曾國欣
國立中央大學太空及遙測研究中心
khtseng@csrsr.ncu.edu.tw
A landslide dam formed in the upstream Matai’an Creek, Hualien, Taiwan during Typhoon Wipha in July 2025. For emergency managers, the instability of this natural barrier and the consequences of sudden outbursts posed an immediate risk to downstream areas. Continuous satellite monitoring was therefore mobilized to quantify the spatio-temporal evolution of the impounded water and unstable sediment.
This talk demonstrates a multi-sensor workflow in an operational emergency response context. Very-high resolution commercial SAR (TanDEM-X and Capella Space) was used to rapidly delineate lake water extent and estimate water volume change under persistent cloud and heavy rain conditions. These radar observations were combined with very-high resolution optical data (Pleiades) to construct stereo-derived Digital Surface Models (DSMs). By differencing these DSMs with pre-event optical and airborne LiDAR elevation data, we quantified the post-breach geomorphic change and sediment thickness deposited along the channel.
The Matai’an Creek results, together with other marine pollution response cases that will be presented in this talk, clearly demonstrate that spaceborne Earth Observation, particularly a fleet of SAR constellations, enables rapid, repeatable, and metric-level observations in urgent situations when field teams cannot safely access the site. The combination of on-demand commercial SAR tasking with open-data missions now provides a robust operational model for near real-time surveillance of high-risk environmental hazards.
關鍵詞：SAR Remote Sensing, Geospatial Intelligence, Situational Awareness
