Evaluation of quantitative Landslide Susceptibility Zonation (LSZ) method for Nilgiri District

Landslide susceptibility zonation (LSZ) in Nilgiri district is the catastrophic natural disasters occurring annually. Inventory map from past occurrence were created and used for LSZ quantitative methods Spatial Multi criteria Evaluation (SMCE) and Artificial Neural Network (ANN). Predictors causing landslides, identified namely rainfall,soil, lineament, geomorphology, geology, drainage, road, railway, land use/land cover, slope and aspect were used as parameters grouped into four factors namely hydrological, anthropogenic, geology and geomorphological in the form of thematic maps. Two statistical approaches namely SMCE and ANN were used to calculate the weights and ratings. GIS was used to prepare landslide hazard zonation map after overlaying the several layers with weights. The hazard zonation map was classified five zones namely very low, low, moderate, high and very high. Landslide hazard maps were validated, using field check and Receiver Operating curve . The LSZ maps of both the methods were compared with the real land with respect to blocks, it showed that ANN method best and also 92% accuracy and forfeited SMCE method.