A knowledge-based approach for C-factor mapping in Spain using Landsat TM and GIS

The cover and management factor (C) in the Universal Soil Loss Equation (USLE), is one of the most important parameters for assessing erosion. In this study it is shown how a knowledge-based approach can be used to optimize C-factor mapping in the Mediterranean region being characterized by unhomogeneous land cover. The study area covers a quarter scene of a Landsat TM in the Madrid region. Via a visual classification of multi-temporal false colour composites, the study area could be divided into major land classes consisting of urban areas, natural vegetation and areas dominated by annual crops. Each class was sampled separately followed by a supervised classification using the maximum likelihood algorithm. With the use of additional data, an improved distinction was made between cover types characterized by different risk levels with respect to the C-factor. For areas with natural vegetation a clear distinction was made between forested and non-forested areas and for areas with annual crops, potential fields with sunflower, fallow land and olive plantations were identified. Vegetated urban areas, of which some were expected to have high C-factors, could not be assigned any C-factor due to the limitations of the USLE