pioneer 1998-2001 preparation and integration of analysis tools towards operational forecast of...

01Preparation and integration of analysis tools towards operational forecast of nutrients in estuaries of European rivers (PIONEER).

Preparation and integration of analysis tools towards operational forecast of nutrients in estuaries of European rivers (PIONEER).

Preparation and integration of analysis tools towards operational forecast of nutrients in estuaries of European rivers (PIONEER).Contract No MAS3-CT98-0170

1 September 1998 – August 2001

Coordinator Hans von Storch, GKSS

Scientific Manager Konstanze Reichert, OceanWaveS GmbH

Participants

Institute for Coastal Research, GKSS - Geesthacht, Germany

Maritime Research Institute – Szczecin, Poland

Technical University Szczecin, Water Environment Engineering Department – Szczecin, Poland

Universitat Politècnica de Catalunya, Laboratorio d'Ingenieria Maritima – Barcelona, Spain

Universidad Politecnica de Valencia, Departamento de Ingenieria Hidraulica y Medio Ambiente – Valencia, Spain

Københavns Universitet, Institute of Geography - København, Denmark

Danish Hydraulic Institute, Water and Environment - Hørsholm, Denmark

Netherlands Institute for Sea Research, Department of Biological Oceanography - Texel, Den Burg, The Netherlands

Nansen Environmental and Remote Sensing Center – Bergen, Norway

Association pour la Recherche et le Développement desMéthodes et Processus Industriels – ARMINES, Ecole Nationale Supérieure des Mines de Paris, Centre de Géostatistique – Fontainebleau, France

Objectives:The main objective of PIONEER is the development of techniques for the day-to-day monitoring, analysis and short-term prediction of nutrient and related suspended matter distributions in estuaries.

Such systems are expected to be routinely operated by management authorities as well as commercial companies in the future.

Brief description The sustainable management of the coastal zone requires routine monitoring and assessment of the status of the ecosystem "coastal zone".

To obtain an efficient and cost-effective system, „intelligent“ observational strategies combined with an analysis software encoding our present knowledge about the dynamics of the considered system are needed.

This software determines the informational value of actual observations, combines all observations with previous forecasts and returns a best guess of the detailed present space-time state.

PIONEER sets up analysis systems for routine day-to-day monitoring, analysis and short-term prediction of nutrient distributions in the Odra and Ebro estuaries.

The project integrates presently available technology and methodology in data management, geostatistical and dynamical data assimilation and numerical modelling in co-operation between scientific institutions, management authorities and commercial companies.

The overall approach of PIONEER parallels the analysis in weather forecasting. Point observations together with a „best guess“ are processed in a data assimilation scheme.

Since data assimilation with respect to nutrients is a new application three schemes of increasing complexity are explored: simple spatial interpolation, geostatistics and dynamical data assimilation.

The systems developed by PIONEER include:

• State-of-the-art data assimilation schemes to determine the actual spatial and temporal distribution of nutrients and water quality parameters.

• Dynamical models to forecast temporal and spatial evolution of water quality parameters for several days and weeks.

• Data management systems for long-term storage and fast exchange.

Plan of presentation

• Data and models for the Ebro estuaryby Joan Pau Sierra

• Data and models for the Odra estuaryby Dorota Dybkowska-Stefek

• Combining data and models for analysis and prediction purposes

by Hans Wackernagel

• Conclusions and Outlookby Hans von Storch

Plan of presentation

• Data and models for the Ebro estuaryby Joan Pau Sierra

• Data and models for the Odra estuaryby Dorota Dybkowska-Stefek

• Combining data and models for analysis and prediction purposes

by Hans Wackernagel

• Conclusions and Outlookby Hans von Storch

01Data and models for the Ebro estuary

by Joan Pau SierraUniversitat Politècnica de Catalunya, Laboratorio d'Ingenieria Maritima – Barcelona, Spain

• Introduction• Field campaigns• Estuary modelling• Plume modelling• Conclusions

Introduction

C apT orto sa

T rab u cad o rA lfacsB ay

F ang arS p it

L a B an y aS p it

2 9 4 0 0 0 2 9 9 0 0 0 3 0 4 0 0 0 3 0 9 0 0 0 3 1 4 0 0 0 3 1 9 0 0 0 3 2 4 0 0 0

L o n g itu d e (U T M )

4 4 9 5 0 0 0

4 5 0 0 0 0 0

4 5 0 5 0 0 0

4 5 1 0 0 0 0

4 5 1 5 0 0 0

4 5 2 0 0 0 0L

M E D IT E R R A N E A N S E A

E bro riv e r

B ud a Is .

E uca lip to s

Field campaigns (I)

• Available data:

– Meteorology (different stations - every 10’)

– Tide level (2 gauges – every 30’)

– Waves (Cap Tortosa – every 3 h)

– River discharge (Tortosa – every 1 h)

– Salinity, SS, nutrients (Tortosa – every month)

Field campaigns (II)

• Field campaigns • (1 per season – estuary & plume):

– Salinity– Temperature– Current profiles (ADP)– Nutrients (nitrate, nitrite, ammonium,

silicates, SRP)– Chlorophyll a

Field campaigns (III)

C apT orto sa

T rabu cado rA lfacsB ay

F ang arS p it

L a B an yaS p it

2 9 4 0 0 0 2 9 9 0 0 0 3 0 4 0 0 0 3 0 9 0 0 0 3 1 4 0 0 0 3 1 9 0 0 0 3 2 4 0 0 0

4 4 9 5 0 0 0

4 5 0 0 0 0 0

4 5 0 5 0 0 0

4 5 1 0 0 0 0

4 5 1 5 0 0 0

4 5 2 0 0 0 0

r 5r 6

r 2r 1

r 5r 6

m 0 1m 0 2

m 0 5m 0 6 m 0 7

m 0 8r 1

r 2r 3

r 5r 8

E bro rive r

B ud a Is .

E uca lip to s

C apT o rto sa

T rabu cado rA lfacsB ay

F an g arS p it

L a B an yaS p it

2 9 4 0 0 0 2 9 9 0 0 0 3 0 4 0 0 0 3 0 9 0 0 0 3 1 4 0 0 0 3 1 9 0 0 0 3 2 4 0 0 0

4 4 9 5 0 0 0

4 5 0 0 0 0 0

4 5 0 5 0 0 0

4 5 1 0 0 0 0

4 5 1 5 0 0 0

4 5 2 0 0 0 0

r 2r 3

r 4r 5

r 6r 7 b

r 5r 6

r 7 aM 1

M 4M 5M 6

M 7M 1

P IO N E E R 2 S am p lin g S tatio n sR iv e r S ta tio n s

S e a S ta tio n s (J u ly 1 0 th , 1 9 9 9 )

S e a S ta tio n s (J u ly 1 1 th , 1 9 9 9 )

S u rfa c e T ra n se c t (Ju ly 1 0 th , 1 9 9 9 )

S u rfa c e T ra n se c t (Ju ly 1 1 th , 1 9 9 9 )

E b ro rive r

B u d a Is .

E u ca lip to s

Measurements. Salt wedge (summer)

0 2 4 6 8 1 0 1 2 1 4 1 6 1 8

D is ta n c e fro m riv e r m o u th (k m )

-1 0 .0

Measurements. Nitrate (spring)

0 2 4 6 8 10 12 14 16 18

Distance from river mouth (km)

Depth(m)

Ebro Estuary Model

Objective:

To set up a model system able to model nutrient concentrations and transport from a monitoring station in the river to the Sea.

Nutrient transport calculated by MIKE 12 river model system

MIKE 12 is a 2 layer one dimensional river model system from DHI water & environment consisting of:• Hydrodynamic module (water level, discharge & velocity)• Advection dispersion module (transport of substances)• Nutrient module with 20 state variables, (Chl-a, plankton C-N-P, detritus C-N-P,

NH4, NO3, PO4, O2, sediment: organic C-N-P, NH4, NO3 & PO4 in pore water)

MIKE 12 Hydrodynamic Model

Date 07-07-1999, Discharge=119.80 m3·s-1

DHI water & environment

Simulation of salinity time series at 11.6 km from the river mouth

upper layer

lower layer

Findings: Hydrodynamic model

• Salt wedge position depending on discharge.

• Simulated salt wedge position agree with observed discharge and positions.

• Salt wedge position sensitive to sea level

Simulated & measured nutrients in salt wedge 6 km from mouthMarch 1999- March 2000

Findings Nutrient transport simulation

The N and P load to the sea was 17.900 tonnes N & 950 tonnes P from 1.03.99 to 15.02.00.

The salt wedge acts like a sink for particulate matter including particulate bound N (PON) and P (POP) due to:

• Increased sedimentation from surface layer after flocculation at PSU > 1. • Input of suspended matter from area covered by salt wedge to surface layer lower than resuspension from same area covered by river bead. Sedimentation of POP and PON from upper to salt wedge was equal to about 30 % of the POP and PON load at Tortosa in 1999.

Plume modelling

Objective:

To set up a model system able to model nutrient concentrations and spreading in the river plume

Ebro plume models

• Hydrodynamic model: TRIM3D (water level and currents) – Reynolds equations

• Dispersion model: LAD3D (plume spreading – salinity) – Advection-diffusion equation

Hydrodynamic simulation of river plume

u=2.2 m/s; =190º; Q=166.6 m3/s

110 120 130 140 150 160 170 180 190

I (x-direction)

Pioneer B

Calibration

1 2 3 4 5 6 7Station num ber

linity

M easured

M odelled

1 5 2 0 2 5 3 0 3 5 4 0

SALIN IDAD (pp t)

1 5 2 0 2 5 3 0 3 5 4 0

SALIN IDAD (pp t)

1 5 2 0 2 5 3 0 3 5 4 0

SALIN IDAD (pp t)

nitrates

silicateSRP

Link between concentration of nutrients and salinity in the Ebro plume(in summer)

Findings of plume modelling

• Plume hydrodynamics depending on wind direction and velocity and bathymetry

• Simulated plume spreading and salinity values agree with observed ones.

• Almost all the nutrients follow a conservative behaviour with salinity (except SRP and silicates in spring)

Conclusions

• Salinity and nutrient concentrations can be suitably predicted in the estuary

• Salinity and plume spreading can also be well simulated

• Nutrients in the plume can be indirectly predicted due to their conservative behaviour with salinity (with some exceptions in spring)

• Continuous nutrient monitoring in the head of the estuary needed in order to get an operational forecast system

Data and models for the Odra estuary

by Dorota Dybkowska-StefekMaritime Research Institute – Szczecin, Poland

The Lower Odra

The Odra Lagoon

PeeneŚwina Dziwna

Kleines Haff

Wielki Zalew

The Odra Estuary

PIONEER

PIONEERThe concept of the

system for forecasting nutrients in the Odra Estuary

The main objective of the system is to predict changes in nutrients distribution in the Odra Lagoon on a basis of nutrient load discharged by the Lower Odra, being estimated from the numerical simulation of flows and transport processes in the river network. The simulation is performed for forecasted hydro-meteorological conditions and forecasted nutrient loads coming from the river upstream. In this simulation inflows of nutrients from the sources located along the Lower Odra and around the Odra Lagoon are taken into account.

MRI PIONEERDATABASE

BOUNDARYCONDITIONS(BC Module)

Informationfor

Polish - German

Monitoring Network

FORECASTModule

Coupled models for Odra Lagoon

TRIM3D/ERSEM

Modelling System for Lower Odra River Network

MODRIM

forecast updating

data aquisition

tailor-made binary file

with BC obtained

for Odra LagoonBC for MODRIM

BC for

TRIM3D/ERSEM

saving of

simulation results

time series creating

THE SYSTEM FOR THE FORECAST OF NUTRIENTS IN THE ODRA ESTUARY

Model system for the Odra estuary

Computational

Parameters:

Dt, Dl, n, cw, ...

Boundary Conditions for

HD Unsteady Flow

Module

Geometry Data Base

of a River Network

HD Unsteady Flow Module of

the MODRIM Model with

Additional Modules AD and SPM

Main Interface for Topology

and Boundary Conditions of

HD Steady Flow Module

Initial Conditions for HD

Unsteady Flow Module

ASCII files

with Results of Run

Initial Conditions

for AD Module

Boundary Conditions

(time series) for AD

Module: Concentrations,

Transport and Inflow of

the Chosen Quality

Parameters

Overall Flowchart of Hydrodynamic (HD), Advection-Diffusion (AD) and Suspended Matter (SPM) Modelling System for a River Network MODRIM

PIONEER

Model system for the lower Odra river MODRIM

PIONEER

Simulation and forecast with the modelling system MODRIM of the

Lower Odra river

The system needs: time series of water level downstream and flow upstream time series of nutrients at the upstream cross-section, averaged load of nutrients from sources located along the

Lower Odra river.

The system produces as results time series of: water levels, flows, velocities, concentration and transport of nutrients, concentration of SPM at every cross-section of the Lower

Odra River network.

PIONEER

35 33 31

282936

272829

171915

162022

212324

4342 40

49505152

Dabie Lake

Z(t) t

57Trzebież

AZSRegalicaPodjuchyGryfino East

Gryfino-WestWiduchowaBielinek

Gozdowice

H-D upper

boundary condition

node number

WESTERN PATHAZS

measuring station or

analyzed cross-section

Parnica

Odra Lagoon

Concentration and flowin inner nodes & cross-sections

Nutrients concentration at the upper boundary

Police

EASTERN PATH

Computational network of the Lower Odra river for nutrients simulation and forecast

PIONEER

Concentration of Nitrogen at t = 144h

0 20 40 60 80 100 120

Eastern path (from Trzebież to Gozdow ice - length [km]

Concentration of Nitrogen at t = 192 h

0 20 40 60 80 100 120

k1_10_050_060_d5

Development in time of nitrogen concentration along the eastern path of the Odra river network

0 20 40 60 80 100 120

PIONEER

Time series of nitrogen concentration in several cross-sections of the Odra river

network

0 24 48 72 96 120 144 168 192 216 240Time [h]

cross-sec. No 1(B1)

cross-sec. No 6(B1)

cross-sec. No 28(B7)

PIONEERSimulation and forecast with coupled models

TRIM3D/ERSEM

The coupled models need: time series of water level at three outlets of the Odra Lagoon to the Baltic

Sea, time series of nutrients concentration and flow at TRZEBIEŻ cross-section, averaged load of nutrients from sources located around the Odra Lagoon.

TRIM3D produces as results time series of: water levels, velocities, concentration of SPM in the points of computational grid of the Odra Lagoon.

ERSEM - based on input data from TRIM3D and sedimentation–resuspension module - produces as results time series of :

averaged nutrients concentration for each box of the Odra Lagoon.

PIONEER

Total number of sea-points: 13840 (92.97%)

Mean/min/max/ box depth: 3.1m/0.1m/12.2m

Size of single grid cell:0.0625 km2

Total area of boxes: 8,650 km2 (92.97%)

Total volume of boxes: 28,262 km3

(95.74%)

PEEN_ACHTER (Box 1)

KLEINES HAFF (Box 2)

WIELKI ZALEW (Box 3)

Division of the Odra Lagoon into three boxes for ERSEM model

PIONEER

Time series of simulated phosphates in Kleines Haff compared with measurements

KLEINES HAFF, year 1997 Phosphates

[mmol/m3]

PIONEER

Time series of simulated phosphates in

three boxes of the Odra Lagoon

year 1997 WIELKI ZALEW, KLEINES HAFF, PEEN_ACHTER

Phosphates [mmol/m3]

THE MRI PIONEER DATABASE is supplied with hydro-meteorological and water quality data from Lower Odra, Odra Lagoon and Ebro Delta from period 1997-2000. The database consists of

Upper Layer with definition of measurement stations

five Lower Layers with data of following types:

Point (one value per time) - Layer_1,

Vector (several values per time) - Layer_2,

Matrix (several vectors per time) - Layer_3,

Moveable_Point (x y,z val_x,val_y,val_z taken in different time and places along the ship route across Odra Lagoon or Ebro Delta) - Layer_4,

Moveable_Vector (vertical profiles taken in different time and places along the ship route across Odra Lagoon or Ebro Delta - Layer_5.

The database is located on the MRI server estua.im.man.szczecin.pl and avaliable via WWW.

results of hydro-meteorological observations carried out by Maritime Research Institute Szczecin (MRI) within the scope of the Odra Estuary monitoring network,

long time series of hydro-meteorological data obtained from Polish governmental Institute of Meteorology and Water Management,

water quality parameters obtained from Voivodeship Inspectorate of Environmental Protection in Szczecin, responsible for surface waters monitoring,

time series of water quality parameters measured at automatic monitoring station Widuchowa,

results of hydro-meteorological and water quality observations carried out by GKSS Research Center within the western part of the Odra Lagoon,

results of hydro-meteorological and water quality measurements done during Polish-German campaigns on the Odra Lagoon.

Data for the Odra Estuary:

PIONEER

THREE POSSIBILITIES OF LOOKING THROUGH

THE DATA STORED IN THE DATABASE:

to create a HTML table of the data,

to display the data as graph and

to extract the data as ASCII files in chosen format.

Preparing and integrating analysis and management tools in a

system suitable for simulation, analysis and prediction.

Empirical knowledge and modelling skills achieved in the

PIONEER project for the Odra Estuary:

PIONEER

The system simulates different hydrodynamic and nutrients events (previous or hypothetical) depending on: concentration of nutrients coming from the river

upstream, inflows of nutrients from the sources located along

the Lower Odra River and around the Odra Lagoon, transformation processes of the nutrients during their

transport towards to the Baltic Sea, changes of hydro-meteorological conditions in the

river part and the Odra Lagoon area.

Scientific achievements

PIONEER

The process of calibrating and validating the system has extended our knowledge about the hydrodynamic conditions and nutrients distribution in the Odra estuary.

In the course of preparing the system the inventory of available and potentially available data have been made; this data partially has been stored in the database, especially constructed for the project purposes.

Scientific achievements

PIONEER

The system is a useful tool to study changes of nutrients in the Odra estuary.

Combining data and models for analysis and prediction purposesby Hans WackernagelAssociation pour la Recherche et le Développement desMéthodes et Processus Industriels – ARMINES, Ecole Nationale Supérieure des Mines de Paris, Centre de Géostatistique – Fontainebleau, France

The Data: The Data: a great variety...

• Chemical/Physical measurements• Automatic stations (few)• « Stations » with repeated

measurements• Ship cruises• Remote sensing• Climate/meteorological (analyzed)

The Knowledge: The Knowledge: different fields..

• Hydrodynamic models• SPM models• Meteorological models• Ecological models

• Hydrodynamic model on small scale 250x250x3 m3 grid• Ecosystem model implemented on large compartments

PIONEERPIONEER - model coupling in the Odra Lagoon

Exchange

Physical models

Ecosystem model

Data AssimilationData Assimilation

Merge data and models using:

Kalman filter (dynamic)

Kriging methods (static) PIO

Sequential Data AssimilationSequential Data Assimilation

Statistical conceptsStatistical concepts

Raw dataObservational errorModel errorSensitivity of results

Conductivity: Conductivity: variograms for two spatial directions

Chlorophyll: Chlorophyll: two variogram models

Chlorophyll: Chlorophyll: conditional simulation (cubic variogram)

Chlorophyll: Chlorophyll: conditional simulation (exponential variogram)

Salinity in the Odra Lagoon

TRIM3D model Kriging along ship channel

Odra Lagoon: Nitrate forecast

Nitrate: rms error of forecast

Conclusions•Both techniques, geostatistical analyses and sequential data assimilation allow for a detailed, time-space variable description and forecast of hydrodynamical and water quality parameters in coastal seas, based on relatively few observations.

•The geostatistical method lends itself to applications in situations with sparse observations in time.

Conclusions and outlookby Hans von StorchInstitute for Coastal Research, GKSS - Geesthacht, Germany

Ebro Estuary• Salinity and nutrient concentrations

can be suitably predicted in the estuary• Salinity and plume spreading can also

be well simulated• Nutrients in the plume can be indirectly

predicted due to their strong correlation with salinity (with some exceptions in spring)

• Continuous nutrient monitoring in the head of the estuary would allow for the installation of an operational forecast system

Simulation of salinity time series at 11.6 km from the river mouth

upper layer

lower layer

•The TRIM3D / ERSEM / MODRIM model system allows the simulation of the hydrodynamics and of the water quality in the lower Odra river and in the Odra Lagoon.

•Essential driving parameters are water levels at the outlets to the Baltic Sea plus the water level and the nutrient influx at the upper boundary of the river segment.

•Data, both on the Odra and the Ebro situation, have been stored in a joint data base, which is accessible through the internet.

Odra Estuary

MRI PIONEERDATABASE

BOUNDARYCONDITIONS(BC Module)

Informationfor

Polish - German

Monitoring Network

FORECASTModule

Coupled models for Odra Lagoon

TRIM3D/ERSEM

Modelling System for Lower Odra River Network

MODRIM

forecast updating

data aquisition

tailor-made binary file

with BC obtained

for Odra LagoonBC for MODRIM

BC for

TRIM3D/ERSEM

saving of

simulation results

time series creating

THE SYSTEM FOR THE FORECAST OF NUTRIENTS IN THE ODRA ESTUARY

Analysis and PredictionTwo techniques have been found to combine robustly and skillfully dynamical knowledge (models) and limited empirical evidence (data): Statistical methods (geostatistics)Sequential data assimilation (Kalman filters).

They differ in complexity, and their application depends on regional expertise and temporal availability of data.

Modeling is improved through data constraints. In particular, water quality modelling is expected to become more realistic when steered intermittently by data.

Odra Lagoon: Nitrate forcast

Outlook

Systematic combination of routine measurements and detailed modelling of hydrodynamics and nutrient dynamics on time scales of days has been shown to be a realistic option for operational monitoring of the hydrodynamic state and water quality of coastal seas.

Such MODAP systems are cost-effective and robust. They allow for useful estimates even if only very little in-situ data are available.

MODAP = Model & data analysis and prediction.

Benefits

• The MODAP system, after developed into a commercial version – will be useful for authorities and commercial companies in analyzing and predicting the nutrient load and water quality in coastal waters.

• The system is portable to other estuaries all over the world.

Coastal monitoring is dividing among many different authorities with often insufficiently staffed scientific R&D resources. Thus, the MODAP concept is not yet widely known among practitioners. It is therefore met with some reservations among responsible authorities.

For demonstration of the utility and cost-effectiveness of MODAP systems, a multi-year demonstration should be established.

Additional R&D efforts will undoubtedly further reduce the costs and increase the robustness, allowing it‘s use in remote areas in a mostly automated set-up.

pioneer 1998-2001 preparation and integration of analysis tools towards operational forecast of...

data management

analysis systems

overall approach of

main objective of pioneer

analysis software

dynamical data assimilation

integration of analysis

coastal research

Documents

nitrogen in the marine environment in marine.pdf · use of...

estuaries estuaries. estuaries an estuary is defined as a...

estuaries: introduction

chapter 7: estuaries · 2018-04-11 · chapter 7 estuaries...

estuaries description and formation of estuaries

physical features of estuaries. basic information estuaries...

estuaries: communities

estuaries 101 a brief introduction to natural and ...

estuaries what regulates sedimentation in estuaries? ·...

environmental ﬂows for sub-tropical estuaries ·...

water estuaries - environment.gov.scot

estuaries where rivers meet the sea. origins and types of...

2016-2017...pioneer pride. pioneer pride. pioneer pride....

cemw - state of the estuary report · narrative sidebar...

estuaries (prandle)

tide-dominated estuaries

freshwater biome, estuaries, nitrogen cycle. aquatic biomes...

nutrients and carbon fluxes in the estuaries of major rivers...

our estuaries

estuaries & salt marshes1 estuaries and salt marshes...