FLASH FLOOD MAPPING BASED ON DATA FROM LANDSAT-8 SATELLITE AND WATER INDICES

Abstract

Flood is a kind of natural disasters that frequently occurred in several areas of Thailand. In September 2020, there was constant raining in Nan Province caused by strong monsoon trough. Such occurrence caused a flood affecting to living of local people and severely damaged agricultural areas. As a result, this research aims to create flash flood map based on data from Landsat-8 Satellite and water indices pre-flood period, flood period, and post-flood period whereas flood occurred in August 2020 of Wiangsa District in Nan Province as case study. The research was conducted by analyzing on data obtained from Landsat-8 Satellite and 2 formats of water indices including: 1) $NDWI$; and 2) $MNDWI$. Subsequently, the results of obtained from those 2 formats were used in evaluating statistical correctness. The results revealed that analysis on data obtained from Landsat-8 Satellite and $NDWI$ was the most appropriate methods and it was appeared that overall accuracy was 92.67% whereas Kappa coefficient was 0.96. The results of this research represented that the methodology of this research could create flood map and specify flood areas efficiently.