Church, T. M., Gallardo, L., et al. (2013). Megacities and large urban agglomerations in the coastal zone: Interactions between atmosphere, land, and marine ecosystems. Ambio, 42(1), 13–28. https://doi.org/10.1007/s13280-012-0343-9

Goeldner-Gianella, L. (2008). Perceptions and attitudes towards de-polderisation in Europe : a comparison of five opinion surveys. Journal of Coastal R, 23(5), 1218–1230.

Grabowski, J. H., Brumbaugh, R. D., Conrad, R. F., Keeler, A. G., Opaluch, J. J., Peterson, C. H., et al. (2012). Economic Valuation of Ecosystem Services Provided by Oyster Reefs. BioScience, 62(10), 900–909. https://doi.org/10.1525/bio.2012.62.10.10

Green, E. P., & Short, F. (2003). World atlas of seagrasses. Prepared by the UIMEP World Conservation Monitoring Centre. University of California Press, Berkeley, USA. (Vol. 47). https://doi.org/10.1515/BOT.2004.029

Griggs, G. B. (2005). The Impacts of Coastal Armoring. Shore & Beach, 73(13), 13–22.

Grobicki, A., Chalmers, C., Jennings, E., Jones, T., & Peck, D. (2016). An Introduction to the RAMSAR Convention on Wetlands. Ramsar Convention Secretariat (7th ed. (p). Gland, Switzerland. Retrieved from https://www.mendeley.com/viewer/?fileId=658dea6f-a739-d995-65b0-e9d2a5a9f4c6&documentId=158ad1b6-0317-3653-ac6d-fb0139511534

References

192

De Groot, R. S., Blignaut, J., Van Der Ploeg, S., Aronson, J., Elmqvist, T., & Farley, J. (2013). Benefits of Investing in Ecosystem Restoration. Conservation Biology, 27(6), 1286–1293. https://doi.org/10.1111/cobi.12158

Guannel, G., Arkema, K., Ruggiero, P., & Verutes, G. (2016). The power of three: Coral reefs, seagrasses and mangroves protect coastal regions and increase their resilience. PLoS ONE, 11(7), 1–22. https://doi.org/10.1371/journal.pone.0158094

Guzmán, J. M., Martine, G., McGranahan, G., Schensul, D., & Tacoli, C. (2009). Population Dynamics and Climate Change. (J. M. Guzmán, G. Martine, G. McGranahan, D. Schensul, & C. Tacoli, Eds.).

Haas, J., Furberg, D., & Ban, Y. (2015). Satellite monitoring of urbanization and environmental impacts—A comparison of Stockholm and Shanghai. International Journal of Applied Earth Observation and Geoinformation, 38, 138–149. https://doi.org/10.1016/j.jag.2014.12.008

de Haas, T., Pierik, H. J., van der Spek, A. J. F., Cohen, K. M., van Maanen, B., & Kleinhans, M. G. (2018). Holocene evolution of tidal systems in The Netherlands: Effects of rivers, coastal boundary conditions, eco-engineering species, inherited relief and human interference. Earth-Science Reviews, 177(October 2017), 139–163. https://doi.org/10.1016/j.earscirev.2017.10.006

Haddad, J., Lawler, S., & Ferreira, C. M. (2016). Assessing the relevance of wetlands for storm surge protection: a coupled hydrodynamic and geospatial framework. Natural Hazards, 80(2), 839–861. https://doi.org/10.1007/s11069-015-2000-7

Hallegatte, S., Green, C., Nicholls, R. J., & Corfee-Morlot, J. (2013). Future flood losses in major coastal cities. Nature Climate Change, 3(9), 802–806. https://doi.org/10.1038/nclimate1979

Haltiner, J. B. ., Boyer, K. E., Williams, G. D., Callaway, J. C., & Zedler, J. (1997). Influence of physical processes on the design, functioning and evolution of restored tidal wetlands in California (USA). Wetlands Ecology and Management, 4(2), 73–91. https://doi.org/10.1007/BF01876230

Hamburg Port Authority. (2006). Concept for a sustainable development of the Tidal Elbe River as an artery of the metropolitan region Hamburg and beyond, 20.

Hanley, M. E., Hoggart, S. P. G., Simmonds, D. J., Bichot, A., Colangelo, M. A., Bozzeda, F., et al. (2014). Shifting sands? Coastal protection by sand banks, beaches and dunes. Coastal Engineering, 87, 136–146. https://doi.org/10.1016/j.coastaleng.2013.10.020

Hansen, K. (2015). Ecosystem functions of tidal marsh soils of the Elbe estuary. University of Hamburg.

Hanson, S., Nicholls, R., Ranger, N., Hallegatte, S., Corfee-Morlot, J., Herweijer, C., et al. (2011). A global ranking of port cities with high exposure to climate extremes. Climatic Change, 104(1), 89–111. https://doi.org/10.1007/s10584-010-9977-4

Harris, D. L., Rovere, A., Casella, E., Power, H., Canavesio, R., Collin, A., et al. (2018). Coral reef structural complexity provides important coastal protection from waves under rising sea levels. Science Advances, 4(2), 1–8. https://doi.org/10.1126/sciadv.aao4350

Hartman, B. D., & Cleveland, D. A. (2018). The socioeconomic factors that facilitate or constrain restoration management: Watershed rehabilitation and wet meadow (bofedal) restoration in the Bolivian Andes. Journal of Environmental Management, 209, 93–104. https://doi.org/10.1016/j.jenvman.2017.12.025

References

193

Hatvany, M. G. (2003). Marshlands: Four centuries of environmental change in the shores of the St. Lawrence. Sainte-Foy: Les Presses de l’Université Laval. Retrieved from https://books.google.be/books?hl=fr&lr=&id=-Hk9eDMLFFkC&oi=fnd&pg=PR17&dq=Marshlands:+Four+centuries+of+environmental+change++in++the++shores++of++the++St.++Lawrence.&ots=g8KIwhVDaK&sig=NPJNqp9LCNN_io912s7jJpcCTB8#v=onepage&q&f=false

He, Y., & Zhang, M. (2001). Study on wetland loss and its reasons in China. Chinese Geographical Science, 11(3), 241–245. https://doi.org/10.1007/s13398-014-0173-7.2

Heap, A., Bryce, S., Ryan, D., Radke, L., Smith, C., Smith, R., et al. (2001). Australian estuaries & coastal waterways: a geoscience perspective for improved and integrated resource management. Australian Geological Survey Organisation. Retrieved from http://www.ga.gov.au/metadata-gateway/metadata/record/gcat_36203

Hill, K. (2015). Coastal infrastructure: A typology for the next century of adaptation to sea-level rise. Frontiers in Ecology and the Environment, 13(9), 468–476. https://doi.org/10.1890/150088

Himes-Cornell, A., Pendleton, L., & Atiyah, P. (2018). Valuing ecosystem services from blue forests: A systematic review of the valuation of salt marshes, sea grass beds and mangrove forests. Ecosystem Services, 30, 36–48. https://doi.org/10.1016/j.ecoser.2018.01.006

Hinkel, J., Lincke, D., Vafeidis, A. T., Perrette, M., Nicholls, R. J., Tol, R. S. J., et al. (2014). Coastal flood damage and adaptation costs under 21st century sea-level rise. Proceedings of the National Academy of Sciences of the United States of America, 66(9), 3292–7. https://doi.org/10.1002/elsc.201300165

Hobbs, R. J., Higgs, E., & Harris, J. A. (2009). Novel ecosystems : implications for conservation and restoration, (August), 599–605. https://doi.org/10.1016/j.tree.2009.05.012

Hoegh-Guldberg, O., Mumby, P. J., Hooten, A. J., Steneck, R. S., Greenfield, P., Gomez, E., et al. (2007). Coral Reefs Under Rapid Climate Change and Ocean Acidification. Science, 318(5857), 1737–1742. https://doi.org/10.1126/science.1152509

Hoeksema, R. J. (2007). Three stages in the history of land reclamation in the Netherlands. Irrigation and Drainage, 56(S1), S113–S126. https://doi.org/10.1002/ird.340

Hoggart, S., Hawkins, S. J., Bohn, K., Airoldi, L., van Belzen, J., Bichot, A., et al. (2015). Ecological Approaches to Coastal Risk Mitigation. Coastal Risk Management in a Changing Climate. https://doi.org/10.1016/B978-0-12-397310-8.00004-X

Hong, P. N. (2001). Reforestation of mangroves after severe impacts of herbicides during the Vietnam war: the case of Can Gio. FAO Unasylva, 207(52), 57–60. Retrieved from http://www.fao.org/docrep/004/y2795f/y2795f11.htm#o

Van Hooidonk, R., Maynard, J., Tamelander, J., Gove, J., Ahmadia, G., Raymundo, L., et al. (2016). Local-scale projections of coral reef futures and implications of the Paris Agreement. Scientific Reports, 6(December), 1–8. https://doi.org/10.1038/srep39666

Hu, K., Chen, Q., & Wang, H. (2015). A numerical study of vegetation impact on reducing storm surge by wetlands in a semi-enclosed estuary. Coastal Engineering, 95, 66–76. https://doi.org/10.1016/j.coastaleng.2014.09.008

References

194

Huguet, J. R., Bertin, X., & Arnaud, G. (2017). Managed realignment to mitigate storm-induced flooding: A case study in La Faute-sur-mer, France. Coastal Engineering, (December 2016), 1–9. https://doi.org/10.1016/j.coastaleng.2017.08.010

IPCC. (2007). Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. In S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, et al. (Eds.), Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press. Retrieved from https://www.ipcc.unibe.ch/publications/wg1-ar4/faq/docs/AR4WG1_FAQ-Brochure_LoRes.pdf

IPCC. (2013). Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press (Vol. AR5). Cambridge, United Kingdom and New York, NY, USA.

Jiang, T. ting, Pan, J. fen, Pu, X. M., Wang, B., & Pan, J. J. (2015). Current status of coastal wetlands in China: Degradation, restoration, and future management. Estuarine, Coastal and Shelf Science, 164, 265–275. https://doi.org/10.1016/j.ecss.2015.07.046

Käkönen, M. (2008). Mekong Delta at the crossroads: more control or adaptation? Ambio, 37(3), 205–212. https://doi.org/10.1579/0044-7447(2008)37[205:MDATCM]2.0.CO;2

King, S. E., & Lester, J. N. (1995). The value of salt marsh as a sea defence. Marine Pollution Bulletin, 30(3), 180–189. https://doi.org/10.1016/0025-326X(94)00173-7

Kirwan, M. L., & Megonigal, J. P. (2013). Tidal wetland stability in the face of human impacts and sea-level rise. Nature, 504(7478), 53–60. https://doi.org/10.1038/nature12856

Kirwan, M. L., Guntenspergen, G. R., D’Alpaos, A., Morris, J. T., Mudd, S. M., & Temmerman, S. (2010). Limits on the adaptability of coastal marshes to rising sea level. Geophysical Research Letters, 37(23), 1–5. https://doi.org/10.1029/2010GL045489

Kirwan, M. L., Temmerman, S., Skeehan, E. E., Guntenspergen, G. R., & Faghe, S. (2016). Overestimation of marsh vulnerability to sea level rise. Nature Climate Change, 6(3), 253–260. https://doi.org/10.1038/nclimate2909

Knapp, K. R., Kruk, M. C., Levinson, D. H., Diamond, H. J., & Neumann, C. J. (2010). The International Best Track Archive for Climate Stewardship (IBTrACS). Bulletin of the American Meteorological Society, 91(3), 363–376. https://doi.org/10.1175/2009BAMS2755.1

Knutson, T. R. (2014). Tropical Cyclones and Hurricanes: Tropical Cyclones and Climate Change. Encyclopedia of Atmospheric Sciences: Second Edition (Second Edi, Vol. 6). Elsevier. https://doi.org/10.1016/B978-0-12-382225-3.00508-9

Knutson, T. R., McBride, J. L., Chan, J., Emanuel, K., Holland, G., Landsea, C., et al. (2010). Tropical cyclones and climate change. Nature Geoscience, 3(February), 157–163. https://doi.org/10.1038/ngeo779

Kobashi, D., & Mazda, Y. (2005). Tidal flow in riverine-type mangroves. Wetlands Ecology and Management, 13(6), 615–619. https://doi.org/10.1007/s11273-004-3481-4

References

195

Koch, E. W., Ackerman, J. D., Verduin, J., & Kuelen, M. Van. (2006). Fluid Dynamics in Seagrass Ecology—from Molecules to Ecosystems. Seagrasses: Biology, Ecology and Conservation, 193–225. Retrieved from http://link.springer.com/chapter/10.1007/978-1-4020-2983-7_8%5Cnhttp://books.google.com/books?id=quMJfqMSFGAC&pg=PR7&lpg=PR7&dq=Fluid+dynamics+in+seagrass+ecology:+from+molecules+to+ecosystems&source=bl&ots=-ZROD5r85p&sig=TW4jGM91lqAXzUJ5H2G2h8harZY&hl=en&s

Koch, E. W., Barbier, E. B., Silliman, B. R., Reed, D. J., Perillo, G. M. E., Hacker, S. D., et al. (2009). Non-linearity in ecosystem services: Temporal and spatial variability in coastal protection. Frontiers in Ecology and the Environment, 7(1), 29–37. https://doi.org/10.1890/080126

Komar, P. D. (1998). Wave erosion of a massive artificial coastal landslide. Earth Surface Processes and Landforms, 23(5), 415–428. https://doi.org/10.1002/(SICI)1096-9837(199805)23:5<415::AID-ESP855>3.0.CO;2-T

Krauss, K. W., Doyle, T. J. T. W., Doyle, T. J. T. W., Swarzenski, C. M., From, A. S., Day, R. H., & Conner, W. H. (2009). Water level observations in mangrove swamps during two hurricanes in Florida. Wetlands, 29(1), 142–149. https://doi.org/10.1672/07-232.1

Krauss, K. W., McKee, K. L., Lovelock, C. E., Cahoon, D. R., Saintilan, N., Reef, R., & Chen, L. (2014). How mangrove forests adjust to rising sea level. The New Phytologist, 202(1), 19–34. https://doi.org/10.1111/nph.12605

Kron, W. (2013). Coasts: The high-risk areas of the world. Natural Hazards, 66(3), 1363–1382. https://doi.org/10.1007/s11069-012-0215-4

Landry, C. E., & Liu, H. (2009). A semi-parametric estimator for revealed and stated preference data-An application to recreational beach visitation. Journal of Environmental Economics and Management, 57(2), 205–218. https://doi.org/10.1016/j.jeem.2008.05.002

Lawler, S., Haddad, J., & Ferreira, C. M. (2016). Sensitivity considerations and the impact of spatial scaling for storm surge modeling in wetlands of the Mid-Atlantic region. Ocean and Coastal Management, 134, 226–238. https://doi.org/10.1016/j.ocecoaman.2016.10.008

Lawrence, P. J., Smith, G. R., Sullivan, M. J. P., & Mossman, H. L. (2018). Restored saltmarshes lack the topographic diversity found in natural habitat. Ecological Engineering, 115(February), 58–66. https://doi.org/10.1016/j.ecoleng.2018.02.007

Ledoux, L., Cornell, S., O’Riordan, T., Harvey, R., & Banyard, L. (2004). Towards sustainable flood and coastal management: Identifying drivers of, and obstacles to, managed realignment. Working Paper - Centre for Social and Economic Research on the Global Environment, 22(1), 1–32. https://doi.org/10.1016/j.landusepol.2004.03.001

Lee, S.-M., Cho, Y.-C., & Lee, C.-S. (2012). Feasibility of seed bank for restoration of salt marsh: a case study around the Gwangyang Bay, southern Korea. Journal of Ecology and Field Biology, 35(2), 123–129. https://doi.org/10.5141/JEFB.2012.016

Lee, S. Y., Primavera, J. H., Dahdouh-Guebas, F., Mckee, K., Bosire, J. O., Cannicci, S., et al. (2014). Ecological role and services of tropical mangrove ecosystems: A reassessment. Global Ecology and Biogeography, 23(7), 726–743. https://doi.org/10.1111/geb.12155

References

196

Leonardi, N., Carnacina, I., Donatelli, C., Ganju, N. K., Plater, A. J., Schuerch, M., & Temmerman, S. (2018). Dynamic interactions between coastal storms and salt marshes: A review. Geomorphology, 301, 92–107. https://doi.org/10.1016/j.geomorph.2017.11.001

Lewis, R. R., & Brown, B. (2014). Ecological Mangrove Rehabilitation A field manual for practitioners, 275.

Li, M. S., & Lee, S. Y. (1997). Mangroves of China: A brief review. Forest Ecology and Management, 96(3), 241–259. https://doi.org/10.1016/S0378-1127(97)00054-6

Lichter, M., Vafeidis, A. T., Nicholls, R. J., & Kaiser, G. (2011). Exploring data-related uncertainties in analyses of land area and population in the “Low-Elevation Coastal Zone” (LECZ). Journal of Coastal Research, 27(4), 757–768. https://doi.org/10.2112/JCOASTRES-D-10-00072.1

Lim, Y.-K., Schubert, S. D., Kovach, R., Molod, A. M., & Pawson, S. (2018). The Roles of Climate Change and Climate Variability in the 2017 Atlantic Hurricane Season Global Modeling &amp; Assimilation O ffice, 2018. Retrieved from https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20180002568.pdf

Liu, H., Zhang, K., Li, Y., & Xie, L. (2013). Numerical study of the sensitivity of mangroves in reducing storm surge and flooding to hurricane characteristics in southern Florida. Continental Shelf Research, 64, 51–65. https://doi.org/10.1016/j.csr.2013.05.015

Loder, N. M., Irish, J. L., Cialone, M. a., & Wamsley, T. V. (2009). Sensitivity of hurricane surge to morphological parameters of coastal wetlands. Estuarine, Coastal and Shelf Science, 84(4), 625–636. https://doi.org/10.1016/j.ecss.2009.07.036

De los Santos, C. B., Godbold, J. A., & Solan, M. (2017). Short-term growth and biomechanical responses of the temperate seagrass Cymodocea nodosa to CO2enrichment. Marine Ecology Progress Series, 572, 91–102. https://doi.org/10.3354/meps12153

Lotze, H. K., Lenihan, H. S., Bourque, B. J., Bradbury, R. H., Richard G. Cooke, Kay, M. C., et al. (2006). Depletion, Degradation, and Recovery Potential of Estuaries and Coastal Seas. Science, 312(5781), 1806–1809. https://doi.org/10.1126/science.1128035

Lovelace, J. K. (1994). Storm-Tide Elevations Produced By Hurricane Andrew Along the Louisiana Coast, August 25-27,1992. U.S. Geological Survey Open-File Report 94-371. Prepared in Cooperation with the Federal Emergency Management Agency.

Lovelock, C. E., Cahoon, D. R., Friess, D. A., Guntenspergen, G. R., Krauss, K. W., Reef, R., et al. (2015). The vulnerability of Indo-Pacific mangrove forests to sea-level rise. Nature. https://doi.org/10.1038/nature15538

Ma, Z., Melville, D. S., Liu, J., Chen, Y., Yang, H., Ren, W., et al. (2014). ECOSYSTEMS MANAGEMENT Rethinking China’s new great wall. Science, 346(6212), 912–914. https://doi.org/10.1126/science.1257258

Maddison, A. (2001). The World Economy: A Millennial Perspective. Oecd, 384. https://doi.org/10.1787/9789264189980-en

Marchand, M. (2008). Mangrove restoration in Vietnam, 34.

Maris, T., Cox, T., Temmerman, S., De Vleeschauwer, P., Van Damme, S., De Mulder, T., et al. (2007). Tuning the tide: Creating ecological conditions for tidal marsh development in a flood control area. In Hydrobiologia (Vol. 588, pp. 31–43). https://doi.org/10.1007/s10750-007-0650-5

References

197

Marois, D. E., & Mitsch, W. J. (2015). Coastal protection from tsunamis and cyclones provided by mangrove wetlands – a review. International Journal of Biodiversity Science, Ecosystem Services & Management, 11(June 2015), 1–13. https://doi.org/10.1080/21513732.2014.997292

Marsooli, R., Orton, P. M., Georgas, N., & Blumberg, A. F. (2016). Three-dimensional hydrodynamic modeling of coastal flood mitigation by wetlands. Coastal Engineering, 111, 83–94. https://doi.org/10.1016/j.coastaleng.2016.01.012

Masters, J. (2012). A Detailed View of the Storm Surge: Comparing Katrina to Camille. Retrieved from https://www.wunderground.com/hurricane/surge_details.asp

Mathieu, L. F., Langford, I. H., & Kenyon, W. (2003). Valuing marine parks in a developing country: A case study of the Seychelles. Environment and Development Economics, 8(2), 373–390. https://doi.org/10.1017/S1355770X03000196

Mattocks, C., & Forbes, C. (2008). A real-time, event-triggered storm surge forecasting system for the state of North Carolina. Ocean Modelling, 25(3–4), 95–119. https://doi.org/10.1016/j.ocemod.2008.06.008

Mazda, Y., Wolanski, E., King, B., Sase, A., Ohtsuka, D., & Magi, M. (1997). Drag force due to vegetation in mangrove swamps. Mangroves and Salt Marshes, 1, 193–199. https://doi.org/10.1023/A:1009949411068

Mazda, Y., Magi, M., Ikeda, Y., Kurokawa, T., & Asano, T. (2006). Wave reduction in a mangrove forest dominated by Sonneratia sp. Wetlands Ecology and Management, 14(4), 365–378. https://doi.org/10.1007/s11273-005-5388-0

McArthur, L. C., & Boland, J. W. (2006). The economic contribution of seagrass to secondary production in South Australia. Ecological Modelling, 196(1–2). https://doi.org/10.1016/j.ecolmodel.2006.02.030

McEvedy, C., & Jones, R. (1978). Atlas of World Population History. New York: NY:Facts on File.

McGee, B. D., Goree, B. B., Tollett, R. W., Woodward, B. K., & Kress, W. H. (2006). Hurrican Rita Surge Data, Southwestern Louisiana and Souteasthern Texas, September to November 2005. U.S. Geological Survey Data Series 220. Retrieved from http://pubs.usgs.gov/ds/2006/220/index.htm

Mcgranahan, G., Balk, D., & Anderson, B. (2006). Low coastal zone settlements, (59), 23–26.

McGranahan, G., Balk, D., & Anderson, B. (2007). The rising tide: assessing the risks of climate change and human settlements in low elevation coastal zones. Environment and Urbanization, 19(1), 17–37. https://doi.org/10.1177/0956247807076960

McIvor, A. L., Möller, I., Spencer, T., & Spalding, M. (2012). Reduction of Wind and Swell Waves by Mangroves. Natural Coastal Protection Series, 1–27. https://doi.org/ISSN 2050-7941.

McIvor, A. L., Spencer, T., Möller, I., & Spalding, M. (2012). Storm Surge Reduction by Mangroves. Natural Coastal Protection Series, 35. Retrieved from http://www.naturalcoastalprotection.org/documents/storm-surge-reduction-by-mangroves

McIvor, A. L., Spencer, T., & Möller, I. (2013). The response of mangrove soil surface elevation to sea level rise. Natural Coastal Protection Series, 1–59.

References

198

McLeod, E., Chmura, G. L., Bouillon, S., Salm, R., Björk, M., Duarte, C. M., et al. (2011). A blueprint for blue carbon: Toward an improved understanding of the role of vegetated coastal habitats in sequestering CO2. Frontiers in Ecology and the Environment, 9(10), 552–560. https://doi.org/10.1890/110004

Mcowen, C., Weatherdon, L., Bochove, J.-W., Sullivan, E., Blyth, S., Zockler, C., et al. (2017). A global map of saltmarshes. Biodiversity Data Journal, 5, e11764. https://doi.org/10.3897/BDJ.5.e11764

Meire, P., Dauwe, W., Maris, T., Peeters, P., Coen, L., Deschamps, M., et al. (2014). Sigma Plan Proves Efficiency. ECSA Bulletin, 62, 19–23.

Meng, W., Hu, B., He, M., Liu, B., Mo, X., Li, H., et al. (2017). Temporal-spatial variations and driving factors analysis of coastal reclamation in China. Estuarine, Coastal and Shelf Science, 191, 39–49. https://doi.org/10.1016/j.ecss.2017.04.008

MFF Pakistan. (2016). A Handbook on Pakistan’s Coastal and Marine Resources. Pakistan: MFF Pakistan.

Middleton, B. A. (2016). Differences in impacts of Hurricane Sandy on freshwater swamps on the Delmarva Peninsula, Mid-Atlantic Coast, USA. Ecological Engineering, 87, 62–70. https://doi.org/10.1016/j.ecoleng.2015.11.035

Millennium Ecosystem Assessment. (2005). ECOSYSTEMS AND HUMAN WELL-BEING: WETLANDS AND WATER Synthesis. (World Resources Institute, Ed.). Washington, DC.

Moller, I., Spencer, T., French, Leggett, D., & Dixon, M. (1999). Wave transformation over salt marshes: A field and numerical modelling study from north Norfolk, England, 411–426. https://doi.org/10.1006/ecss.1999.0509

Möller, I., Kudella, M., Rupprecht, F., Spencer, T., Paul, M., van Wesenbeeck, B. K., et al. (2014). Wave attenuation over coastal salt marshes under storm surge conditions. Nature Geoscience, 7(10), 727–731. https://doi.org/10.1038/ngeo2251

Morris, J. T., Sundareshwar, P. V., Nietch, C. T., Kjerfve, B., & Cahoon, D. R. (2002). Responses of Coastal Wetlands to Rising Sea Level. Ecology, 83(10), 2869. https://doi.org/10.2307/3072022

Morris, R. L., Konlechner, T. M., Ghisalberti, M., & Swearer, S. E. (2018). From grey to green: Efficacy of eco-engineering solutions for nature-based coastal defence. Global Change Biology, 24(5), 1827–1842. https://doi.org/10.1111/gcb.14063

Muis, S., Verlaan, M., Winsemius, H. C., Aerts, J. C. J. H., & Ward, P. J. (2016). A global reanalysis of storm surges and extreme sea levels. Nature Communications, 7(May), 11969. https://doi.org/10.1038/ncomms11969

Muis, S., Verlaan, M., Nicholls, R. J., Brown, S., Hinkel, J., Lincke, D., et al. (2016). An intercomparison of two datasets of extreme sea levels and resulting global flood risk. Earth’s Future. https://doi.org/10.1002/eft2.197

Mumby, P. J., & Steneck, R. S. (2018). Paradigm Lost: Dynamic Nutrients and Missing Detritus on Coral Reefs. BioScience, 68(7), 487–495. https://doi.org/10.1093/biosci/biy055

Murray, A. (2017). Natural Flood Management. Adopting ecosystem approaches to managing flood risk, (February). Retrieved from https://www.foe.ie/download/pdf/natural_flood_management_a_study_for_friends_of_the_earth_february_2017.pdf

References

199

Murray, N. J., Clemens, R. S., Phinn, S. R., Possingham, H. P., & Fuller, R. A. (2014). Tracking the rapid loss of tidal wetlands in the Yellow Sea. Frontiers in Ecology and the Environment, 12(5), 267–272. https://doi.org/10.1890/130260

Narayan, S., Beck, M. W., Reguero, B. G., Losada, I. J., Van Wesenbeeck, B., Pontee, N., et al. (2016). The effectiveness, costs and coastal protection benefits of natural and nature-based defences. PLoS ONE, 11(5), 1–17. https://doi.org/10.1371/journal.pone.0154735

Narayan, S., Beck, M. W., Wilson, P., Thomas, C. J., Guerrero, A., Shepard, C. C., et al. (2017). The Value of Coastal Wetlands for Flood Damage Reduction in the Northeastern USA. Scientific Reports, 7(1), 9463. https://doi.org/10.1038/s41598-017-09269-z

Neumann, B., Vafeidis, A. T., Zimmermann, J., & Nicholls, R. J. (2015). Future Coastal Population Growth and Exposure to Sea-Level Rise and Coastal Flooding - A Global Assessment. PLoS ONE, 10(3), e0118571. https://doi.org/10.1371/journal.pone.0118571

Newman, W. S. (1982). Isostatic adjustment. In Beaches and Coastal Geology (pp. 497–498). Boston, MA: Springer US. https://doi.org/10.1007/0-387-30843-1_244

Nicholls, R. J., & Cazenave, A. (2010). Sea Level Rise and Its Impact on Coastal Zones. Science, 328(2010), 1517–1520. https://doi.org/10.1126/science.1185782

Nicholls, R. J., & Small, C. (2002). Improved estimates of coastal population and exposure to hazards released. Eos, 8(2), 301–305. https://doi.org/10.1029/2002EO000216

Nicholls, R. J., Hanson, S., Herweijer, C., Patmore, N., Hallegatte, S., Corfee-Morlot, J., et al. (2007). Ranking of The World’s Cities Most Exposed to Coastal Flooding Today Executive Summary. Organisation for Economic Cooperation and Development.

Nicholls, R. J., Hanson, S., Herweijer, C., Patmore, N., Hallegatte, S., Corfee-Morlot, J., et al. (2008). Ranking port cities with high exposure and vulnerability to climate extremes: exposure estimates. OECD Environment Working Papers (Vol. 1). https://doi.org/10.1787/011766488208

Nicholls, R. J., Brown, S., Goodwin, P., Wahl, T., Lowe, J., Solan, M., et al. (2018). Stabilization of global temperature at 1.5°C and 2.0°C: implications for coastal areas. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 376(2119), 20160448. https://doi.org/10.1098/rsta.2016.0448

Niu, Z. G., Gong, P., Cheng, X., Guo, J. H., Wang, L., Huang, H. B., et al. (2009). Geographical characteristics of China’s wetlands derived from remotely sensed data. Science in China, Series D: Earth Sciences, 52(6), 723–738. https://doi.org/10.1007/s11430-009-0075-2

O’Brien, K. R., Waycott, M., Maxwell, P., Kendrick, G. A., Udy, J. W., Ferguson, A. J. P., et al. (2017). Seagrass ecosystem trajectory depends on the relative timescales of resistance, recovery and disturbance. Marine Pollution Bulletin, (September), 0–1. https://doi.org/10.1016/j.marpolbul.2017.09.006

Ondiviela, B., Losada, I. J., Lara, J. L., Maza, M., Galván, C., Bouma, T. J., & van Belzen, J. (2014). The role of seagrasses in coastal protection in a changing climate. Coastal Engineering, 87, 158–168. https://doi.org/10.1016/j.coastaleng.2013.11.005

References

200

Oosterlee, L., Cox, T. J. S., Vandenbruwaene, W., Maris, T., Temmerman, S., & Meire, P. (2018). Tidal Marsh Restoration Design Affects Feedbacks Between Inundation and Elevation Change. Estuaries and Coasts, 41(3), 613–625. https://doi.org/10.1007/s12237-017-0314-2

Parés-Ramos, I., Álvarez-Berríos, N., & Aide, T. (2013). Mapping Urbanization Dynamics in Major Cities of Colombia, Ecuador, Per, and Bolivia Using Night-Time Satellite Imagery. Land, 2(1), 37–59. https://doi.org/10.3390/land2010037

Pascal, N., Allenbach, M., Brathwaite, A., Burke, L., Le Port, G., & Clua, E. (2016). Economic valuation of coral reef ecosystem service of coastal protection: A pragmatic approach. Ecosystem Services, 21, 72–80. https://doi.org/10.1016/j.ecoser.2016.07.005

Passeri, D. L., Long, J. W., Plant, N. G., Bilskie, M. V., & Hagen, S. C. (2018). The influence of bed friction variability due to land cover on storm-driven barrier island morphodynamics. Coastal Engineering, 132(February 2017), 82–94. https://doi.org/10.1016/j.coastaleng.2017.11.005

Paul, B. K. (2009). Why relatively fewer people died? The case of Bangladesh’s cyclone sidr. Natural Hazards, 50(2), 289–304. https://doi.org/10.1007/s11069-008-9340-5

Pendle, M. (2013). Estuarine and coastal managed realignment sites in England selected case studies.

Pendleton, L., Donato, D. C., Murray, B. C., Crooks, S., Jenkins, W. A., Sifleet, S., et al. (2012). Estimating Global “Blue Carbon” Emissions from Conversion and Degradation of Vegetated Coastal Ecosystems. PLoS ONE, 7(9). https://doi.org/10.1371/journal.pone.0043542

Pennings, S. C., & Bertness, M. D. (2000). Salt Marsh Communities. In Marine Community Ecology (pp. 289–316). https://doi.org/1605352284

Perillo, G. M. E., Wolanski, E., Cahoon, D. R., & Brinson, M. M. (Eds.). (2009). Coastal wetlands : an integrated ecosystem approach (First edit). https://doi.org/10.1016/j.annfar.2004.08.007

Pethick, J. (2002). Estuarine and tidal wetland restoration in the United Kingdom: Policy versus practice. Restoration Ecology, 10(3), 431–437. https://doi.org/10.1046/j.1526-100X.2002.01033.x

Pethick, J., & Orford, J. D. (2013). Rapid rise in effective sea-level in southwest Bangladesh: Its causes and contemporary rates. Global and Planetary Change, 111, 237–245. https://doi.org/10.1016/j.gloplacha.2013.09.019

Phan, L. K., van Thiel de Vries, J. S. M., Stive, M. J. F., & de Vries, Jaap S M van Thiel, Stive, M. J. F. (2015). Coastal Mangrove Squeeze in the Mekong Delta. Journal of Coastal Research, 31(2), 233–243. https://doi.org/10.2112/JCOASTRES-D-14-00049.1

Pierik, H. J., Cohen, K. M., Vos, P. C., van der Spek, A. J. F., & Stouthamer, E. (2017). Late Holocene coastal-plain evolution of the Netherlands: the role of natural preconditions in human-induced sea ingressions. Proceedings of the Geologists’ Association, 128(2), 180–197. https://doi.org/10.1016/j.pgeola.2016.12.002

Pontee, N. (2013). Defining coastal squeeze: A discussion. Ocean and Coastal Management, 84, 1–4. https://doi.org/10.1016/j.ocecoaman.2013.07.010

Pranzini, E. (2018). Shore protection in Italy: From hard to soft engineering … and back. Ocean and Coastal Management, 156, 43–57. https://doi.org/10.1016/j.ocecoaman.2017.04.018

References

201

Pranzini, E., Wetzel, L., & Williams, A. T. (2015). Aspects of coastal erosion and protection in Europe. Journal of Coastal Conservation, 19(4), 445–459. https://doi.org/10.1007/s11852-015-0399-3

Principe, P. P., Bradley, P., Yee, S. H., Fisher, W. S., Johnson, E. D., Allen, P., & Campbell, D. E. (2012). Quantifying Coral Reef Ecosystem Services, EPA/600/R-(January), 158.

Rangel-Buitrago, N., Williams, A. T., & Anfuso, G. (2018). Hard protection structures as a principal coastal erosion management strategy along the Caribbean coast of Colombia. A chronicle of pitfalls. Ocean and Coastal Management, 156, 58–75. https://doi.org/10.1016/j.ocecoaman.2017.04.006

Rao, N. S., Carruthers, T. J. B., Anderson, P., Sivo, L., Saxby, T., Durbin, T., et al. (2013). An economic analysis of ecosystem-based adaptation and engineering options for climate change adaptation in Lami Town, Republic of the Fiji Islands. A technical report by the Secretariat of the Pacific Regional Environment Programme. Retrieved from http://ian.umces.edu/pdfs/ian_report_392.pdf

Reddy, S. M. W., Guannel, G., Griffin, R., Faries, J., Boucher, T., Thompson, M., et al. (2016). Evaluating the role of coastal habitats and sea-level rise in hurricane risk mitigation: An ecological economic assessment method and application to a business decision. Integrated Environmental Assessment and Management, 12(2), 328–344. https://doi.org/10.1002/ieam.1678

Rego, J. L., & Li, C. (2009). On the importance of the forward speed of hurricanes in storm surge forecasting: A numerical study. Geophysical Research Letters, 36(7), 1–5. https://doi.org/10.1029/2008GL036953

Reguero, B. G., Bresch, D. N., Beck, M. W., Calil, J., & Meliane, I. (2014). Coastal Risks, Naure-Based Defenses and the Economics of Adaptation: An Application in the Gulf of Mexico, USA. Coastal Engineering, 1–15. https://doi.org/http://dx.doi.org/10.9753/icce.v34.management.25

Reguero, B. G., Beck, M. W., Bresch, D. N., Calil, J., & Meliane, I. (2018). Comparing the cost effectiveness of nature- based and coastal adaptation : A case study from the Gulf Coast of the United States, 1–24. https://doi.org/10.17605/OSF.IO/D6R5U.

Reise, K. (2005). Coast of change: Habitat loss and transformations in the Wadden Sea. Helgoland Marine Research, 59(1), 9–21. https://doi.org/10.1007/s10152-004-0202-6

Resio, D. T., & Westerink, J. J. (2008). Modeling the physics of storm surges. Physics Today, 61(9), 33–38. https://doi.org/10.1063/1.2982120

RESTORE. (2017). Restoring the Mississippi River Delta. Retrieved from http://mississippiriverdelta.org/files/2017/11/MRD_PriorityProjectReport.pdf

Rodriguez, E., Morris, C. C., & Belz, J. J. (2006). A global assessment of the SRTM performance. Photogrammetric Engineering and Remote Sensing, 72(3), 249–260. https://doi.org/10.14358/PERS.72.3.249

Rovere, A., Stocchi, P., & Vacchi, M. (2016). Eustatic and Relative Sea Level Changes. Current Climate Change Reports, 2(4), 221–231. https://doi.org/10.1007/s40641-016-0045-7

Rozé, F., & Lemauviel, S. (2004). Sand dune restoration in North Brittany, France: A 10-year monitoring study. Restoration Ecology, 12(1), 29–35. https://doi.org/10.1111/j.1061-2971.2004.00264.x

References

202

Rupp-Armstrong, S., & Nicholls, R. J. (2007). Coastal and Estuarine Retreat: A Comparison of the Application of Managed Realignment in England and Germany. Journal of Coastal Research, 236, 1418–1430. https://doi.org/10.2112/04-0426.1

Rupprecht, F., Möller, I., Paul, M., Kudella, M., Spencer, T., van Wesenbeeck, B. K., et al. (2017). Vegetation-wave interactions in salt marshes under storm surge conditions. Ecological Engineering, 100, 301–315. https://doi.org/10.1016/j.ecoleng.2016.12.030

San Francisco Bay Joint Venture. (2018). San Francisco Bay Joint Venture A partnership working to protect wetlands for the benefit of wildlife and people in the Bay Area. Retrieved June 27, 2018, from http://www.sfbayjv.org/about-goals.php

Sandi, S., Rodriguez, J., Saintilan, N., Riccardi, G., & Saco, P. (2018). Rising tides, rising gates: the complex ecogeomorphic response of coastal wetlands to sea-level rise and human interventions. Advances in Water Resources, 114, 135–148. https://doi.org/10.1016/j.advwatres.2018.02.006

Sasmito, S. D., Murdiyarso, D., Friess, D. A., & Kurnianto, S. (2016). Can mangroves keep pace with contemporary sea level rise? A global data review. Wetlands Ecology and Management, 24(2), 263–278. https://doi.org/10.1007/s11273-015-9466-7

Saunders, M. I., Leon, J. X., Callaghan, D. P., Roelfsema, C. M., Hamylton, S., Brown, C. J., et al. (2014). Interdependency of tropical marine ecosystems in response to climate change. Nature Climate Change, 4(8), 724–729. https://doi.org/10.1038/nclimate2274

Schepers, L., Kirwan, M., Guntenspergen, G., & Temmerman, S. (2017). Spatio-temporal development of vegetation die-off in a submerging coastal marsh. Limnology and Oceanography, 62(1), 137–150. https://doi.org/10.1002/lno.10381

Schmitt, K. (2012). Mangrove planting, community participation and integrated management in Soc Trang Province, Viet Nam. Sharing Lessons on Mangrove Restoration. Retrieved from http://www.mangrovesforthefuture.org/assets/Repository/Documents/Call-for-Action-and-Proceedings-from-2012-Colloquium-Mamallapuram-India.pdf

Schueler, K. (2017). Nature-based solutions to enhance coastal resilience. Retrieved from https://publications.iadb.org/bitstream/handle/11319/8526/Nature_based_solutions_to_enhance_coastal_resilience.pdf?sequence=1&isAllowed=y

Schuerch, M., Spencer, T., Temmerman, S., Kirwan, M. L., Wolff, C., Lincke, D., et al. (2018). Future response of global coastal wetlands to sea-level rise. Nature, 561(7722), 231–234. https://doi.org/10.1038/s41586-018-0476-5

Scott, D. B., Frail-Gauthier, J., & Mudie, P. J. (2014). Coastal Wetlands of the World. (Cambirdge University Press, Ed.). https://doi.org/10.1017/CBO9781107296916

Seavitt, C. (2013). Yangtze River Delta Project. Retrieved August 21, 2018, from https://scenariojournal.com/article/yangtze-river-delta-project/

Sengupta, D., Chen, R., & Meadows, M. E. (2018). Building beyond land: An overview of coastal land reclamation in 16 global megacities. Applied Geography, 90(May 2017), 229–238. https://doi.org/10.1016/j.apgeog.2017.12.015

References

203

Sheng, Y. P., Lapetina, A., & Ma, G. (2012). The reduction of storm surge by vegetation canopies: Three-dimensional simulations. Geophysical Research Letters, 39(20), 1–5. https://doi.org/10.1029/2012GL053577

Shepard, C. C., Crain, C. M., & Beck, M. W. (2011). The protective role of coastal marshes: A systematic review and meta-analysis. PLoS ONE, 6(11). https://doi.org/10.1371/journal.pone.0027374

de Sherbinin, A., Schiller, A., & Pulsipher, A. (2007). The vulnerability of global cities to climate hazards. Environment and Urbanization, 19(1), 39–64. https://doi.org/10.1177/0956247807076725

Shoemaker, C. M., Ervin, G. N., & DiOrio, E. W. (2017). Interplay of water quality and vegetation in restored wetland plant assemblages from an agricultural landscape. Ecological Engineering, 108(May), 255–262. https://doi.org/10.1016/j.ecoleng.2017.08.034

Short, F. T., & Neckles, H. A. (1999). The effects of global climate change on seagrasses. Aquatic Botany, 63(3–4), 169–196. https://doi.org/10.1016/S0304-3770(98)00117-X

SigmaPlan. (2017). Over het SigmaPlan. Retrieved August 17, 2017, from http://sigmaplan.be/nl/over-het-sigmaplan/

Sigren, J. M., Figlus, J., Armitage, A. R., Barone, D. A., McKenna, K. K., Farrell, S. C., et al. (2014). Coastal sand dunes and dune vegetation: restoration, erosion, and storm protection. Shore & Beach, 82(4), 5.

Sigren, J. M., Figlus, J., Highfield, W., Feagin, R. A., & Armitage, A. R. (2018). The Effects of Coastal Dune Volume and Vegetation on Storm-Induced Property Damage: Analysis from Hurricane Ike. Journal of Coastal Research, 341(1), 164–173. https://doi.org/10.2112/JCOASTRES-D-16-00169.1

Silva, R., Martínez, M. L., Odériz, I., Mendoza, E., & Feagin, R. A. (2016). Response of vegetated dune-beach systems to storm conditions. Coastal Engineering, 109, 53–62. https://doi.org/10.1016/j.coastaleng.2015.12.007

Simas, T., Nunes, J. ., & Ferreira, J. . (2001). Effects of global climate change on coastal salt marshes. Ecological Modelling, 139(1), 1–15. https://doi.org/10.1016/S0304-3800(01)00226-5

Simenstad, C., Reed, D., & Ford, M. (2006). When is restoration not? Incorporating landscape-scale processes to restore self-sustaining ecosystems in coastal wetland restoration. Ecological Engineering, 26(1), 27–39. https://doi.org/10.1016/j.ecoleng.2005.09.007

Small, C., & Nicholls, R. J. (2003). A Global Analysis of Human Settlement in Coastal Zones. Journal of Coastal Reserach, 19(3), 584–599.

Smolders, S., Plancke, Y., Ides, S., Meire, P., & Temmerman, S. (2015). Role of intertidal wetlands for tidal and storm tide attenuation along a confined estuary: a model study. Natural Hazards and Earth System Sciences Discussions, 3(5), 3181–3224. https://doi.org/10.5194/nhessd-3-3181-2015

Spalding, M. D., Blasco, E., & Field, C. D. (1997). World Mangrove Atlas. The International Society for Mangrove Ecosystems. https://doi.org/10.1017/S0266467498300528

Spalding, M. D., Ravilious, C., & Green, E. P. (2002). World Atlas of Coral Reefs (Prepared a, Vol. 44). University of California Press, Berkeley, USA. https://doi.org/10.1016/S0025-326X(01)00310-1

References

204

Spalding, M. D., Kainuma, M., & Collins, L. (2010). World Atlas of Mangroves. (Earthscan, Ed.). London: Published with ISME, ITTO and project partners FAO, UNESCO-MAB, UNEP-WCMC and UNU-INWEH. Retrieved from https://books.google.be/books?id=wzSCkulW9SQC

Spalding, M. D., McIvor, A. L., Beck, M. W., Koch, E. W., Möller, I., Reed, D. J., et al. (2013). Coastal ecosystems: A critical element of risk reduction. Conservation Letters, 7(3), 293–301. https://doi.org/10.1111/conl.12074

Spalding, M. D., McIvor, A. L., Tonneijck, F., Tol, S., & van Eijk, P. (2014). Mangroves for coastal defence. Guidelines for coastal managers & policy makers.

Spalding, M. D., Ruffo, S., Lacambra, C., Meliane, I., Hale, L. Z., Shepard, C. C., & Beck, M. W. (2014). The role of ecosystems in coastal protection: Adapting to climate change and coastal hazards. Ocean and Coastal Management, 90, 50–57. https://doi.org/10.1016/j.ocecoaman.2013.09.007

Stanturf, J. A., Goodrick, S. L., & Outcalt, K. W. (2007). Disturbance and coastal forests: A strategic approach to forest management in hurricane impact zones. Forest Ecology and Management, 250(1–2), 119–135. https://doi.org/10.1016/j.foreco.2007.03.015

Stark, J. (2016). Effects of intertidal ecosystems on estuarine hydrodynamics and flood wave attenuation : a multi-scale study. https://doi.org/10.13140/RG.2.2.24404.19840

Stark, J., Van Oyen, T., Meire, P., & Temmerman, S. (2015). Observations of tidal and storm surge attenuation in a large tidal marsh. Limnology and Oceanography, n/a-n/a. https://doi.org/10.1002/lno.10104

Stark, J., Plancke, Y., Ides, S., Meire, P., & Temmerman, S. (2016). Coastal flood protection by a combined nature-based and engineering approach: Modeling the effects of marsh geometry and surrounding dikes. Estuarine, Coastal and Shelf Science, 175, 34–45. https://doi.org/10.1016/j.ecss.2016.03.027

Storlazzi, C. D., Elias, E., Field, M. E., & Presto, M. K. (2011). Numerical modeling of the impact of sea-level rise on fringing coral reef hydrodynamics and sediment transport. Coral Reefs, 30(SUPPL. 1), 83–96. https://doi.org/10.1007/s00338-011-0723-9

Suman, D. O. (2019). Mangrove Management. In Coastal Wetlands (pp. 1055–1079). Elsevier. https://doi.org/10.1016/B978-0-444-63893-9.00031-9

Sun, G., Ranson, K. J., Kharuk, V. I., & Kovacs, K. (2003). Validation of surface height from shuttle radar topography mission using shuttle laser altimeter. Remote Sensing of Environment, 88(4), 401–411. https://doi.org/10.1016/j.rse.2003.09.001

Sutton-Grier, A. E., Wowk, K., & Bamford, H. (2015). Future of our coasts: The potential for natural and hybrid infrastructure to enhance the resilience of our coastal communities, economies and ecosystems. Environmental Science and Policy, 51, 137–148. https://doi.org/10.1016/j.envsci.2015.04.006

Sutton-Grier, A. E., Gittman, R. K., Arkema, K. K., Bennett, R. O., Benoit, J., Blitch, S., et al. (2018). Investing in natural and nature-based infrastructure: Building better along our coasts. Sustainability (Switzerland), 10(2), 1–11. https://doi.org/10.3390/su10020523

Syvitski, J. P. M. (2005). Impact of Humans on the Flux of Terrestrial Sediment to the Global Coastal Ocean. Science, 308(5720), 376–380. https://doi.org/10.1126/science.1109454

References

205

Syvitski, J. P. M. (2008). Deltas at risk. Sustainability Science, 3(1), 23–32. https://doi.org/10.1007/s11625-008-0043-3

Syvitski, J. P. M., & Saito, Y. (2007). Morphodynamics of deltas under the influence of humans. Global and Planetary Change, 57(3–4), 261–282. https://doi.org/10.1016/j.gloplacha.2006.12.001

Syvitski, J. P. M., Kettner, A. J., Overeem, I. I., Hutton, E. W. H., Hannon, M. T., Brakenridge, G. R., et al. (2009). Sinking deltas due to human activites. Nature Geoscience, 2(September), 6–11. https://doi.org/10.1038/ngeo629

Tack, F. M. G., De Pauw, N., Du Laing, G., & Rousseau, D. (2007). Contaminants in natural and constructed wetlands: Pollutant dynamics and control. Science of the Total Environment, 380(1–3), 1–2. https://doi.org/10.1016/j.scitotenv.2007.02.018

Tateishi, R., Hoan, N. T., Kobayashi, T., Alsaaideh, B., Tana, G., & Phong, D. X. (2014). Production of Global Land Cover Data – GLCNMO2008. Journal of Geography and Geology, 6(3). https://doi.org/10.5539/jgg.v6n3p99

Temmerman, S., & Kirwan, M. L. (2015). Building land with a rising sea. Science. https://doi.org/10.1126/science.aac8312

Temmerman, S., Govers, G., Wartel, S., & Meire, P. (2004). Modelling estuarine variations in tidal marsh sedimentation: Response to changing sea level and suspended sediment concentrations. Marine Geology, 212(1–4), 1–19. https://doi.org/10.1016/j.margeo.2004.10.021

Temmerman, S., De Vries, M. B., & Bouma, T. J. (2012). Coastal marsh die-off and reduced attenuation of coastal floods: A model analysis. Global and Planetary Change, 92–93, 267–274. https://doi.org/10.1016/j.gloplacha.2012.06.001

Temmerman, S., Meire, P., Bouma, T. J., Herman, P. M. J., Ysebaert, T., & De Vriend, H. J. (2013). Ecosystem-based coastal defence in the face of global change. Nature, 504(7478), 79–83. https://doi.org/10.1038/nature12859

Tessler, Z. D., Vörösmarty, C. J., Grossberg, M., Gladkova, I., Aizenman, H., Syvitski, J. P. M., & Foufoula-Georgiou, E. (2015). Profiling risk and sustainability in coastal deltas of the world. Science, 349(6248), 638–643.

Teuchies, J., Vandenbruwaene, W., Carpentier, R., Bervoets, L., Temmerman, S., Wang, C., et al. (2013). Estuaries as Filters: The Role of Tidal Marshes in Trace Metal Removal. PLoS ONE, 8(8), 1–11. https://doi.org/10.1371/journal.pone.0070381

Thampanya, U., Vermaat, J. E., Sinsakul, S., & Panapitukkul, N. (2006). Coastal erosion and mangrove progradation of Southern Thailand. Estuarine, Coastal and Shelf Science, 68(1), 75–85. https://doi.org/10.1016/j.ecss.2006.01.011

The World Bank. (2017). Implementing nature-based flood protection Principles and implementation guidance, 7–31. Retrieved from http://documents.worldbank.org/curated/en/739421509427698706/pdf/120735-REVISED-PUBLIC-Brochure-Implementing-nature-based-flood-protection-web.pdf

Tian, B., Wu, W., Yang, Z., & Zhou, Y. (2016). Drivers, trends, and potential impacts of long-term coastal reclamation in China from 1985 to 2010. Estuarine, Coastal and Shelf Science, 170, 83–90. https://doi.org/10.1016/j.ecss.2016.01.006

UNEP-WCMC. (2006). In the front line: shoreline protection and other ecosystem services from mangroves and coral reefs. United Nations Environmental Programme- World Conservation Monitoring Centre International Coral Reef Action Network, (24), 40. https://doi.org/10.1089/bfm.2011.0071

References

206

United Nations. (2012). World Population Prospects: the 2010 Revision. Waste Management Research, 27(8), 800–812. https://doi.org/10.1553/populationyearbook2010s77

United Nations Department of Economic and Social Affairs Population Division. (2016). The World’s Cities in 2016. Data Booklet (ST/ESA/SER.A/392). Retrieved from www.unpopulation.org.

UT BATTELLE LLC. (n.d.). LandScan frequently asked questions. Retrieved November 16, 2016, from http://web.ornl.gov/sci/landscan/landscan_faq.shtml

Vafeidis, A. T., Boot, G., Cox, J., Mcfadden, L., & Nicholls, R. J. (2005). The diva database documentation, 1–33.

Valiela, I. (2006). Global Coastal Change. Environmental Conservation (Vol. 33). Oxford, UK: Blackwell Publishing. https://doi.org/DOI: 10.1017/S0376892907253652

Valiela, I., Bowen, J. L., & York, J. K. (2001). Mangrove Forests: One of the World’s Threatened Major Tropical Environments. BioScience, 51(10), 807. https://doi.org/10.1641/0006-3568(2001)051[0807:MFOOTW]2.0.CO;2

Valiela, I., Kinney, E., Culbertson, J., Peacock, E., & Smith, S. (2009). Global Loss of Coastal Habitats Rates, Causes and Consequences. In C. M. Duarte (Ed.), Global Loss of Coastal Habitats Rates, Causes and Consequences (pp. 108–142).

Vermaat, J. E., & Thampanya, U. (2006). Mangroves mitigate tsunami damage: A further response. Estuarine, Coastal and Shelf Science, 69(1–2), 1–3. https://doi.org/10.1016/j.ecss.2006.04.019

Vitousek, S., Barnard, P. L., Fletcher, C. H., Frazer, N., Erikson, L., & Storlazzi, C. D. (2017). Doubling of coastal flooding frequency within decades due to sea-level rise. Scientific Reports, 7(1), 1–9. https://doi.org/10.1038/s41598-017-01362-7

Voortman, H. G., Van Gelder, P. H. a J. M., & Vrijling, J. K. (2003). Risk-Based Design of Large-Scale Flood Defence Systems. Coastal Engineering 2002 Solving Coastal Conundrums Proceedings of the 28th International Conference. https://doi.org/10.1142/9789812791306_0199

Vos, P. C. (2015). Origin of the Dutch coastal landscape. Universiteit Utrecht.

Vuik, V., Jonkman, S. N., Borsje, B. W., & Suzuki, T. (2016). Nature-based flood protection: The efficiency of vegetated foreshores for reducing wave loads on coastal dikes. Coastal Engineering, 116, 42–56. https://doi.org/10.1016/j.coastaleng.2016.06.001

Vuik, V., van Vuren, S., Borsje, B. W., van Wesenbeeck, B. K., & Jonkman, S. N. (2018). Assessing safety of nature-based flood defenses: Dealing with extremes and uncertainties. Coastal Engineering, 139(April), 47–64. https://doi.org/10.1016/j.coastaleng.2018.05.002

Wamsley, T. V., Cialone, M. a., Smith, J. M., Ebersole, B. a., & Grzegorzewski, A. S. (2009). Influence of landscape restoration and degradation on storm surge and waves in southern Louisiana. Natural Hazards, 51(1), 207–224. https://doi.org/10.1007/s11069-009-9378-z

Wamsley, T. V., Cialone, M. A., Smith, J. M., Atkinson, J. H., & Rosati, J. D. (2010). The potential of wetlands in reducing storm surge. Ocean Engineering, 37(1), 59–68. https://doi.org/10.1016/j.oceaneng.2009.07.018

Wang, W., Liu, H., Li, Y., & Su, J. (2014). Development and management of land reclamation in China. Ocean & Coastal Management, 102, 415–425. https://doi.org/10.1016/j.ocecoaman.2014.03.009

References

207

Webster, P. J., Holland, G. J., Curry, J. A., Chang, H.-R., Sciences, A., Webster, P. J., et al. (2005). Changes in Tropical Cyclone Number , Duration , and Intensity in a Warming Environment. Sciences, Atmospheric, 437(1996), 2003–2006. https://doi.org/10.1126/science.1116448

van Wesenbeeck, B. K., Mulder, J. P. M., Marchand, M., Reed, D. J., De Vries, M. B., De Vriend, H. J., & Herman, P. M. J. (2014). Damming deltas: A practice of the past? Towards nature-based flood defenses. Estuarine, Coastal and Shelf Science, 140, 1–6. https://doi.org/10.1016/j.ecss.2013.12.031

van Wesenbeeck, B. K., de Boer, W., Narayan, S., van der Star, W. R. L., & de Vries, M. B. (2017). Coastal and riverine ecosystems as adaptive flood defenses under a changing climate. Mitigation and Adaptation Strategies for Global Change, 22(7), 1087–1094. https://doi.org/10.1007/s11027-016-9714-z

White, A. T., Vogt, H. P., & Arin, T. (2000). Philippine coral reefs under threat: The economic losses caused by reef destruction. Marine Pollution Bulletin, 40(7), 598–605. https://doi.org/10.1016/S0025-326X(00)00022-9

Wilkinson, C., Lindén, O., Cesar, H., Hodgson, G., Rubens, J., & Strong, A. E. (1999). Ecological and Socioeconimoc Impacts of 1998 Coral Mortality in the Indian Ocean: An ENSO Impact and a Warning of Future Change? Ambio, 28(2), 188–196. https://doi.org/10.5363/tits.6.3_37

Wolanski, E., & Elliott, M. (2015). Estuarine Ecohydrology: An Introduction. Elsevier Science, Amsterdam, 322pp.

Wolff, W. J. (1993). Netherlands-Wetlands. Hydrobiologia, 265(1–3), 1–14. https://doi.org/10.1007/BF00007260

Woodruff, J. D., Irish, J. L., & Camargo, S. J. (2013). Coastal flooding by tropical cyclones and sea-level rise. Nature, 504(7478), 44–52. https://doi.org/10.1038/nature12855

World Population History. (n.d.). World Population. Retrieved September 4, 2017, from http://worldpopulationhistory.org/map/306/mercator/1/0/25/

Yepsen, M., Baldwin, A. H., Whigham, D. F., McFarland, E., LaForgia, M., & Lang, M. (2014). Agricultural wetland restorations on the USA Atlantic Coastal Plain achieve diverse native wetland plant communities but differ from natural wetlands. Agriculture, Ecosystems and Environment, 197, 11–20. https://doi.org/10.1016/j.agee.2014.07.007

Ysebaert, T., Yang, S. L., Zhang, L., He, Q., Bouma, T. J., & Herman, P. M. J. (2011). Wave attenuation by two contrasting ecosystem engineering salt marsh macrophytes in the intertidal pioneer zone. Wetlands, 31(6), 1043–1054. https://doi.org/10.1007/s13157-011-0240-1

Ysebaert, T., van der Hoek, D. J., Wortelboer, R., Wijsman, J. W. M., Tangelder, M., & Nolte, A. (2016). Management options for restoring estuarine dynamics and implications for ecosystems: A quantitative approach for the Southwest Delta in the Netherlands. Ocean and Coastal Management, 121, 33–48. https://doi.org/10.1016/j.ocecoaman.2015.11.005

Ysebaert, T., Jansen, H. M., & Poelman, M. (2017). Coastal protection and seafood security in Bangladesh : Food for Thought Nature-based solutions at the land-ocean interface : Building with Nature Introduction The coastal area of Bangladesh Ecosystem-based coastal protection. In Ocean/Blue Economy for Bangladesh Workshop (pp. 22–23). Dhaka.

References

208

Zhang, K., Liu, H., Li, Y., Xu, H., Shen, J., Rhome, J., & Smith, T. J. (2012). The role of mangroves in attenuating storm surges. Estuarine, Coastal and Shelf Science, 102–103, 11–23. https://doi.org/10.1016/j.ecss.2012.02.021

Zhao, H., Cui, B., Zhang, H., Fan, X., Zhang, Z., & Lei, X. (2010). A landscape approach for wetland change detection (1979-2009) in the Pearl River Estuary. Procedia Environmental Sciences, 2(5), 1265–1278. https://doi.org/10.1016/j.proenv.2010.10.137

Zhao, Q., Bai, J., Huang, L., Gu, B., Lu, Q., & Gao, Z. (2016). A review of methodologies and success indicators for coastal wetland restoration. Ecological Indicators, 60, 442–452. https://doi.org/10.1016/j.ecolind.2015.07.003

Zhu, M. S., Sun, T., & Shao, D. D. (2016). Impact of Land Reclamation on the Evolution of Shoreline Change and Nearshore Vegetation Distribution in Yangtze River Estuary. Wetlands, 36, 11–17. https://doi.org/10.1007/s13157-014-0610-6