Spatial Disaggregation of Agricultural Production Data: An Econometric Approach using Minimum Cross Entropy

The objective of this paper is to propose a spatial disaggregation procedure that will be useful in assessing economic and environmental impacts of the last CAP reform of EU agriculture(e.g., the study of water pollution due to agriculture). The proposed disaggregation method uses agricultural data in conjunction with biophysical processes to break down agricultural FADN data available at the regional level to 100m× 100m pixel spatial units. When the agricultural data are more precisely located, they can be more efficiently used to evaluate environmental policies. The proposed model is a two-step procedure. First, we estimate a land use model using a Multinomial Logit (MNL) model. Second, we disaggregate the observed FADN regional land use shares using a Cross Entropy (CE) approach, and taking the first step predictions as priors. This procedure is applied to the Picardie region.