Flooding mapping detection and urban affectation using Google Earth Engin

Abstract

Floods are a phenomenon that can be triggered by river overflow or heavy rainfall. In this context, detecting flooded areas is crucial to document affected zones in urban environments over time. This study focuses on the development of a model based on automatic extraction of flood map images using the Synthetic Aperture Radar (SAR) of Sentinel-1 from the online Google Earth Engine (GEE) platform, specifically for the metropolitan city of Iquitos in Peru. The methodology involved mapping the flooding extent occurred over a seven-year period (2015-2021) to create a probability map of occurrences. Subsequently, identified flood areas were validated using river levels from a two-stage gauge, revealing a positive correlation. The probability map of occurrences was then superimposed on a basemap, identifying the affectation of 14.7 km of roads, 130 schools, and 91 hospitals. These findings can provide significant information for decision-making related to disaster prevention and management.