Mapeamento de inundações urbanas com uso de radar de abertura sintética para o município de Marabá-PA

Abstract

The Amazon region is characterized by its vast hydrographic network and high hydrological variability, increasing its susceptibility to extreme events such as floods. In the municipality of Marabá, Pará, located at the confluence of the Tocantins and Itacaiúnas rivers, unregulated occupation of floodplain areas and inadequate urban infrastructure exacerbate the recurrence of flooding, directly impacting the population as well as economic and social sectors. Given this scenario, the main objective of this research was to map urban floods in Marabá using Synthetic Aperture Radar (SAR) data from the Sentinel-1 satellite. The methodology was based on a multitemporal analysis of SAR images from 2018 to 2023, pre-processed on the Google Earth Engine platform. Geometric and radiometric corrections were applied, along with adaptive thresholding techniques for flood extent segmentation. The results were validated using hydrological data from the National Water Agency (ANA) and meteorological data from the National Institute of Meteorology (INMET). The findings indicate that 2020 and 2022 recorded the largest total flooded areas, reaching 12.82 km² and 12.52 km², respectively. The most affected regions are concentrated in low-altitude areas with deficient infrastructure, highlighting the exposure of urban centers to flood risks. Additionally, the results align with SDG 6 (Clean Water and Sanitation), SDG 11 (Sustainable Cities and Communities) by providing insights to enhance urban resilience, and SDG 13 (Climate Action) by addressing the intensification of extreme weather events in the Amazon region. This study highlights the need to disseminate knowledge on radar-based flood mapping so that these techniques can effectively contribute to territorial management and disaster response in the Amazon.