Evaluation of RUSLE and PESERA models for predicting soil erosion losses in the first year after wildfire in NW Spain

The accelerated erosion that occurs after wildfire in forest ecosystems is a major factor affecting the sustainability of these environments. Soil erosion prediction models are crucial for estimating fire impacts and for planning post-fire soil stabilization measures. We evaluated the performance of the Revised Universal Soil Loss Equation (RUSLE) and the Pan-European Soil Erosion Risk Assessment (PESERA) models by comparing the soil losses predicted by the models and the soil losses measured in the first year after wildfires in NW Spain. Two alternative RUSLE approaches were also considered: i) the use of reduction coefficients for R and C factors and ii) the application of a soil erodibility reduction factor to take into account soil surface stone cover. The RUSLE model overestimated the observed erosion rates in all cases. The PESERA model performed reasonably well in predicting soil losses, but needs a better characterization of soil burn severity in the soil inputs (crusting and erodibility) because of the high sensitivity of the model to these parameters. In both cases, there is an urgent need for an appropriate assessment of the impact on soil burn severity in the erosional process. This seems critical if these models are to be used for post-fire soil stabilization activities planning. (C) 2016 Elsevier B.V. All rights reserved.