Assessment of shallow landslide susceptibility using artificial neural networks in Jabonosa River Basin, Venezuela

Landslides represent one of the most morphodynamic processes that affect the steep lands, and may destroy croplands as well as urban and industrial development. Landslide risk analysis can help government agencies to select suitable locations for development schemes and plan mitigation measures in unstable landslide-prone areas. This study describes an approach for assessing the landslide risk potential, mainly for shallow landslides, with reference to Jabonosa river basin in the Venezuelan Andes using artificial neural networks (ANNs), specifically a Multilayer Perceptron with backpropagation learning algorithm. The approach developed uses a wide range of parameters of slope instability derived from digital elevation models (DEMs), remote sensing imagery and documentary data. Around 90% overall accuracy produced by the ANN technique were found promising for future studies in landslide susceptibility zonation.