Flood Extent Mapping From The Sentinel-1 And Sentinel-2 Data In Sri Lanka

Sri Lanka is an important node in the "The Belt and Road" initiative of our country.The country is located in the tropics,located in the northern part of the India ocean.It has abundant precipitation,and the river network is dense.The topography of the central part is high,and the terrain is flat around.Due to its unique geographical location and geological characteristics,the southwest region of the country and the southwest of the central mountainous area are affected by the wind surface,the climate is moist,and the heavy rainfall often occurs in the rainy season,which often causes serious flood disasters.In order to better cope with the possible future flood disaster in Sri Lanka,it is necessary to strengthen the prediction of strong rainfall and flood disaster prevention in Sri Lanka.Because of its macroscopic,comprehensive,rapid and dynamic characteristics,remote sensing technology is an important means to extract the flooded area of flood and waterlogging,and has been widely used in flood and flood disaster.With the launch and use of Sentinel-1/2 satellites,higher resolution remote sensing data can be free to use,which brings convenience to the extraction of flood disasters.This paper was collected the Sentinel-1/2 data before and during the 2017,studied the spectral characteristics of the water body in these data,and analyzed the advantages and disadvantages of the water extraction method based on the Sentinel1/2data in the specific environment of the Sri Lanka flood disaster.Besides,it has took the flood disaster situation in 2017 of the southern part of Matara as an example which is one of the administrations of Sri Lanka,the Sentinel1/2 data is used to extract and analyze the flooded area,and to improve the actual problems encountered,as well as to propose a flood submerge extraction process suitable for Sri Lanka.The main contributions and innovations of this article are:(1)In this paper,the Sentinel-1/2 data of Sri Lanka dry season were collected and water extraction was carried out.The applicability and advantages anddisadvantages of the optical data extraction methods under different conditions such as cloudless,cloudless,and urban environment were analyzed.The NDWI index method for Sentinel-2 water extraction and the empirical threshold method for Sentinel-1 water extraction flood are the most suitable water extraction methods for flood disasters.(2)In this paper,we use the FMask algorithm to detect the cloud of Sentinel-2image and remove the cloud area,which reduces the water extraction error caused by the cloud coverage area.(3)This paper collected Sentinel-1 and Sentinel-2 image data from the south of the Matala administrative region during the catastrophic flooding in 2017.The flooded inundation area of the area was extracted using the appropriate method analyzed in this paper,and the extraction of flooded areas in Sri Lanka was proposed.Process.The NDWI method can accurately extract the submerged area.The radar image threshold method is not as good as optical data,but it is still supplemented by good data in the cloud.