[M-AG32] Satellite Land Physical Processes Monitoring at Medium/High/Very High Resolution

[MAG32-P01] Monitoring Unstable Soil in Watersheds Using High-Resolution Satellite Stereo Imagery: A Case Study in Butangbunasi River, Taiwan

*Tee-Ann Teo¹、Ko-Hsin Mei¹ (1.NYCU)

キーワード：Unstable soil, watershed monitoring, high-resolution triple images, digital surface model

Unstable soil represents a significant accumulation of sediment in a transient or unstable equilibrium state on slopes or riverbeds within a watershed. The management of unstable soil is crucial. Therefore, the development of a monitoring system utilizing satellite remote sensing technology combined with historical spatial data is an important aspect of unstable soil management. Remote sensing techniques, such as LiDAR point clouds or stereoscopic optical imagery, can be employed for sediment volume analysis by generating digital elevation models. Due to the large coverage capability of satellite stereo imagery, it can be effectively applied to large-scale sediment volume analysis. This study utilizes stereo satellite imagery from different periods to monitor unstable soil in watershed areas. The main tasks include (1) generating high-resolution digital surface models (DSM) through image-dense matching and (2) calculating sediment change by subtracting pre- and post-event digital surface models. The study area is the Butangbunasi Watershed in Kaohsiung City, Taiwan. Two Pleiades tri-stereo datasets were used to generate digital surface models for January and November 2023. The difference of DSM was obtained by subtracting these two digital surface models, and the influence of vegetation height was eliminated using the normalized difference vegetation index to derive unstable sediment volume change. The total volume of unstable soil in the entire watershed is the sum of sediment accumulation, with 1.1178 million cubic meters deposited upstream and 0.4451 million cubic meters in the middle and downstream riverbeds, resulting in a total unstable soil volume of +1.5629 million cubic meters from January to November. In summary, this study demonstrates that high-resolution stereo satellite images are capable of monitoring large-area unstable soil. The findings provide valuable insights for effective watershed management and contribute to the development of strategies for mitigating the impact of unstable soil on the watershed environment.