International Journal of Research and Scientific Innovation (IJRSI) |Volume IX, Issue XII, December 2022|ISSN 2321-2705
Kumarayapa Y A A, Bandara A S
Department of Electronics, Wayamba University of Sri Lanka, Kulitapitiya, Sri Lanka
Abstract: Although, there are high population densities living around some of the major river basins due to the fertility of the areas, even in the case of minor floods, these settlements are constantly at risk due to flow of this excess water to the Dry Residential Areas. When such over-flooding occurs in the river basins, it is important to have real-time remote mapping, causality determination, and analyzing method to prevent disasters and to deploy emergency response teams to rescue human lives. Nevertheless, the use of aerial images for geographical mapping of such critical areas is difficult and unclear due to the cloudy atmosphere that exists with heavy rainfall and other environmental disturbances. To circumvent such obstacles, we propose a novel method that employs Sentinel 1 SAR data, as well as a speckle filter, to further refine the critical flood events over a free-defined period. Compared to the existing optical sensing methods (MODIS, Landsat etc.), the proposed method gives more accuracy in data analysis and prediction. In land cover classification test, it achieves 94.41% accuracy. The method combined with user-friendly GEE-SAR based platform, foremost over normal aerial photograph analysis as it avoids any environmental disturbances. Thus, the general public here without computer literacy can be informed timely about the threats via phones, etc. Hence, this research study with our novel risk identification method and formulated applet will protect those lives near the rivers. The proposed methodology based approach can be used for flood daunted river basins exist anywhere in the world
Keywords: Flood mapping, Remote Sensing, Synthetic Aperture Radar (SAR), speckle filter, image feature extraction
I. INTRODUCTION
Floods are natural disaster that occurs regularly in many parts of the world. Generally, a flood occurs when the water level of a river exceeds its maximum water level, and this excess water flowing into dry areas is called a flood [1]. Floods are the most unfavorable structure of natural risks in each local and global context. This is actually in phrases of every loss of existence and property damage. In Sri Lanka too, floods are more common than any other natural hazard [2]. When considering the early major flood hazard in Sri Lanka, 14-16 May 2021 Heavy rains and winds resulted in flooding in several districts with Colombo, Gampaha, and Galle being the worst affected with a high range of damages and displacements. A whole of 43,493 humans was affected in all districts [3]. In the month of May 2021, this flash flood occurred due to Cyclone Yaas which originated in the Bay of Bengal. According to Sri