講演情報
[PS-01-10]Improving remote Urban Flood Mapping leveraging Global Building Data and Dual-Polarisation SAR
*Simeon Wagener1,2, Tomohito Yamada1, Xiaoxiang Zhu2 (1. Graduate School of Engineering, Hokkaido University, 2. Data Science in Earth Observation, Technical University of Munich)
キーワード：
SAR、urban flood mapping、building orientation、dual-polarisation、sensor transferability
Urban flood mapping in built-up environments remains a critical challenge for disaster response and risk assessment. Existing SAR-based methods systematically exclude urban areas due to ambiguous backscatter signatures caused by complex building morphology, leaving precisely the most exposed populations unmapped. This study addresses this gap by proposing a building orientation-informed, dual-polarisation framework that adapts polarisation selection to local urban geometry, unlocking flood signals in areas where conventional approaches fail. Leveraging freely available global building datasets and multi-mission SAR data, the methodology is designed for transferability across sensors, geographic regions, and data-scarce environments. The expected outcome is a scalable, physics-grounded tool for building-scale urban flood detection, directly supporting post-disaster damage assessment, emergency response, and parametric insurance applications.