Assessment of soil vulnerability to erosion by a fuzzy logic model

Abstract

A simplified fuzzy logic model to assess soil vulnerability to degradation processes, such as soil erosion, is presented. The model is based on the use of fuzzy logic to operate with explanatory continuous variables esti­mated from remote sensing images and digi­tal elevation models. It combines explana­tory variables concerning the state of vege­tation cover and topographic parameters. The estimations are made by a decision sys­tem implemented in a geographic informa­tion system (GIS). Output vulnerability maps can be updated with new satellite im­ages. The study area is located on the south­east Iberian Peninsula, characterized by a semiarid Mediterranean climate. It was found high vulnerability for all estimations in relation to the expected maximum and minimum thresholds. Responses of the vulnerability changes in relation with changes of existing classes of potential erosion (based on the USLE) were analysed. It was considered a high correlation between changes in classes of soil erosion and changes in vulnerability estimations.