Shan et al. (2026) Spatial-temporal dynamics of meteorological and groundwater drought in Northwest China: Propagation, threshold, recovery time, drivers

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-01-06
• Authors: Jianan Shan, Rui Zhu, Zhenliang Yin, Chunshuang Fang, Rong Li, Ganlin Zhou
• DOI: 10.1016/j.ejrh.2025.103090

Research Groups

• State Key Laboratory of Ecological Safety and Sustainable Development in Arid Lands, National Cryosphere Desert Data Center, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou, Gansu, China
• National-Local Joint Engineering Research Center of Technologies and Applications for National Geographic State Monitoring, Key Laboratory of Science and Technology in Surveying & Mapping, Gansu Province, Faculty of Geomatics, Lanzhou Jiaotong University, Lanzhou, Gansu, China
• Shaanxi Normal University, Xian, Shaanxi, China
• Institute of Arid Meteorology, China Meteorological Administration, Lanzhou, Gansu, China

Short Summary

This study investigated the spatial-temporal dynamics of meteorological and groundwater drought propagation, thresholds, recovery times, and driving factors in Northwest China from 1960 to 2024. It found that extreme meteorological droughts predominantly trigger groundwater droughts, with climate change and human activities being the strongest drivers, highlighting the need for differentiated groundwater management.

Objective

• To understand the propagation mechanism from meteorological to groundwater drought for improved groundwater management and drought early warning.
• To analyze the changes in meteorology, hydrology, and drought conditions in Northwest China from 1960 to 2024.
• To determine the propagation characteristics, thresholds, and recovery times of meteorological-groundwater drought.
• To quantitatively assess the driving factors of droughts.

Study Configuration

• Spatial Scale: Northwest China (NWC), encompassing six provinces (Shaanxi, Gansu, Ningxia, Qinghai, Xinjiang, and western Inner Mongolia), divided into 20 sub-regions. Data were resampled to a 0.25° spatial resolution.
• Temporal Scale: 1960 to 2024 (65 years), with data processed at a monthly temporal scale.

Methodology and Data

• Models used:
  • Drought Indices: Standardized Precipitation Index (SPI), Groundwater Drought Index (GDI).
  • Propagation Analysis: Run theory, Convergent Cross Mapping (CCM), Copula function, Bayesian network.
  • Driver Analysis: XGBoost-SHAP (eXtreme Gradient Boosting with SHapley Additive exPlanations), Cross-Wavelet Analysis, Spearman correlation coefficient, Geographical detector, Local Moran's I.
  • Data Processing: STL (Seasonal-Trend decomposition using Loess) for GPP.
• Data sources:
  • Precipitation: 1-kilometer monthly precipitation dataset for China (1901–2023) from the National Tibetan Plateau Data Center.
  • Terrestrial Water Storage Anomaly (TWSA): BNML_TWSA dataset (machine-learning-based reconstruction, 1960–2022) and GRACE RL06 mascon product (2023–2024).
  • Soil Moisture, Snow Water Equivalent, Canopy Water: Global Land Data Assimilation System (GLDAS).
  • Gross Primary Production (GPP): CMIP6 (16 GCMs), "Long-term (1982–2018) global gross primary production dataset based on NIRv", and MODIS (Terra Gross Primary Productivity).
  • Atmospheric Circulation Factors: Arctic Oscillation (AO), El Niño - Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Pacific Decadal Oscillation (PDO), and Atlantic Multidecadal Oscillation (AMO).
  • Topographic Factors: 1-kilometer resolution Digital Elevation Model (DEM) and derived slope data.
  • Meteorological Factors: 1-kilometer monthly mean temperature and potential evapotranspiration datasets for China.
  • Land Surface Elements: Normalized Difference Vegetation Index (NDVI) and Global Land Evaporation Amsterdam Model (GLEAM) surface soil moisture (SM).
  • Human Activity Factors: Downscaled gridded global dataset for Gross Domestic Product (GDP) per capita, High-Resolution Gridded Dataset for China’s Monthly Sectoral Water Use (domestic water withdrawal), global gridded population dataset (POP), and Multi-temporal land use remote sensing monitoring dataset of China (CNLUCC).
  • Validation Data: In-situ groundwater level data from the National Tibetan Plateau Data Center and the "GWscn1km" monthly gridded groundwater level dataset for China.

Main Results

• Meteorological drought (SPI) showed a significant increasing trend, while groundwater storage anomaly (GWSA) and Groundwater Drought Index (GDI) exhibited a significant decreasing trend, particularly after 2008.
• The average number of meteorological-groundwater drought events (2.29 events) was lower than meteorological droughts (68.85 events) but higher than groundwater droughts (12.48 events). Compound events had the shortest average duration (2.29 months) and lowest severity (3.94).
• The average propagation time (PT) from meteorological to groundwater drought in Northwest China was 4.69 months.
• The average probabilities of meteorological drought triggering mild, moderate, severe, and extreme groundwater droughts were 30.27 %, 20.60 %, 9.63 %, and 5.50 %, respectively.
• The propagation threshold for triggering groundwater drought was predominantly governed by extreme meteorological drought, accounting for 55.69 % of the study area.
• The average recovery time for compound meteorological-groundwater droughts was 3.05 months, which was longer than that for individual meteorological (1.91 months) or groundwater (2.14 months) droughts.
• The El Niño - Southern Oscillation (ENSO) had the strongest influence on groundwater drought.
• The interaction between climate change and human activities contributed the largest average share (64 %) to the dynamics of meteorological-groundwater drought.
• Digital Elevation Model (DEM), precipitation (Pre), soil moisture (SM), and Gross Domestic Product (GDP) were identified as the primary driving factors across different categories.

Contributions

• Provided a comprehensive analysis of meteorological-to-groundwater drought propagation characteristics, thresholds, recovery times, and driving factors in Northwest China, addressing a significant research gap in arid and semi-arid regions.
• Introduced a novel definition of 'compound meteorological-groundwater drought event' to better capture the full system response from deep water deficit initiation to shallow water recovery.
• Quantified the contributions of various driving factors using the XGBoost-SHAP framework, offering a robust method for interpreting model outputs and identifying dominant environmental and anthropogenic influences.
• Utilized a machine-learning-based reconstructed Terrestrial Water Storage Anomaly (TWSA) dataset (BNML_TWSA) to extend the analysis period from 1960 to 2024, overcoming limitations of satellite data availability.
• Emphasized the importance of establishing a 'meteorological-groundwater' cascade drought early warning system and implementing differentiated water resource management strategies in arid and semi-arid regions.

Funding

• Strategic Priority Research Program of the Chinese Academy of Sciences (XDB0720103)
• National Natural Science Foundation of China (42161018)
• Youth Innovation Promotion Association CAS (2021424)
• Science and Technology Project of Gansu Province (23ZDKA017, 23ZDFA018, 24YFWA019)

Citation

@article{Shan2026Spatialtemporal,
  author = {Shan, Jianan and Zhu, Rui and Yin, Zhenliang and Fang, Chunshuang and Li, Rong and Zhou, Ganlin},
  title = {Spatial-temporal dynamics of meteorological and groundwater drought in Northwest China: Propagation, threshold, recovery time, drivers},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2025.103090},
  url = {https://doi.org/10.1016/j.ejrh.2025.103090}
}