| Popis: |
Wildland fires are one of the major causes of ecosystem degradation, especially in semiarid climates, where the erosion hazard is high. The identification of potential erosion zones is typically difficult as it requires expensive field and laboratory work. This paper proposes a methodology based on remote sensing and GIS techniques, which permits speedy identification of erosional areas in a semi-automatic way, tested in a large burn scar in south-eastern Spain. Inputs were slope, aspect, and fire severity. In order to obtain the latter a new method has been proposed, based on the difference in NDVI between two images (acquired before and after the fire event). Combining these maps in a GIS, a Forest Intervention Priority map (FIP) is produced, which identifies areas of high erosion potential. Field work was conducted to assess the method. Results indicate that the applied methodology reliably predicted the extent of very severe fire and, further, was generally useful for identifying sites of significant erosion. Additional work is required to refine: (1) remotely sensed fire severity thresholds, particularly for other Mediterranean forest systems and substrate conditions; and (2) associated mapping tools for informing post-fire management applications. |
