DefinePK

Abstract

Using different resolution data, a number of scientists have developed landslide inventories after Kashmir earthquake (7.6 Mw), which triggered thousands of landslides in Muzaffarabad and Balakot. However, the impact of image spatial resolution on landslide inventory is less explored. This study aims to evaluate the impact of image spatial resolution on landslide inventories. Five different resolution images including, Worldview-2, SPOT-5, Sentinel-2, ASTER and Landsat-8 were used to compute landslide inventory for the 2005 Kashmir earthquake affected area. Support Vector Machine (SVM) classifier was used to classify the acquired images and derived landslide inventory. Number of landslides and covered area from different sensors were computed. Accuracy assessment of the developed inventories was performed for each of developed landslide inventory. The result shows that with increase in image spatial resolution the number and area of landslides increases. The recently lunched freely available data of Sentinel-2 with spatial resolution of 10 m is good source of dataset for regional based landslide inventory developing and landslide susceptibility mapping. The study shall assist the researchers to select an optimal remote sensing images for developing landslide inventory; and subsequent hazard assessment and mitigation.