Permafrost and underground ice are key solid water resources and engineering geological carriers. The high ice content permafrost in the Qinghai Tibet Engineering Corridor is extremely sensitive to degradation response, which restricts the accurate simulation of permafrost degradation by climate models and is not conducive to the scientific formulation of engineering disaster prevention and control measures in vulnerable areas of the cryosphere.
This study is based on 1158 sets of measured data from underground boreholes (measured data, engineering survey borehole data, and shared data from public literature and scientific data centers). Two machine learning algorithms, random forest and decision tree, are introduced, combined with 13 environmental factors such as annual average temperature, ground temperature, altitude, slope, vegetation index, etc., to achieve spatial simulation of underground ice content at a resolution of 200m along the entire corridor. The results show that 50% of the areas within the Qinghai Tibet Engineering Corridor are developed with high ice content permafrost; The annual average temperature, ground temperature, active layer thickness, altitude, and precipitation are the dominant factors controlling the spatial differentiation of underground ice, as detailed in Fan et al. (2026).
This study can provide basic background data for the refined assessment of regional underground ice reserves, and also provide scientific support for ecological protection and engineering construction design in permafrost areas.
| # | number | name | type |
| 1 | XZ202501ZY0139 | other | |
| 2 | 2024M753468 | other | |
| 3 | XDB0950200 | Strategy Priority Research Program (Category B) of Chinese Academy of Sciences | |
| 4 | SKLFSE-ZQ-202308 | other |
This work is licensed under
CC BY 4.0 (Creative Commons Attribution 4.0 International License).
CSTR:11738.11.ncdc.permafrost.db7317.2026
10.12072/ncdc.permafrost.db7317.2026
