Probabilistic Techniques, GIS and Remote Sensing in Landslide Hazard Mitigation: A Case Study from Sikkim Himalayas, India

Landslides form one of the most devastating natural hazards in the Sikkim State of India. This natural hazard alone causes severe damage to properties and human lives. The identified conditioning factors include adverse rock types (mica schist, phyllite, granite gneiss and calc schist), multiple joint sets, active tectonism, and very high annual precipitation (3539 mm). The triggering factors are mainly rapid pore pressure built up, seismic activity and anthropogenic interference. This paper evaluates the landslide hazard zones using information theory (I) and regression analysis (R) in GIS environment. In all, 14 variables are identified as conditioning and triggering factors and accordingly probabilistic prediction maps are prepared individually by both the methods. The results thus generated are compared and classified into three slope instability zones viz. low (−0.143 < Ij < 0.02 & 0.38 < Rj < 0.55), medium (−0.02 < Ij < 0.103 & 0.55 < Rj < 0.73) and high ( Ij > 0.13 & Rj > 0.73) on the basis of histogram distribution. Further, these probabilistic prediction maps are compared with the actual landslide map generated from recent satellite data (January 2002) for the accuracy of prediction. The generated hazard maps agree with the observed landslide incidences. Thus, the proposed methodology can be utilized effectively in landslide hazard zonation studies.