Datasets for A 10-year daily 0.0625° multi-layer soil temperature and moisture dataset over China derived from a cascaded LightGBM machine learning framework (CSTSMv1.0)

The China Soil Temperature and Soil Moisture version 1.0 (CSTSMv1.0) dataset, is a daily 0.0625°product for multiple depths (10, 20, 40, and 80cm) from 2015 to 2024. The dataset is generated by a cascaded machine learning framework based on LightGBM to simultaneously estimate multi-layer soil temperature and soil moisture (at the depth of 10, 20, 40, and 80 cm) by integrating in situ observations from more 3000 staions, CLDAS and ERA5L reanalysis, GLASS leaf area index, topographic and CSDL soil properties data. The generated long-term daily soil temperature and soil moisture product at 0.0625° resolution demonstrates high accuracy across different soil layers and climate regimes.