Analysis of Landslides Triggered by the 2015 Gorkha Earthquake, Nepal

Abstract

On 25 April 2015, 7.8 Mw earthquake “Gorkha earthquake” struck central Nepal. Its epicenter was located 77 km northwest of Kathmandu near Barpak village in Gorkha District. The focus of the Gorkha earthquake was at a depth of approx. 15 km (considered shallow and therefore more damaging). On May 12, 2015, a major aftershock of 7.3 Mw followed the main shock. The Gorkha earthquake and its aftershocks caused thousands of human casualties and extensive damages, and triggered a large number of landslides of various types, including highly disrupted shallow slides, rock falls, and large-scale avalanches. An avalanche at Mount Everest killed at least 19 and another huge avalanche in the Langtang Valley killed about 350 people. The widespread landslides occurrence is posing a great threat to post-earthquake reconstruction. Field investigations and image interpretation identified thousands of landslides in affected districts of central Nepal. This provides a basis for better understanding and illustrating the distribution pattern of landslides triggered by the Gorkha earthquake and of the hazards related to them. The coseismic landslides were analyzed with respect to landslide causing and triggering factors, in order to characterize the spatial characteristics of landslides and to highlight the future hazard and risk. Slope angle and geology are the most significant parameters and potential ground acceleration values reflect an anomalous correlation with respect to landslide locations. The frequency ratio method was adopted for landslide susceptibility or hazard modelling. A landslide hazard map was generated and classified into five categories: very low, low, medium, high, and very high. 39.1% of the most affected areas belongs to very low and low classes with corresponding 5.9% of the inventoried landslides. Medium hazard zone makes up 37.1% of the area with 29.7% of the landslides. The rest of the area is classified into high and very high levels, which makes up 23.7% of the area with corresponding 64.3% of the total landslides. The verification of results indicates satisfactory agreement between the presumptive hazard map and the existing data on landslide locations. Consequently, there is a good correlation between areas defined as representing “high” hazard and the known landslides. The hazard map not only reveals the likelihood of future landslides and debris flows. It is also helpful for the relocation of displaced people and for reconstruction strategies.