Author(s): Maulana Ibrahim Rau; Natsu Miura; Daisuke Nohara; Atriyon Julzarika; Tsuyoshi Yamaguchi; Yoriyuki Yamada; Natsuki Yoshikawa
Linked Author(s): Daisuke Nohara
Keywords: DTM satellite imageries flood analysis steep-slope regions LiDAR
Abstract: High-resolution Digital Terrain Models (DTMs) are crucial for flood modeling, particularly in accurately representing small channels in steep-slope terrains. This study focuses on assessing the capability of remote-sensing-based DTMs (RS-DTMs) to reproduce channel bathymetry and evaluate their applicability in flood analysis within steep topographic regions. High-resolution satellite imagery from WorldView-2 (50 cm) was used to construct DTMs that were compared to LiDAR-based survey terrain data and the Geospatial Information Authority of Japan (GSI)’s LiDAR DTM. A numerical simulation-based flood model was employed to evaluate the hydrographs from the channels, with accuracy assessed using Nash-Sutcliffe Efficiency (NSE). In addition, the coefficient of determination (R2) and root mean square error (RMSE) were used to assess the comparison between RS-DTM and UAV-LiDAR-based data for ground elevation. One-at-a-Time (OAT) sensitivity analysis and tornado diagram were also analysed to evaluate the impact of using RS-DTM on hydrologic characteristics in flood modeling. Results indicated that all indices -- R2, RMSE from channel bathymetric analyses, and NSE from flood model evaluations -- showed that the DTM derived from WorldView-2 produced results that aligned well with site survey UAV-based LiDAR measurements. While the DTM exhibited sensitivity to hydrographic characteristics, the findings demonstrated that RS-DTMs were a viable and cost-efficient alternative for flood modeling in steep-slope areas, providing acceptable accuracy for representing channel bathymetry and supporting practical flood management applications.
Year: 2025