A framework for improved near real-time flood mapping

ICEYE has been a leader in the mapping and monitoring of global floods for the insurance sector and governments over the last two years. Current operational flood monitoring is based on the large-scale and systematic availability of synthetic aperture radar (SAR) data from the small satellite constellation deployed and operated by ICEYE. The main advantages of SAR images are that they provide synoptic views over wide areas, day and night and in all-weather conditions. However, SAR can be less suitable for providing flood extent information in dense urban areas and under tree canopy cover. In addition, SAR-based flood depth generation methods struggle to provide accurate depth estimations in steep terrain. There is currently a demand to aid observational flood models with physically-based flood modeling in urban areas.

Most operational real-time flood estimates are based on predictions of discharges at river flow monitoring stations using 1D hydrological models. 2D inundation models are computationally expensive and thus require special tooling for creating rapid flood maps. In this presentation, ICEYE will describe a framework that can be used for improving the robustness and accuracy of near real-time flood predictions.