GIS modelling of land degradation in Northern-Jordan using landsat imagery

An empirical model based on high resolution spatial and temporal remotely sensed data offers the ability to assess the degradation impacts of changes in land cover in a spatial context. In an attempt to assess the impacts of changing land cover on soil, a GIS-based erosion model has been developed to predict annual soil loss by water in northern Jordan. This model uses the Revised Universal Soil Loss Equation (RUSLE). Spatially distributed static (topographic and soil) parameters for this model are extracted from a regional GIS. The dynamic (vegetation cover) parameter is estimated from the land cover maps, derived by digital processing of multi-resolution, multi temporal Landsat MSS (14. 9. 1972), and TM (28. 8. 1992). Mapping of vegetation cover was carried out by applying TM-Linear Mixture Modeling and NDVI, while mapping of fallow lands was carried out by both on-screen digitizing and sketch mapping in the field. The image difference technique was used in the change detection analysis. The erosion model predicts an increase in the amount of soil loss in the study area from 1972 to 1992, as a result of land cover changes. It was concluded that the degradation of the soil in the study area, observed during the last two decades, was caused by effects of these land cover changes. The study is not intended to present detailed or semi-detailed results, with the use of Landsat 30-meters spatial resolution images covering an area of 600 km², the model defines endangered areas, further detailed studies needs to be conducted for specific areas.