Error-Reduced Digital Elevation Model of the Qinghai-Tibet Plateau using ICESat-2 and Fusion Model

Area of Science:

• Geosciences
• Remote Sensing
• Geomorphology

Background:

• The Qinghai-Tibet Plateau (QTP) is crucial for studying Earth's surface processes.
• Open-access Digital Elevation Models (DEMs) for the QTP are noisy and inconsistent due to terrain, canopy, and snow.
• Glacier regions in the QTP require frequent updates due to dynamic changes.

Purpose of the Study:

• To develop a highly accurate and consistent digital elevation model for the QTP.
• To address limitations of existing DEMs in complex environments like the QTP.
• To create an updated terrain representation for the QTP for 2021.

Main Methods:

• Fusion of open-access DEMs with high-accuracy photons from the Ice, Cloud, and land Elevation Satellite-2 (ICESat-2).
• Incorporation of snow cover and canopy height data.
• Development of an ensemble learning fusion model for multi-sensor elevation data integration.

Main Results:

• Creation of HQTP30, the most accurate 2021 terrain model for the QTP.
• HQTP30 demonstrated superior accuracy compared to high-resolution imagery, UAV DEMs, control points, and ICESat-2.
• Achieved a Mean Absolute Error (MAE) of 0.71m in non-glacier regions and reduced MAE by 4.35m in glacier regions compared to COPDEM.

Conclusions:

• HQTP30 represents a significant advancement in QTP terrain modeling.
• The fusion approach effectively overcomes noise and inconsistency issues in DEMs.
• HQTP30 offers versatile applicability and improved accuracy for diverse QTP environments, including dynamic glacier regions.