Authors: Yasin Wahid Rabby*, University of Tennessee, Yingkui Li, Department of Geography, University of Tennessee, Knoxville
Topics: Hazards, Risks, and Disasters, Geographic Information Science and Systems, Geomorphology
Keywords: Landslide, Google Earth, Participatory Mapping, Landslide Detection, Landsat8
Session Type: Paper
Start / End Time: 8:00 AM / 9:40 AM
Room: Napoleon B3, Sheraton 3rd Floor
Presentation File: No File Uploaded
Landslide is one of the most common hazards that cause loss of human lives and infrastructural damages around the world. Landslide inventory mapping is of critical importance to mitigate these damages. This study aims to test the feasibility of using various techniques to map landslide inventory for Chittagong Hilly Areas in Bangladesh. We first identified 161 landslides based on the interpretation of Google Earth imagery. Then, participatory field mapping was conducted in selected areas where landslides were reported from newspapers and governmental reports. We mapped 654 landslides in the field using this method. The field-identified landslides were used to validate the landslides identified using multi-temporal satellite imagery. We used a supervised classification to classify landslides in the Rangmati Municipality based on multi-temporal (Pre and Post Event: June 10-15, 2017 Landslides) Landsat8 imagery. We also used ASTER digital elevation model (DEM) to refine the classified landslide sites. The landslides mapped using these three methods in the Rangamati Municipality are compared to check if the relatively coarse resolution multi-temporal Landsat imagery is suitable for the landslide inventory mapping of the whole Chittagong Hilly Areas.