Modelling soil erosion hazard by using a fuzzy knowledge-based approach

Land degradation as the result of water erosion represents a serious environmental threat. The processes are complex and depend in many determined on turn by a set of characteristics. knowledge-based exploratory models are designed to investigate the susceptibility to erosion by incorporating expert knowledge, in the sense that information on soil properties and/or landscape elements assumed to control accelerated soil erosion are incorporated into the modelling process. Fuzzy boundaries are applied to rank the landscape factors used in modelling the likelihood of an area to be affected by different erosion degrees. Although the model provides qualitative estimations, it reveals very useful to explore indicators-causes-processes relationships. In addition, it allows to test the importance of individual landscape elements related to soil erosion, and select those suggested as best predictors.