Indicators in time, space and multiple domains: lessons from applying an integrated assessment tool for agricultural systems.

In Integrated Assessment and Modelling (IAM) quantitative simulation models are frequently used to compute indicators on several dimensions of sustainability. These quantitative simulation models can be derived from different disciplines and formalisms and can operate on different temporal and spatial scales. In the SEAMLESS project, a set of such models were integrated into an assessment tool for agricultural systems that targets ex-ante assessment of policy and technology changes. Processes at field, farm, regional and EU market level were captured in biophysical and bio-economic models. An example of the integrated use of these models for indicator assessment will be presented. Based on SEAMLESS several lessons can be drawn with respect to the definition and calculation of indicators. First, the spatial and temporal coverage and integration of existing statistical data sources is crucial to come to a comprehensive assessment and to derive indicators on the different aspects of sustainability. This is not a trivial task as concepts and scales have different meanings in different domains and because of data availability. Second, through typologies and statistical techniques, it is possible to reach a high spatial and temporal resolution of the calculation of indicators. For example, by developing biophysical and farm typologies, diversity in different data sources can be captured and these typologies can be linked through statistical techniques. Hereby many different indicators can be calculated EU-wide for agricultural systems in agri-environmental zones with relatively homogenous conditions for farming. Third, several quantitative models can be used in a standardized and homogenized way to provide consistent indicators across scales if their data requirements and key model outputs have been aligned. Finally, transparency is crucial in relation to the definition, calculation, and delivery of indicators in integrated assessments. On the one hand this requirement is met by developing a software framework that allows the user to explore the calculated indicators. On the other hand we hope to facilitate transparency and to maximise the impact of our efforts by making models and data available under open source conditions