Impact Assessment of Community Vulnerability in Cyclone Bulbul and Analyzing Change Detection Impacts by Geospatial Techniques

Keywords: Vulnerability Ranking, Remote Sensing, Hazard LULC, Impact Analysis, Change Detection
Abstract Type: Poster Abstract

Abstract

Cyclones have been one of the most destructive natural catastrophes for the last four decades, with the damage varying according to a wide range of conditions. This study developed and evaluated an approach for assessing cyclone impacts on a coastal community of Barguna. The study used a semi-structured questionnaire and non-probability sampling techniques. Five PRA sessions were held with 100 participants, each group consisting of 20 people. Via mental mapping method, people's perceptions of natural disasters were assessed. A weighted scale of 20 semi-structured questions has been divided into 4 categories, where n=20; 1-10: issue number (i/n). Before and after phase cloud-free satellite images were used to assess the results and compare the vulnerability of communities. After accuracy assessment of image classification, homology detection via SVM algorithm was followed for impact assessment. Findings indicate that 35% people have direct knowledge of disasters, while 85% have heard about cyclones. Among the 34 vulnerability issues listed by the respondents, 19 were marked as very high. Agriculture related vulnerability issues were severe in people's vulnerability index. Geospatial analysis indicated that dense vegetation was mostly affected (82.75%). Dense vegetation damage accounted for 8.2% of overall study area's damage (20%). On top of that, agriculture sustained major damage (10.67% of the overall 20% damage), correlating with the results of the people's vulnerability index. Despite several government and community-based initiatives for disaster risk management, coastal communities' socio-economic vulnerability has not changed yet. Thus, locally led adaptation can be a good adaptation strategy.