Crop simulation models are powerful tools both in research and in rural planning. Applications depend
on the availability of input data and on their suitability to model requirements. In agroecosystem
simulations oriented to rural planning (e.g. leaching evaluation at regional scale), multi-year crop
rotations are needed. Crop rotations can be retrieved by Common Agricultural Policy (CAP) databases,
where single crops and surface areas are yearly recorded at the cadastral level. However, simulations at
the regional scale require an abatement of the very large number of rotations reconstructed from the
crops series over the same field in a given time window in order to alleviate time-consuming simulation
runs and to better explain model outputs. Moreover, the lack of information over a part of the total
agricultural area must be filled using the available data in order to run complete geographical
simulations. An approach aimed at upscaling crop rotations from the field to the regional scale and
reducing the number of crop rotations while maintaining the representativeness of local agricultural
systems and the linkage to the geographic features is proposed. The approach takes also into account
that cropping systems simulations will be carried out over long time periods (decades), therefore
requiring repeated rotations cycles.
