Wang et al. (2026) Divergent mountain runoff dynamics but declining per capita freshwater availability across the Third Pole by mid-21st century

Identification

• Journal: npj Climate and Atmospheric Science
• Year: 2026
• Date: 2026-01-07
• Authors: Lei Wang, Junshui Long, Deliang Chen, N. Li, Xiuping Li, Tandong Yao
• DOI: 10.1038/s41612-025-01313-4

Research Groups

• State Key Laboratory of Tibetan Plateau Earth System, Environment and Resources (TPESER), Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
• University of Chinese Academy of Sciences, Beijing, China
• Department of Earth System Science, Tsinghua University, Beijing, China

Short Summary

This study quantifies historical and future per-capita freshwater supply and river runoff dynamics across the Third Pole's climate domains, revealing that while runoff dynamics diverge (increasing in monsoon, non-significantly declining in westerlies), per-capita freshwater availability is projected to decline significantly across all domains by mid-21st century due to rapid population growth.

Objective

• To quantify the changes in historical and future per-capita freshwater supply in the Third Pole (TP) and accurately depict the response patterns of river runoff in the TP’s three different climate domains to climate change.

Study Configuration

• Spatial Scale: The Third Pole region, encompassing 12 major river basins divided into three climate domains: westerlies (Amu Darya, Syr Darya, Indus, Tarim, Heihe, Shule rivers), Indian monsoon (Brahmaputra, Ganges, Mekong, Salween rivers), and westerlies‒monsoon transition (Yangtze, Yellow rivers). Analysis covers mountain-outlet basins and their hydrologically-dependent downstream regions. Climate data has a 0.5° spatial resolution.
• Temporal Scale:
  • Historical: 1960–2016 for runoff observations; 1960–1970 for past per-capita water supply (APWS); 2006–2015 for present APWS.
  • Future: 2017–2100 for runoff trends; 2000–2100 for glacier mass change; 2030–2050 for near-future APWS; 2080–2100 for far-future APWS.
  • Climate Data Reference: 2001–2015 for ERA5-Land.

Methodology and Data

• Models used:
  • Deep Learning Model: Long Short-Term Memory (LSTM) for simulating monthly runoff, evapotranspiration (ET), and glacier mass change.
  • Cryosphere-hydrology models: VIC-glacier model (for reconstructing natural runoffs of Amu Darya and Syr Darya rivers).
  • Glacier Evolution Model: Python Glacier Evolution Model (PyGEM) (for future glacier mass balance projections).
• Data sources:
  • Satellite: Randolph Glacier Inventory (RGI) data (1970s, 2000s, Version 6).
  • Observation: In-situ annual mountain-outlet runoff records from the Ministry of Water Resources in China (MWRC), Department of Hydrology and Meteorology in Nepal (DHM), Pakistan Water & Power Development Authority (WAPDA). Monthly China Meteorological Administration (CMA) precipitation and discharge data.
  • Reanalysis: ERA5-Land historical (2001–2015) precipitation, air temperature, and monthly ET.
  • Projections: Corrected and downscaled CMIP6 (Coupled Model Intercomparison Project Phase 6) outputs from five General Circulation Models (GCMs: GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL) under Shared Socioeconomic Pathway (SSP) scenarios SSP245 (moderate emission) and SSP585 (high emission) from Inter-Sectoral Impact Model Intercomparison Project 3b (ISIMIP3b). Future glacier mass balance dataset (Rounce et al., 2020). Historical and future population data (CIESIN, Gao, Jones et al.).

Main Results

• By the end of the 21st century (2017–2100), mountain runoff in the TP's monsoon domain and westerlies‒monsoon transition domain is projected to increase significantly (p < 0.05), whereas runoff in the westerlies domain will experience a non-significant decline.
• This divergence is attributed to the balance between precipitation minus evapotranspiration (P-ET) and glacier melt:
  • In monsoon and transition domains, rapid increases in P-ET largely offset future glacier meltwater decline, leading to increased runoff.
  • In the westerlies domain, increased precipitation is mostly offset by increased ET, resulting in an insignificant trend in P-ET, and relatively stable glacial meltwater before 2100 leads to largely unchanged mountain runoff.
• Long-term (1960–2100) mountain runoff changes show contrasting patterns:
  • Westerlies domain: Runoff initially rises from 234 ± 25 mm (1960–1970) to 266 ± 13 mm (2006–2015), then falls in the near future (2030–2050) under both SSP245 and SSP585.
  • Monsoon domain: Runoff initially falls by approximately 6.5% from 450 ± 27 mm (1960–1970) to the present, then consistently rises to 621 ± 15 mm (SSP245) and 684 ± 24 mm (SSP585) in the far future.
  • Westerlies‒monsoon transition domain: Shows a consistent increase from 129 ± 24 mm (1960–1970) to 172 ± 17 mm (SSP245) and 175 ± 11 mm (SSP585) in the far future.
• Annual per-capita water supply (APWS) uniformly decreased from the past (1960–1970) to the present (2006–2015) by -53%, -58%, and -29% across the three climate domains, primarily due to consistent and rapid population increase.
• In the near future (2030–2050), APWS shows slight increases for the monsoon and transition domains but a decline for the westerlies domain.
• The arid westerlies domain's APWS is projected to sharply decline from above 2400 m³ per person per year to below the chronic water scarcity level (1000 m³ per person per year) from 1960–1970 to 2030–2050. The monsoon domain's APWS is projected to drop from above 1000 m³ per person per year to approaching 500 m³ per person per year in the near future.

Contributions

• Challenges the prevailing paradigm that future runoff across the entire Third Pole will follow a unidirectional increasing trend, emphasizing the complex and divergent responses across different climate domains.
• Provides a pan-regional, coherent, and observation-constrained quantification of historical and future per-capita freshwater supply dynamics, addressing a critical gap in existing literature.
• Highlights the dominant role of rapid population growth in driving declining per-capita freshwater availability, even in regions with increasing total runoff, offering crucial insights for water security.
• Offers vital scientific information for re-evaluating traditional water resource management models, guiding adaptive strategies, and supporting evidence-based policy-making to achieve Sustainable Development Goal 6 in Asia.

Funding

• Second Tibetan Plateau Scientific Expedition and Research Program (2024QZKK0400)
• National Key R&D Program of China (2024YFF0808602)
• Tsinghua University (100008001)

Citation

@article{Wang2026Divergent,
  author = {Wang, Lei and Long, Junshui and Chen, Deliang and Li, N. and Li, Xiuping and Yao, Tandong},
  title = {Divergent mountain runoff dynamics but declining per capita freshwater availability across the Third Pole by mid-21st century},
  journal = {npj Climate and Atmospheric Science},
  year = {2026},
  doi = {10.1038/s41612-025-01313-4},
  url = {https://doi.org/10.1038/s41612-025-01313-4}
}