the use of carbon cycle modelling for the production of ......marine and atmospheric research was...

The use of carbon cycle modelling for the production of ecosystem accounts A joint study between CSIRO Marine and Atmospheric Research and the

Bureau of Meteorology

Contributing to the Australian Government National Plan for Environmental Information initiative

The use of carbon cycle modelling for the production of ecosystem accounts

ISBN: 978-0-642-70651-5

Environmental Information Programme

Bureau of Meteorology

Email: [email protected]

www.bom.gov.au/environment

Citing this publication

Haverd, V., P.R. Briggs, N.J. Grigg, M.R. Raupach and R.E. Mount, 2014. The use of carbon cycle modelling for

the production of ecosystem accounts. Prepared by CSIRO Marine and Atmospheric Research and the

Environmental Information Services Branch, Bureau of Meteorology, Canberra, Australia

www.bom.gov.au/environment.

With the exception of logos or where otherwise noted, this report is licensed under the Creative Commons

Australia Attribution 3.0 Licence. The terms and conditions of the licence are at:

www.creativecommons.org/licenses/by/3.0/au

Cover: Tasmanian Blue Gum (Eucalyptus globulus), Cape Tourville, Tasmania © Richard Mount. The Creative

Commons licence does not apply to this material. If you wish to further use this material you must contact the

copyright owner directly to obtain permission.

http://www.bom.gov.au/environment/

http://www.creativecommons.org/licenses/by/3.0/au

i

National Environmental Information


Abbreviations .......................................................................................................................... i

1 Introduction ....................................................................................................................... 1

1.1 Background ........................................................................................................ 1 1.2 Purpose ............................................................................................................. 1 1.3 Rationale ............................................................................................................ 2 1.4 Scope ................................................................................................................. 3

2 Methodology ..................................................................................................................... 4

2.1 Concepts ............................................................................................................ 4 2.2 Data sources ...................................................................................................... 5 2.3 Methods ............................................................................................................. 7 2.4 Accounting frameworks .................................................................................... 12

3 Results ........................................................................................................................... 13

4 Discussion ...................................................................................................................... 29

4.1 What can the results tell us? ............................................................................ 29 4.2 What can’t the results tell us? (constraints and uncertainties) .......................... 30 4.3 Working with temporal and spatial scale .......................................................... 31 4.4 Future directions .............................................................................................. 31

5 Conclusion...................................................................................................................... 37

References ........................................................................................................................... 39

List of tables

Table 1: Different units for quantifying amounts of carbon (C) .................................................. 7

List of figures

Figure 1: Relationship of Net Ecosystem Production (NEP) to Net Primary Production (NPP), Gross Primary Production (GPP), and Net Biome Production (NBP). ........................... 5

Figure 2: Location of observations used for BIOS2 constraint and evaluation .......................... 7

Figure 3: Comparison of the BIOS2 model domain with a land mask derived from Geoscience Australia’s GEODATA TOPO 2.5M framework layer. ................................ 9

Figure 4: Regionalisations used in this study. ........................................................................... 10

Figure 5: Net Ecosystem Production for the Australian continent. ............................................ 14

Figure 6: Net Ecosystem Production for the Murray–Darling Basin. ......................................... 15

Figure 7.1: Net Ecosystem Production aggregated for the Australian spatial domain. .............. 16

Figure 7.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated for the Australian spatial domain. .................................................................................. 16

Figure 7.3: Total landscape carbon stock aggregated for the Australian spatial domain. .......... 17

Figure 8.1: Net Ecosystem Production aggregated for Australian States and Territories spatial domains. ............................................................................................................ 18

Contents

ii



Figure 8.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated for Australian States and Territories spatial domains. ................................................... 18

Figure 8.3: Total landscape carbon stock aggregated for Australian States and Territories spatial domains. ............................................................................................................ 19

Figure 9.1: Net Ecosystem Production aggregated for the Murray–Darling Basin spatial domain. ......................................................................................................................... 20

Figure 9.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated for the Murray–Darling Basin spatial domain. ................................................................ 20

Figure 9.3: Total landscape carbon stock aggregated for the Murray–Darling Basin spatial domain. ......................................................................................................................... 21

Figure 10.1: Net Ecosystem Production aggregated for NRMs intersecting the Murray–Darling Basin. ............................................................................................................... 22

Figure 10.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated for NRMs intersecting the Murray–Darling Basin. ....................................... 23

Figure 10.3: Total landscape carbon stock aggregated for NRMs intersecting the Murray–Darling Basin. ............................................................................................................... 24

Figure 11.1: Net Ecosystem Production aggregated for the RECCAP regional domains. ......... 25

Figure 11.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated for the RECCAP regional domains. ............................................................ 25

Figure 11.3: Total landscape carbon stock aggregated for the RECCAP regional domains. ..... 26

Figure 12.1: Net Ecosystem Production aggregated for the IBRA Terrestrial Ecoregions. ........ 27

Figure 12.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated for the IBRA Terrestrial Ecoregions. ........................................................... 27

Figure 12.3: Total landscape carbon stock aggregated for the IBRA Terrestrial Ecoregions. .... 28

Figure 13: The Australian carbon budget 1990–2011 ............................................................... 29

i



Abbreviations

ABARES Australian Bureau of Agricultural and Resource Economics and Sciences

ABS Australian Bureau of Statistics

ACT Australian Capital Territory

ANU Australian National University

AVHRR Advanced Very High Resolution Radiometer

C carbon

CABLE Community Atmosphere Biosphere Land Exchange

CASA-CNP Carnegie-Ames-Stanford Approach, Carbon, Nitrogen, Phosphorus

CO2 carbon dioxide

CSIRO Commonwealth Scientific and Industrial Research Organisation

d day

DIICCSRTE Department of Industry, Innovation, Climate Change, Science, Research and Tertiary Education

fPAR fraction Photosynthetic Absorbed Radiation

g gram

Gg gigagram

GHG greenhouse gas

GIS Geographic Information System

GPP Gross Primary Production

Gt gigatonne

ha hectare

IBRA Interim Biogeographic Regionalisation for Australia

kg kilogram

km kilometre

kt kilotonne

LAI Leaf Area Index

m metres

MDB Murray–Darling Basin

Mg megagram

Mt megatonne

NBP Net Biome Production

NEP Net Ecosystem Production

NPEI National Plan for Environmental Information

ii



NPP Net Primary Production

NRM Natural Resource Management

NSW New South Wales

NT Northern Territory

Pg petagram

PMSEIC Prime Minister’s Science, Engineering and Innovation Council

ppm parts per million

Qld Queensland

RECCAP REgional Carbon Cycle Assessment and Processes

SA South Australia

SEEA System of Environmental-Economic Accounting

SLI Soil-Litter-Iso

SOC soil organic carbon

T Tonne

TAS Tasmania

TERN Terrestrial Ecosystem Research Network

Tg teragram

VAST Vegetation and Soil–carbon Transfer

VIC Victoria

WA Western Australia

yr year


1.1 Background

Established in mid-2010, the National Plan for Environmental Information (NPEI) initiative is

an Australian Government programme for improving the quality and accessibility of

environmental information. It is being jointly implemented by the Commonwealth Department

of the Environment and the Bureau of Meteorology (the Bureau).

A key activity for delivering on the NPEI’s objectives is the development of the Bureau’s

environmental accounting capability. This includes several components such as guides,

standards, tools, and exemplars. Research and development is required to support the

development of evidence bases, conceptual models, and methodologies for environmental

accounting. In turn, this will be the basis for producing a series of credible, legitimate,

relevant, and enduring environmental accounts.

Since 2010, the Bureau has collaborated with several agencies and organisations to conduct

a series of ‘demonstrator’ or ‘pilot’ projects. A joint study between the Bureau and CSIRO

Marine and Atmospheric Research was initiated in 2013 to examine the use of carbon cycle

modelling for the production of ecosystem accounts.

The Bureau defined the study brief with assistance and advice from CSIRO. As the producer

of the scientific component of the report’s contents, CSIRO responded to the brief through

analysis and some interpretation of the results. This report provides a description of the study

and its methodology, the results, and a brief discussion.

1.2 Purpose

The primary purpose of this study was to determine the potential of carbon flux modelling for

measuring the capacity of ecosystems to provide ecosystem services.

The secondary purpose was to demonstrate and strengthen the use of environmental

accounting in Australia. In this way, the study aimed to:

• road test the Guide to environmental accounting in Australia (Bureau of Meteorology

2013a) and the associated Environmental account framing workbook (Bureau of

Meteorology 2013b);

1 Introduction


• pilot concepts relevant to ecosystem capital accounting, including sustainability

indicators, and ecosystem services accounting, including primary productivity and

ecosystem potential;

• contribute to the demonstrator work undertaken by the Bureau of Meteorology and the

Murray–Darling Basin Authority; and

• provide an example of environmental accounting for the NPEI initiative.

1.3 Rationale

1.3.1 Why measure carbon?

Monitoring changes in biospheric carbon stocks is important for the following reasons.

1. Net ecosystem production and carbon stores underpin many of the ecosystem goods and

services that humans depend upon. These include provisioning services (e.g.,

agricultural production, timber production, and biomass for energy production), regulation

services (e.g., pH buffering in soils, water and air quality improvement, waste

decomposition, and local-scale climate regulation) and cultural services (e.g., recreational

areas).

2. Net biospheric carbon accumulation is a large component of Australia’s total carbon

budget (Haverd et al., 2013a).

3. The ability of the landscape to accumulate carbon, particularly after drought or other

disturbance events, is a good indicator of ecosystem integrity and resilience.

Overall, the study is a small, but important, first step towards testing components of methods

for measuring and estimating ecosystem accounting with a long-term goal of potentially

producing environmental accounts about ecosystem functioning, including degradation and

enhancement or restoration.

1.3.2 What is the role of ecosystem accounting?

Growth in Australia’s population and economy draws from and impacts upon Australia’s

natural capital. Considerable reductions in Australia’s natural capital are identified in every

State of the Environment report (e.g., farm land salination, spread of weeds and pests,


nutrient pollution, loss of biodiversity) and, if further losses are to be averted, ways must be

found to decouple economic progress from ecological depletion and degradation.

Several approaches can be considered to improve the stewardship of natural resource use,

but their implementation and management require relevant, credible, up-to-date, and

comprehensive information on the condition of Australia’s natural resources and ecosystems.

A particularly important role for environmental accounts is to complement economic accounts

and enable depletion and degradation of natural resources to be embedded in political and

economic decision-making. Common features of environmental accounting frameworks

include sub-accounts for stocks and flows of biomass or carbon (for food, fuel, or other

produce), stocks and flows of water resources, and ecosystem functioning (i.e., the

continued capacity of the ecosystem to reproduce and maintain itself and to provide goods

and services to humans).

1.4 Scope

The accounting of carbon stocks in this study encompasses changes in stocks attributable to

the combined effects of rising carbon dioxide (CO2) and variable climate. The carbon cycle is

tightly coupled to the water and energy cycles in this analysis. Impacts of disturbance (e.g.,

fire and land use change) are not included and are the subject of ongoing investigation.


2.1 Concepts

Biospheric carbon stocks (in soil, biomass and litter) are simulated continentally using the

BIOS2 land-surface modelling environment described below. The native spatial resolution is

0.05o (approximately 5 km) and the model is run at an hourly time step for the fast

biophysical process and a daily time step for the biogeochemical processes. For this study,

results are aggregated to annual time resolution and various spatial domains specified in

Section 2.3.2 below. Results are reported for a century (1911–2011) which is the time period

from which meteorological data are available. A long time period is required to contextualise

the results because the turnover time for biospheric carbon in Australia is around 70 years

(Haverd et al., 2013b).

While fluxes (flows) and stores (stocks or pools) of carbon associated with multiple stores are

simulated, the presentation of results is restricted to two lumped quantities, as follows:

1. Net Ecosystem Production (NEP) is defined as Net Primary Production (NPP) minus

the heterotrophic respiration flux that would occur without the influences of fire,

transport (by river and dust), and harvest. In other words, NEP is the flux, or net rate

of accumulation of carbon in the biosphere in the absence of disturbance, and is the

sum of net fluxes into the soil, litter, and vegetation carbon stocks. NEP can be

positive (net accumulation) or negative (net loss). Figure 1 illustrates the relationship

of NEP to NPP, Gross Primary Production (GPP) and Net Biome Production (NBP).

2. Total stock of ecosystem carbon is the sum of carbon stored in the soil, litter and

vegetation, and is dominated by the soil term. It is a slowly varying quantity because

the turnover time for soil carbon is long (approximately 100 years). This carbon stock

(pool) is estimated based on observations made in the field at specific locations and

extrapolated across the continent using the model to estimate fluxes in and out of the

stock.

2 Methodology


Source: Kirschbaum et al., 2001.

Figure 1: Relationship of Net Ecosystem Production (NEP) to Net Primary Production (NPP),

Gross Primary Production (GPP), and Net Biome Production (NBP).

2.2 Data sources

2.2.1 Key input (forcing) data

BIOS2 is run (forced) using gridded meteorological data, soil properties and vegetation cover

at 0.05° spatial resolution, which are described briefly below. Further details of the

meteorological data and soil properties appear in Appendix 2 of Haverd et al., 2013b.

Meteorology: The meteorological data comprise daily gridded rainfall, temperature, vapour

pressure and solar irradiance surfaces from the Bureau of Meteorology’s Australian Water

Availability Project dataset (Grant et al., 2008; Jones et al., 2009). Data are downscaled from

daily to hourly time steps (on the half-hour) using a weather generator within BIOS2.

Soil: Soil information is taken from the McKenzie and Hook (1992) and McKenzie et al.

(2000) interpretations of the 725 principal profile forms (soil types) mapped in the Digital

Atlas of Australian Soils (Northcote et al., 1960; Northcote et al., 1975).


Vegetation cover: Vegetation cover was prescribed using fPAR (fraction Photosynthetic

Absorbed Radiation) estimates obtained from the AVHRR (Advanced Very High Resolution

Radiometer) record (1990–2006) (Donohue et al., 2009), with an annual climatology being

used outside of the period of data availability. Total fPAR was partitioned into persistent

(mainly woody) and recurrent (mainly grassy) vegetation components, following the

methodology of Donohue et al. (2009) and Lu et al. (2003). The Leaf Area Index (LAI) for

woody and grassy components were estimated from the fPAR components by Beer’s Law

(e.g., Houldcroft et al., 2009). Grassy LAI was partitioned between C3 and C4 components

according to the proportion of all grass species that are C4 species, as estimated by

Hattersley (1983).1

2.2.2 Datasets for parameter estimation and model evaluation

Several types of observations were used for parameter estimation (i.e., the model’s settings,

‘constraints’ and ‘training’ data) and model evaluation (model testing). For each data type,

less than 30 per cent of the data was used in parameter estimation. All datasets were used in

their entirety for model evaluation. The datasets, described in detail in Haverd et al. 2013a,

are:

• streamflow records for 416 unregulated catchments were obtained from (a) 231 of 232 for

southeastern Australia (Vaze et al., 2011); and (b) 185 of 719 Australia-wide (Zhange et

al., 2011);

• land–atmosphere exchanges of energy and CO2 from 12 OzFlux sites;

• litterfall, above-ground biomass, above-ground fine litter, and soil carbon observations

from the Vegetation and Soil–Carbon Transfer (VAST) database (Barrett, 2001) and

additional biomass data compiled by Raison et al. (2003); and

• previous estimates of regional carbon budget components for three forest ecosystems—

Victorian Eucalyptus regnans forests (2,324 km2); New South Wales Coastal Corymbia

maculata forests (58 km2) and Queensland poplar-box (Eucalyptus populnea) woodlands

(2,812 km2).

Locations of the observations are shown in Figure 2.

1 C3 and C4 are the two different pathways that plants use to capture CO2 during photosynthesis. The

pathways are associated with different environmental conditions and plant growth requirements (e.g., light, temperature and moisture). The majority of Australian grasses have been assigned as having the C3 or C4 photosynthetic pathway (see Hattersley 1983).


Figure 2: Location of observations used for BIOS2 constraint and evaluation

2.3 Methods

2.3.1 Units of measurement

Carbon fluxes (NEP) and stocks are presented here in units of gC m-2 d-1 and gC m-2

respectively.

Table 1: Different units for quantifying amounts of carbon (C)

1 Mg (megagram) = 1,000 kg = 1 t (tonne) 1 molC m-2

= 120 kgC ha-1

1 Gg (gigagram) = 106 kg = 1 kt (kilotonne) 1 molCO2 m-2

= 440 kgCO2 ha-1

1 Tg (teragram) = 109 kg = 1 Mt (megatonne) 1 gC m-2

= 10,000 gC ha = 10 kgC ha

1 Pg (petagram) = 1012 kg = 1 Gt (gigatonne) 1 gC m

-2 d

-1 = 10,000 gC ha d

-1 =

10 kgC ha d-1

= 3,650 kgC ha yr-1

= 3.6 tC ha yr

-1 = 3.6 MgC ha yr

-1

1 kg m-2

= 10 t ha-1

1 tC = 3.667 tCO2 (using respective molecular weights of 12:44)

1 molC = 12 gC 1 tC ha-1

= 8.33 molC m-2

Note: kg = kilogram, m = metre, ha = hectare, g = gram, yr = year, d = day


2.3.2 Geography

The model domain for BIOS2 includes approximately 96 per cent of the Australian continent

(black areas in Figure 3). Excluded areas (in yellow) are those for which soil parameter data

were unavailable: salt lakes, salt pans, inland water, and some coastal features.

Shapefiles for several different regionalisations were converted to 0.05° rasters in ArcGIS

and masked against the model domain, leaving the yellow-coloured areas in Figure 3. These

were used to generate time series of spatial averages for each regionalisation detailed

below.

2.3.2.1 Australian States and Territories

Title: GEODATA TOPO 2.5M 2003, ‘framework’ layer

Custodian: Commonwealth of Australia (Geoscience Australia)

Metadata: www.ga.gov.au/metadata-

gateway/metadata/record/gcat_60804/GEODATA+TOPO+2.5M+2003

Documentation: www.ga.gov.au/image_cache/GA2657.pdf

Note: For efficiency GEODATA TOPO 10M 2002 was used as the overlay for graphics.

2.3.2.2 Murray–Darling Basin

Title: Murray–Darling Basin Boundary – Water Act 2007

Custodian: Commonwealth Department of the Environment

Metadata: www.environment.gov.au/metadataexplorer/full_metadata.jsp?docId={41F867A7-

789E-49CC-97A4-42378A430829}

2.3.2.3 Natural Resource Management Zones

Title: Natural Resource Management Region Boundaries

Custodian: Commonwealth Department of the Environment

Metadata: www.environment.gov.au/metadataexplorer/full_metadata.jsp?docId={C5C90620-

9FEC-4541-83CA-BFA75C783CFC}

Note: Only NRMs within the Murray–Darling Basin are considered in this study.

2.3.2.4 RECCAP Regions

Title: REgional Carbon Cycle Assessment and Processes regions for Australia

http://www.ga.gov.au/metadata-gateway/metadata/record/gcat_60804/GEODATA+TOPO+2.5M+2003

http://www.ga.gov.au/metadata-gateway/metadata/record/gcat_60804/GEODATA+TOPO+2.5M+2003

https://www.ga.gov.au/image_cache/GA2657.pdf

http://www.environment.gov.au/metadataexplorer/full_metadata.jsp?docId=%7b41F867A7-789E-49CC-97A4-42378A430829%7d

http://www.environment.gov.au/metadataexplorer/full_metadata.jsp?docId=%7b41F867A7-789E-49CC-97A4-42378A430829%7d

http://www.environment.gov.au/metadataexplorer/full_metadata.jsp?docId=%7bC5C90620-9FEC-4541-83CA-BFA75C783CFC%7d

http://www.environment.gov.au/metadataexplorer/full_metadata.jsp?docId=%7bC5C90620-9FEC-4541-83CA-BFA75C783CFC%7d


Custodian: Continental Biogeochemical Cycles Group, CSIRO Marine and Atmospheric

Research

Documentation: Haverd et al. 2013a (www.biogeosciences.net/10/851/2013/bg-10-851-

2013.html)

Note: The 6 RECCAP regions for Australia are a simple aggregation of the 18 agro-climatic

classes of Hutchinson et al. (2005).

2.3.2.5 Terrestrial Ecoregions

Title: Terrestrial Ecoregions in Australia

Custodian: Commonwealth Government Department of the Environment

Documentation: www.environment.gov.au/parks/nrs/science/bioregion-framework/terrestrial-

habitats.html

Note: The seven ecoregions are an aggregation of the Interim Biogeographic Regionalisation

for Australia (IBRA), version 7.

Figure 3: Comparison of the BIOS2 model domain with a land mask derived from Geoscience

Australia’s GEODATA TOPO 2.5M framework layer.

Yellow areas are the parts of the Australian landmass not modelled by BIOS2.

http://www.biogeosciences.net/10/851/2013/bg-10-851-2013.html

http://www.biogeosciences.net/10/851/2013/bg-10-851-2013.html

http://www.environment.gov.au/parks/nrs/science/bioregion-framework/terrestrial-habitats.html

http://www.environment.gov.au/parks/nrs/science/bioregion-framework/terrestrial-habitats.html


Figure 4: Regionalisations used in this study.

Original shapefile definitions (lines) are overlaid on 0.05° rasters (colours). White areas are not

modelled due to lack of soil information.

2.3.3 Modelling

Biospheric carbon fluxes and stores were derived using BIOS2 (Haverd et al., 2013b),

constrained by multiple observation types, and forced using remotely-sensed vegetation

cover. BIOS2 is a fine-spatial-resolution (0.05°) offline modelling environment built on

capability developed for the Australian Water Availability Project (King et al., 2009; Raupach.

et al., 2009). It includes a modification of the Community Atmosphere Biosphere Land

Exchange (CABLE) land surface scheme (Wang et al., 2011b), incorporating the Soil-Litter-

Iso (SLI) soil model (Haverd and Cuntz, 2010) and the Carnegie-Ames-Stanford Approach,

Carbon, Nitrogen, Phosphorus (CASA-CNP) biogeochemical model (Wang et al., 2010).

BIOS2 parameters are constrained and predictions are evaluated using multiple observation

sets from across the Australian continent, including streamflow from 416 gauged catchments,

eddy flux data (CO2 and H2O) from 12 OzFlux sites, litterfall data, and data on soil, litter and

biomass carbon pools (Haverd et al., 2013b).

States andTerritories

NSW

NT

WA

SA

VIC

TAS

ACT

QLD

Ecoregions

Deserts and Xeric Shrublands

Tropical and

Subtropical Moist Broadleaf Forests

Mediterranean Forests,

Woodlands, and Scrub

Temperate Broadleaf and

Mixed Forest

Montane

Grasslands and

Shrublands

Temperate

Grasslands,

Savannas and Shrublands

Tropical and Sub-tropical

Grasslands, Savannas, and

Shrub-lands

RECCAP Tropics

Savanna

Desert

Mediterranean

Cool

Temperate

Warm

Temperate

MDB withNRMs

South West(Qld)

MaranoaBalonne

Border Rivers

Border Rivers-Gwydir

Namoi

Western

Central West

Lachlan

Murrumbidgee

ACTMurray

North East(VIC)Goulburn

Broken

NorthCentralWimmera

Mallee

LowerMurray Darling

SA MurrayDarlingBasin

SA Arid

South East (SA)


CABLE consists of five modules (Wang et al., 2011a):

1. the radiation module which describes direct and diffuse radiation transfer and

absorption by sunlit and shaded leaves

2. the canopy micrometeorology module which describes the surface roughness length,

zero-plane displacement height, and aerodynamic conductance from the reference height

to the air within canopy or to the soil surface;

3. the canopy module includes the coupled energy balance, transpiration, stomatal

conductance, and photosynthesis of sunlit and shaded leaves;

4. the soil module describes heat and water fluxes within soil and snow at their respective

surfaces; and

5. the ecosystem carbon module which accounts for the respiration of stem, root and soil

organic carbon decomposition.

In BIOS2, the default CABLE v1.4 soil and carbon modules were replaced respectively by

the SLI soil model (Haverd and Cuntz, 2010) and the CASA-CNP biogeochemical model

(Wang et al., 2010). Modifications to CABLE, SLI, and CASA-CNP for use in BIOS2 are

detailed in Haverd et al., (2013b).

Nitrogen and phosphorous cycles in CASA-CNP were disabled and land management was

not considered explicitly; however, BIOS2 is driven by remotely-sensed vegetation cover and

parameters and uncertainties were estimated using multiple observation types spanning the

entire bioclimatic space, including managed lands. These two factors mitigate against the

exclusion of potentially important processes. Moreover, model structural errors incurred by

process omission are incorporated in the model-observation residuals, which are propagated

through to uncertainties in model predictions (Haverd et al., 2012).

In this study, BIOS2 simulations were extended back in time to 1911, to assess the effects of

changing climate and atmospheric CO2 on NPP and NEP. CASA-CNP carbon pools were

initialised by spinning the model 200 times over a 39-year period (with initial C-pools updated

at the end of each spin-cycle) using NPP generated with atmospheric CO2 fixed at the pre-

industrial value of 280 parts per million (ppm), and 1911–1949 meteorology, corresponding

to the earliest available rainfall and temperature data from the Bureau of Meteorology’s

Australian Water Availability Project dataset (Grant et al., 2008; Jones et al., 2009).

Following spin-up, the 1911–2011 simulation was performed using actual deseasonalised

atmospheric CO2 (from the Law Dome ice core prior to 1959 [MacFarling Meure et al., 2006],


and from global in-situ observations from 1959 onward [Keeling et al., 2001] with repeated

1911–1949 meteorology prior to 1911 and actual meteorology thereafter).

Uncertainty in BIOS2 predictions (all uncertainties hereafter expressed as onestandard

deviation), due to parameter uncertainty and uncertainty in forcing data were estimated

separately and combined in quadrature to give total uncertainty, as described by Haverd et

al. (2012). To obtain uncertainties in model predictions associated with parameter

uncertainties in a parameter set p, the parameter covariance matrix C was projected onto the

variance in the prediction. Uncertainties in model predictions associated with forcing

uncertainties were estimated as the absolute change in prediction associated with

perturbations to forcing inputs. NEP uncertainty estimates also include the uncertainty due to

an assumed 20 per cent uncertainty in the partial derivative of NPP with respect to

atmospheric CO2 concentration.

2.3.4 Presentation of results

Results are presented as annual time series of NEP and total biospheric carbon stock, as

well as 10-year running mean NEP, converted to units of percentile rank. The results are

defined according to the accounting framework (see next).

2.4 Accounting frameworks

The accounting framework adopted in this study is the System of Environmental-Economic

Accounting (SEEA) Central Framework (European Commission et al. 2012) and SEEA

Experimental Ecosystem Accounting (European Commission et al. 2013). The scoping and

specification for the prototype ecosystem-style account will be documented in an account

specification based on the Environmental accounting framing workbook (Bureau of

Meteorology 2013b) and the Guide to environmental accounting in Australia (Bureau of

Meteorology 2013a). The tables and charts presented in the results below are specified

according to SEEA structures, but have not yet been confirmed as meeting SEEA standards.

This study has essentially produced data in an accounting format, but further work is needed

to convert them into a publishable account. The prototype accounts are essentially carbon

stocks and flows as they relate to the fluxes of carbon into and out of the ecosystem carbon

pools across Australia.


Figure 5 gives an overview of the spatial and interannual variability of Net Ecosystem

Production (NEP) across the Australian continent. Across most of Australia, variability in NEP

is strongly related to rainfall, although there are time-lags (of the order of months to years) in

the relationship. For example, the high NEP in south-eastern Australia in 2010 coincides with

high rainfall and hence high NPP. This high rainfall and NPP persist in 2011 but the NEP is

reduced due to the increased heterotrophic respiration associated with higher litter-fall and

elevated soil moisture content.

The remaining figures disaggregate the information in Figure 5 spatially and temporally. Each

type of time series gives information at a different temporal scale:

1. The annual time series of NEP show the interannually varying uptake and release of

carbon as the biosphere responds to interannual variations in weather.

2. The 10-year running mean percentile rank time series show the biospheric carbon

accumulation response to decadal climate variability. The units of percentile rank allow

time series with different amplitudes of variation to be compared on the same scale. For

an example, see the comparison of precipitation, NPP, and NEP in Haverd et al., 2013a.

3. The time series of total carbon stock, which is dominated by the contribution of the soil,

vary at time scales of multiple decades, and reflect trends in drivers, especially rainfall,

temperature, and CO2.

3 Results


Figure 5: Net Ecosystem Production for the Australian continent.


Figure 6: Net Ecosystem Production for the Murray–Darling Basin.


Figure 7.1: Net Ecosystem Production aggregated for the Australian spatial domain.

Figure 7.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated for

the Australian spatial domain.


Figure 7.3: Total landscape carbon stock aggregated for the Australian spatial domain.


Figure 8.1: Net Ecosystem Production aggregated for Australian States and Territories spatial

domains.


Australian States and Territories spatial domains.


Figure 8.3: Total landscape carbon stock aggregated for Australian States and Territories

spatial domains.


Figure 9.1: Net Ecosystem Production aggregated for the Murray–Darling Basin spatial domain.


the Murray–Darling Basin spatial domain.


Figure 9.3: Total landscape carbon stock aggregated for the Murray–Darling Basin spatial

domain.


Figure 10.1: Net Ecosystem Production aggregated for NRMs intersecting the Murray–Darling

Basin.


Figure 10.2: Percentile rank of Net Ecosystem Production (10-year running mean) aggregated

for NRMs intersecting the Murray–Darling Basin.


Figure 10.3: Total landscape carbon stock aggregated for NRMs intersecting the Murray–

Darling Basin.


Figure 11.1: Net Ecosystem Production aggregated for the RECCAP regional domains.


for the RECCAP regional domains.


Figure 11.3: Total landscape carbon stock aggregated for the RECCAP regional domains.


Figure 12.1: Net Ecosystem Production aggregated for the IBRA Terrestrial Ecoregions.


for the IBRA Terrestrial Ecoregions.


Figure 12.3: Total landscape carbon stock aggregated for the IBRA Terrestrial Ecoregions.


4.1 What can the results tell us?

These results can be used to detect impacts of rising CO2 and climate variability on carbon

accumulation, but not the impacts of disturbance (e.g., land-clearing and bushfires).

The changes can be interpreted in an integrated way, insofar as carbon stock change can be

linked to changes in resource (energy and water) availability. Attribution of NEP to soil, litter

and vegetation compartments is also an output of the model. Further, the biospheric

components of the carbon budget can be linked to the entire carbon budget, as indicated in

Figure 13, reproduced from Haverd et al., 2013a.

Source: Haverd et al., 2013a.

Figure 13: The Australian carbon budget 1990–2011.

Classes of change suitable for accounting purposes may be defined by identifying changes

in a particular carbon stock (e.g., biomass) which perturb the state beyond a threshold, such

that a transition between ecosystem types occurs (e.g., rainforest > savannah; savannah >

4 Discussion


grassland; closed forest > open forest); however, identification of such thresholds is outside

the scope of this study, and detection of such changes would require the inclusion of a

dynamic vegetation module in BIOS2.

Total Ecosystem Potential could be assessed by establishing the maximum carrying

capacity, assuming potential vegetation cover, deduced, for example, from a map of inferred

pre-European vegetation cover.

Results for total carbon stock could be used to populate the ‘Primary reservoirs of biocarbon’

in the ‘Carbon stock and flow account framework’ presented by Adjani et al. (2013).

4.2 What can’t the results tell us? (constraints and uncertainties)

BIOS2 was used to estimate NEP (defined as biospheric carbon accumulation in the

absence of disturbance) under the assumption of current vegetation cover. Therefore, the

results do not tell us about changes in stocks due to disturbance, particularly land use and

land cover change.

BIOS2 uncertainties were evaluated in Haverd et al., (2013b). Uncertainties (1 standard

deviation) in annual NEP are around 0.1 g C m-2 d-1, and around 0.05 g C m-2 d-1 for

decadally averaged NEP. These are dominated by contributions from uncertainties in forcing

data, namely prescribed vegetation cover and meteorology, and particularly their effect on

initial conditions. In contrast, uncertainties in long-term carbon stocks are dominated by

parameter error (which implicitly includes model structural errors) and are about 70 per cent

of the total stock.

Parameter and model structural errors are likely to be reduced by future model developments

enabling the range of data constraints to be extended via a process of model–data fusion.

Uncertainties due to forcing data are less amenable to reduction, with results being sensitive

to artefacts in the forcing data. For example, solar radiation data are only available from

1990, with a climatology being used prior to the period of data availability. This is not obvious

in most of the results because carbon accumulation is limited by water resources rather than

energy across most of Australia; however, in Tasmania, the artefact in the radiation forcing

data is clearly evident (Figure 8.1), with an onset of high interannual variability in NEP

occurring in 1990.


4.3 Working with temporal and spatial scale

4.3.1 Temporal scale

High variability at monthly and annual time scales means longer-term averages are required

as a measure of ecosystem integrity. On the other hand, shifts in seasonality (e.g., due to

changing annual rainfall patterns) can be inferred from monthly data. Total carbon stocks

vary over much longer time scales of multiple decades.

Hindcasts can be repeated when improved methods and data sources are developed.

4.3.2 Spatial scale

The BIOS2 outputs are equally valid at any size aggregation greater than or equal to 0.05°

(approximately 5 km) which is the spatial resolution of the meteorological inputs; however,

these uncertainty estimates do not account for spatial variation in driver uncertainty, with

temporal and spatial variation in the sparseness of precipitation observations being a

particular concern.

4.4 Future directions

4.4.1 Consideration of disturbance: land use change and fire

An important evolution of this study will be to include the effects of disturbance on biospheric

carbon accumulation. This will require the model to be developed from one which is forced

externally by remotely-sensed vegetation cover, to one which prognoses vegetation cover

internally, allowing it to respond dynamically to fire, land-clearing, re-afforestation, and

changes in land management practices. Further, data-sets which provide continental

coverage of disturbance history (fire, land cover, and land use change) will be required as

model input.

4.4.2 Consideration of limiting resources: water, nutrients, or energy

This study could be extended to identify whether the limiting resource for NEP is water,

nutrients, or energy. While carbon, water and energy cycles are linked in BIOS2 (water and

energy balances not presented here), nutrient cycles have not been considered in the current

study; however, nitrogen and phosphorous cycles are encoded within BIOS2 and could be

enabled in future work.


4.4.3 Quantifying natural capital

An extension of this study to produce natural capital results would require carbon stocks to

be related to a baseline threshold. For example, the State of the Environment report (State of

the Environment 2011 Committee 2011) pointed to critical soil organic carbon (SOC) levels.

Maintenance of SOC above these levels is important for soil fertility, water-holding capacity

and pH buffering. Similarly, vegetation cover could be related to a level below which the

ecosystem flips to a different type, or accelerates other forms of degradation such as

erosion.

Extension of this study would require the identification of such thresholds, and the model

would need to be developed to allow for dynamic vegetation cover.

4.4.4 Natural capital degradation

The 2011 State of the Environment report (State of the Environment 2011 Committee 2011)

points to a need for identifying the state of soil carbon, where it has been degraded, trends in

condition and explanatory factors. The study presented here starts to identify such factors,

but it would be useful to extend it to account for the effect of management, especially

grazing. In the fuller system of national accounts, this would allow us to identify and assess

trade-offs between regional economic return, and impacts on ecosystem health.

Degradation could be quantified by including ecosystem responses to disturbance. In

particular being able to predict the rates of recovery following disturbance will allow us to

assess the impact of changing disturbance frequency and/or intensity on carbon stocks. This

is especially relevant to quantifying the threshold below which the capacity to deliver

ecosystem services is being degraded.

4.4.5 How can this study inform measures of resilience?

SEEA Experimental Ecosystem Accounting is focused on achieving an integrated view of

ecosystem functioning. Ideally, this should be based on as direct measures or estimates of

ecosystem functioning as possible; one such measure is resilience.

Resilience is defined as ‘the capacity of a system to absorb disturbance and reorganise so

as to retain essentially the same function, structure, and feedbacks’ (Walker and Salt, 2006;

Walker and Salt, 2012). It is an important attribute of social-ecological systems, and one that

can only be assessed by taking a whole-of-system approach. Given that carbon dynamics in

the terrestrial ecosystem provide a key underpinning to many ecosystem goods and


services, any assessment of resilience will need to draw on knowledge of the carbon stocks

and flows as developed in this study; however, these quantities by themselves are

insufficient to make conclusions about resilience, and require an interpretation that is

integrated into the broader SEEA framework.

The SEEA Central Framework (European Commission et al., 2012) considers individual

environmental assets, each in isolation (e.g., timber resources), whereas a key feature of the

SEEA Experimental Ecosystem Accounting framework (see Section 2.5; European

Commission et al., 2013) is that it takes a broader ecosystem perspective that includes the

interactions between environmental assets. This extended framework offers the prospect of

being able to make inferences about system resilience.

Given Australia’s history of floods, droughts, fires, land use change and other sources of

disturbance, it is helpful to be able to assess the resilience of key environmental assets and

ecosystem services to such disturbances. Furthermore, accounting for links between

ecological, social and economic systems allows the possibility of characterising the resilience

of important social and economic benefits to ecosystem change (e.g., resilience of rural

economies to drought). It is important to note that resilience is not always desirable (e.g., a

degraded ecosystem may be resilient to attempts to restore it to its previous condition).

Assessments of ‘specified resilience’ identify the assets or services of interest and assess

their resilience in the face of specified disturbances (e.g., the resilience of coastal

communities to cyclones and storm surges). Assessments of ‘general resilience’ recognise

that there are some attributes of social-ecological systems that confer resilience to a range of

unspecified shocks or disturbances. For example, ‘response diversity’, or the diversity of

responses to environmental change among species contributing to the same ecosystem

function, has been identified as a key contributor to general resilience (Elmqvist et al., 2003).

Characterising general resilience is more challenging than specified resilience, so focusing

on specified resilience would be an appropriate place to start when interpreting SEEA work.

There are many well-identified challenges in using ecosystem accounting frameworks to

assess resilience. In a review of the SEEA Central Framework, Walker and Pearson (2007)

outlined some of these challenges. Of particular interest is the fact that dynamic, interactive

processes in natural ecosystems are not well captured by any linear, compartmentalised

system of accounting. Feedbacks and flow-on effects cannot be adequately accounted for,

and linear accounting frameworks can overlook the existence of critical thresholds in

ecosystems that, once crossed, lead to collapse (and so a stock near such a threshold can


be over-valued as a small perturbation in that stock will lead to a rapid and potentially

irreversible loss).

The SEEA Experimental Ecosystem Accounting Framework (European Commission et al.

2013) acknowledges many of the challenges identified by Walker and Pearson (2007) and its

broader scope and system-wide emphasis provides some opportunity to include resilience

considerations when developing and interpreting ecosystem accounts. The underlying

modelling method used in this study can be developed to make progress in this direction.

The Prime Minister’s Science, Engineering and Innovation Council stressed the tight

interconnections between carbon, energy and water in Australia: ‘These exchanges connect

stationary energy systems, water systems, land systems and food production, transport

systems, built environments, industrial systems, and the ecosystems upon which all these

aspects of the human enterprise are based.’ (PMSEIC 2010) The report authors

recommended that ‘comprehensive, rigorous and transparent accounting for energy, water

resources, greenhouse gas emissions and carbon stocks will enable administrative systems

to identify and avoid perverse effects’, and recommended initiatives that focus on building

resilience in landscapes, towns and cities.

The terrestrial carbon accounts described in this report are derived from a dynamic model

that explicitly accounts for many important linkages identified in the PMSEIC (2010) report

(Haverd et al., 2013b). For example, water balance accounts are included as a necessary

underpinning to calculating different terms in the terrestrial carbon budget (e.g., water cycle

constraints on NPP, or freshwater fluxes of dissolved organic carbon). Explicitly included in

the broader Australian Terrestrial Carbon Budget (Haverd et al., 2013a) are fossil fuel

emissions which are readily linked to activity in the energy sector. Furthermore, disturbances

such as land use change and fire could be readily accommodated in the model, again

providing the opportunity for self-consistent accounting that links across sectors. Such a

capacity can also be extended to include nutrient dynamics.

An outcome of such an integrating framework is the opportunity to report on how climate–

carbon–water–energy–nutrient–vegetation–disturbance links are evolving over time. This is

important for assessing resilience, where the interacting consequences over time are of more

relevance than stand-alone snapshots of individual environmental assets. The Stockholm

Resilience Centre maintains a ‘Regime Shifts Database’ (www.regimeshifts.org) that

provides examples of the characteristics of social-ecological regime shifts and the suite of

variables needed to monitor and understand such shifts. For example, the ‘forests to

savannahs’ regime shift is driven by interactions between land use change, fire, global

http://www.regimeshifts.org/


climate change, and local climate-vegetation feedbacks, and in turn affects key ecosystem

processes such as primary production, water cycling, soil formation, nutrient cycling, and

related provisioning, regulating and cultural services. The likelihood of an environmental

asset undergoing a regime shift is a measure of its resilience.

Maintaining a set of ecosystem accounts that track and report on these interactions serves

two functions:

1. provision of the necessary data to better characterise and understand such regime shifts

when they occur; and

2. opportunity to provide early warnings when systems are approaching known thresholds.

An approach used in the resilience community is to identify either known thresholds or

‘thresholds of potential concern’ in social-ecological systems, and devise measuring and

monitoring programmes to be able to track where the system is relative to these thresholds

(Walker and Salt, 2012). The maintenance of integrated ecosystem accounts contributes to

that monitoring capacity. The 2011 State of Environment report (State of the Environment

2011 Committee 2011) includes some discussion on resilience and where possible identifies

thresholds relevant to Australian ecosystems. For example, critical thresholds for soil organic

carbon, pH, and surface vegetation cover are described for Australian soils.

The terrestrial carbon accounts developed for this study currently do not include important

feedbacks between climate and vegetation; however, this is under development. Such a

capacity for including feedbacks of this kind would be helpful for identifying where there are

desirable and undesirable feedback loops at work in the landscape (e.g., feedback loops

fuelling desertification). This knowledge is important because not all shocks to a system

come from externally imposed disturbance such as fire or flood, but rather are the result of

slowly changing variables that trigger a rapid shift due to changes in internal system

feedbacks (Scheffer et al., 2001).

Finally, in the long run there is the prospect of tracking interactions and trade-offs between

different capitals. It is helpful to know what capitals are contributing to resilience and adaptive

capacity in the face of disturbances and be able to monitor where, for example, economic

activity in a region is viable only because that activity is drawing down social and natural

capital stocks (rather than living off the surplus flowing from those stocks).


4.4.6 Potential extensions and applications

Users of this work may broadly include government agencies, researchers, and academics,

international agencies, policy-makers and environment and community groups. The potential

users represent a variety of perspectives: from users and managers of the account

phenomena (i.e., ecosystems) to decision makers, researchers, and analysts.

More specifically, the study may contribute to the development of carbon stock accounts as

part of the work underway between the former Department of Industry, Innovation, Climate

Change, Science, Research and Tertiary Education (DIICCSRTE, now in the Department of

the Environment), the Australian Bureau of Statistics (ABS) and the Australian National

University (ANU).

A national trial that extends the current work also has potential to contribute to work being

conducted by the Australian Bureau of Agricultural and Resource Economics and Sciences

(ABARES) and the Terrestrial Ecosystem Research Network (TERN).

Further, the study may contribute to the work underway with the Murray–Darling Basin

Authority, link to the Wentworth Group/NRM Regional Environmental Accounting trials, and

to the ABS Land Accounts.

As a demonstrator of environmental accounting in Australia, the study will be of interest to

the Commonwealth Department of the Environment, in particular with reference to its

potential contributions to State of the Environment reporting and Sustainability reporting.


The purpose of this study was to determine the potential of carbon flux modelling for

measuring the capacity of ecosystems to provide ecosystem services. The value of doing so

lies in the central role of carbon in all life forms and living systems. By measuring the fluxes

of carbon in and out of ecosystems it is possible to obtain fundamental information about

their functioning and their status. A further feature of carbon modelling is that, by necessity, it

must integrate information about water and energy cycles. A central proposition of this study

is that placing this combined information into an accounting format will enhance the

management of ecosystems by tracking whether they, and the ecosystem services that flow

from them, are being maintained or degraded over time. If this can be shown to be a valid

proposition, then continental-wide carbon flux modelling could become an important source

of nationally-consistent data relevant to the improved management of Australia’s natural

capital.

This study has made progress towards this objective by clarifying the temporal and spatial

scales that CSIRO’s BIOS2 land surface model is currently capable of delivering. The model

was used to estimate daily NEP (defined as biospheric carbon accumulation in the absence

of disturbance) for the past 101 years under the assumption of current vegetation cover. It is

important to note that the results do not yet tell us about shorter term changes in carbon

stocks due to ecosystem disturbance, particularly land use and land cover change. The study

identified that the uncertainties in the BIOS2 model result are well described and are likely to

be reduced in the future.

An important part of the study was to see if the carbon flux data could be formatted to meet

the requirements of the SEEA Experimental Ecosystems Accounting carbon accounting

model. The study showed that the modelled total carbon stock could be used to populate the

‘Primary reservoirs of biocarbon’ table in the ‘Carbon stock and flow account framework’

presented by Adjani et al. (2013) and consistent with the SEEA Experimental Ecosystems

Accounting carbon accounting model. Further work is necessary to ensure the tables meet

accounting rules.

With respect to data aggregation, it would be appropriate to interpret the BIOS2 study results

at any aggregation of approximately 5 km; however, uncertainties associated with the

variable distribution of precipitation data are a constraining factor and drive the interpretation.

The result is that, in general, areas with well-defined precipitation can achieve aggregation at

5 Conclusion


NRM regional scales. In summary, the data can, with a few exceptions, be aggregated by

NRM Region, Murray–Darling Basin, State, Territory, and the entire continent and it is

appropriate to interpret NEP at monthly (seasonal), annual and decadal time-scales.

Options to further develop the methods used in this study are being actively pursued by the

CSIRO. To add in ‘disturbance’ drivers (fire, land cover and land use change), the model

requires further data-sets which provide continental coverage of disturbance history. To

report on natural capital and changes in natural capital (e.g., degradation or restoration)

requires the identification of thresholds of change. Extensions of this study would require the

identification of such thresholds, and the model would need to be developed to allow for

dynamic vegetation cover.

The study presents a valuable discussion on the contribution of carbon modelling to

estimating ecosystem and social-ecological resilience. The discussion argued that, when

seeking to estimate resilience, carbon-cycle modelling makes a useful contribution, but is not

sufficient on its own and additional information is needed. Pathways were identified for

extending the work and for interpreting the results as they relate to resilience within the

context of the SEEA.

Overall, the study has made a useful contribution to understanding the potential for using

carbon flux modelling as a data source for ecosystem accounting in Australia, particularly

with regard to important technical issues. The study also sets out pathways for the future

development of the approach. Finally, the results themselves give a glimpse of the changes

in ecosystem functioning across the entire continent over the past 100 years; in effect

providing long-term average climatologies of background ecosystem growth regimes as

driven by, among other things, water, solar inputs, atmospheric CO2 concentrations, and

temperature.


Ajani JI, Keither H, Blakers M, Mackey BG, King HP 2013, 'Comprehensive carbon stock and flow accounting: a national framework to support climate change mitigation policy', Ecological Economics, vol. 89, pp. 61–71.

Barrett DJ 2001. NPP multi-biome: VAST calibration data, Oak Ridge National Laboratory Distributed Active Archive Centre, Oak Ridge, Tennessee.

Bureau of Meteorology 2013a, Guide to environmental accounting in Australia, Environmental

Information Programme Publication Series no. 3, Bureau of Meteorology, Canberra.

Bureau of Meteorology 2013b, Environmental account framing workbook, Bureau of Meteorology,

Canberra.

Donohue RJ, McVicar TR Roderick ML 2009, ‘Climate-related trends in Australian vegetation cover as inferred from satellite observations, 1981–2006.’ Global Change Biology, vol. 15, issue 4, pp.

1025–39.

Emqvist T, Folke C, Nystrom M, Peterson G, Bengtsson J, Walker B, Norberg J 2003, 'Response diversity, ecosystem change, and resilience’, Frontiers in Ecology and the Environment, vol. 1, issue 9, pp. 488–94.

European Commission, Organisation for Economic Co-Operation and Development, United Nations and World Bank 2013, System of Environmental–-Economic Accounting 2012: experimental ecosystem accounting, white cover publication, pre-edited text subject to official editing, viewed

18 July 2013, unstats.un.org/unsd/envaccounting/eea_white_cover.pdf

European Commission, Food and Agriculture Organisation, International Monetary Fund, Organisation for Economic Co-operation and Development, United Nations and World Bank 2012, System of Environmental–-Economic Accounting Central Framework, United Nations, New York.

Grant I, Jones D, Wang W, Fawcett R, Barratt D 2008, 'Meteorological and remotely sensed datasets for hydrological modelling; a contribution to the Australian Water Availability Project, catchment-scale hydrological modelling and data assimilation (CAHMDA-3), International Workshop on Hydrological Prediction: Modelling, Observation and Data Assimilation, Melbourne.

Hattersley PW 1983, ‘The distribution of C3-grass and C4-grasses in Australia in relation to climate’, Oecologia, vol. 57, issues 1–2, pp. 113-28.

Haverd V, Cuntz, M 2010, ‘Soil-Litter-Iso: a one-dimensional model for coupled transport of heat, water and stable isotopes in soil with a litter layer and root extraction’, Journal of Hydrology,

vol. 388, issues 3–4, pp. 438–55.

Haverd V, Raupach MR, Briggs PR, Canadell JG, Davis SJ, Law RM, Meyer CP, Peters GP, Pickett-Heaps C, Sherman B 2013a, ‘The Australian terrestrial carbon budget’, Biogeosciences, vol.

10, pp. 1–19.

Haverd V, Raupach MR, Briggs PR, Canadell JG, Isaac P, Pickett-Heaps C, Roxburgh SH, van Gorsel E, Viscarra Rossel RA, Wang Z 2012, ‘Multiple observation types reduce uncertainty in Australia's terrestrial carbon and water cycles’, Biogeosciences Discuss., vol. 9, issue 9, pp.

12,181–258.

Haverd V, Raupach MR, Briggs PR, Canadell JG, Isaac P, Pickett-Heaps C, Roxburgh SH, van Gorsel E, Viscarra Rossel RA, Wang Z 2013b, ‘Multiple observation types reduce uncertainty in Australia's terrestrial carbon and water cycles’, Biogeosciences, vol. 10, issue 3, pp. 2011–40.

Houldcroft CJ, Grey WMF, Barnsley M, Taylor CM, Los SO, North PRJ 2009, ‘New vegetation albedo parameters and global fields of soil background albedo derived from MODIS for use in a climate model’, Journal of Hydrometeorology, vol. 10, issue 1, pp. 183–98.

References

http://unstats.un.org/unsd/envaccounting/eea_white_cover.pdf


Hutchinson MF, McIntyre S, Hobbs RJ, Stein JL, Garnett S, Kinloch J 2005, 'Integrating a global agro-climatic classification with bioregional boundaries in Australia', Global Ecology and Biogeography, vol. 14, issue 3, pp. 197–212.

Jones DA, Wang W, Fawcett R 2009, 'High-quality spatial climate data-sets for Australia', Australian Meteorological and Oceanographic Journal, vol. 58, pp. 233–48.

Keeling, CD, Piper SC, Bacastow RB, Wahlen M, Whorf TP, Heimann M, Meijer HA 2001, 'I. Global aspects', Exchanges of atmospheric CO2 and 13CO2 with the terrestrial biosphere and oceans from 1978 to 2000, Scripps Institution of Oceanography, San Diego.

King EA, Paget MJ, Briggs PR, Trudinger CM, Raupach MR 2009, 'Operational delivery of hydro-meteorological monitoring and modeling over the Australian continent, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 2, issue 4, pp. 241–

49.

Kirschbaum MUF, Eamus D, Gifford RM, Roxburgh SH, Sands PJ 2001, 'Definitions of some ecological terms commonly used in carbon accounting', CRC for Greenhouse Accounting NEE Workshop Proceedings, Canberra, 18–21 April 2001.

Lu H, Raupach MR, McVicar TR, Barrett DJ 2003, 'Decomposition of vegetation cover into woody and herbaceous components using AVHRR NDVI time series', Remote Sensing of Environment,

vol. 86, issue 1, pp. 1–18.

MacFarling Meure CM, Etheridge D, Trudinger C, Steele P, Langenfelds R, van Ommen T, Smith A, Elkins J, 2006, 'Law Dome CO2, CH4 and N2O ice core records extended to 2000 years BP', Geophysical Research Letters, vol. 33, issue 4, pp. 1–4.

McKenzie NJ, Hook J 1992, Interpretation of the Atlas of Australian Soils, Technical report no 94,

CSIRO Division of Soils, Canberra.

McKenzie NJ, Jacquier DW, Ashton LJ Creswell HP 2000, Estimation of soil properties using the Atlas

of Australian Soils, CSIRO Land and Water, Canberra.

Northcote KH, Beckmann GG, Bettenay E, Churchward HM, Van Dijk DC, Dimmock GM, Hubble GD, Isbell RF, McArthur WM, Murtha GG, Nicolls KD, Paton TR, Thompson CH, Webb AA and Wright MJ 1960, Atlas of Australian Soils, sheets 1–10, with explanatory data, CSIRO

Australia and Melbourne University Press, Melbourne.

Northcote KH, Hubble GD, Isbell RF, Thompson CH, Bettenay E 1975, A Description of Australian Soils, CSIRO, Canberra, pp. 1–170.

Prime Minister’s Science, Engineering and Innovation Council 2010, Challenges at Energy–Water–Carbon Intersections, PMSEI, Canberra.

Raison J, Keith H, Barrett DJ, Burrows B, Grierson P 2003, ‘Spatial estimates of biomass in ‘mature’ native vegetation’, Sustainability Science, issue 1, vol. 1, pp. 91–106.

Raupach MRR, Briggs PR, Haverd V, King EA, Paget M, Trudinger CM 2009, Australian water availability project (AWAP): CSIRO marine and atmospheric research component, final report

for phase 3, CSIRO and Bureau of Meteorology, Canberra.

Scheffer M, Carpenter S, Foley JA, Folke C, Walke, B 2001, ‘Catastrophic shifts in ecosystems’, Nature, vol. 413, issue 6,856, pp. 591–6.

State of the Environment 2011 Committee 2011, Australia State of the Environment 2011, Independent report to the Commonwealth Government Minister for Sustainability, Environment, Water, Population and Communities, Department of Sustainability, Environment, Water, Population and Communities, Canberra.

Vaze J, Chiew FHS Perraud JM, Viney N, Post D, Teng J, Wang B, Lerat J, Goswami M 2011, ‘Rainfall-runoff modelling across southeast Australia: datasets, models and results’, Australian Journal of Water Resources, vol. 14, issue 2, pp. 101–13.

Walker BH, Salt D 2006, Resilience thinking: sustaining ecosystems and people in a changing world,

Island Press, Washington DC, USA.


Walker BH, Pearson L 2007, ‘A resilience perspective of the SEEA’, Ecological Economics, vol. 61,

issue 4, pp. 708–15.

Walker BH, Salt D 2012, ‘Resilience practice: building capacity to absorb disturbance and maintain function’ Island Press, Washington DC,USA.

Wang WL, Dungan J, Hashimoto H, Michaelis AR, Milesi C, Ichii K, Nemani RR 2011a, ‘Diagnosing and assessing uncertainties of terrestrial ecosystem models in a multimodel ensemble experiment: 2. Carbon balance’, Global Change Biology, vol. 17, issue 3, pp. 1367–78.

Wang YP, Kowalczyk E, Leuning R, Abramowitz G, Raupach MR, Pak B, van Gorsel E, Luhar A 2011b, ‘Diagnosing errors in a land surface model (CABLE) in the time and frequency domains’, Journal of Geophysical Research-Biogeosciences, vol. 116, issue G1.

Wang YP, Law RM and Pak B 2010, ‘A global model of carbon, nitrogen and phosphorus cycles for the terrestrial biosphere’, Biogeosciences, vol. 7, issue 7, pp. 2261–82.

Zhang YQ, Viney N, Chen Y, Li HY 2011, Collation of streamflow dataset for 719 unregulated Australian catchments, Water for a Healthy Country National Research Flagship, CSIRO,

Canberra.

the use of carbon cycle modelling for the production of ......marine and atmospheric research was...

Documents