講演情報
[U16-P04]Accuracy Assessment of ICESat-2 ATL03 Photon Data over Flat Terrain in the Noto Peninsula, Japan: A Preliminary Study toward Future Satellite-Based Topographic Monitoring
*金田 真一1 (1.アジア航測株式会社)
キーワード：
衛星高度計ライダー、地形データ、精度検証、ICESat-2
Background and Motivation:
Continuous topographic monitoring using space-borne photon-counting LiDAR is increasingly essential for tracking global geomorphological changes, such as sedimentation, erosion, and crustal deformation. As part of ongoing R&D efforts to develop next-generation satellite laser altimetry systems, rigorous evaluation of current sensor performance is critical for effective disaster impact assessment and long-term environmental management. This preliminary study assesses the elevation accuracy of NASA’s ICESat-2 ATL03 (global geolocated photon) data over flat terrain, with a focus on its potential as a reliable baseline for future monitoring frameworks.
Methods:
The study area is the Noto Peninsula, Japan, which offers diverse but locally flat and open terrain suitable for validation. We analyzed ATL03 datasets from 2019 to 2024. To minimize uncertainties from vegetation and built structures, flat and open areas were selected by visually cross-referencing ICESat-2 ground tracks with high-resolution aerial imagery. Photon-derived ellipsoidal heights were converted to orthometric heights using the Japanese Geoid Model 2011 (GSIGEO2011) for direct comparison with the 5-m resolution Digital Elevation Model (DEM) provided by the Geospatial Information Authority of Japan (GSI).
Analysis of Noise Characteristics:
Photon-counting LiDAR performance is highly sensitive to the signal-to-noise ratio (SNR). Visual inspection of ATL03 along-track profiles revealed substantial differences between daytime and nighttime acquisitions. Daytime data showed elevated solar background noise, which hindered reliable ground photon identification. In contrast, nighttime observations exhibited high SNR with minimal background noise. Therefore, this accuracy assessment was restricted to nighttime data to establish a high-quality performance baseline.
Results and Discussion:
Preliminary comparisons indicate that nighttime ATL03 elevations achieve an accuracy of approximately 1 m relative to the GSI 5-m DEM in selected flat regions. This discrepancy is likely due to the ~17-m diameter laser footprint averaging surface heights over a larger area than the DEM grid, along with residual effects from photon classification. Although the current data density was insufficient for robust change detection in the Noto region, these results highlight the sensor’s fundamental precision under optimal conditions.
Conclusion:
This analysis confirms that nighttime ICESat-2 ATL03 data can provide a high-precision topographic reference, supporting its role in future missions for monitoring dynamic geomorphological processes.
Continuous topographic monitoring using space-borne photon-counting LiDAR is increasingly essential for tracking global geomorphological changes, such as sedimentation, erosion, and crustal deformation. As part of ongoing R&D efforts to develop next-generation satellite laser altimetry systems, rigorous evaluation of current sensor performance is critical for effective disaster impact assessment and long-term environmental management. This preliminary study assesses the elevation accuracy of NASA’s ICESat-2 ATL03 (global geolocated photon) data over flat terrain, with a focus on its potential as a reliable baseline for future monitoring frameworks.
Methods:
The study area is the Noto Peninsula, Japan, which offers diverse but locally flat and open terrain suitable for validation. We analyzed ATL03 datasets from 2019 to 2024. To minimize uncertainties from vegetation and built structures, flat and open areas were selected by visually cross-referencing ICESat-2 ground tracks with high-resolution aerial imagery. Photon-derived ellipsoidal heights were converted to orthometric heights using the Japanese Geoid Model 2011 (GSIGEO2011) for direct comparison with the 5-m resolution Digital Elevation Model (DEM) provided by the Geospatial Information Authority of Japan (GSI).
Analysis of Noise Characteristics:
Photon-counting LiDAR performance is highly sensitive to the signal-to-noise ratio (SNR). Visual inspection of ATL03 along-track profiles revealed substantial differences between daytime and nighttime acquisitions. Daytime data showed elevated solar background noise, which hindered reliable ground photon identification. In contrast, nighttime observations exhibited high SNR with minimal background noise. Therefore, this accuracy assessment was restricted to nighttime data to establish a high-quality performance baseline.
Results and Discussion:
Preliminary comparisons indicate that nighttime ATL03 elevations achieve an accuracy of approximately 1 m relative to the GSI 5-m DEM in selected flat regions. This discrepancy is likely due to the ~17-m diameter laser footprint averaging surface heights over a larger area than the DEM grid, along with residual effects from photon classification. Although the current data density was insufficient for robust change detection in the Noto region, these results highlight the sensor’s fundamental precision under optimal conditions.
Conclusion:
This analysis confirms that nighttime ICESat-2 ATL03 data can provide a high-precision topographic reference, supporting its role in future missions for monitoring dynamic geomorphological processes.