Mapeamento das áreas agrícolas inundadas na cheia do ano 2023 usando dados Sentinel-1 e Sentinel-2 no Distrito de Boane

Abstract

Natural disasters are phenomena that have been occurring frequently throughout the world. Mozambique in particular has been the target of floods, high waters and cyclones, with a tendency to worsen in recent years due to climate change, geographical position and uncontrolled urbanization. Floods have been a major cause of losses in the country's agricultural and other sectors, and knowledge of their true extent can help in managing their effects at a national level. Mapping techniques such as Geographic Information System (GIS), Google Earth Engine, among others, make it possible to visualize the extent of floods. The objective of this work was to remotely map the agricultural areas flooded in 2023 in the Boane District using the Google Earth Engine platform to generate images of the flooded area through the Sentinel-1 constellation that used the method of differentiating images of pre and post flood dates to extract the flooded area in the district and then validate the accuracy presenting a total accuracy of 81.52% and a Kappa index of 0.7588 compared to the Copernicus EMS map that is classified as a good result according to the Landis and Kock table in the Kappa test, also through the Environmental System Resourch Institute (ESRI) that using Sentinel-2 provides images of use and coverage of the region for the reclassification and extraction of the agricultural area of the Boane district using the QGIS application. During the February 2023 floods in Boane District, a total area of 620 hectares was flooded, corresponding to about 8.4% of the total cultivated area in this District. This study demonstrated that the use of satellite imagery and GIS are useful to retrieve information on the impact of past floods and thus be able to predict potential impacts in real time using similar methodologies.