Sentinel data for flood disaster monitoring and assessment: case studies in Romania

Abstract

Floods represent one of the most significant natural disaster in Romania. The paper describes the main stages of image processing, for flood disaster monitoring and assessment in radar domain: the processing chain involves the stages starting with image calibration, speckle filtering, segmentation/histogram analysis and terrain corrections. The results of satellite data processing were integrated in a GIS environment to obtain standardized, cartographic products. The satellite-based flood monitoring includes: mapping the maximum flood extent, flooded area classification, flood evolution maps etc., while the damage assessment covers: flooded areas, affected population, categories of land use and communication infrastructure. The paper presents recent flood case studies of March 2018, in the upper part of the Olt basin, located in the central part of Romania, based on Sentinel-1 (S1) satellite imageries and ancillary in-situ data (hydrological and meteorological measurements). The S1 satellite provide high temporal frequency (daily) and good spatial resolution (10-11 m), which have been used as a primary data source for flood disaster monitoring (S1) and damage assessment (Corine Land Cover, 2018). In order to map the flood extent, the Synthetic Aperture Radar (SAR) sensor of S1 is used due to all-weather monitoring capability and its large coverage.