Chen et al. (2026) High-resolution, multi-depth mapping of soil bulk density and pH in China's forests using machine learning

Identification

• Journal: Earth system science data
• Year: 2026
• Date: 2026-06-30
• Authors: Jizhen Chen, Zihao Fan, Tao Liu, Wenfa Xiao, Qiwu Sun, Xiangyang Sun, Zhilin Huang
• DOI: 10.5194/essd-18-4451-2026

Research Groups

• Key Laboratory of Forest Ecology and Environment of National Forestry and Grassland Administration, Chinese Academy of Forestry, Beijing 100091, China
• Hubei Zigui Three Gorges Reservoir Forest Ecosystem Observation and Research Station, Zigui 443600, China
• Department of Earth System Science, Ministry of Education Key Laboratory for Earth System Modeling, Institute for Global Change Studies, Tsinghua University, Beijing 100084, China
• Research Institute of Forestry, Chinese Academy of Forestry, Beijing 100091, China
• College of Forestry, Beijing Forestry University, Beijing 100083, China

Short Summary

This study presents high-resolution (90 m), forest-specific maps of soil bulk density and pH across China using machine learning. The results provide a spatially explicit dataset to support forest soil inventories, carbon accounting, and assessments of soil acidification risk.

Objective

• To establish a nationwide forest soil profile database.
• To generate high-resolution digital maps of forest soil BD and pH.
• To provide a spatially explicit dataset for regional Earth system and land surface models.

Study Configuration

• Spatial Scale: Nationwide (China) with 90 m resolution.
• Temporal Scale: Data collected between 2018 and 2023, with long-term annual means and growing season averages calculated from monthly records spanning 2003 to 2023.

Methodology and Data

• Models used: Quantile Regression Forests (QRF) algorithm for spatial prediction and uncertainty quantification.
• Data sources: Soil profiles sampled through extensive field surveys, environmental covariates derived from various datasets (e.g., Sentinel-2 imagery, MODIS products).

Main Results

• High-resolution maps of forest soil BD and pH across China's forests at 90 m resolution.
• Model performance achieved model efficiency coefficients (MEC) ranging from 0.78 to 0.89 for BD and 0.83 to 0.87 for pH in cross-validation, indicating strong capability to capture spatial variability of forest soil properties.

Contributions

• The study provides a comprehensive forest soil database for China, comprising 4356 profiles and 11 873 samples across five standardized depth intervals.
• The high-resolution maps contribute to the GlobalSoilMap initiative and provide critical baseline data for estimating forest soil carbon stocks across China.

Funding

• This research was supported by [list projects, programs, and reference codes that funded this research].

Citation

@article{Chen2026Highresolution,
  author = {Chen, Jizhen and Zhang, Xin and Fan, Zihao and Liu, Tao and Xiao, Wenfa and Sun, Qiwu and Sun, Xiangyang and Huang, Zhilin},
  title = {High-resolution, multi-depth mapping of soil bulk density and pH in China's forests using machine learning},
  journal = {Earth system science data},
  year = {2026},
  doi = {10.5194/essd-18-4451-2026},
  url = {https://doi.org/10.5194/essd-18-4451-2026}
}