This dataset provides a high-precision geothermal spatial simulation product for the heterogeneity of the annual average ground temperature (MAGST) spatial distribution of permafrost in the Greater and Lesser Khingan Mountains in Northeast China. The dataset is constructed based on multi-source observation and machine learning framework, and comprehensively utilizes the measured data of ground temperature at a depth of 15 meters from deep (>20 meters) and shallow boreholes in the region. The study selected environmental factors such as precipitation (PRE), surface melting index (TDD), and terrain position index (TPI) as key predictive variables, and used the Random Forest (RF) regression algorithm to simulate and generate an annual average ground temperature distribution map with a spatial resolution of 1 km in the study area. This data effectively reconstructed and expanded the spatial pattern information of ground temperature in Northeast China.
Basic sample: Integrated 15 meter depth ground temperature measurement data from all deep and shallow boreholes in the region (covering 104 stations).
Predictive variables: Precipitation (PRE), Surface Melting Index (TDD), Topographic Location Index (TPI), etc.
Data integration: Clean and integrate borehole temperature measurement data as training and validation sets.
Model construction: Using random forest regression algorithm to establish a nonlinear relationship model between ground temperature and environmental factors.
Spatial mapping: Apply the trained model to the entire region to generate a MAGST spatial distribution map.
Quality control: Ensure the fitting accuracy of the model at the site scale through cross validation.
This data is modeled using machine learning methods to calculate confusion matrix and overall accuracy. The results show that the model has high consistency.
| # | number | name | type |
| 1 | 2022FY100700 | Survey of Permafrost Conditions and Freeze-Thaw Damage in the High-Latitude Regions of Northeast China | Basic Resource Survey Project |
This work is licensed under
CC BY 4.0 (Creative Commons Attribution 4.0 International License).
CSTR:11738.11.NCDC.NIEER.DB7272.2026
10.12072/ncdc.nieer.db7272.2026
