Wang et al. (2025) Acceleration of diverging runoff trends on the Third Pole

Identification

• Journal: Communications Earth & Environment
• Year: 2025
• Date: 2025-11-17
• Authors: Lei Wang, Xiuping Li, Arthur Lutz, Santosh Nepal, Deliang Chen, Tandong Yao, Fengge Su, Lan Cuo, Zhijun Yao, Yinsheng Zhang, Zhidan Hu, Jingheng Huang, Mei Hou, Ruishun Liu, Junshui Long, Chenhao Chai, Zhaofei Liu, Ahmad Bashir, Sonu Khanal, He Sun, Yong Nie, Yongqiang Zhang, Tao Wang
• DOI: 10.1038/s43247-025-02854-5

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
• Faculty of Geosciences, Department of Physical Geography, Utrecht University, Utrecht, Netherlands
• International Water Management Institute (IWMI), Nepal office, Kathmandu, Nepal
• Department of Earth System Science, Tsinghua University, Beijing, China
• Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China
• State Key Laboratory of Water Cycle and Water Security, China Institute of Water Resources and Hydropower Research, Beijing, China
• Pakistan Agriculture Research Council, Islamabad, Pakistan
• Institute of Mountain Hazards and Environment, Chinese Academy of Sciences, Chengdu, China

Short Summary

This study quantifies long-term (1960-2016) divergent runoff trends in the Third Pole's major rivers, revealing significant increases in westerlies-dominated rivers and insignificant declines in monsoon-dominated rivers, with these contrasting changes remarkably accelerating post-1997 due to atmospheric circulation shifts and cryospheric melt.

Objective

• To quantify and explain the long-term historical changes in mountain-outlet river runoff across the Third Pole (TP) from a coherent, regional perspective, considering different climatic domains and the influence of large-scale atmospheric circulations and cryospheric changes.

Study Configuration

• Spatial Scale: Third Pole (TP) region, focusing on the mountain-outlet basins of 12 major rivers categorized into westerlies-dominated (Tarim, Shule, Heihe, Amu Darya, Syr Darya, Indus), Indian monsoon-dominated (Brahmaputra, Ganges, Mekong, Salween), and westerlies–monsoon transition (Yangtze, Yellow) domains.
• Temporal Scale: 1960–2016, with a specific analysis of changes before and after the turning year of 1997.

Methodology and Data

• Models used:
  • Cryosphere–hydrological models (VIC-glacier, WEB-DHM) for reconstructing natural runoff and components in specific basins (Amu Darya, Syr Darya, Brahmaputra, Yangtze).
  • Deep learning model: Long Short-Term Memory (LSTM) for reconstructing historical basin-scale glacier mass changes.
  • Volume-area scaling approach for updating glacier area and volume in cryosphere-hydrology models.
• Data sources:
  • Runoff: In-situ observations from discharge gauges (Ministry of Water Resources in China, Department of Hydrology and Meteorology in Nepal, Pakistan Water & Power Development Authority, Global Runoff Data Center, Scientific-Information Center of the Interstate Commission for Water Coordination in Central Asia). Linear statistical methods for Mekong and Salween.
  • Precipitation (P) & Evapotranspiration (ET): Ensemble-based products derived from multiple observation-constrained long-term datasets (e.g., GPCC, GPCP, CMAP, ERA5, JRA-55, PERSIANNCDR, Princeton Global Forcings for P; GLDAS, GLEAM, MaET for ET).
  • Air Temperature (tair): Ensemble-based product from GHCN_CAMS, ERA5-Land, and NCEP-NCAR reanalysis 1.
  • Glacier data: Randolph Glacier Inventory (RGI6.0) for glacier outlines and Rounce et al. (2020) for glacier mass balance projections.
  • Total Water Storage Change (ΔS): GRACECSR, GLDAS-catchment, and LiTWSA.
  • Climatic indices: Zonal index for the Northern Hemisphere and Indian monsoon index for June–September.

Main Results

• Divergent Runoff Trends (1960-2016): Mountain-outlet runoff showed contrasting trends. Westerlies-dominated rivers (e.g., Indus, Amu Darya, Tarim) experienced significant increases (e.g., Upper Indus: 22.7 ± 8.4 × 10^8 cubic meters per decade). Monsoon-dominated rivers (e.g., Ganges, Brahmaputra) showed insignificant declines.
• Acceleration Post-1997: The contrasting runoff changes accelerated markedly after 1997.
  • Westerlies-dominated rivers: The rate of total runoff increase nearly tripled (17.6 ± 8.5 millimeters per decade, or 6.4 ± 3.1 percent per decade) during 1997–2016, compared to 1960–2016.
  • Monsoon-dominated rivers: An abrupt decline in runoff (−50.1 ± 21.2 millimeters per decade, or −13.0 ± 5.5 percent per decade) was observed since 1997.
  • Westerlies–monsoon transition rivers (Yellow and Yangtze): Runoff reversed to an increasing trend (14.4 ± 11.9 millimeters per decade, or 11.4 ± 9.4 percent per decade) between 1997 and 2016.
• Driving Forces: Runoff changes are largely explained by varying precipitation (P) and evapotranspiration (ET).
  • Westerlies domain: Significant increases in P and ET led to an increase in P-ET, consistent with runoff increases. Post-1997, steeper P and ET increases contributed to the accelerated runoff growth.
  • Monsoon domain: An insignificant decrease in P and a significant increase in ET led to a decrease in P-ET. Post-1997, a sharp decrease in P combined with an insignificant ET increase resulted in a significant P-ET and runoff decline.
  • Total Water Storage Change (ΔS), including glacier melting and groundwater depletion, plays a non-negligible role, especially after 1997, in modulating freshwater availability and explaining discrepancies between runoff and P-ET trends.
• Runoff Components: For the Amu Darya (westerlies domain), natural runoff increase was attributed to significant increases in glacier melt (35.1%), rainfall runoff (34.6%), and groundwater flow (30.3%). For the Brahmaputra (monsoon domain), the simulated runoff decline (1997-2016) was mainly due to significant declines in rainfall runoff (72.5%) and snow melt (20.4%).

Contributions

• Provides a coherent, regional quantification of long-term historical runoff changes across the Third Pole, addressing previous knowledge gaps.
• Identifies and quantifies the remarkable acceleration of divergent runoff trends in westerlies- and monsoon-dominated regions after 1997.
• Attributes these changes to shifts in large-scale atmospheric circulations (enhanced westerlies, weakened Indian summer monsoon), the P-ET balance, and the non-negligible role of total water storage changes, including cryospheric melt.
• Highlights the critical implications for regional water, ecology, and food security, emphasizing the need for proactive adaptations.
• Employs a robust methodology combining extensive in-situ observations, advanced cryosphere-hydrological models, and deep learning techniques to overcome data scarcity in high-mountain regions.

Funding

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

Citation

@article{Wang2025Acceleration,
  author = {Wang, Lei and Li, Xiuping and Lutz, Arthur and Nepal, Santosh and Chen, Deliang and Yao, Tandong and Su, Fengge and Cuo, Lan and Yao, Zhijun and Zhang, Yinsheng and Hu, Zhidan and Huang, Jingheng and Hou, Mei and Liu, Ruishun and Long, Junshui and Chai, Chenhao and Liu, Zhaofei and Bashir, Ahmad and Khanal, Sonu and Sun, He and Nie, Yong and Zhang, Yongqiang and Wang, Tao},
  title = {Acceleration of diverging runoff trends on the Third Pole},
  journal = {Communications Earth & Environment},
  year = {2025},
  doi = {10.1038/s43247-025-02854-5},
  url = {https://doi.org/10.1038/s43247-025-02854-5}
}