9:30 AM - 9:45 AM
[HDS11-02] Comparison of Slope Deformations Identified in Time-Series InSAR Images and Airborne LiDAR Data: A Case Study in the Minobu Mountains Area, Japan
Keywords:Landslide, ALOS-2, Bitemporal DEM analysis
Airborne LiDAR data (LP data) and differential interferometric SAR (InSAR) using SAR satellites are effective methods for detecting slope deformation associated with mass movements (Yagi et al., 2024; Usami et al., 2024; Yamazaki, 2025). However, as most previous studies have focused on representative cases that are easily detectable by either technology, a systematic investigation is needed to clarify which types of slope deformation can be effectively captured by LP data and InSAR and which remain challenging to detect.
In this study, we analyze the characteristics of slope deformations detectable by each method by comparing time-series InSAR (TS-InSAR) images, published by the Geospatial Information Authority of Japan (Ishimoto et al., 2024), with multi-temporal LP data in the Minobu Mountains region, where multiple sets of LP data have been acquired (Fig. 1).
As of February 2025, we have compared the elevation changes derived from LP data acquired in 2018 and 2022 with the TS-InSAR images. For target slope A (northwest-facing), where sagging landforms, such as small scarps, have been identified, the TS-InSAR images from the descending right-looking orbit indicate displacement away from the satellite. However, the LP data do not reveal any significant elevation changes (Fig. 2). The displacement velocity estimated from the TS-InSAR images is 0.9 cm/yr in the direction away from the satellite. Considering that the elevation error of the digital elevation model (DEM) generated from LP data is approximately 40-60 cm (Sato et al., 2004), InSAR appears to be more suitable for detecting such subtle displacements.
In target slope B (east-facing), where a clear failure terrain is evident in aerial photographs, the LP data captured subsidence of the upper part of the failure terrain (Fig. 3). However, in the TS-InSAR images from the ascending right-looking orbit, the entire area appears uniformly colored, making it difficult to distinguish this deformation. The affected area covers approximately 27,000 square meters, whereas each InSAR pixel is approximately 8,100 square meters (pixel spacing 90 m) due to multilooking. This suggests that localized deformations may have been spatially smoothed out, indicating that LP data are more effective for capturing such localized changes.
During the presentation, we will also report additional analysis results incorporating longer-term LP data.
Acknowledgments
The LP data were provided by the Chubu Regional Development Bureau, Ministry of Land, Infrastructure, Transport and Tourism. The time-series InSAR images from the Geospatial Information Authority of Japan were obtained from the Satellite SAR Ground Deformation Monitoring Data Download Service.
References
Ishimoto et al. (2024) J. Geo. Inf. Aut. Japan, https://doi.org/10.57499/JOURNAL_137_03
Sato et al. (2004) J. Jap. Soc. Pho. Rem. Sensing, https://doi.org/10.4287/jsprs.43.4_13
Usami et al. (2024) Remote Sens., https://doi.org/10.3390/rs16152687
Yagi et al. (2024) J. Jap. Land. Soc., https://doi.org/10.3313/jls.61.77
Yamasaki (2025) J. Jap. Land. Soc., https://doi.org/10.3313/jls.62.20
In this study, we analyze the characteristics of slope deformations detectable by each method by comparing time-series InSAR (TS-InSAR) images, published by the Geospatial Information Authority of Japan (Ishimoto et al., 2024), with multi-temporal LP data in the Minobu Mountains region, where multiple sets of LP data have been acquired (Fig. 1).
As of February 2025, we have compared the elevation changes derived from LP data acquired in 2018 and 2022 with the TS-InSAR images. For target slope A (northwest-facing), where sagging landforms, such as small scarps, have been identified, the TS-InSAR images from the descending right-looking orbit indicate displacement away from the satellite. However, the LP data do not reveal any significant elevation changes (Fig. 2). The displacement velocity estimated from the TS-InSAR images is 0.9 cm/yr in the direction away from the satellite. Considering that the elevation error of the digital elevation model (DEM) generated from LP data is approximately 40-60 cm (Sato et al., 2004), InSAR appears to be more suitable for detecting such subtle displacements.
In target slope B (east-facing), where a clear failure terrain is evident in aerial photographs, the LP data captured subsidence of the upper part of the failure terrain (Fig. 3). However, in the TS-InSAR images from the ascending right-looking orbit, the entire area appears uniformly colored, making it difficult to distinguish this deformation. The affected area covers approximately 27,000 square meters, whereas each InSAR pixel is approximately 8,100 square meters (pixel spacing 90 m) due to multilooking. This suggests that localized deformations may have been spatially smoothed out, indicating that LP data are more effective for capturing such localized changes.
During the presentation, we will also report additional analysis results incorporating longer-term LP data.
Acknowledgments
The LP data were provided by the Chubu Regional Development Bureau, Ministry of Land, Infrastructure, Transport and Tourism. The time-series InSAR images from the Geospatial Information Authority of Japan were obtained from the Satellite SAR Ground Deformation Monitoring Data Download Service.
References
Ishimoto et al. (2024) J. Geo. Inf. Aut. Japan, https://doi.org/10.57499/JOURNAL_137_03
Sato et al. (2004) J. Jap. Soc. Pho. Rem. Sensing, https://doi.org/10.4287/jsprs.43.4_13
Usami et al. (2024) Remote Sens., https://doi.org/10.3390/rs16152687
Yagi et al. (2024) J. Jap. Land. Soc., https://doi.org/10.3313/jls.61.77
Yamasaki (2025) J. Jap. Land. Soc., https://doi.org/10.3313/jls.62.20