Abstract

Information concerning erosion and deposition patterns from various soil redistribution processes is crucial for undertaking conservation of natural resources and for understanding landscape development. After an overview of erosion and deposition modelling on arable land, an interactive, menu-driven system (TBGIS) was developed within ARC/INFO GIS to generate DEM, extract terrain attributes, estimate terrain-controlled erosion potential, and simulate corresponding erosion and deposition patterns. Using a range of available options implemented in TBGIS, derived terrain attributes allowed a critical examination of the performance of the algorithms and consideration of the implications of the observed variability in results.
Topographic data and soil samples were collected from two study sites, Butsford Barton in Devon, UK, and Yesa in northern Spain. The topographic data were used as input to generate relative erosion and deposition patterns from tillage erosion and water erosion processes on the fields using TBGIS. The soil samples were analysed for caesium-137. The measured distributions of caesium-137 inventories were used as indicators of the net effects of soil redistribution processes occurring on the fields over last four decades.
Results show that the measured caesium-137 distributions are more similar to the predicted overall erosion and deposition patterns from tillage erosion on both sites. But, there is evidence that water erosion is active in some parts of the fields. Three important conclusions can be made. Firstly, the developed software (TBGIS) has made the simulation of soil redistribution processes on arable land more transparent, flexible, and realistic. It also allows important spatial patterns of DEM-derived terrain attributes to be interactively examined along with their statistical characteristics. Secondly, the study has demonstrated the sensitivity of terrain-based erosion and deposition modelling to the quality of DEM, the algorithms used for extracting terrain attributes, and the diffusion methods adopted for routing. Thirdly, the study has shown that the effects of tillage erosion on soil redistribution must be taken into account in any attempt to derive quantitative erosion and deposition data from caesium-137 detection on arable land.