Harmonised Modelling Tools for
Integrated River Basins Management

Work package 5:

Integrated Assessment & the Policy Science Interface

Workshop Report

on 4th Harmoni-CA/WP5 Policy Workshop:

Review of model-based tools with regard to the interaction of water management and agriculture

Author

Guido M. Bazzani

Deliverable no. D 5.1c and 5.4c

29. July 2007

Review of model-based tools with regard to the interaction of water management and agriculture

Guido M. Bazzani

Contact authors email

<G.Bazzani@ibimet.cnr.it>

Commissioned by Harmoni-CA

Deliverable number: D 5.1c and 5.4c

www.harmoni-ca.info/products

Harmoni-CA is a research project supported by the European Commission under the Fifth Framework Programme and contributing to the implementation of the Key Action "Sustainable Management and Quality of Water" within the Energy, Environment and Sustainable Development. EVK1-CT-2002-20003.

www.harmoni-ca.info

Disclaimer:

This report is the sole responsibility of the author(s) and does not represent the opinion of the European Commission, nor is the European Commission responsible for any use that might be made of the information appearing herein.

Index

List of abbreviations.. VI
List of tables and figures.. VIII
Executive summary.. IX
1. Introduction.. 1
2. The Water Framework Directive and the Common Agricultural Policy.. 4
2.1 Water use and agriculture. 4
2.2 The Common Agricultural Policy reform... 8
2.3 The Water Framework Directive. 9
2.4 The CAP and the WFD.. 11
3. Categorisation of existing tools and models related to water and agriculture. 14
3.1 Agronomic tools. 15
3.2 Hydrology and water quality tools. 18
3.3 Land use and economic tools. 24
3.4 Web portal on water-related tools. 32
4.. Case studies: examples of good application.. 37
4.1 ARID cluster and WaterStrategyMan.. 37
4.2 The German Elbe River Basin case study.. 39
4.3 The case study of the Piave River Basin in Italy.. 40
5.. Conclusions.. 43
References.. 47

Annex 1: EU-funded projects related to water and agricultural modelling.. 54
Annex 2: The Common Agricultural Policy.. 56
Annex 3: Measures to implement the WFD and the CAP.. 59
Annex 4: Monetary valuation.. 61
Annex 5: A categorisation of tools and models for water management and agriculture 62

List of abbreviations [1]

List of tables and figures

Figure 1: Example of a conceptual model of water management in the agricultural system.. 6

Table 1: Gap between scientists, policy-makers and water managers. 2
Table 2: WFD policy objectives, agri-environmental instruments and possible effects on water and agriculture 12
Table 3: Examples of rules associated with nodes in a hydrological model 19
Table 4: European web portal, including toolbox for water 33
Table 5: Types of models used in SEAMLESS. 36
Table 6: EU-funded projects related to water and agricultural modelling. 55
Table 7: Good agricultural and environmental practice. 60
Table 8: Monetary valuation. 61
Table 9: Information sheet on agricultural and water related models. 63

Executive summary

Objective

This document is primarily directed at water and agricultural managers who wish to find out about agricultural policy reform and existing models and tools that support joint water and agricultural management processes. It may also be beneficial to water authorities, water irrigation and reclamation boards, farmers associations and NGOs. The goals of the document are manifold:

• to clarify the policy context, paying special attention to the recent reform of the Common Agricultural Policy (CAP) and its interaction with the Water Framework Directive (WFD);
• to categorise existing tools and models related to water and agriculture;
• to encourage stakeholders to use models and tools by offering examples of good application;
• to suggest how models and tools should be used to support a participatory implementation process of the WFD in agriculture.

Interaction between the Common Agricultural Policy and the Water Framework Directive

Agriculture plays a central role in water management, since it is a significant utiliser of water resources in Europe, accounting for around 30% of total use. In southern Europe, water is a fundamental agricultural input, with irrigation accounting for over 50% of demand. Furthermore, agriculture and forestry, which cover more than three-quarters of the area of the European Union, play a key role in determining the rural economy and environmental quality. Environmental pressures due to agricultural activities, including water pollution, are often viewed as a serious problem in many areas.

Due to an agreement on the Mid-Term Review of the CAP reached in June 2003, the Common Agricultural Policy (CAP) has adopted a new model for European agriculture. This model reflects the multifunctional role played by farming, integrating economic viability, food safety, social balance and environmental concerns. The key elements of the reform, which completely changes the way in which the EU supports its farm sector, are: adoption of a single farm payment for EU farmers aid paid to producers will no longer be dependent on type of production, thus decoupling aid from, thus removing the causes of indirect environmental damage; enforcement of cross-compliance, which makes direct payments conditional on specific requirements; strengthening Rural Development (RD) policy, into which instruments directly concerned with environmental outcomes are inserted.

The WFD and RD policy both adopt planning processes. However, important differences arise between them, of which scale and time could cause major problems. Win-win situations for both policies require close attention in the implementation phase; representatives from different authorities should interact to create a common vision of agricultural and environmental problems.

Categorisation of existing tools and models

This review focuses on over fifty available tools that have a wider application, have been developed and applied in the context of EU-funded projects, or that present specific aspects of interest to support the implementation of the WFD. To categorise and summarise these tools, a framework based on four critical dimensions (spatial scale, irrigation measures, agricultural measures and economic analysis) is adopted. The tools are categorised into three classes:

• agronomic tools;
• hydrology and water quality tools;
• economic and land use tools.

This review demonstrates how most tools tend to shift towards a new class, embracing holistic integrated Decision Support Systems (DSS). This follows the observed trend in modelling, which is changing from monodisciplinary to multidisciplinary approaches, in which several domains are integrated.

The non-exhaustive list can support the identification of models and tools to address specific issues, including the correct consideration of domains, scales and types of measures. A table in the annex summarises the relevant information and includes each tools website and contact details.

Further information is provided, referring interested readers to a dedicated web portal that offers comprehensive, shared European information, such as WISE, the Water Information System for Europe. Since innovation in computational technology will further enhance the frontier of modelling, web portals are a key instrument for water and agricultural managers to access updated information.

Examples of good application

Three examples of good application are presented. The selected case studies have been carried out recently in different European water basins in the context of EU-funded projects. The projects are:

• WaterStrategyMan in the ARID cluster
• the German Elbe River Basin
• the Piave River Basin in Italy.

The case studies give an insight into relevant questions such as: How can tools improve transparency in decision-making? How can tools support water managers in their present tasks? How can various management options be compared in terms of their ecological, economic and social impact?

Models and tools to support the participatory implementation process of the WFD

The final section of the review explores considerations of how models and tools should be used to support the implementation process of the WFD.

Good practice in model application deserves continual attention. In fact, the credibility and impact of the information and insight that modelling aims to generate are highly dependent on the quality of the modelling exercise.

The political process-oriented nature of water management requires an adaptive approach, which should also consider how models and tools should be used. A multilevel approach in modelling is recommended. At a higher level, conceptual models can support the definition of a common conceptual framework for cross-disciplinary work involving authorities, managers, stakeholders and researchers. At lower levels, quantitative models can support specific analyses exploring the strength of interactions and the sensitivity of the system to changes. The information produced, clarifying the areas of greatest uncertainty and influence on system evolution under different assumptions, can help prioritise field research and support water and agricultural management. However, it cannot offer definitive solutions, since uncertainty and subjectivity are intrinsic to all decision processes, which tools can reduce but not eliminate.

Since water pollution caused by agricultural activities is not specific to irrigated agriculture, the design of policies capable of increasing water quality, while preserving the economic and social sustainability of agricultural systems, requires a clear understanding of the complex relation at different scales from the field/farm up to region catchments and basins. Such understanding can be favoured by a combination of existing tools used by multidisciplinary teams of well-trained people with adequate data in the participatory process.

The definition of packages of integrated measures addressing quality and quantity issues can be supported by tools. However, there is no general solution, since measures are case-dependent. Various tools include specific routines to support the economic analysis requested by the WFD in one or more of the addressed aspects, ranging from the economic analysis of water use, cost recovery and pricing schemes, up to the evaluation of benefits to the environment and society, and eventually of disproportionate cost. Nonetheless, the support provided by such tools is highly dependent on how they are used at present, it is incumbent on the Member States to decide how such analyses and valuations should be carried out; no common guidelines are defined.

The translation of the experience gained in research to the management domain will probably be a long process, in which support by the scientific community should be considered essential. In order to bridge the gap to real-world decision processes, the following recommendations are deemed important:

• a multilevel approach in modelling that integrates conceptual and quantitative models should be adopted;
• articipatory modelling involving end users and stakeholders in modelling exercises should be implemented to enable them to understand and review the various assumptions and their implications for the modelling results;
• training and capacity-building activities are necessary at the local level;
• models and tools need to be adapted to local pre-established approaches, not vice versa;
• models and tools should be viewed as Information and Communication Technology (ICT) components of a structured and flexible approach.

1. Introduction

This report follows the 2^nd Policy Workshop of Harmoni-CA/WP5: Interaction of the Common Agricultural Policy and the Implementation of the Water Framework Directive at European and Regional Levels, which was held on 4-5 April 2005 in Brussels. The Harmoni-CA Document: HCA-WP5-2005-Re05/Final Version-Deliverable No. D 5.3.2 20.04.05 by Ilke Borowski and Johannes Heeb summarises: The workshop gave 30 agricultural and water managers from European, national and regional authorities the chance to get their hands on tools/models which may support their management activities during the implementation of the European Water Framework Directive and the Common Agricultural Policy. During the workshop an intensive exchange between the developers of the presented tools/models took place. Seven existing and easily accessible tools/models were presented and evaluated according to management questions which were defined by the participants with regard to the changing conditions under the WFD and the CAP. During the workshop the discussion of the seven presented tools/models showed that most of the tools/models can support the joint water and agricultural management processes in some regard. However there was no tool/model identified which was considered fully capable to sufficiently answer all questions raised by the policy makers with regards to the interaction of CAP and WFD. Also, no combination of the presented tools/models was able to do so. The main gaps of the tools/models were identified in terms of the general integration of different domains like water quality, quantity, ecology but especially of economic evaluation means (Borowski and Heeb, 2005, p. 3).

The report includes a to-do list for tool/model developers, which contains a summary of the demands identified during the workshop from the policy side:

1. Supporting participation - The existing tools/models can support and facilitate participation. But stakeholders have to be involved more intensively in the tool/model development process to bridge the knowledge gap between them, the river basin managers and the scientists (or in other words to make better use of existing knowledge).
2. Supporting economic assessments - There is a clear need for tools/models to support integrated economic assessments of policy measures. Existing tools/models should be integrated with existing agro-economic tools/models.
3. Transferability and applicability - Models should be developed to make a region-specific evaluation of policy measures possible. This, however, means that rather than having a few general tools/models, there is a need for a large number of specific, regional tools/models; a need which requires more resources. Tools/models should facilitate conflict and synergy management between CAP and WFD.
4. Integration - Tools/models should be developed to make more systemic work possible, integrating ecological, economical and social issues at qualitative and quantitative levels.
5. Simulation of behaviour - Tools/models are needed to simulate farmers behaviour in response to policy measures. This would help to work out concrete action plans and select the most sustainable development scenario based on the simulation results.
6. Motivate policy makers to use the tools/models - Tools/models are still mostly used in the academic world. In order to convince policy makers to use tools/models in their daily work, information about the abilities, the validity, the level of uncertainty and the conditions of applicability of the tools/models have to be clearer to the practitioners. Tools/models should make clear what added value they can provide to policy makers working in integrated and participator river basin management in the CAP WFD context. Access to the tools/models should be made easier for policy makers: promising options are toolboxes as developed, e.g. in the BMW project or the Harmoni-CA toolbox (www.harmoni-ca.info). In addition to technical solutions, more work needs to be done to understand the working constraints of policy makers so that tools/models can be adapted to fit their work.

The difficulties that model-based tools have in supporting practical water management are further explored in Borowski and Hare (2007).

The existing gap between scientists and policy-makers is confirmed by Mysiak et al. (2006) (see Table 1):

Table 1: Gap between scientists, policy-makers and water managers

(adapted from Mysiak et al., 2006)

The present report offers a broader review on tools and models, addressing both water management and agricultural management issues. It also responds to the mission of Harmoni-CA/WP5 to establish and maintain dialogue between tool/model developers and policy-makers to improve the use of tools/models in management processes.

The report is organised as follows:

• he first section briefly analyses water use in agriculture, and endeavours to clarify the policy context, introducing the Common Agricultural Policy (CAP), the Water Framework Directive (WFD) and existing interactions between the two policies.
• he second section categorises existing tools and models related to water management and agriculture.
• he third section describes three examples of good tool application.
• The final section considers how tools should be used to support a participatory implementation process of the WFD, and provides conclusions.

The review includes fives annexes that address specific issues:

1. A list of EU-funded projects concerning tools for water and agriculture.
2. The CAP
3. Measures to implement the WFD and the CAP
4. Monetary valuation
5. An information sheet on tools and models for water management in agriculture

2. The Water Framework Directive and the Common Agricultural Policy

2.1 Water use and agriculture

Agriculture is a significant user of water resources in Europe, accounting for around 30% of total use. This is particularly true in the Mediterranean region (including Greece, Italy, Portugal and Spain), where irrigated agriculture, although it covers only 20-25% of the total cultivated surface, accounts for about 50% of total agricultural production and 45-81% of the total water demand[2].

Furthermore, agriculture and forestry, which cover over three-quarters of the area of the European Union, play a key role in determining the rural economy and the environment.

Agricultural systems changed radically in the second half of the previous century. Diversified evolution patterns emerged and unforeseen negative consequences developed over time. On the one side, specialisation of holdings took place in both livestock and arable sectors, based on farm enlargement and capital intensive farming methods, which exerted great pressure on the environment; on the other side, decline and abandonment occurred.

A variety of driving forces that underlie farmers management decisions influenced this process: technological development, changes in market conditions, alterations in the costs of labour, land and other factors of production, modification of the socio-cultural environment, structural changes and a range of different public policies, including the CAP itself.

In the market economy, the comparative profitability of farming, which is highly influenced by the CAP, forced farmers to pursue financial objectives. Consequentially, traditional practices were often abandoned in favour of intensive ones: the substitution of labour with capital and of organic with industrial inputs increased farm productivity at the cost of environmental sustainability. As a consequence, water pollution due to agricultural activities is often considered a serious problem in many areas.

However, agriculture is not a homogeneous sector. In most regions, farms that adopt intensive management practices exerting great environmental pressure coexist with others that are environmentally friendly. In many cases, the latter group is dwindling in size, although only marginally in economic terms. This implies the need for targeted interventions in order to focus on appropriate management at the farm level. This would lead to favouring long-term sustainability through the balanced development of rural areas.

Devising policies that are not only capable of reducing consumption and increasing water quality but also of preserving the economic and social sustainability of agricultural systems requires a clear understanding of the complex relation between water and agriculture. This task can be enhanced by constructing conceptual models that favour interdisciplinary dialogue and the definition of a common vision among experts, stakeholders and decision-makers. Estimating the effects of a policy measure requires the identification of the causal links between the implementation of the measure and its ultimate impact on human activities and the environment. A conceptual framework for examining such links can be found in the DPSIR (Driving forcePressureStateImpactResponse) approach proposed by the European Environment Agency (EEA) for the description of environmental pathways. In a policy-making context, this approach can help policy-makers to conceptualise and structure their decisions on potential alternative measures, according to the cause-effect relationships.

Figure 1 is an example of a conceptual model of water and agriculture.

- The model identifies the qualitative and quantitative dimensions of water.

- Point and diffuse sources of pollution are connected with agriculture and farming, although they are not specific to irrigated agriculture; there is strong evidence that in many regions rain-fed agriculture and intensive animal breeding without land are the main causes of environmental impact, e.g. nitrates and other pollutants.

- Concerning agricultural irrigation, it is important to understand how farmers use water and make decisions in response to external stimuli. From an economic perspective, water is a production factor, which substantially enlarges farmers sets of choices in terms of available crops and processes (Ward and Michelsen, 2002). Irrigation has other important effects, such as the higher quality of production and risk reduction due to uncertain and unstable climate conditions, particularly for fruit, vegetables and other high-value crops. Furthermore, water enables farmers to standardise production over space and time, which is becoming a stringent market requirement, while the quantitative increase of production is merely a secondary effect.

- Water demand in agriculture is linked to agronomic factors (Doorenbos et al. 1979). On the one hand, crops have a different effect on evapotranspiration, resistance to stress and water/yield response. On the other hand, agricultural practices, cropping patterns and irrigation methods can significantly influence water requirements at the farm level and in terms of soil characteristics.

Figure 1: Example of a conceptual model of water management in the agricultural system

- Climatic conditions impact on water demand and availability in many ways. For instance, temperature influences plants evapotranspiration, water needs and rain distribution, and quantity is a key factor regarding water availability and water quality, since scarcity is generally linked to a higher concentration of pollutants in the resource.

- Economic factors operate at both macro and micro levels. At the macro level, the rate of growth of gross domestic product (GDP), the distribution of income in society and population growth are key drivers for demand in food, which can be differentiated into quality and quantity, and into perceptions of and the demands placed on the rural environment. For instance, demand for recreation in rural areas, and hence for environmental quality, is higher in rich societies where basic needs have already been satisfied. At the micro level, prices and costs at the farm level, which vary between crops and farms, as well as investment and operating costs for irrigation equipment influence farmers choices.

- Social and institutional factors play a central role in water management and should be considered with due care. All stakeholders represented by Farmers Unions, Water Users Associations, Water Authorities, Water Boards and NGOs should be involved in the participatory process of the implementing the WFD.

- Policy factors, which represent a response of society to impact, play a wide range of roles. There are many instruments that model water and agriculture (see section 2.2).

- Agriculture is a multifunctional activity that should be assessed in a multicriteria framework. Relevant dimensions are:

- the economic dimension: quantified at the micro level by the income produced by farm type and at the macro level by the contribution of the aggregate sector to GDP.

- the social dimension: represented not only by employment induced but also by the cultural aspects linked to rural life,

- the environmental dimension, which is particularly complex: positive and negative indirect effects are both present. On the one hand, positive externalities are that appreciated landscapes[3] depend on water availability in agriculture, biodiversity conservation and soil protection. On the other hand, soil erosion, soil and water pollution and subsidence sometimes pose serious problems.

- the dimension of food production: this traditional objective of agricultural activity still plays an important role. This needs to be assessed properly by considering the high value of European production with reference to food quality and safety, due to the strict rules enforced by the CAP. For example, genetically modified crops (GMO crops) cannot be cultivated due to precautionary principle; there are limits to pesticide concentrations on fruit and vegetables, which are much more stringent for organic farming, while imported crops and food do not follow the same requirements.

- Water distribution networks deserve specific attention. Many water distribution networks are old, operating under very poor conditions with low maintenance, if at all. In such cases, high losses (up to 50%) are frequent. Measures focusing on water management at the farm level cannot address this problem. However, water savings could be achieved by investing in infrastructure.

- Demand should be satisfied using water derived from legal sources, such as irrigation networks and authorised basins, drills and rivers. However, hidden sources are a real option in many situations. The use of hidden water as an alternative to legal sources is particularly relevant, considering that economic instruments (e.g. water prices, water quotas, etc.) can only be applied to the legal uses and the cost of enforcing an effective monitoring policy to reduce illegal behaviour could be extremely high.

- The groundwater recharge effects of agricultural irrigation are another critical issue to be considered. If water is not polluted, runoff and percolation are a positive externality. In such cases, the environmental benefits induced by an increase in irrigation efficiency (e.g. due to a shift from surface to sprinkler irrigation) or the reduction of irrigation itself may be questionable.

2.2 The Common Agricultural Policy reform

Since its establishment, the whole concept of the CAP and its method of operationalisation have greatly changed through successive reforms. The integration of environmental issues into other EU policies, such as the CAP, is part of a long process dating back to the 1980s. It is legally bound in the Treaty of the European Community (TEU): Environmental protection requirements must be integrated into the definition and the implementation of (all) the Community policies and activities in particular with a view to promoting sustainable development[4]. The Agreement on the Mid-Term Review of the CAP, reached in June 2003, gives form to a new European model for agriculture, reflecting the multifunctional[5] role that farming plays, and integrating economic viability, food safety, social balance and environmental concerns. The key elements of the reform, which completely changes the way in which the EU supports its farming sector, are:

1. Adoption of a Single Farm Payment (SFP) for EU farmers: aid paid to producers will no longer be dependent on type of production, thus decoupling it from or removing the causes of indirect environmental damage[6].

2. Introduction of a cross-compliance rule, which makes direct payments conditional on adherence to specific requirements[7].

3. Strengthening of the RD policy, in which instruments directly concerned with environmental outcomes are inserted[8].

The previous elements are described in further detail in Annex 2.

The new more decentralised model of agricultural policy grants Member States greater freedom by reducing the risk of distorting competition or renationalising the CAP. The reform should secure multifunctional, sustainable and competitive agriculture throughout Europe[9].

2.3 The Water Framework Directive

The 2000/60/EC Directive, known as the Water Framework Directive (WFD), establishes a framework for common action in the field of water policy in Europe. Its main goal is to raise the water quality of nearly all bodies of water to a good status by 2015. The directive is an important element of European environmental policy.

In the 1990s, economic instruments were increasingly suggested and recommended to enhance the sustainability of the environment. They found their full legitimacy in the Rio Declaration on Environment and Development by the United Nations in 1992. The WFD has a similar emphasis. In fact, it calls for the application of economic principles, i.e. the polluter pays principle, and cost recovery for water services, with the inclusion of environmental and resource costs. The directive clearly recommends economic approaches and tools, and the consideration of economic instruments to achieve its environmental objectives in the most effective manner.

The WFD identifies a strict time schedule for key tasks aimed to develop and implement River Basin Management Plans (RBMP). Moreover, article 5 of the WFD requires that each Member State shall ensure that an economic analysis of water use is undertaken for each river basin district or part of an international river basin district falling within its territory, according to the technical specifications set out in Annexes II and III. This should be completed within four years of the Directive coming into force.

The principle of the cost recovery (Art. 9) of water services, including supply, environmental and resource costs[10] should be adopted in accordance with the polluter pays principle.

Furthermore, Art. 9 requires that Member States shall ensure by 2010 that water pricing policies provide adequate incentives for water users to use water resources efficiently, thereby contributing to the environmental objectives of the WFD.

Economic analysis can also justify derogation, including the designation of a water bodys status, (Art. 4) if disproportionate costs can be demonstrated. The affected groups ability to pay should be considered in the cost assessment.

What emerges from the previous articles is that different aspects are addressed, ranging from the economic analysis of water use, cost recovery and pricing schemes, to the evaluation of benefits to the environment and society, and eventually disproportionate costs.

The following activities, which involve economic components, should be undertaken in the implementation process:

• An economic analysis of water use with the main objective of assessing the importance of water to the economy and the socio-economic development of river basin districts (RBD). Tradeoffs between socio-economic development and water protection should be estimated and made available.
• The dynamics of the RBD should be adequately investigated and the baseline scenario defined. Forecasts of the key economic drivers likely to influence water pressure should be made and the water status should be given.
• An assessment of the current levels of cost recovery for water services, considering financial, environmental and resource components, and how they are distributed among water users.
• A cost-effectiveness analysis of the main measures to be considered in the development of the RBMP.

The economic valuation of benefits is a challenging task, since benefits are many and very different in nature. A reduction of water treatment costs is a market benefit that can easily be valued. Since the goal of the WFD is to achieve higher water quality, which translates into more environmental services, many benefits are of a non-market nature. Non-market benefits could be linked to use, such as an increase in open-access recreation, but they may also be values that individuals simply relate to higher environmental quality.

Until now, Member States are left to operationalise the form of analyses and valuations. No common rules or methodologies have yet been defined. Hence, the Member States are free to adopt the procedure they wish to apply.

For further information on economic valuation, see Annex 4.

Finally, the WFD also requires an integrated participatory water resources policy, which high-quality computer-based tools could support.

2.4 The CAP and the WFD

Both the WFD and the RD adopt planning processes. However, important differences emerge, of which scale could pose a major problem: the WFD identifies River Basins (RB) as the proper planning scale, whereas the CAP follows a national/regional approach[11]. Win-win situations for both policies require that great attention is paid in the implementation phase. They also require representatives from different authorities to interact, in order for them to construct a common vision of agricultural and environmental problems.

The first column of Table 2 shows different policy objectives of the WFD: the quality objective identified in the good water status; the quantity objective related to a reduction in water demand and the economic objective related to the recovery of costs for services. Specific policy instruments are identified in the second column: some are compulsory and others are not, some belong to agricultural policy (A) and others to water policy (W), others still are a combination of both. The main effects are reported in the final column.

Table 2: WFD policy objectives, agri-environmental instruments and possible effects on water and agriculture

Many of the previous instruments require the definition of specific measures (for more detailed information on agricultural measures, see Annex 3). Many agricultural measures aim to wholly or partly improve or protect water quality (e.g. agricultural measures to reduce the use of pesticides and fertilizers) or water quantity (e.g. measures to reduce irrigation). Specific actions and measures could involve land use activities, requiring changes in land use and management (e.g. development of low-input farming systems, such as organic farming, changing from arable to grassland, investments to improve the state of irrigation infrastructure and allowing farmers to shift to improved irrigation techniques). Some instruments, such as buffer strips, crop diversity and rotation, are agronomic in nature; others, such as taxes on pollutant input, are economically-oriented.

Past experience suggests the adoption of an approach that combines schemes. Such an approach would integrate broad and shallow measures seeking widespread basic services with other narrow and deep measures that pursue clearly defined, more demanding, targeted benefits in particular priority areas.

When combined schemes are adopted, given the systemic nature of the environment, it is impossible to isolate the specific contribution of a specific measure to a given objective; instead, multicriteria valuation is used to identify the advantages and disadvantages of alternative schemes. The WFD requires that the cost-effectiveness of such alternatives should be considered, which implies rather complex economic analysis.

It is essential to integrate and coordinate between the two policies at the implementation level, in order to reduce or avoid the risk of conflict, and to take advantage of the potential of RD programmes to deliver the WFDs objectives. Future RD budgets suggest that some kind of prioritisation of objectives will be necessary, since funds are limited. On the other hand, close attention should be paid to the design of the RBMP, since it can also have a huge impact on RD. In this respect, the WFD comprises tools that can be used to mitigate conflicts: public participation, economic analyses and derogations. Alternatives should be considered when the most cost-effective measures involve disproportionate costs for rural communities, which could put the sustainability of agricultural activity at risk.

3. Categorisation of existing tools and models related to water and agriculture

There are so many tools related to water and agriculture that it is virtually impossible to offer an exhaustive review. The approach adopted in this report is to focus on available tools that have a wider application, that have been developed and applied in the context of EU-funded projects or that provide specific aspects of interest to support the implementation of the WFD.

All of the investigated tools are summarised in the table shown in Annex 5, giving the following information: tool name, brief description, spatial scale, irrigation measures, agricultural measures, economic analysis, type of measures included, link to the tools website, contact details, further characterisations and previous applications.

To categorise and summarise the tools, we adopt a framework based on four critical dimensions: spatial scale, followed by the implementation of irrigation and agricultural measures, and concluding with an economic analysis.

To synthesise the highly inhomogeneous information from the previous four dimensions, we have adopted a simplified classification approach. The spatial scale represents the geographic unit of the tool in many cases, such a dimension is not unique; the predefined levels are: farm, catchment, sub-basin, basin, region, State and Europe. Two distinct dummies (YES/NO) identify the tools capability to analyse irrigation and agricultural measures. A three-position scale is adopted to classify the economic analysis; the code reflects the economic approach adopted in the tool: MI = micro economic, MA = macro economic, OA = other approach.

Tools can embed one or more models, which are a simplified representation of a system (or process or theory) intended to enhance our ability to understand, predict, and possibly control the behaviour of the system (Neelamkavil, 1988).

A few decades ago, tools were mainly monodisciplinary, used to describe relatively simple problems in specific domains of the economy, agronomy, ecology, hydrodynamics, groundwater or surface water quality. In the past decade, we gradually moved towards more multidisciplinary approaches to problem-solving, in which several domains are integrated.

Due to its demand to integrate groundwater, surface water, ecological and economic aspects of water management at the river basin scale (holistic approach), and due to the explicit requirement to study the impact of alternative measures (human interventions), the WFD seems to support this trend of exploiting more sophisticated, integrated models.

By following the adoption of the integrated water resources management paradigm, in which public participation is a key element, the WFD moreover explicitly calls for public participation and active stakeholder involvement in the process of water resources management, and hence also in the modelling process (Pahl-Wostl, 2002). There is widespread agreement that stakeholder involvement does not imply active participation in the technical modelling itself, but rather appears as a demand to be able to understand and review the various assumptions and their implications for the modelling results (Refsgaard et al. 2005).

The tools have been classified into three classes, considering the original core of the software:

• agronomic tools
• hydrology and water quality tools
• economic and land use tools

Most tools, however, are developing towards a new class, embracing holistic integrated DS.

3.1 Agronomic tools

This group contains agronomic models to study crop yield response to water and chemicals, plant protection models and tools to support the process of feed planning and budgeting. The observed trend in this group of models and tools is also towards stronger integration.

Such tools are even more relevant in agriculture-intensive areas, where the environmental impact of rural development plans, manure and fertilizer management and of agrotechniques involving the use of agrochemicals can be assessed.

There are many versions of software available to model, and therefore predict, the agronomical, environmental and economic consequences of the complex interactions between crop management, soil and the atmosphere. Few of these guarantee a simple approach; most utilise complex relationships, e.g. models that allow the simulation of stochastic scenarios (Acutis et al., 2000; Peralta and Stckle, 2001), which can be useful in the estimation of probabilities associated with the occurrence of events. For further references, see: Donatelli et al., 1997; Johnsson et al., 2002; Lewis et al., 2003; Ten Berge et al., 2000; Wolf et al., 2003.

Many of these tools were developed in the United States of America. Examples include:

The EPIC Erosion Productivity Impact Calculator model (Williams 1990; Williams 1995) is a widely used simulation tool for agricultural policy analysis. The model, originally developed by the US Department of Agriculture (USDA), is now maintained by the Texas A&M Blacklands Research Center. EPIC is a field-scale model that can be adapted to a large range of crop rotations, management practices and environmental conditions. The model was designed to assess the impact of soil erosion on crop productivity (Williams, Jones, and Dyke 1984). The new version is called Environmental Policy Integrated Climate (Mitchell et al. 1996), reflecting the evolution of the tool. Exemplary applications include estimations of soil erosion from water and wind, and climate change impacts on crop yield and soil erosion. EPIC is also used as part of Agricultural Policy/Environmental eXtender APEX, a tool for managing whole farms or small watersheds to obtain maximum production efficiency and maintain environmental quality. Examples of its application include: terrace systems, grass waterways, strip cropping, buffer strips/vegetated filter strips, crop rotation, fertilizers, irrigation, liming, furrow diking, drainage and waste management (feed yards, dairies with or without lagoons).

Another interesting tool is CropSyst (Stckle et al., 2003), which is a cropping system simulation model, distributed free of charge. The model was developed as an analytic tool to study the effect of cropping system management on productivity and the environment. The model simulates the soil water budget, the soil-plant nitrogen budget, crop canopy and root growth, dry matter production, yield, residue production and decomposition and erosion. Management options include: cultivar selection, crop rotation (including fallow years), irrigation, nitrogen fertilization, tillage operations and residue management. A link to economic and risk analysis models is under development. Four input data files are required to run CropSyst: location, soil, crop and management files. The separation of files makes it easier to link CropSyst simulations to Geographical Information System (GIS)[12] software. CropSyst provides a platform for simulating crop rotations, an automatic management events scheduler and the possibility to run multiple simulations in connection with a GIS. All these characteristics make CropSyst ideal for scenario simulations; other models simulate crop growth processes with more detail but have lower or no flexibility in specifying routine management techniques.

CREAMS Chemicals, runoff and erosion from agricultural management systems is a field-scale model for predicting runoff, erosion and chemical transport from agricultural management systems. It is applicable to field-sized areas.

GLEAMS, Groundwater Loading Effects of Agricultural Management Systems, can simulate edge-of-field and bottom-of-root-zone loadings of water, sediment, pesticides and plant nutrients from complex climate-soil-management interactions. The tool provides estimates of the impact of management systems (e.g. planting dates, cropping systems, irrigation scheduling and tillage operations) on the potential for chemical movement. GLEAMS can be useful in long-term simulations for pesticide screening in soil/management. It can track the movement of pesticides with percolated water, runoff and sediment. The upward movement of pesticides and plant uptake are simulated with evaporation and transpiration. Degradation into metabolites is also simulated for compounds that have potentially toxic bi-products. Erosion in overland flow areas is estimated using a modified Universal Soil Loss Equation. Erosion in chemicals and deposition in temporary impoundments, such as tile outlet terraces, are used to determine sediment yield at the edge of the field[13].

AGNPS, the Agricultural Non-Point Source Pollution Model, is a computer model developed jointly by the USDA Agricultural Research Service and the Natural Resources Conservation Service to predict non-point source pollutant loads within agricultural watersheds.

The integration of agronomic models with GIS reflects recent developments.

The Agricultural Production System Simulator, or APSIM, is another well-known tool to simulate biophysical processes in farming systems. The tool, developed in Australia by the Commonwealth Scientific and Industrial Research Organisation (CSIRO), integrates models derived from fragmented research efforts, enabling comparisons to be made on a common platform. The system includes plant, soil and management modules. These modules consider a diverse range of crops, pastures and trees, soil processes including water balance, N and P transformations, soil pH[14], erosion and a full range of management controls. APSIM resulted from a need for tools that provide accurate predictions of crop production in relation to climate, genotype, soil and management, while addressing long-term resource management issues. It relates to the economic and ecological outcomes of management practices in the face of climate risk.

There are many interesting developments in Europe, including:

MicroLEIS DSS, A Land Evaluation Decision Support System for Agricultural Soil Protection, is a computer-based set of tools for the orderly arrangement and practical interpretation of land resources/agricultural management data, developed and applied in Spain. The DSS is an agro-ecological system. Its major characteristics are: data and knowledge engineering through the use of a variety of databases and innovative modelling techniques; the scaling-up of process knowledge from the micro scale to the landscape scale; land evaluation; use of monthly meteorological data and standard information; an integrated agro-ecological approach, combining biophysical data with agricultural management experience; software development for PC platforms, web- and GIS-based versions.

CROPWAT is a Decision Support System (DSS) used for irrigation planning and management, developed by the Land and Water Development Division of the FAO. It is intended to be a practical tool to help agro-meteorologists, agronomists and irrigation engineers perform standard calculations for evapotranspiration and crop water use studies and, more specifically, to devise and manage irrigation schemes. It allows the assessment of production under rain-fed conditions or deficit irrigation. Standard crop data are included in the program, and climatic data can be obtained for 144 countries through the CLIMWAT database. CLIMWAT is a climatic database that can be used in combination with the computer program CROPWAT. It allows the immediate calculation of crop water requirements, irrigation supply and irrigation scheduling for various crops for a range of climatological stations worldwide.

SIMIS, the Scheme Irrigation Management Information System, is a tool designed to facilitate the management tasks of irrigation schemes. This program, developed by the FAO, is not limited to water aspects, but covers all major issues of day-to-day management activities, including the control of maintenance, accounting, water fees and other relevant tasks. FAO advice states that SIMIS should no longer be supported.

The creation of web systems to promote the dissemination of agronomic information is a promising element to support resource management. Unfortunately, however, the following interesting applications do not have an English version:

PlanteInfo is an information and decision support system for farmers and agricultural advisers, developed and applied in Denmark. Most of the information is generated dynamically with models using frequently updated databases. A subscription system enables personalised information to be accessed. For example, since the geographic location of a user's home is known after login, local weather data are used for model calculations. PlanteInfo can store users previously entered information, such as fields, crops and actions, for future use. PlanteInfo also has a public version, which contains the facilities visible in the menu.

IRRINET is an example of agrometeorology that integrates data from different sources, up to personalised, guided irrigation scheduling for farmers. Guided irrigation scheduling is based on water balance models at the plot level, based on local weather data and information on specific agro-techniques. The service has been employed in Northern Italy since 1985, and yields an average 20% reduction in water consumption.

3.2 Hydrology and water quality tools

Models describing water flows, water quality and ecology are being developed and applied in increasing number and variety. The observed trend in the past decade is towards increasingly sophisticated computerised systems, integrating watershed processes that operate at different spatial and temporal scales, simulation models and decision-making approaches.

These tools have been developed for a variety of purposes, such as the prevention of water shortages (drought), surpluses (floods) and water impairment (pollution). The complexity of model-based water management has extended even further, as integrated system dynamics and stochastic simulations models, known as ecohydrological models, represent the state of the art of formal modelling (Reca, 2001). Ecohydrology combines the study of hydrological, biogeochemical and ecological processes and their interrelations in soil and water bodies. This type of model for a river catchment contains a hydrological module as its basic element and a vegetation sub-model, which usually includes further sub-models for biogeochemical cycles (carbon, nitrogen, phosphorus) with an appreciable level of complexity. The previous sub-models are usually combined to include interaction and feedback between processes, such as water and nutrient drivers for plant growth, water transpiration by plants, nutrient transport with water, etc.

Many earlier tools adopted nodal network approaches, which are a common framework for considering water allocation problems (see, for example, McKinney et al., 1999; Rosegrant et al., 2000; Merritt et al., 2004; Letcher et al., in press; Jakeman and Letcher, 2003; Fedra and Jamieson, 1996). In this type of model framework, a river basin is represented as a series of nodes. Nodes represent points where extraction and other activities that impact the stream are aggregated and modelled for a region. Regions refer to land or users attached to a node. Depending on the problem addressed by the model, regions may be defined by physical boundaries (e.g. sub-catchment areas) or by social, economic, technical and political boundaries. One example of this type of boundary is the property areas of irrigators who extract along a reach of a stream between two nodes. Flows are generally routed from upstream nodes to downstream nodes. Thus the impact of upstream land and water use activities on downstream users is modelled.

Examples of rules associated with nodes in a hydrological model are given in Table 3:

Table 3: Examples of rules associated with nodes in a hydrological model

The reliability of process-based models, their flexibility and level of integration have improved in the past decade (Krysanova et al., 2005; Quinn et al., 2004; Hattermann et al., 2004; He, 2003; Krysanova, Hattermann, Wechsung, in press).

The United States Environmental Protection Agency (USEPA) supports and recommends that state and federal agencies use a set of models available within a framework called Better Assessment Science Integrating Point and Non-point Sources (BASINS). Created in 1996 with subsequent releases in 1998, 2001, and 2004, the tool is a multipurpose environmental analysis system designed for use by agencies to conduct watershed and water quality-based studies. BASINS integrates environmental data, analytical tools and modelling programs to support the development of cost-effective approaches to watershed management and environmental protection, including Total Maximum Daily Loads (TMDLs).

BASINS 4.0 and BASINS 3.1 represent two co-existing versions. In BASINs 3.1, the watershed loading/water quality embedded model is the Soil Water Assessment Tool (SWAT) (Arnold et al., 1993; Di Luzio et al., 2002), developed by the United States Department of Agriculture-Agriculture Research Service (USDA-ARS). The SWAT model was created in an attempt to simulate processes as physically and realistically as possible. Most of the model inputs are physically based (that is, based on readily available information). It is important to note that SWAT is not a parametric model with a formal optimisation procedure (as part of the calibration process) to fit any data. Instead, a few important variables that are not well defined physically, such as the runoff curve number and Universal Soil Loss Equation (USLE) cover and the management factor (such as the C factor, which quantifies the vulnerability of specific land use to water erosion), can be adjusted to provide a better fit. SWAT has been applied in numerous hydrological and/or non-point source pollution studies (http:// www.brc.tamus.edu/swat/swat-peer-reviewed.pdf), including in Europe.

BASINS 4.0 uses an open source GIS software architecture, which can easily be shared with other GIS software. BASINS 4.0 includes all of the functionalities of BASINS 3.1, except the AGWA[15] and SWAT models. Furthermore, it has the ability to quickly develop and create "plug-in" functions to update or enhance the watershed analysis process and monitor the environment. Both tools adopt QUAL2K, the River and Stream Water Quality Model, as their water quality model.

MIKE BASIN (MB) is a commercial, versatile GIS-based water resource and environmental modelling package produce by DHI[16]. It provides a simple, yet powerful framework for managers and stakeholders to address multi-sectoral allocation and discharge issues in a river basin. MB represents all elements of water resource modelling: users, reservoirs, hydropower, surface water, groundwater, rainfall-runoff and water quality[17]. By default, MB aims to study water allocation within a basin; however, a water quality option and a module for simulating groundwater can also be selected. The first step in setting up a river basin model is to define an underlying model river network upon which the node references can be based. Overlays with other features used in ArcView outside MB can also be performed, e.g. shaping files of river systems, thematic maps, etc. Water demand as processes/activities (irrigation area and public/industrial water supply) can be incorporated into the model. The model allows users to define the priorities of river diversions and water extractions (water rights) from multiple reservoir systems, as well as priorities for water allocation to multiple usages from individual extraction points. The model output comprises information on the performance of each individual reservoir and irrigation scheme within the entire simulation period, illustrating the magnitude and frequency of water shortages. The Groundwater module consists of a simple physical model of an aquifer, which is conceptualised as a linear reservoir that exchanges water with water users and surface water bodies. Water balance is analysed according to pumping, recharge, seepage from rivers and discharge to rivers. The first three are assigned by the user as time series. Furthermore, time series of river flow at all nodes are simulated, enabling users to determine the combined impact of selected schemes on river flows. In order to tailor MB to the requirements of the WFD, attributes for representing the range of treatment options at all point sources are currently being incorporated, and the model itself has been extended by a water quality module. The soil erosion assessment module uses two well-known methods for simple source erosion assessments: the Universal Soil Loss Equation (USLE) and the Revised Universal Soil Loss Equation (RUSLE). Non-point pollution, including N and P-pollution from mainly agricultural areas, may be simulated in different ways at a more advanced level using the Daisy model.

RIBASIM River Basin Simulation Model is a generic model package to analyse the behaviour of river basins under various hydrological conditions, produced by Delft Hydraulics. The software includes a graphical user interface, a database, a simulation program and a tool for the analysis of results. The model package is a comprehensive, flexible tool that links the hydrological water inputs at various locations to specific water users in the basin. The tool describes a basin in terms of water sources and uses. It performs simulations of water allocation along a certain time horizon. It can be useful to identify possible water use conflicts between different types of users, such as farmers or industries, to study the sustainable development of the river basin itself and to plan adequate measures to solve conflicts or generally improve the status of the water resource.

WaterWare, Water Resources Management Information System, is an integrated, model-based information and decision support system for water resources management. WaterWare Release 5.1 is fully accessible from the internet. This DSS can address a wide range of issues, such as: determining the limits of development; evaluating the impact of new environmental legislation; deciding what, where and when new resources should be developed; assessing the environmental impact of water-related development; formulating strategies for river and groundwater pollution-control schemes, etc. The tool supports the integration of databases, GIS, simulation models, optimisation models and analytical tools into a common, easy-to-use framework.

SWIM, the Soil and Water Integrated Model, is a simulation tool for hydrological cycles, erosion, vegetation growth and nutrient transport in watershed at the meso scale, developed in Germany. The aim of SWIM is to analyse climate change and the impact of land use change on hydrology and water quality at the regional scale. The tool is based on two previously developed models, SWAT (Arnold et al., 1993) and MATSALU, which is a system of four coupled models developed for the Matsalu Bay watershed in Estonia (Krysanova et al., 1989). SWIM includes modules from both predecessors, endeavouring to combine their advantages (hydrological submodel and GRASS interface from SWAT; spatial disaggregation scheme and nutrient modules from MATSALU), and to avoid over parameterisation. A simplified EPIC approach (Williams et al., 1984) is used to simulate all crops and natural vegetation, using parameter values for each plant type from the database.

SPAW, Soil-Plant-Air-Water, simulates the daily hydrology of agricultural fields and ponds, including wetlands, lagoons and reservoirs. The objective of the SPAW model is to understand and predict agricultural hydrology and its interaction with soils and crop production, without the undue burden of computation time and input details. Field hydrology is represented by daily climatic descriptions of rainfall, temperature and evaporation; a layered soil profile with automated water characteristics; annual crop growth; and management with crop rotation and irrigation. Pond, lagoon and wetland simulations that have agricultural watershed fields or producer operations as their water source provide daily inundation levels, controlled by multiple input and depletion processes. Data input and file selection are by graphical screens. Simulation results are both tabular and graphical. Typical applications include analyses of crop water status, deep seepage, wetland inundation duration and frequency, lagoon designs and water supply reservoir reliability.

SWIM, Soil Water Infiltration and Movement, is a tool for simulating the infiltration, evapotranspiration and redistribution of water, developed by CSIRO Land and Water in Australia. The overall purpose of the model is to address issues relating to the soil water and solute balance. As such, it is a research tool that can be integrated into laboratory and field studies concerned with soil water and solute transport. It is also eminently suitable for management and education.

In the Netherlands, the RIZA (Rijksinstituut voor Integraal Zoetwaterbeheer en Afvalwaterbehandeling) has developed a set of coherent models to support the policy management of groundwater resources in the country. These models compute the hydrological effects of interventions on saturated and unsaturated zones, as well as the effects on agriculture, drinking water supply and nature.

LEACHM, the Leaching Estimation and Chemistry Model, is a suite of simulation models describing the water and chemical regime in the soil root zone, developed and applied in Australia. The suite consists of four simulation models and several utilities. The simulation models utilise numerical solution schemes to simulate vertical water and chemical movement. They differ in their description of chemical equilibrium, transformation and degradation pathways. Water regimes, pesticides, nitrogen and phosphorus and salinity in calcareous soils are simulated.

Ecohydrological models are highly relevant to studying the impact of agricultural on water systems. For instance, they can be used to quantify diffuse losses of N and P. The EU-funded project EUROHARP compared nine different methodologies and created a total of 17 study catchments across gradients in European climate, soils, topography, hydrology and land use. These methodologies are applicable at the catchment scale and are currently used by European research institutes to inform policy-makers at national and international levels.

Low Flows 2000 is a catchment-based water resource decision support tool for the United Kingdom. It is a decision support tool designed to estimate river flows at ungauged sites and to support the development of catchment and regional water resources. It is the standard software system used by the Environment Agency and the Scottish Environment Protection Agency to provide estimates of river flows, as represented by annual and monthly flow duration statistics, for any river reach in the UK.

NIRAMS, the Nitrogen Risk Assessment Model for Scotland, is a GIS-based model to calculate losses of nitrogen from diffuse pollution. Estimates of annual residual N (by crop) are leached to surface and groundwaters by hydrological flows and routed to the stream system. The system uses nationally available data sets on land use, soils, topography and meteorology. It was developed within the ArcView GIS. The model was delivered to the Scottish Executive Environment and Rural Affairs Department and the Scottish Environmental Protection Agency.

NL-CAT, Nutrient Losses at catchment scale, is a tool capable of simulating phosphorus and nitrogen losses in both the soil and surface water, developed and applied in the Netherlands to replace the previous model stone (Wolf et al., 2003). The model chain was constructed to evaluate the impact of different types of measures on the improvement of the surface water quality at the catchment scale. The surface water model consists of the two important key components, the Soil Water Atmosphere Plant (SWAP) and the Agricultural NItrogen Model (ANIMO), in combination with a surface water quantity and quality module. The SWAP module generates hydrological input to the ANIMO module, and simulates the nutrient cycle in soil and the nutrient leaching to groundwater and surface waters. Surface water discharges are simulated by a specific module, while simulation of surface water quality processes and retention within a (large) catchment is performed by the nutrient cycles in small surface waters (NUSWALITE) module. An erosion module based on the modified and revised Universal Soil Loss Equations is implemented in a GIS environment. NL-CAT is capable of simulating phosphorus and nitrogen losses in both the soil and surface water.

DRIPS, Drainage runoff input of pesticides in surface water, is a GIS-based DSS, developed on behalf of the German EPA (Environmental Protection Agency/Umweltbundesamt) for the exposure assessment of agriculturally used pesticides in surface waters. The tool estimates the quantity of pesticide input from non-point sources via surface runoff, tile drainage and spraydrift. Furthermore, the resulting predicted environmental concentration of pesticides in surface waters (PECsw) can be calculated, considering the mean daily inputs of substances into river basins, characterised by their daily discharge. A graphical user interface (GUI) was created to provide users of the DSS with easy access to the models algorithms. Model parameters can be modified by users in order to generate customised scenarios, predicting PECs for a choice of field crops, orchards or vineyards. Results are available as grid cell maps for the territory of Germany. The model aims to enable users to predict expected pesticide concentrations in river basins, thereby creating estimates of the probability of a quality target being exceeded. The results, illustrated in maps, can be used to identify hot spots of diffuse pesticide input. The water management industry can base its own measurements on these findings (Rpke, Bach and Frede, 2004).

TRK is a tool developed in Sweden for nitrogen (N) and phosphorus (P) gross and net load calculations, retention and source apportionment. The TRK system supports calculations of concentration and area losses of diffuse sources (for N from arable land using the dynamic soil profile model SOILNDB), calculations of the water balance (using the distributed dynamic HBV[18] model) and N transport and retention processes in water (using the HBV-N model). The results are presented in the GIS, and source apportionment is carried out for each sub-basin as well as for whole river basins. Results from the system have been used for international reports on transport to the sea, the assessment of the reduction of the anthropogenic load on the sea and for guidance on effective measures to reduce the load on the sea at the national scale. The tool is applied in Nordic countries and Sweden.

WetSpa, Water and Energy Transfer between Soil, Plants and Atmosphere, is a GIS-based hydrologic model that simulates hydrological processes continuously, both in space and time. Its aim is to simulate the hydrological behaviour of catchments with regard to flood prediction, land use and climate change scenario analysis and water management. The pollutant transport module aims to simulate the transport of phosphorus through a catchment. Domains covered include hydrology, land use, climate change, water management and phosphorus transport.

Other commercial tools for modelling water resources are distributed by:

EMS-I Environmental Modeling System Inc, including the Groundwater Modeling System (GMS software), Surface-water Modeling System (SMS software), and Watershed Modeling System (WMS software). For further information, see: http://www.ems-i.com/index.html.

BOSS International offers hydrology and hydraulics software - http://www.bossintl.com/.

3.3 Land use and economic tools

Another important group is represented by spatial models, where integration is again the paradigm. Integration of economic and ecological information in a spatial context is recognised as a valuable approach for strategic policy development and decision-making (e.g. Tiwari et al., 1999). Several conceptual frameworks encompassing the science, methods of capturing data and responses, and the human-biophysical dimension of problems have recently become the focal point of interdisciplinary research and analysis.

Lambin et al. (2000) categorise land use (LU) and land cover change (LC) models into five classes:

1. Empirical-statistical: uses multivariate statistical analyses of relevant factors to identify the causes/drivers of LU change. This type of model suffers from obvious limitations, since there is no reason why any statistically identified set of LC factors will apply robustly (or at all) outside the region or space/time slice from which they were identified.

2. Stochastic: a transition matrix quantifying probability describes how different LU types change from one to another. Since probabilities are determined from an analysis of the past, they can suffer from the empirical problems of a lack of robustness/generality, and of modelling data rather than the world.

3. Optimisation: LU change is modelled as a process, whereby one measure is optimised (e.g. income per unit land). The approach assumes that actors behave rationally and are in search of the best solution.

4. Dynamic process-based: a more complex type of model that uses representations of social, economic and biophysical processes to simulate LC. Such complexity can require substantial effort to develop, verify and validate the data. However, the models can overcome the problems of previous approaches, and provide more insight.

5. Integrated: this type of model combines the characteristics of the above four categories.

In recent years, there has been a rapid expansion of interest and research into spatial decision support systems (SDSS), to which all the previous methods are applied. To varying degrees, these approaches attempt to

- capture the system dynamics;

- deliver outputs as spatial data that define biophysical, economic and social constraints;

- use new methods to translate factor layers into standardised inputs for problem criteria definition;

- use new methods to capture uncertainty in the ranking of alternatives.

The consideration of socioeconomic decisions and impact components requires the inclusion of decision models that need to represent the key land use, water use and management decisions made in the catchments, e.g. agricultural production decisions, industrial and urban water use decisions, reforestation and urbanisation decisions. The specific decisions to be simulated and the types of models used to represent these decisions will depend on the spatial and temporal scales at which these decisions are to be modelled as well as on the types of activities present in the catchments. For example, even where extractive uses such as irrigation direct from the stream are considered, this decision may be modelled differently depending on whether the decision is posed as a short-run decision, considering capital to be constrained, or a long-run decision where capital investment decisions are included in the model. Additionally for some issues a representative farm model, simulating decisions by an individual farm, may be used, whereas for larger scale studies, or studies where trade-offs between different industry users are to be considered, aggregated regional production models may be used. In either case, it is the relevant land and water use decisions that are being represented. Letcher et al. (in press).

A key element is the separation of the economic theory into two branches: macro and micro.

Macroeconomics is the study of aggregated variables addressing the state of the whole economy. The focus is on price levels, employment levels, economic output in real and monetary terms, the quantity of money in the economy, overall consumption, savings, investment, wage levels, etc. The aim is to provide an insight into guiding the formation of economic policy. The following tools implement macro-economic theory in operational models:

AGLINK is a multi-country and commodity dynamic model of world agriculture, developed by the OECD Secretariat in close co-operation with its Member States. The overall design of the model focuses primarily on the potential medium-term influence of agricultural policy on agricultural markets. AGLINK is a partial equilibrium model, primarily of major OECD commodity markets. AGLINK estimates supply, demand and prices. Non-agricultural sectors are not modelled, and are treated as exogenous to the model. Since 2004, this modelling system has been enhanced by the FAOs development of the COmmodity SImulation Model (COSIMO), which represents agricultural sectors in a large number of developing countries. The AGLINK-COSIMO modelling system is currently one of the most comprehensive partial equilibrium models for global agriculture. The model is one of the tools used in the generation of baseline projections underlying the OECD-FAO Agricultural Outlook.

CAPRI the Common Agricultural Policy Regional Impact Analysis is an agricultural sector model covering both the whole of EU27 and Norway at the regional level (250 regions) and global agricultural markets. The following environmental indicators are covered by CAPRI: balances for N, P and K; emissions of ammonia, methane and N2O; global warming potentials. CAPRI users work in research institutions and EU Commission services.

WATSIM the World Agricultural Trade Simulation Model is a recursive-dynamic, spatial world trade model for agricultural commodities. In its current version, it covers 12 regions and 29 commodities. Simulations run from the year 2000 to 2010. Policies covered are ad-valorem and specific tariffs, tariff rate quotas (TRQs), safeguards (flexible tariffs), export subsidies and production quotas. These policies are explicitly modelled by formulating the model as an MCP (mixed complementarity problem). The most important application of WATSIM is the medium-term analysis of trade policy changes.

GTAP The Global Trade Analysis Project is one of the most relevant projects in this field. GTAP is a global network of researchers and policy-makers who are conducting a quantitative analysis of international policy issues with the goal of improving the quality of the quantitative analysis of global economic issues within an economy-wide framework.

Micro-economics explores economic agents, in this case farmers, and their behaviour and interrelation. The interpretation of a systems behaviour is an important classification criteria:

Neoclassical economic theory describes decision-makers (e.g. farmers) as income maximisers, which implies a homogeneity assumption in preferences (Varela-Ortega et al., 1998; Iglesias et al., 2004).

In an alternative multicriteria (MC) perspective, the farmers decision-making processes are simultaneously driven by various attributes related to economic, social, cultural and natural dimensions (Amador et. al, 1998; Romero, 1989). Such attributes include the maximisation of income, the minimisation of risk, the maximisation of leisure time, the minimisation of managerial problems, etc. (Sumpsi et al. 1996). A decision-maker tries to satisfy all these criteria at the same time, and the choice is a compromise reached despite conflicting objectives (Berbel and Rodriguez, 1998; Gomez-Limn et al., 2002).

Other approaches follow a systems behaviour paradigm, including: agent-based modelling, cellular automata and related models that focus on complex spatial interactions; network theory and dynamical systems.

The previous approaches are adopted by different tools working at the micro scale, where a farm model describes the production possibilities (e.g. agricultural practices and management options) by engineering production functions that follow an approach, known as the primal representation of technology. With this approach, inputs and outputs including externalities are expressed in physical terms. This implies the possibility to measure and evaluate the effects of different sources of change on the environment as well as on the socio-economic dimension. The approach allows changes in soil and irrigation management practices, changes in the use of farm input and natural resources, and how agri-environmental measures and environmental regulations may constrain actions taken by farmers (Yaron and Dinar, 1982; Flichman et al., 1995; Doppler et al., 2002; Iglesias et al., 2003; Yang et al., 2003; Bazzani et al., 2005). An alternative approach is positive mathematical programming (PMP), which derives unknown production functions from observed choices through the statistical calibration of the model (Howitt, 2001). PMP has generated numerous applications and extensions at different investigation levels, several of which are reported in Heckelei and Britz (2005).

The EU WADI project (The sustainability of European irrigated agriculture under Water Directive and Agenda 2000) is a recent example of economic modelling in agriculture to analyse the sustainability of irrigated agriculture in Europe in the context of post-Agenda 2000 CAP Reform and of the Water Framework Directive (website: http://www.uco.es/grupos/wadi/). The research, conducted in five European countries, adopts a MC model-based approach, focusing on representative farms. Scenarios are used to anticipate the possible future state, related to macro-economic and institutional conditions. Irrigated agriculture is assessed in terms of sustainability, taking into account the multi-functional dimension. A set of indicators offers an insight into economic, social and environmental dimensions. The agricultural model is composed of a collection of microeconomic mathematical programming models, each representing the optimising farmers behaviour at the farm level. Simulation results are aggregated according to farm localisation, type and size at higher territorial levels. The conclusion of the WADI project deserves attention: Future demands for agricultural economists and operational researchers imply a demand for models capable of generating detailed microanalyses based upon diversity in farming systems. Such diversity appears to be greater than ever, as conventional agriculture faces up to organic farming, and as various technologies (GMOs, etc.) compete for consumer markets; global equilibrium may not reflect diversity in the adaptation of strategies and the impact of measures. For this reason, we need to encourage the development and adoption of simple models of irrigated agriculture that are capable of simulating changing policy scenarios and measuring the impact of such changes on social, economic and environmental indicators. Agricultural economic models may help us to understand the value and cost of water and the multiple links between uses, consumption and pollution, by simulating the way in which these aspects are connected through the farming systems and their adaptation to external scenarios modelling can support the definition of institutional measures that will facilitate the achievement of objectives (e.g. markets, quotas, economic support), by identifying incentive-and institutionally compatible alternatives and supporting their choice through simulations Models could play a major role in the analysis and management of conflict of water use as well as in the analysis of the potential benefits of cost-sharing. Berbel (2004) pg. 217.

A recent contribution analyses the implications and applications for the European WFD of hydro-economic modelling in river basin management (Heinz et al., 2007). Four main applications are discussed:

• Economic, social (e.g. employment) and environmental impacts of water pricing.
• Cost-effectiveness analysis of integrated management strategies for pollution control.
• River basin modelling to analyse resource and environmental costs.
• Participatory planning.

To support policy, the socio-economic sub-system needs to include a component that describes the relevant social and economic impact of change in other system components, such as the impact on farm income and employment. In some cases, local impacts are aggregated and transferred to a higher scale model to assess the impact on the regional economy. Again, the scale and range of impacts to be considered dictates the type of modelling approach used. The choice (and subsequent implementation of the chosen option) represents the final phase of the decision-making process. This stage can be supported by multi-criteria decision analysis (MCDA). Once the criteria for evaluation have been determined, options can be compared and assessed against their expected impacts using Multi-Criteria Analysis (MCA) evaluation techniques.

The European school of MCDA (Roy, 1996) has created extensive literature and diverse methodologies for the application of MCDA, including the ELECTRE[19] family. Alternative approaches to MCDA are the Analytical Hierarchy Process (AHP) (Saaty, 1980) on the one hand, and an array of cognitive methods on the other. These approaches address the decision-making process in detail, and deal with a limited, clearly defined set of alternatives[20].

Many computer-based approaches have been developed to deliver MCDA, or elements thereof, in a range of forms, e.g. MACBETH[21], Intelligent Decision System IDS[22] routines in IDRISI[23] or MULINO-DSS, and more recently in NetSyMod (Network Analysis - Creative System Modelling - Decision Support), which is an operational tool that aims to support and guide decision-makers at each step of the overall decision-making process, from problem conceptualisation to the choice of the best policy to solve it.

Many existing tools have added socio-economic modules to their environment, and economic studies have also ensued (see: Ward et al., 1996; Rosegrant e al., 2000; Cai and Rosegrant, 2004; Reimund et al., in press). Examples include:

The Decision Support for Irrigation (DSIrr) is a support tool for scenario analysis using simulation approaches and bio-economic optimisation models. Scenarios evaluate a full range of water development and management options, taking account of multiple and competing uses of water. Scenarios can also describe agricultural policy or other exogenous drivers. A multi-scale modelling approach is adopted, and different types of farms can be aggregated to create a model network. The individual models are specialised to address specific questions and reflect heterogeneity between farms. Farm preferences can be described using a multi-attribute approach. Changes in irrigation and farm practices are considered. Standard procedures assess the impact of alternative water pricing schemes and of reduction in water allotment. The tool allows the CAP to be simulated, including decoupling subsidies, eco-conditionality and agri-environmental measures. However, case-by-case implementation is required. A set of indicators is quantified to assess socio-economic and environmental impacts at farm and regional levels. The tool, which was developed and applied in Italy, can easily be linked to other models.

FARMIS assesses the impact of the Mid-term Review policy reform on German agriculture by simulation approaches and optimisation models based on the Farm Accountancy Data Network (FADN)[24]. The model network includes different models that are updated and combined with each other as necessary. The individual models are specialised to address specific questions (for example, regional impacts, farm impacts, market impacts). Comparative-static analysis is conducted, where the main purpose is to analyse the sectoral effect of policy measures addressing the farm level.

The Interactive Component Modelling System (ICMS) is a framework in which to embed models and tools for the analysis and presentation of environmental options for environmental managers. It applies scenarios as model inputs, which take into account achievable policy/management drivers and socio-economic capacity for change, as well as uncontrollable system shocks. The DSS computes indicators, covering a spectrum of biophysical and socioeconomic impacts for each scenario as model outputs. The tool was developed in Australia and is being applied in Asia.

The Integrated Water Resource Assessment and Management (IWRAM) software contains two modelling toolboxes that utilise a nodal network structure for catchment analysis: a Biophysical Toolbox (erosion, streamflow, crop) and an Integrated Modelling Toolbox, which links models of household decision-making to allow for the consideration of socio-economic and environmental trade-offs of many development and policy scenarios. Previous applications were performed in Northern Thailand.

The Spatial Decision Support System (SDSS) provides watershed economic analysis by maximising the profit of a representative farm assumed to cover the whole watershed with the constraints of production technology, resource, sediment control objectives and sustainable utilisation. The tool includes two major types of models: static and dynamic. Each model type supports variations in plant growth, grazing and ranch operations. Upland erosion is estimated through RUSLE2 and the sediment yield of a watershed is estimated from upland erosion and sediment delivery ratios. The tool was developed and is being applied in the United States of America.

WaterStrategyManDSS (WSM DSS) is a GIS-based Decision Support System developed for the WaterStrategyMan Project. It aims to assess the state of a water resource system in terms of sources, usage, water cycles (pathways) and environmental quality. In addition, it evaluates the effects of actions on the basis of different scenarios, alternatives and policies. Water allocation is performed according to a set of demand and supply priorities, reflecting pricing systems, social preferences, environmental constraints and development priorities. WSM DSS includes the following types of management options: supply enhancement options, which are intended to increase available water quantities during drought, are concerned with structural interventions that attempt to enhance water supply; demand management options, aiming to decrease water demand through various conservation techniques and use limitations; socio-economic measures to mitigate impacts, such as pricing and changes in regional development; methods to produce management strategies through combinations of control measures seeking optimum and efficient solutions. Applications have been carried out in the Mediterranean region.

LADSS, the Land Allocation Decision Support System, is a farm-scale land use planning tool being developed in the United Kingdom to assist in the case-based investigation of policy and environmental change impacts on land use systems. The tool supports strategic, farm-scale, land use planning by suggesting possible combinations of land use to meet multiple objectives. LADSS provides a framework within which the financial, social and environmental consequences of changes in land use can be evaluated. It acts as a channel for technology transfer from land use scientists to land managers, and facilitates the inclusion of practitioner knowledge into models of land use systems.

The DSS for the Elbe River Water Quality Management includes models, spatial and non-spatial data and analysis tools under a user-friendly GIS-based interface. The tool confronts decision-makers with possible measures, as well as multiple management objectives. DSS helps water managers to formulate policy for river basin management and to take appropriate action to realise policy objectives. Furthermore, the DSS is especially suited to support participative decision-making.

The management of Regional German River Catchments (REGFLUD) is a project that addresses the problem of diffuse pollution. The overall objective is to develop and apply multi-criteria scientific methods to set up a DSS aimed at reducing diffuse pollution in river catchments, subject to economic feasibility and social acceptability. A model network is constructed, consisting of an agricultural sector model, a water balance model and a denitrification model that enables policy measure analyses to be performed. Alternative agri-environmental measures were carried out by a benefit-cost approach based on interviews regarding the social acceptability of alternative measures. Alternative forms of farm management regarding nutrient surplus were analysed as a guide for water management.

AQUATOOL is a generalised DSS for water resource planning and operational management, developed in Spain. The model simulates the operational management of the system on a monthly basis. It is responsible for water allocation to water uses, and considers the connected use of surface water and groundwater. Operating policies are defined by the following variables: target, minimum and maximum volumes of reservoirs, inter-reservoir relationships and priorities of use, minimum flow in rivers, flow requirements for hydroelectric plants, targeted water demand for each agricultural, industrial and domestic area and demand priorities used in water allocation.

AgriBMPWater is a tool designed to compare Best Management Practices (BMPs) in terms of environmental efficiency, economic cost and potential acceptability by farmers. It adopts negotiation-based implementation methods.

DANUBIA is a DSS currently under development in the Global Change in the Hydrological Cycle (GLOWA)[25] - Danube project. Upon completion, DANUBIA will be able to simulate water-related issues of environmental management under ecological, economic and cultural aspects, such as flood risk and protection, agriculture, water quality and quantity, tourism and water, as well as water and climate. It will examine the sustainability of the proposed solution scenarios.

A tool that adopts a different approach is PlayAgriPoliS, a policy simulation game implementing an agent-based model that establishes a link between agricultural policy reform and structural changes. Although PlayAgriPoliS is currently under development, a beta version is currently available for download.

The Water Evaluation and Planning system (WEAP) is a user-friendly software tool applied in the USA that follows an integrated approach to water resource planning. The tool supports the identification and evaluation of the impact of climate change on water for agriculture, and entails alternative strategies to study the cost of water in watersheds. The tool supports the assessment of water supply augmentation through an inter-basin transfer within a firm yield analysis and the development of supply and demand balances.

SIGRIA, the Information System on Water Management for Irrigation, is an integrated set of tools developed in Italy by the Istituto Nazionale di Economia Agraria (INEA) for the Ministry of Agriculture and for regional governments. SIGRIA is used: to plan new infrastructure, including dams and water distribution networks; identify and solve competitive situations for water by different users; mitigate the effects of annual droughts; develop and implement new policies on water pricing. At the local level, SIGRIA is a key component of information systems for irrigation managed by reclamation consortia. SIGRIA is based on GIS technologies and includes: a detailed map on land cover/use, identifying all irrigated areas and crops grown per season; an irrigation suitability map; all irrigation water networks, from the water source (dam river, etc.) to the farm or group of farms, including a database on the technical features of the network; extensive databases on the features of irrigation (i.e. crops, farms, costs) at the local level; a linear programming model on the optimum allocation of resources for irrigation for certain areas, at farm and basin levels, and a model on local water requirements by crops, depending on soil and weather data.

3.4 Web portal on water-related tools

Water managers, decision-makers, extension agents, regulatory agencies, planning organisations, consultants, students, researchers and environmental groups pose a number of questions, such as:

What new models are available that I may not be aware of?

What modifications and new versions have been added to my favourite model?

Are new user interfaces, general databases or other time-saving devices available?

Which applications can I take as a reference to address my problem in water management or water policy?

Consistent and comprehensive answers to the previous questions are essential for the practical and wide-spread application of models and tools.

Many existing models have been used or developed in the context of projects funded by the EU in recent years, and others are soon forthcoming (see Annex A for a list of relevant projects).

The importance of spreading existing information to all actors, updating it and making it easily available is crucial. For this reason, the EU has funded specific projects to create a European-wide comprehensive, shared web-based data and information management system for water, including river basins, with the primary purpose of consolidating information for potential users (Table 4).

Table 4: European web portal, including toolbox for water

Other water-related web resources are:

The Register of Ecological Models (REM), a meta-database for existing mathematical models in ecology. It is a co-operative service of the University of Kassel and the GSF - National Research Center for Environment and Health) -
http://eco.wiz.uni-kassel.de/ecobas.html

The Hydroarchive website is an open source toolbox for model optimisation, evaluation and hydrological modelling, including model codes for rainfall runoff modelling - http://www.sahra.arizona.edu/software/index_main.html.

The Hydrologic Modeling Inventory is maintained through a cooperative effort between the Bureau of Reclamation and Texas A&M University - http://hydrologicmodels.tamu.edu/

The West National Technology Support Center (WNTSC) for the USDA Natural Resources Conservation Service (NRCS) -
http://www.wsi.nrcs.usda.gov/products/W2Q/H&H/Tools_Models/tool_mod.html

The Archives of Models and Modeling Tools[26] of the Surface-water quality and flow Modeling Interest Group (SMIG) of the US Department of the Interior | US Geological Survey -
http://smig.usgs.gov/SMIG/model_archives.html

The Surface Water and Water Quality Models Information Clearinghouse (SMIC) is a database of the descriptions and features of environmental surface water and water quality models, and abstracts of projects using such models -
http://smig.usgs.gov/SMIC/

A quite different tool is AQUASTAT the FAO's global information system on water and agriculture. This information system consists of many databases, including: dams, institutions, river sediment yields, the state of water resources and agricultural water use by country and region, climate, a global map of irrigated areas - http://www.fao.org/nr/water/aquastat/main/index.stm

As far as water and agricultural models and tools are concerned, two new projects are relevant:

The first, Open Modelling Interface and Environment (OpenMI), aims to create a standard for model linkage in the water domain. It is being developed by the EU-funded HarmonIT project[27]. Specific objectives are that the mechanisms design should:

• be applicable to new and existing models;
• impose as few restrictions as possible on the modellers freedom;
• be applicable to most, if not all, time-based simulation techniques;
• require a minimum of change to the program code of existing applications;
• keep the costs, skills and time required to migrate an existing model to a minimum;
• be easy to use;
• not unreasonably degrade performance.

The assumption is that integrated catchment management requires integrated analysis that can be supported more effectively by integrated modelling systems. These can be developed and maintained b