Sattar et al. (2026) Drought dynamics and climatic drivers in the Tarim Basin using remote sensing indices and pixel-wise machine learning

Identification

• Journal: Scientific Reports
• Year: 2026
• Date: 2026-09-18
• Authors: Mutallip Sattar, Alim Abbas, Sardar Parhat, Alimujiang Yasen, Muhemaiti Wahafu, Akida Salam, Batur Bake
• DOI: 10.1038/s41598-026-72142-5

Research Groups

• Mutallip Sattar
• Alim Abbas
• Sardar Parhat
• Alimujiang Yasen
• Muhemaiti Wahafu
• Akida Salam
• Batur Bake

Short Summary

This study investigates drought dynamics in the Tarim Basin, China's largest inland arid region, using remote sensing indices and machine learning. The results reveal that vapor pressure deficit (VPD) is the primary driver of both landscape-scale drought and agricultural water stress.

Objective

• Analyze spatio-temporal characteristics of drought in the Tarim Basin from 2000 to 2024.
• Quantify long-term trends and seasonal variations of remote sensing indices.
• Identify dominant environmental drivers through pixel-wise Random Forest with permutation importance.
• Examine the role of large-scale climate oscillations in modulating drought variability.

Study Configuration

• Spatial Scale: The study area covers approximately 530,000 km² in southern Xinjiang, northwestern China (35°–42°N, 74°–90°E).
• Temporal Scale: The study period spans from 2000 to 2024.

Methodology and Data

• Models used: Random Forest regression model.
• Data sources:
  • MOD11A2 Terra Land Surface Temperature and Emissivity 8-Day Global 1 km datasets for TVDI calculation.
  • MOD16A2 evapotranspiration product for CWSI calculation.
  • Meteorological raster data from the Resource and Environmental Science Data Platform (RESDC) for climate variables.

Main Results

• The spatial distribution of TVDI shows significant spatial heterogeneity, with extreme drought dominating 51.5% of the basin.
• CWSI exhibits an "oasis-concentrated" pattern, with extreme drought being the dominant category.
• VPD dominates both TVDI and CWSI, contributing 72–75% to spatio-temporal drought variation.

Contributions

• This study provides a comprehensive assessment of drought dynamics in the Tarim Basin using remote sensing indices and machine learning.
• The results highlight the importance of VPD as a primary driver of drought in arid regions.

Funding

• Not specified.

Citation

@article{Sattar2026Drought,
  author = {Sattar, Mutallip and Abbas, Alim and Parhat, Sardar and Yasen, Alimujiang and Wahafu, Muhemaiti and Salam, Akida and Bake, Batur},
  title = {Drought dynamics and climatic drivers in the Tarim Basin using remote sensing indices and pixel-wise machine learning},
  journal = {Scientific Reports},
  year = {2026},
  doi = {10.1038/s41598-026-72142-5},
  url = {https://doi.org/10.1038/s41598-026-72142-5}
}