INTEGRATING COASTAL ZONE MANAGEMENT AND RIVER BASIN MANAGEMENT, AN APPLICATION OF GIS FOR THE RIVER ELBE MANAGEMENT (GERMANY)

(1) C. Nunneri, (2) J. Hoffmann

(1) GKSS Forschungszentrum, Geesthacht, (DE)
(2) Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Berlin (DE)

Abstract
The integrated view of the coastal waters and the river catchment as one system requires the integration of River Basin Management (IRBM) and Coastal Zone Management (CZM), as targeted in the Water Framework Directive (WFD).
This paper presents a methodological approach of computing nutrient inputs from the catchment to the coastal zone and its practical application to the Elbe river and the North and Wadden Sea as an example. Available information about sources, fluxes and concentration levels of different compounds flowing from the catchment to the coastal zone through the river systems is collected and integrated in natural science models.
By connecting socio-economic tools (participation, scenarios and management policies) with nutrient transport and ecosystem models, possible solutions (management measures) for reducing the nutrient inputs and the consequent eutrophication in the coastal area can be analysed. Much has been done towards nutrient emission reduction during the last decades: in the Elbe basin N-emissions were reduced by about 56% and P-emission by about 70%. A further reduction might be perceived as unnecessary and will surely entail relatively high costs. Therefore the final goal will be that of providing decision-makers of a cost-effectiveness-testing tool for emission-reduction policies (measures).

1 Introduction
The understanding and management of water-quality related issues in the coastal areas requires an interdisciplinary approach aiming to analyse the interwoven processes typical of intensively populated regions subject to multiple uses (and inputs). A key analysis and modelling characteristic of changes in the coastal zone is the multiple links to the relevant natural and anthropogenic processes in the related river catchments (e.g. nutrient input triggered by land-use). In this context the catchment -coast continuum reality is a strong reason for the adoption of an integrated assessment approach. The integration and connection of Integrated River Basin Management (IRBM) and Integrated Coastal Zone Management (ICZM) is one of the ambitious challenges of the Water Framework Directive (EC, 2000) and is one of the goals of the EUROCAT Project , of which this study, dealing with eutrophication of the coastal waters, is part.
Such an integrated assessment aims not only to investigate the main natural processes that play a role in environmental issues, but also implicitly addresses the socio-economic component deeply shaping and driving the processes at the basis of environmental changes. The ultimate aim of integrating natural science findings with socio-economic analysis, is to delineate a management strategy ideally harmonising the different needs (and related uses) that characterise catchment and coast in order to find an optimum between ecosystem use and ecosystem conservation. In this paper the methodology used for analysis in EUROCAT is shown, together with some results of the analysis applied to the Elbe catchment.

2 The Elbe case Study
Altogether the Elbe study area comprises the river basin of 148,268 km2 and includes the coastal strip shown in figure 1 (as delimited by the ERSEM model boxes).

Fig. 1: The Elbe catchment and its coastal zone, the German Bight. The delineation of the coastal zone within the REBCAT study is based primarily on the boundary conditions of the applied models: the relevant boxes (COCOA based) for the application of the ecosystem model ERSEM are given with the numbers 59, 69 and 78. In addition the map shows the coastal strip in a width of 10 km in yellow color, representing the jurisdiction of the EU water framework directive (WFD) on coastal waters.

The Elbe River Basin (148,268 km²) covers large parts of two central European countries, namely the Czech Republic and Germany. About two thirds of the catchment area belong to Germany and one third to the Czech Republic. A total population of 24 millions lives in the catchment area and comprises 58% of the Czech and 22.9% of the German population.
The Elbe basin stretches through different geographical regions from middle mountains to large flatlands and lowlands, presenting numerous geographical regions and landscapes that show an almost undisturbed nature of inter-regional importance for plants and animals otherwise endangered.
A variety of socio-economic conditions characterises the catchment, especially due to the changing political and economic conditions after German reunification in 1990.
At present about 61% of the catchment is used for agriculture, 29% is covered by forests and 6% is urban area. Different economic activities such as agriculture, drinking water supply, industry and tourism represent often conflicting interests.
The MONERIS model describes the input of nutrients from the cachment to the coastal zone taking into account different diffusion paths (Behrendt et al., 2000).
The delineation of the coastal zone impacted by the Elbe river is based primarily on the boundary conditions of the model ERSEM, which describes the ecological status of the coastal area (Lenhart, 2001). The receiving sea basin of the river Elbe is the south-eastern corner of the North Sea, namely the German Bight. The coastal surroundings of the German Bight are a rural agricultural area with low population density and tourism as dominating economic force. Land reclamation and coastal defence have historically been essential for survival in this area, thus being part of the cultural heritage. Since the 1990s wind farms for electric power generation have become a major issue for spatial and environmental planning, in some cases becoming regionally the second most important economic sector behind tourism (e.g. Kannen et al., 2000).

3 Methodological Approach
The methodology adopted for analysis is comprised of a scoping framework, the DPSIR-approach (EEA, 2003 Turner et al. 1998), the use of qualitative future scenarios and two models, one for computing the fluxes of nutrients to the coastal zone (MONERIS), and one to compute the ecological effects in the coastal waters (ERSEM). The focus is set on the coastal zone, being the coastal waters the receiving body of emissions in river systems and specifically nutrient emissions from the catchment and the resulting eutrophication of the North Sea coastal waters are the key issues (see also Colijn et al. 2002). Based on different scenarios, MONERIS will be applied for testing the effects (on the eutrophication of the coastal area) of measures for nutrient emission reduction (applied to the catchment). Finally a Multi Criteria Analysis based on expert judgement is foreseen, in order to rank possible management options based on costs of realisation and effects on the ecosystem (including ecological side-effects), thus providing a supporting tool for decision-making.

3.1 The DPSIR-approach and the role of scenarios
The DPSIR-approach of the European Environment Agency is the analytical tool selected to handle complex interactions between the socio-economic (humankind processes) and the natural system (ecosystem processes). Human activities may cause some impact on the (coastal) environment and potentially damage the (coastal) ecological integrity. Such complex human-ecosystem interactions can be observed by dividing them into five variables: (1) Drivers and (2) Pressures resulting by socio-economic development; (3) State of and (4) Impact on the environment and finally (5) societal Response (policy measures) to such unwanted impacts (table 1). Following the DPSIR scheme, a methodological approach for integrated assessment (IA) starts from the analysis of the current state of the system. Then successive steps look for causality chains, the drivers and pressures that may change the present state (the current P and D might be different from the future ones), the possible political alternatives, as well as institutions and mechanisms for policy and management responses.
Scenarios are used for assessing possible socio-economic development paths, which are the origin of drivers and pressures (e.g. change in land-use intensity or agricultural production patterns). Thus ecosystems will experience different (intensities of) pressures, and, consequently, impacts on their integrity and functionality under alternative scenarios. Furthermore scenarios represent alternative futures, in which society might think and behave differently. Social mentality, i.e. changed societal values and lifestyle is the key for connecting socio-economic and environmental processes: environmental risk perception and therewith willingness to pay for reducing it (interpretation of the precautionary principle) will considerably influence acceptance, economic feasibility and willingness to pay for emission reduction measures. Three scenarios will be considered for the Elbe catchment, respectively giving priorities to economic growth, full implementation of environmental legislations and nature protection "at nearly any cost". Those scenarios, the relative "desired quality level" of the environment and the ecosystem and the connection with ecological indicators are described in Nunneri et al., 2002.

DRIVER
PRESSURE
STATE
IMPACT
RESPONSE
Agriculture
Nutrient input (N,P)
Genetic and species Diversity Changing
N-P ratio
Imbalance in genetic and species diversity
New farming methods (e.g. fertilisers management)
Land use
Sedimentation/ Erosion

Changed coastal Morphologyl
Morphological change of sea bed and the coastline
Adopt new land use methods (e.g. bufferzones)
Industrialization
Discharge of compounds
Changed concentration of chemical compounds
Toxicity, Diseases
Implement (int.) national policy measures
Urbanization
Paved areas, Sewage, Waste water
Changed coastal morphology , N-P emissions, coastal dynamics
Changes of the ecosystem (e.g. morphdynamics, species composition)
Implement urban and waste planning strategies

Table 1. Major DPSIR variables for the Elbe catchment. Note that there are also relevant drivers and pressures in the coastal areas (e.g. fishery), which are not included in this table.

3.2 Attempts towards an harmonised model system
An essential part of the project is to analyse the response of the coastal seas to past, present and future changes in fluxes of nutrients from the catchment.
Through a GIS application the gathered information on emissions, their pathways and their effects on coastal waters can be used for further analysis in order to address the following questions:
1) What fluxes of nutrients are flowing through different diffusion paths to the coastal zone?
2) What are the physical, chemical, and biological controls, both natural and anthropogenic, of these fluxes?
3) What are the feedbacks of changes in drainage basins on human society and on biogeochemical cycles?

In order to address these questions, the following models are used:
- MONERIS for modeling nutrient emissions in the Elbe basin (Behrendt et al. 2000, 2002).
- ERSEM for modeling the ecosystem changes in the coastal zone (Lenhart 2001)

The outcome of the MONERIS model is used as an input to the ERSEM model in terms of fluxes. MONERIS can help to identify nutrient diffusion paths, thus highlighting more effective policy and management measures in order to achieve the desired standards.
By applying the ERSEM model, the effects of hypothetical reduction of nutrients for the eutrophication in the coastal waters can be shown by load response curves and GIS created maps for the different emission pathways. The link of the dynamic model ERSEM with the steady state model MONERIS requires an artificial resolution of year cycles (based on 5 years means) in order to derive a seasonality on the base of monthly values. The interface between ERSEM and MONERIS consistsmin a transfer function between river loads and concentrations in the sea (transfer of nutrients through the Elbe estuary), also taking into account the retention functions of different Nitrogen species.

3.2.1 MONERIS
The model MONERIS (MOdelling Nutrient Emissions in RIver Systems) was developed for the investigation of the nutrient inputs via various point sources and diffuse sources through different pathways in German river basins (Behrendt et al. 2000). Basic input information entering the model comprises data on river flow, water quality of the investigated basins and GIS-integrating digital maps as well as statistical information for different administrative levels and scales.
Whereas the inputs of municipal wastewater treatment plants (WWTPs) and industrial discharges enter the river system directly, the sum of nutrient inputs into the surface waters from diffuse sources is built from a variety of runoff components taking different pathways in the form of surface runoff, base flow and interflow. The distinction of these individual components is necessary in order to distinguish the emission concentrations of nitrogen (N) and phosphorus (P) and the related processes. The MONERIS model takes seven pathways into account and provides estimates for the following specific inputs (see Fig. 2):
- Point sources
- Atmospheric deposition
- Erosion
- Surface runoff
- Urban areas
- Tile drainage areas
- Groundwater

Fig. 2: Pathways and processes considered in the model MONERIS (Behrendt et al. 2000)

With regard to diffuse sources the distinction of the individual components is not sufficient for the material inputs coupled to surface runoff and the interflow. With surface runoff, inputs of dissolved substances via surface runoff as well as entries of bound nutrients and suspended particulate matter via erosion must be distinguished. Further it has to be considered that the processes coupled to surface runoff depend on the nature of the area. Accordingly, surface runoff and coupled input from paved urban areas must be quantified separately.
Interflow can originate both under natural conditions and through human activities. In particular, inputs from tile drainage must be considered separately. The quantification of the input of substances via natural interflow and the drains is particularly complex. On the one hand, no model results on the share of interflow of the total runoff for all German river catchments is available. On the other hand, there is a lack of data and models to determine the areas drained by tiles in the German river catchments. During this study, an attempt will be made to estimate the proportion of tile-drained areas in the German catchment areas. In addition to the inputs from the tile-drained areas, all other subsurface flows will be summarized in the groundwater inputs. That means that the groundwater paths contain to regional different part also the inputs via natural interflow.

In the diffuse inputs, various transformations-, loss and retention processes play an important role. The actual knowledge of these processes and the existing databases (restricted to medium and large river basins) do not allow to quantify and forecast the nutrient inputs in relation to their source.
The final output of MONERIS is an estimate of annual nutrient load in the river at the outlet of the study catchment, which is equal to the emissions into the river via point and diffuse sources minus the estimated nutrient retention and loss within the river system.
One example for the present situation of nitrogen emissions via different pathways is given in Fig. 3. Note that the entire Elbe basin is regionalized according to the coordination regions of the water framework directive.
Thus MONERIS can help managers to identify pathways that contribute significantly to nutrient loads and should be targeted for management practices aimed at nutrient emission reduction. Combined with geographic information in a GIS, it can help identify hot spots within the catchment -- particular areas that, due to a combination of high potential emission and a susceptibility to efficient transport, contribute nutrients significantly more that other areas.
Once MONERIS has been calibrated for a particular catchment, it can be used to develop management scenarios. For example, a manager may be interested in knowing the reduction of nutrient emissions into the river under a scenario of erosion control.

Fig. 3: Nitrogen emissions during the period 1998-2000 via various pathways

3.2.2 ERSEM
The ecosystem model ERSEM (European Regional Seas Ecosystem Model) was developed to simulate the ecosystem dynamics of the North Sea. The model simulates the annual cycles of carbon, nitrogen, phosphorus and silicon in the pelagic and benthic food webs of the North Sea. The box model combines hydrodynamical and ecological processes into one model with the same resolution in space and time. The model is forced by irradiance and temperature data, suspended matter concentration, hydrodynamical information for advection and diffusion, data on atmospheric nutrient input to the North Sea as well as by inorganic and organic river load data (Fig. 4).
The biological part of the model consists of an interlinked set of modules, describing the biological and chemical interactions between the state variables. A general description of the model is given by Baretta et al. (1995) and Lenhart (2001).
The model covers an area of 577 620 km2 and a volume of 51047 km3 in total. The northern and central parts of the North Sea are divided into 1 by 2 degree boxes. For resolving the horizontal gradients in the coastal areas the spatial resolution was increased to boxes of 0.5 by 1 degree. In this study the ERSEM boxes 78, 69 and 59 were chosen thus covering the Elbe estuary as well as the northern part of the German Bight and Wadden Sea (Fig. 4). This coastal area is nearly identical with the OSPAR regions O-II-3D of the Greater North Sea.

In this way, the model is finely resolved in the coastal area, but also represents the central and northern North Sea with sufficient resolution. In order to account for the effects of stratification in regions outside the coastal areas, the water column can be divided into an upper box (0 - 30 m) and a lower box (30 m - bottom) (Lenhart et al., 1995, 1997). To represent the ecosystem dynamics in the coastal region with its highly variable conditions the relevant information on the morphology has been provided and the transport processes have been parameterised on the scale of the box setup of the model for the years 1988 and 1989. In addition to the transport forcing, realistic forcing is provided also as time series of daily values for radiation and suspended matter concentration. For the atmospheric nitrogen input a constant load is applied to the entire model domain. For the reduction scenarios, as planned within EUROCAT, daily nutrient loads are used for the model simulation as calculated by Lenhart & Pätsch (2003).
In general, the ERSEM model allows for an adequate level of complexity in its biological and biochemical representation for the scenario simulations. In addition, its highly variable forcing permits to follow the anthropogenic signal against the background of the natural variability of the ecosystem. (for more details on the model setup see Lenhart et al., 1997 and Lenhart, 2001).


Fig. 4: Schematic overview on the ecosystem model ERSEM. - The rectangles represent modules which are connected by intercompartmental fluxes, indicated by arrows. The circles express the prescribed forcing. (Lenhart 2001)

4 Past and present nutrient fluxes from the catchment
Nitrogen emissions into the Elbe river basin were about 102 kt/a N in the period 1998-2000 and thus 128 kt/a N, or 56 % lower than in the period 1983-1987. The target of the 50 % reduction of nitrogen loads from Germany into the North Sea was probably achieved only within the catchment area of the Elbe river. The main cause for the decrease of the nitrogen emissions into the river systems was the large reduction of nitrogen discharges from point sources by 78 %. On the other hand the estimated decrease of diffuse emissions was only about 40 %. The input via groundwater (38%) and tile drainages (24%) are the dominant pathway in the period 1998-2000. The share of point sources in nitrogen emissions amounts to about 21 %. The contributions of erosion, surface runoff and atmospheric deposition to the total nitrogen input are low and amount to about 4% only for each of these pathways.
The total phosphorus emissions into the Elbe river basin were about 5.53 kt/a P in the period 1998-2000. Compared with the period 1983-1987, the phosphorus emissions were reduced by about 12.7 kt/a P or 70 %. The target of a 50 % reduction of the phosphorus loads into the North sea was reached. Again the decrease of phosphorus emissions is mainly caused by a 90 % reduction of point emissions. The decrease of diffuse phosphorus emissions was larger than for nitrogen, which is caused by a 59 % reduction of the emissions from urban areas. In spite of the enormous reduction of phosphorus discharges from point sources these sources remain the dominant pathway of phosphorus emissions with 27 % in the period 1998-2000. Among the diffuse pathways, emissions by erosion dominate and represent 26 % of the total input.
The comparison of nutrient emissions in the period 1983-1987 to 1998-2000 shows a reduction of the total amounts and also distinct displacements from point sources to diffuse sources. With regard to diffuse phosphorus emissions the pathway of erosion (+43%) and overland flow (+30%) gain more importance, while the proportions from urban areas (-63%), atmospheric deposition (-46%) and groundwater (-24%) are still decreasing. In the past years the point sources obtained a high reduction, mainly due to point sources like WWTPs (-89%) and industrial inputs (-94%).
Also the nitrogen emissions reveal similar trends with an increasing importance of erosion (+31%). Groundwater and drainages are still the main contributors, even though they decreased by 36% and 51 % respectively.
The main part of all diffuse emissions is caused by agriculture, being the nutrient surplus on agricultural areas is one of the most important factors. The regionalization of nutrient surpluses shows that the P-surplus is in general 2-4 kg/(ha·a) P, only some areas in the Tide-Elbe show lower values of < 2 kg/(ha·a) P.
The N-surplus is in general 40-60 kg/(ha·a) N, with higher values in the Tide Elbe of about 80-100 kg/(ha·a) N and even up to 120 kg/(ha·a) N. Due to the reunification of Germany (1989/1990) and structural changes in agriculture, the N-surplus was reduced during the period 1990-1993 to the level of the 1950's. Since then surpluses are slowly increasing. In the past years N-surpluses kept constant with values of approximately 60 kg/(ha·a) N. A further nutrient reduction presents a challenge for policy- and decision-makers, namely that of choosing policies that will lead to an effective management of nutrient emissions, both in terms of costs and results.

5 Outlook: What management Responses?
The philosophy of the EUROCAT approach is to deliver scientific tools and concepts for an integrated management of catchments and coastal seas. The strategy for user involvement and dissemination will make use of GIS-supported interactive tools for presentation and selection of data and results which can be used by the project partners, other users and the general public. This exchange of information is provided by an external server with the GIS data base linked to the official EUROCAT website (see Appendix 1 for details). The crucial choice of optimal measures (or measure packages) for river basin management will employ a multi criteria analysis (MCA), in which the expected effects of reduced emissions upon the coastal-zone ecosystem (ERSEM results) will be related with the expected costs of reduction and other connected non-monetary benefits (or costs). An essential role in evaluating the effect of management measures is played by a participatory approach involving interest groups (Governmental and Non-governmental Institutions) situated in the Elbe catchment (Nunneri & Hofmann 2003, in preparation). The participatory part of the project is essential for keeping the whole project on a policy-relevant track and determining possible conflicts, feasible measures and criteria for suggesting optimal management solutions. Agriculture has been mentioned as the sector that should next be addressed by reduction measures, if the target is that of a further reduction of the nutrient emissions (especially of Nitrogen). Nonetheless further connection to WWTPs and improvement of the smallest TPs represents also an issue, especially in the Czech part of the catchment (e.g. Behrendt & Schmoll, 2003). If on the one hand agriculture is the main polluter regarding diffuse nutrient emissions, it also represents the sector that receives the maximum pressure for reduction. According to the interviewed interest groups, the costs of reduction should be shared in such a way that they are bearable for the society as a whole. In this context a point of major concern is the lack of communication between policy-makers and key-stakeholders, which would allow for compromises and win-win solutions. Technical feasibility, political and public acceptance and high cost-effectiveness are mentioned as the main criteria after which a measure should be evaluated and chosen, while absolute costs play a secondary role, at least under a certain threshold. The (public) acceptance of measures for nutrient reduction have to find its way through compromises and social equity, allowing for win-win solutions among different groups of interests and balanced spatial division of costs and benefits. All those criteria will provide the basis for the final MCA analysis for evaluating possible measures for nutrient reduction.

6 Acknowledgments
This research was funded by the European Commission under the Fifth Framework Programme in the context of the project EUROCAT (project No. EVK1-CT-2000-00044). This is gratefully acknowledged by the authors.

Appendix 1: Dissemination strategy and user involvement

The link from the EUROCAT website (http://www.iia-cnr.unical.it/EUROCAT/project.htm) to an external server with the GIS data base (http://145.253.133.76/eurocat/index_inside.html) is easy to activate by clicking the button Maps and Data on the EUROCAT website. Besides other catchments the necessary GIS database is available for viewing and downloads of maps for the Elbe river and its coastal zone.
The Eurocat "database" is structured by documents and files in a "common" filesystem. DISPLAY of the "filetree": The files are presented in explorer-like mode. The difference to "normal" filesystems is that in the interest of clearines files of certain types and names are hidden in the filetree. This holds especially for the GIS-data where only one out of the set of a shape-fileset (shp, shx, dbf, prj, xmeta) is visualized. DISPLAY of document/file-contents: Most documents can be viewed alternatively in a the internal Frame of the web-page.
In addition to the available data and maps further possibilities to view and download data and maps are prepared for partners of the EUROCAT project. To get this the partners have to use the same LOGIN/PASSWD as for other restricted areas of the EUROCAT site.
Downloading data differs between "normal" and map documents. To download documents use the common "right-mouse-button"-functions of your browser. To download map-layers use the "download-button" in the mapping client.
The maps of the catchment sites can be opened by clicking the "xxxMap"-directories. The map themes inside a map-presentation are split into the quasi-static "internal" map-layers of the map-projects and the highly dynamic "external" map-layers which are visible in the 'BaseMaps' and 'ProjectMaps'-directories of the file-tree. The handling of the "internal" layers is similar to normal digital desktop GIS. The "external" layers can be included into the "internal" map-views by the user one by one. Map-Data-downloads are restricted to these data by the moment. Data owners can restrict the use of these data to VIEWING within the accompanying XML-METADATA-File.
To ensure data security the webmaster has established 3-levels of application logins. With the basic "public" entry to the database you are able to view the data & maps in the directories BaseData and Basemaps. The content of these directories is supervised by the webmaster. Further data- and map-directories may exist for every catchment (The ProjectData- and ProjectMaps-directories are prepared for every one) but are visible only after "Partner login". The content inside these directories is managed by the project-partners directly themselves. To upload (or change) the login must be changed to "upload login". After successful login an icon occurs at the right side of every directory so that the partner is allowed to upload data.

References