Mapping erosion from space

Soil erosion by water is the most important land degradation problem worldwide. Spatial information on erosion is required for defining effective soil and water conservation strategies. Satellite remote sensing can provide relevant input to regional erosion assessment. This thesis comprises a review on how satellite data have been used previously for evaluating water erosion. Options include erosion detection and the assessment of controlling factors such as topography, soil, and vegetation. Due to the complexity of erosion processes, regional differences, and scale dependency, remote sensing techniques and data integration methods should be tailored to regional characteristics. However, often erosion models are applied without questioning their assumptions, which is aggravated by a general lack of validation. On the basis of several case studies in different tropical regions, this thesis presents new methodologies for qualitative erosion mapping using satellite data. Automatic identification of large erosion gullies from optical imagery was performed for an area in the Brazilian Cerrados. For the same area, an approach using multi-resolution and multi-temporal satellite data accurately identified erosion risk patterns related to sheet and rill erosion. For the Colombian Eastern Plains expert knowledge assisted erosion risk mapping by selecting and combining relevant parameters to be extracted from available data sources. For the West Usambara Mountains of Tanzania, accurate evaluation of erosion risk was possible using only remotely-sensed information on vegetation and slope. It was concluded that flexible, qualitative, region-dependent approaches that effectively use data from spaceborne sensors, can provide good means for quick and accurate regional mapping of erosion risk.