Using SPOT images and field sampling to map burn severity and vegetation factors affecting post forest fire erosion risk

International audienceRunoff and erosion rates are known to increase substantially after a major forest fire. Erosion control measures therefore need to be put into place quickly after a large fire, and determining where to locate the measures requires accurate mapping of post fire erosion risk. Burn severity can be determined from field observations, but these are costly and time consuming. Satellite imagery is an alternative to quickly map burn severity for erosion mapping purposes. Post fire erosion decreases as forest vegetation recovers and this is related to both pre fire vegetation characteristics and soil properties. The objectives of this study were to test the use of SPOT multispectral images for mapping burn severity, pre fire vegetation density, and longer term (2.5 years) vegetation recovery. Indices tested include the Normalized Burn Ratio (NBR), Normalized Difference Vegetation Index (NDVI), Differenced Normalized Burn Ratio (dNBR), and Differenced Normalized Difference Vegetation Index (dNDVI). Indices were compared to field data gathered immediately after the fire and about 2.5 years later. The multi-temporal indices (dNBR and dNDVI) were more useful for burn severity mapping in the heterogeneous forest-scrubland-vineyard environment where bedrock and vineyard surfaces were confused with burned areas when using single images (NBR and NDVI). All of the post fire indices showed traces of the fire 2.5 years later; this was confirmed using Analysis of Variance where differences in indices were related to original ground cover (pine forest, mixed forest, mostly bedrock and vineyards) and the burn scar. Pre fire vegetation cover was less successfully mapped using the NDVI according to the initial field observations. Similarly, none of the post fire indices were able to distinguish differences in N-S vegetation recovery revealed by field measurements of understory vegetation height and cover; N facing slopes had deeper finer soils and these more favourable conditions led to greater vegetation growth than on S facing slopes and topslope convexities. It is suggested that relationships between topography and soil properties can be useful for mapping both soil erodibility and post fire vegetation recovery