Author's personal copy Runoff and water erosion modelling using WEPP on a Mediterranean cultivated catchment Damien Raclot * , Jean Albergel IRD, Laboratoire d’e ´tude des Interactions Sol-Agrosyste `me-Hydrosyste `me, UMR LISAH, Agro.M-INRA-IRD, 2 place Viala, 34060 Montpellier Cedex 1, France Received 30 July 2005; accepted 6 July 2006 Available online 28 September 2006 Abstract Soil erosion due to water is a major environmental problem in many parts of the world. Most of Mediterranean countries are con- cerned because of their specific climate and soils sensitivity, but also because of the recent intensification of human activities and agri- cultural practices. Accurate estimation of soil water erosion for various land-use and climate scenarios is so an important key to define sustainable management policies. In the last decades, several studies have been carried out to build models suitable for quantifying soil erosion. Among these models, the Water Erosion Prediction Project (WEPP, Flanagan, D.C., Nearing, M.A., 1995. USDA-Water Ero- sion Prediction Project: Hillslope profile and watershed model documentation. NSERL Report 10, USDA-ARS National Soil Erosion Research Laboratory, West Lafayette, IN, USA.) is a physically based, distributed-parameter model that has been developed and mainly validated in USA. Only few studies have investigated its applicability to environmental conditions that differs from those where the model was developed. The aim of this work is to test the efficiency of WEPP model to predict soil erosion at catchment scale in a Med- iterranean semi-arid area. Continuous simulations have been conducted between 1995 and 2002 on an cultivated experimental catchment located upstream from a hill reservoir (Kamech catchment, 2.45 km 2 , Cap Bon, Tunisia) where runoff and soil erosion measurements are available at the outlet. Comparison between predictions and measurements shows significant differences. Processes related to seasonal effects (as cracking soils) are pointed out as a weakness of WEPP model for Mediterranean conditions. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Water erosion prediction project; Soil erosion; Runoff; Hill reservoir; Mediterranean catchment; Tunisia 1. Introduction Accurate estimation of soil erosion due to water is very important for the assessment of potential soil loss from cul- tivated lands and the evaluation of the loss of water storage capacity in reservoirs due to sediment deposition (Amore et al., 2004). Due to its modest data demands and transparent model structure, the Universal Soil Loss Equation (USLE) remains nowadays the most popular and widely used model for prediction of water erosion hazard and planning of soil conservation measures (Sonneveld and Nearing, 2003). This lumped empirical erosion model was adopted in 1958 by the Soil Conservation Service in USA. After, it was first updated in 1978 using numerous experiment data sets collected in the USA (Wischmeier and Smith, 1978) and then derived in multiples models such as the Revised Universal Soil Loss Equation-RUSLE (Renard et al., 1998) or the Modified Universal Soil Loss Equa- tion-MUSLE (Williams, 1975) in order to improve the pre- dictive capability of the USLE. Nevertheless all these models are based on empirical coefficients which must to be accurately evaluated from field observations for each specific situation (crop type, protection, climate, ...). In recent decades, several studies have been carried out to build physically based models. One of them is the Water Erosion Prediction Project (WEPP, Nearing et al., 1989; 1474-7065/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.pce.2006.07.002 * Corresponding author. Tel.: +33 4 99 61 22 51; fax: +33 4 67 63 26 14. E-mail address: raclot@ensam.inra.fr (D. Raclot). www.elsevier.com/locate/pce Physics and Chemistry of the Earth 31 (2006) 1038–1047