Zhang et al. (2026) Inverse modeling and coupled analysis of soil moisture content and organic carbon in oasis areas of arid regions

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-08-27
• Authors: Zihan Zhang, Jinjie Wang, Jianli Ding, Jinming Zhang, Liya Shi, Xiangyu Ge
• DOI: 10.1016/j.agwat.2026.110710

Research Groups

• College of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi, China
• Xinjiang Institute of Technology, Aksu, China

Short Summary

This study employed a deep learning approach integrating multi-source data to achieve multi-layer inversion of soil moisture and soil organic carbon in the Wei-Ku Oasis region. The results indicate that the CNN-LSTM model achieves the best performance in multi-depth inversion, with significant spatial heterogeneity and vertical differentiation of the soil water-carbon relationship.

Objective

• Investigate the synergistic evolution mechanisms of soil water and carbon in arid regions.
• Develop a deep learning approach integrating multi-source data for multi-layer inversion of soil moisture and soil organic carbon.

Study Configuration

• Spatial Scale: Regional scale, focusing on the Wei-Ku Oasis region in Xinjiang, China.
• Temporal Scale: Long-term observation from 2017 to 2025, with field-measured samples collected in 2017, 2018, 2021, 2022, and 2024.

Methodology and Data

• Models used:
  • Convolutional Neural Network (CNN)
  • Long Short-Term Memory (LSTM) network
  • CNN-LSTM hybrid model
  • Geographically Weighted Regression (GWR)
  • Bivariate Moran's I
  • Coupling coordination degree model
• Data sources:
  • Multi-source remote sensing data from Sentinel-1 and Landsat series
  • Field-measured soil moisture and organic carbon samples

Main Results

• The CNN-LSTM model achieves the best performance in multi-depth inversion, with the highest accuracy at the 0–10 cm layer (SMC: R2=0.72; SOC: R2=0.74).
• Significant spatial heterogeneity and vertical differentiation of the soil water-carbon relationship.
• Strong spatial clustering between SMC and SOC at different soil profile depths.

Contributions

• This study provides methodological support for multi-depth remote sensing inversion of soil properties in arid regions.
• Enhances understanding of water-carbon interaction mechanisms in oasis ecosystems.

Funding

• This research was funded by the National Natural Science Foundation of China (Grant No. 42161073).
• The Xinjiang University Research Fund (Grant No. XJU2020004).

Citation

@article{Zhang2026Inverse,
  author = {Zhang, Zihan and Wang, Jinjie and Ding, Jianli and Zhang, Jinming and Shi, Liya and Ge, Xiangyu},
  title = {Inverse modeling and coupled analysis of soil moisture content and organic carbon in oasis areas of arid regions},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110710},
  url = {https://doi.org/10.1016/j.agwat.2026.110710}
}