Bulk density datasets using pedotransfer functions in the Qinghai-Tibetan Plateau.

Area of Science:

• Soil Science
• Environmental Science
• Geospatial Analysis

Background:

• Accurate soil bulk density (BD) data is vital for understanding soil compaction, element stocks, and ecological processes.
• Measuring BD across large, complex regions like the Qinghai-Tibetan Plateau (QTP) is challenging due to labor-intensive methods.
• Existing data may not capture the fine-scale spatial heterogeneity crucial for ecological modeling.

Purpose of the Study:

• To develop a high-resolution (90m) soil bulk density (BD) grid dataset for the Qinghai-Tibetan Plateau (QTP).
• To integrate soil pedotransfer functions and machine learning for accurate BD prediction.
• To provide a reliable dataset for ecological research and environmental management in the QTP.

Main Methods:

• Utilized soil pedotransfer functions combined with machine learning algorithms.
• Generated a 90m resolution grid dataset for soil bulk density across the QTP.
• Employed rigorous validation techniques to assess map accuracy and uncertainty.

Main Results:

• Successfully generated a high-resolution soil bulk density (BD) dataset for the QTP.
• Achieved Root Mean Square Error (RMSE) values between 0.144 and 0.190, indicating high accuracy.
• The dataset effectively captures the spatial heterogeneity of BD across the region.
• Uncertainty quantification was rigorously assessed for reliability.

Conclusions:

• The new BD dataset provides a robust foundation for ecological studies in the QTP.
• High spatial and depth resolution makes the data suitable for element storage and soil water transport modeling.
• Informs regional ecological restoration, environmental protection, and sustainable development strategies.