Generating Substantially Complete Landslide Inventory using Multiple Data Sources: A Case Study in Northwest Himalayas, India

Abstract

Landslide inventory contains basic information about landslides such as location, classification, morphometry, volume, run-out distance, activity, date of occurrence, damages caused etc. In most landslide inventory maps only spatio-temporal distribution of slope failures is shown. A complete portrayal of landslide inventory, both in number of landslides mapped and associated attributes such as morphometry, classification etc. is a must as it has bearing on the estimation of landslide susceptibility, hazard and risk. Despite of its enormous importance, landslide inventory is rarely found to be complete. This can be attributed to limited availability of requisite data and the inherent uncertainties in landslide inventory mapping from different sources and importantly the insufficient field inputs. In the present work in the Mandakini valley of Uttarakhand Himalayas attempts have been made to prepare a substantially complete landslide inventory, both in terms of number of landslides and attributes, using multifarious data sources substantiated with thorough field checks. Supplementary historical data sources along with the conventional data sources have been used to prepare the inventory. The approach followed includes pre-field mapping of landslides from different Earth Observation (EO) data followed by interpretation of archival records and detailed field surveys. The data available for the period 1962 to 2013 were used to build up the inventory database in GIS. A total of 644 landslides have been mapped and attributed using conventional EO data and also the different supplementary data sources for an area covering 400 sq. km. An inventory of 151 landslides prepared from EO data have been confirmed and updated during the field study. Finally, the detailed landslide inventory prepared during the course of this study was analysed to understand the behaviour of landslides and role of triggering factors in susceptibility mapping and the geo-environmental conditions governing the magnitude of the event.