Bai et al. (2026) Variation characteristics of water reserves in high mountains and surrounding areas in Arid Central Asia

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-01-16
• Authors: Bing Bai, Ping Yue, Xueyuan Ren, Q Zhang, Tiejun Zhang, Xin Zhang
• DOI: 10.1016/j.ejrh.2026.103138

Research Groups

• Lanzhou Institute of Arid Meteorology, Chinese Meteorological Administration, China
• Gansu Province Key Laboratory of Arid Climate Change and Reducing Disaster, China
• Meteorological Administration Key Laboratory of Arid Climate Change and Reducing Disaster, Lanzhou 730020, China
• Lanzhou Regional Climate Center, Lanzhou 730020, China

Short Summary

This study investigated terrestrial water storage (TWS) trends in Arid Central Asia and its high-mountain regions from 2003 to 2021, revealing contrasting responses to climate change and human activities, with significant TWS decline in the Tianshan Mountains but increases in the Kunlun and Qilian Mountains.

Objective

• To quantify the spatiotemporal variability of terrestrial water storage anomaly (TWSA) at monthly, seasonal, and interannual scales across Arid Central Asia and its high-mountain regions, and to reveal contrasting TWS responses under co-directional and counter-directional changes in climate driving and ecological vegetation regulation.

Study Configuration

• Spatial Scale: Arid Central Asia (ACA), including the Tianshan Mountains (TSM), Kunlun Mountains (KLM), Qilian Mountains (QLM) (collectively ACA-alp), and adjacent non-alpine areas such as the Oases in the Hexi Corridor (OHC) and the Tarim Basin (TB). High mountains are defined as areas above 3500 meters above sea level.
• Temporal Scale: January 2003 to December 2021 (19 years).

Methodology and Data

• Models used:
  • Hydrology Tiled ECMWF Scheme for Surface Exchanges over Land (ECMWF HTESSEL) (implicitly used in ERA5-Land data).
  • Theil-Sen slope estimator for trend calculation.
  • Mann-Kendall test for trend significance.
  • Structural Equation Modeling (SEM) using the "Semopy" library in Python for driving factor analysis.
• Data sources:
  • Satellite: GRACE Level 03 dataset (Gauer et al., 2023) for terrestrial water storage anomaly (TWSA), with data gap filling using Multi-channel Singular Spectrum Analysis (M-SSA) and DDK7 decorrelation filter. Spatial resolution: 1° × 1°.
  • Reanalysis: ERA5-Land monthly data for air temperature, precipitation, and evapotranspiration (ET) from Google Earth Engine (GEE). Spatial resolution: 0.1° × 0.1°.
  • Baseline: TWSA calculated as monthly departures from the 2004–2009 climatological monthly means.

Main Results

• Climate Trends (2003–2021):
  • Temperature: Arid Central Asia (ACA) warmed at an average rate of 0.026 K/year (0.26 K/decade). High-mountain regions (ACA-alp) warmed at 0.023 K/year (0.23 K/decade). Qilian Mountains (QLM) showed the fastest warming at 0.037 K/year (0.37 K/decade), followed by Kunlun Mountains (KLM) at 0.027 K/year (0.27 K/decade), and Tianshan Mountains (TSM) at 0.021 K/year (0.21 K/decade).
  • Precipitation: ACA showed a slight increase of 0.9 mm/decade. ACA-alp exhibited a stronger positive trend of 19 mm/decade. QLM increased by 16 mm/decade, KLM by 7 mm/decade, while TSM decreased by -4 mm/decade.
  • Evapotranspiration (ET): ACA showed a weak decreasing trend of -6 mm/decade, while ACA-alp increased by 8 mm/decade. QLM increased by 9 mm/decade, KLM by 6.2 mm/decade, and TSM by 6 mm/decade.
  • Precipitation-Evapotranspiration (P-ET) Balance: TSM showed a water-storage deficit trend of -28 mm/decade. KLM had a deficit trend of -13 mm/decade. QLM showed a moisture surplus.
• Terrestrial Water Storage Anomaly (TWSA) Trends (2003–2021):
  • Overall: ACA TWSA decreased by -35.6 mm/year. ACA-non-alp decreased by -54.2 mm/year. ACA-alp increased by 18.6 mm/year.
  • Mountain Ranges:
    • Tianshan Mountains (TSM): Experienced a significant decrease in TWSA of -111.3 mm/year. This region transitioned from a water surplus to a deficit in 2014 across all seasons, with seasonal decrease rates ranging from -27.2 mm/year to -28.2 mm/year.
    • Kunlun Mountains (KLM): Showed an increase in TWSA of 78.1 mm/year. This region transitioned from a deficit to a surplus in 2010, with seasonal increase rates ranging from 18.9 mm/year to 19.9 mm/year.
    • Qilian Mountains (QLM): Showed an increase in TWSA of 51.8 mm/year. This region transitioned from a deficit to a surplus in 2015-2017, with seasonal increase rates ranging from 12.5 mm/year to 13.4 mm/year.
  • Non-Alpine Areas:
    • Tarim Basin (TB): Decreased by 28.6 mm/year, indicating persistent water scarcity.
    • Oases in the Hexi Corridor (OHC): Increased by 10.9 mm/year, attributed to human activities and conservation policies.
• Driving Factors:
  • Normalized Difference Vegetation Index (NDVI) was the primary direct contributor to TWS. Positive effects in QLM (path coefficient 0.37) and KLM (0.61), but a strong negative effect in TSM (-0.37).
  • Temperature had significant negative effects in QLM and KLM, but precipitation had positive effects. In TSM, precipitation showed a negative effect.
  • ET had no direct impact on TWS but indirectly modulated the effects of temperature and precipitation.
• Lake and Runoff Changes (Qilian Mountains):
  • Qinghai Lake surface area expanded by 2.89 × 10⁶ m²/year (2.89 km²/year) from 1986 to 2022.
  • Heihe River annual runoff averaged 1.89 × 10⁹ m³ and increased by approximately 6 × 10⁶ m³ per decade.
  • Shiyang River annual runoff averaged 7.3 × 10⁸ m³ and increased by approximately 3.8 × 10⁷ m³ per decade.

Contributions

• Provides a systematic, multi-timescale analysis of terrestrial water storage anomalies across Arid Central Asia and its high-mountain regions using GRACE data, offering an integrated view of all water compartments.
• Reveals distinct and contrasting TWSA trends and their underlying climatic and vegetation drivers across different mountain ranges (e.g., Tianshan decline vs. Kunlun/Qilian increase).
• Highlights that explanations relying solely on precipitation and evapotranspiration are insufficient for mountain TWS, emphasizing the role of snow and glacier melt.
• Demonstrates that human activities, with sound planning and effective institutions, can positively influence watershed water retention and storage capacity, challenging the notion that human intervention always depletes water resources in arid regions.

Funding

• National Natural Science Foundation of China (Project No. U2142208)
• Foundation of Key Talent Program Projects in Gansu Province (Project No. 2023RCXM37)
• Natural Science Foundation of Gansu Province (24JRRA725, 24JRRA726, 25JRRA848)

Citation

@article{Bai2026Variation,
  author = {Bai, Bing and Yue, Ping and Ren, Xueyuan and Zhang, Q and Zhang, Tiejun and Zhang, Xin},
  title = {Variation characteristics of water reserves in high mountains and surrounding areas in Arid Central Asia},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.103138},
  url = {https://doi.org/10.1016/j.ejrh.2026.103138}
}