DEM based Extraction of LS factor: integrate channel networks and convergence flow

LS is a very important factor in erosion evaluation. Slope gradient and slope length is used to calculate LS factor. However a major limitation is the difficulty in extracting the LS factor at regional landscape scales. The geographic information systembased (GIS-based) methods which have been developed for estimating the slope length for USLE and RUSLE model also have limitations. he unit contributing area-based estimation method (UCA) converts slope length to unit contributing area for considering two-dimensional topography, however is not able to predict the different zones of soil erosion and deposition. The flowpath and cumulative cell length-based method (FCL) overcomes this disadvantage but does not consider channel networks, flow convergence and divergence flow in threedimensional topography. The purpose of this research was to overcome these limitations and extend the FCL method through inclusion of channel networks, convergence and divergence flow. We developed LS-TOOL in Microsoft’s .NET environment using C# with a user-friendly interface. Comparing the LS factor calculated with the three methodologies (UCA, FCL and LSTOOL). LS-TOOL delivers improved results. In particular, LSTOOL uses breaks in slope identified from the DEM to locate soil erosion and deposition zones, channel networks, convergence and divergence flow areas. Slope length and LS factor values generated using LS-TOOL correspond more closely with the reality of the Xiannangou catchment than results using UCA or FCL. The LSTOOL algorithm can automatically calculate slope length, slope steepness, L factor, S factor, and LS factors, providing the results as ASCII files which can be easily used in some GIS software. This study is an important step forward in conducting accurate largescale erosion evaluation.