Sensitivity Analysis on GIS Procedures for the Estimate of Soil Erosion Risk
Author: Mendicino, G.
Source: Natural Hazards, Volume 20, Numbers 2-3, November 1999 , pp. 231-253(23)
Abstract:Land degradation by water erosion is one of the major issues in the field of environmental planning. Soil erosion and sediment transport are spatially distributed processes, and their evaluation can be easily realised by means of the use of Geographic Information Systems. The greater availability of digital and geo-lithological data managed and stored inside GIS has implied the development of techniques and procedures aimed at the definition of the spatial prediction of erosion and deposition rates across a catchment. In this paper using terrain data measured on an experimental basin in southern Italy, a sensitivity analysis on different GIS-based methodologies for the estimate of a Length-Slope factor has been developed with the aim of determining which of these is more reliable for spatial erosion risk assessment. Specifically, by using the unit stream power theory, different estimates are shown, depending on the scheme adopted to represent the hydrological and topographic three-dimensional effects inside the Length-Slope factor. The performances of the procedures analysed have been evaluated through the information content of the corresponding spatial distributions, estimated as an entropy measure. The results obtained have shown that among the approaches utilised to describe the routing of the surface runoff along the hillslope profiles, the two-dimensional scheme appeared to be more realistic both on diverging and converging surfaces. Such a scheme, during the computational phases also aimed at distinguishing the areas of the basin experiencing net erosion from those areas experiencing net deposition, resulted, being in accordance with on-site investigations.
Document Type: Regular Paper
Affiliations: Dipartimento di Difesa del Suolo, Universitá della Calabria, C.da S.Antonello, 87040 Montalto Uffugo (CS), Italy
Publication date: November 1, 1999