Tong et al. (2026) Sensitivity of Vegetation Greenness to Multi-Depth Soil Moisture on the Mongolian Plateau: Nonlinear Responses Revealed by RF–SHAP

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Remote Sensing
• Year: 2026
• Date: 2026-08-05
• Authors: Siqin Tong, Shubo Zhang, Xueliang Yang, Jinyuan Ren, Altantuya Dorjsuren, Sainbuyan Bayarsaikhan, Gang Bao, Xiaojun Huang, Yuhai Bao
• DOI: 10.3390/rs18152600

Research Groups

Not specified

Short Summary

This study evaluates the sensitivity of vegetation (NDVI) to soil moisture across different depths on the Mongolian Plateau from 1982 to 2022, finding that sensitivity is highest in the subsurface layer and has increased over time due to vegetation greening.

Objective

• To investigate the vertical differentiation, nonlinear threshold responses, and temporal evolution of vegetation sensitivity to soil moisture in the Mongolian Plateau.

Study Configuration

• Spatial Scale: Mongolian Plateau (arid and semi-arid regions)
• Temporal Scale: 1982–2022

Methodology and Data

• Models used: Random Forest (RF) combined with SHAP (Shapley Additive Explanations)
• Data sources: Long-term NDVI data and ERA5-Land multilayer soil moisture reanalysis products

Main Results

• Vertical Sensitivity: Sensitivity follows the order: subsurface layer (7–28 cm) > surface layer (0–7 cm) > deep layer (28–100 cm).
• Nonlinear Response: A significant nonlinear relationship exists where vegetation sensitivity intensifies under low-moisture conditions, indicating threshold effects.
• Spatial Distribution: High sensitivity is primarily located in typical grassland and desert-steppe regions.
• Temporal Trend: Vegetation sensitivity to soil moisture has generally increased over the last 40 years, especially in semi-arid regions.
• Driving Mechanisms:
  • Vegetation greening (NDVI trend) acts as the strongest positive driver, creating a feedback loop of enhanced water dependence.
  • An "interlayer substitution effect" occurs where improved surface soil moisture reduces the reliance on deeper soil water.

Contributions

• Systematically characterizes the spatiotemporal dynamics and driving mechanisms of vegetation sensitivity using a multilayer soil moisture perspective, offering insights into ecosystem stability and adaptive strategies for arid and semi-arid regions.

Funding

Not specified

Citation

@article{Tong2026Sensitivity,
  author = {Tong, Siqin and Zhang, Shubo and Yang, Xueliang and Ren, Jinyuan and Dorjsuren, Altantuya and Bayarsaikhan, Sainbuyan and Bao, Gang and Huang, Xiaojun and Bao, Yuhai},
  title = {Sensitivity of Vegetation Greenness to Multi-Depth Soil Moisture on the Mongolian Plateau: Nonlinear Responses Revealed by RF–SHAP},
  journal = {Remote Sensing},
  year = {2026},
  doi = {10.3390/rs18152600},
  url = {https://doi.org/10.3390/rs18152600}
}