Huimin et al. (2025) The response of meteorological drought to extreme climate in the water-receiving area of the Tao river diversion project in China

Identification

• Journal: Scientific Reports
• Year: 2025
• Date: 2025-11-26
• Authors: Hou Huimin, D. C. Lu, Dongmeng Zhou, Junhong Bai, Feng Guo, Changjie Chen, Junde Wang, Yufei Cheng, Zhenan Bao, Haohao Li
• DOI: 10.1038/s41598-025-26162-2

Research Groups

• School of Civil and Hydraulic Engineering, Lanzhou University of Technology, Lanzhou, China
• State Key Laboratory of Ecological Safety and Sustainable Development in Arid Lands Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou, China
• Gansu Provincial Hydrology and Water Resources Center, Lanzhou, China
• Gansu Academy for Water Conservancy, Lanzhou, China

Short Summary

This study analyzed the spatiotemporal variations, interrelationships, and driving factors of meteorological drought and extreme climate events in the water-receiving area of the Tao River Diversion Project, China. It found a persistent drying trend since 1988, primarily driven by annual total precipitation, cold days, and summer days, with most extreme climate factors exhibiting complex nonlinear influences and critical thresholds on drought.

Objective

• To create a multidimensional portrait of the dynamics of extreme climate and meteorological drought by combining a suite of relevant indices.
• To decipher the interrelationships and lag effects between meteorological drought and extreme climate using wavelet coherence and cross-wavelet analysis.
• To pinpoint the dominant drivers and unravel the nonlinear influence of extreme climate on drought through an integrated XGBoost-SHAP framework.

Study Configuration

• Spatial Scale: Water-receiving area of the Tao River Diversion Project in central Gansu Province, China, with a spatial resolution of 0.1° × 0.1°.
• Temporal Scale: Annual data from 1985 to 2018 (34 years).

Methodology and Data

• Models used:
  • Meteorological Drought Index: Standardized Precipitation Evapotranspiration Index (SPEI12).
  • Trend and Mutation Analysis: Linear regression, Sen’s slope test, Mann–Kendall test, Innovative Trend Analysis (ITA), Sequential Mann–Kendall test.
  • Drought Characterization: Run theory.
  • Interrelationships and Lag Effects: Wavelet Coherence (WTC), Cross Wavelet Transform (XWT).
  • Driving Factor Analysis: Pearson correlation analysis, XGBoost (eXtreme Gradient Boosting) Regressor (optimized via Bayesian optimization), SHAP (SHapley Additive exPlanations) framework.
• Data sources:
  • SPEI: CHM_Drought dataset (https://zenodo.org/records/14634774).
  • Extreme climate data: National Cryosphere Desert Data Center of China (http://www.ncdc.ac.cn), based on 12 ETCCDI indices.

Main Results

• The SPEI12 index in the study area showed a non-significant downward (drying) trend at a rate of −0.04 per decade from 1985 to 2018, with a significant mutation point identified in 1988. Aridification progressively intensified after 1992.
• Drought frequency exhibited a spatial pattern of being lower in the northwest and higher in the southeast. Southeastern droughts were more frequent but of shorter duration, while southern and western regions experienced less frequent but more prolonged and severe droughts.
• Extreme temperature indices showed significant warming trends: summer days (SU) increased by 4.4 days per decade, warm nights (TN90p) by 3.9% per decade, warm days (TX90p) by 2.3% per decade, and warm spell duration index (WSDI) by 1 day per decade. Cold nights (TN10p) decreased significantly by −1.5% per decade.
• Among precipitation indices, only the Simple Daily Intensity Index (SDII) increased significantly by 0.2 (mm/day) per decade. Other precipitation indices showed insignificant upward trends.
• SPEI exhibited a negative correlation and lagged response to extreme temperature indices (SU, TN90p, TX90p, WSDI), with significant resonance periods typically under 5 years. Conversely, SPEI showed a positive correlation and led extreme precipitation indices (TN10p, TX10p, SDII, R20mm, R95pTOT, PRCPTOT), with some displaying resonance over longer time scales (over 5 years).
• The main drivers of meteorological drought, quantified by SHAP analysis, were annual total precipitation (PRCPTOT, 23.35%), cold days (TX10p, 18.93%), and summer days (SU, 14.75%).
• PRCPTOT and SU showed linear relationships with SPEI12 (PRCPTOT alleviates drought, SU intensifies it). Most other extreme climate indices exhibited complex nonlinear relationships and critical thresholds; for example, consecutive wet days (CWD) below 8.58 days contributed to humidification, while above this threshold, it promoted aridification.

Contributions

• Provides a comprehensive, multidimensional analysis of meteorological drought and extreme climate dynamics, including their interrelationships and lag effects, in a vulnerable region.
• Utilizes an advanced integrated XGBoost-SHAP framework to identify dominant drivers and quantify their nonlinear influences and critical thresholds on meteorological drought, offering insights beyond traditional linear correlation methods.
• Offers actionable recommendations for regional water resource management, including adaptive agricultural practices, leveraging water storage, and integrating machine learning-derived thresholds into drought early warning systems.
• Highlights the critical role of annual precipitation, cold days, and summer days as key drivers of meteorological drought in the study area.

Funding

• Gansu provincial key research and development program (23YFFA0018)
• Gansu Province Water Conservancy Research and Planning Project (23GSLK044)

Citation

@article{Huimin2025response,
  author = {Huimin, Hou and Lu, D. C. and Zhou, Dongmeng and Bai, Junhong and Guo, Feng and Chen, Changjie and Wang, Junde and Cheng, Yufei and Bao, Zhenan and Li, Haohao},
  title = {The response of meteorological drought to extreme climate in the water-receiving area of the Tao river diversion project in China},
  journal = {Scientific Reports},
  year = {2025},
  doi = {10.1038/s41598-025-26162-2},
  url = {https://doi.org/10.1038/s41598-025-26162-2}
}