Linking hydrological, infinite slope stability and land-use change models through GIS for assessing the impact of deforestation on slope stability in high Andean watersheds

In the Ecuadorian Andes, episodic slope movements comprising shallow rotational and translational slides and rapid flows of debris and soil material are common. Consequently, not only considerable financial costs are experienced, but also major ecological and environmental problems arise in a larger geographical area. Sediment production by slope movement on hillslopes directly affects sediment transport and deposition in downstream rivers and dams and morphological changes in the stream channels. In developing countries world-wide, slope movement hazards are growing: increasing population pressure and economic development force more people to move to potentially hazardous areas, which are less suitable for agriculture and rangelands. This paper describes the methods used to determine the controlling factors of slope failure and to build upon the results of the statistical analysis a process-based slope stability model, which includes a dynamic soil wetness index using a simple subsurface flow model. The model provides a time-varying estimate of slope movement susceptibility, by linking land-use data with spatially varying hydrologic (soil conductivity, evapotranspiration, soil wetness) and soil strength properties. The slope stability model was applied to a high Andean watershed (Gordeleg Catchment, 250 ha, southern Ecuadorian Andes) and was validated by calculating the association coefficients between the slope movement susceptibility map of 2000 and the spatial pattern of active slope movements, as measured in the field with GPS. The proposed methodology allows assessment of the effects of past and future land-use change on slope stability. A realistic deforestation scenario was presented: past land-use change includes a gradual fragmentation and clear cut of the secondary forests, as observed over the last four decades (1963-2000), future land-use change is simulated based on a binary logistic deforestation model, whereby it was assumed that future land-use change would continue at the same rate and style as over the last 37 years (1963-2000). (C) 2002 Elsevier Science B.V. All rights reserved.