Satellite-based Study Unveils Rapid Groundwater Depletion in High Mountain Asia, Threatening Water Security
23 Jan 2026
A recent satellite-based study published in Environmental Research Letters has revealed alarming declines in groundwater storage across High Mountain Asia (HMA), often called the "Asian Water Tower". This vital water source, which underpins agricultural irrigation, urban water supply, and ecological security for hundreds of millions of people in more than a dozen downstream countries, is depleting at an alarming rate of approximately 24.2 billion tonnes per year.
The research, led by Prof. WANG Shudong at the Aerospace Information Research Institute, Chinese Academy of Sciences (AIRCAS) addressed the challenges of data scarcity, and complex terrain in the region. The team developed a novel AI-powered assessment by integrating multi-source satellite observations, Earth system modeling, and explainable artificial intelligence. This allowed them to reconstruct two decades of groundwater storage (GWS) change, identified dominant drivers, and conducted forward-looking risk scenario analyses.
The study reveals that, from 2003 to 2020, roughly two-thirds of HMA experienced declines in groundwater storage. The most significant decreases occurred in densely populated, irrigation-intensive downstream basins such as the Ganges–Brahmaputra, Indus, and Amu Darya. Meanwhile, some high-elevation inland regions exhibited localized recovery. Climatic factors were responsible for nearly half of the observed GWS variability, with cryospheric playing a prominent role. However, human water withdrawals, especially in downstream irrigation zones, have increasingly contributed to the depletion of groundwater, a trend that accelerated after 2010.
Projections suggest that downward trend will continue under current water-use practices. While glacier melt may temporarily slow losses in some areas around the 2060s, this "buffer" is unsustainable, and a faster decline could follow. If water-use practices remain unchanged, a more rapid decline is anticipated, putting downstream agricultural zones at even greater risk.
The study used a knowledge-guided data-driven assessment framework that use multi-sensor satellite data to retrieve GWS changes over the past twenty years. The framework employs a lightweight Transformer architecture to capture hydrologic memory and lag effects in mountainous catchments. Explainable machine-learning techniques were also used to provide mechanistically consistent attribution of the inferred changes. The results were cross-validated against thousands of groundwater well measurements and independent datasets, enhancing the reliability of the conclusions.
Basin extent and hydrological characteristics across High Mountain Asia, and the lightweight deep-learning framework employed in this study. (Image by AIRCAS)
This innovative framework effectively integrates remote sensing with physical knowledge and interpretability, overcoming challenges posed by complex terrain and incomplete human water-use data. The study emphasizes the urgency of addressing the growing water-security crisis in HMA, calling for immediate actions such as water-saving measures, efficient water management, managed aquifer recharge, and ecosystem restoration to enhance the region's resilience to water shortages.
Groundwater storage changes and projected future trends in major basins across High Mountain Asia. (Image by AIRCAS)
The research was led by AIRCAS in collaboration with multiple domestic and international partners and was supported by the National Key R&D Program of China and the National Natural Science Foundation of China (NSFC) Key Program.
Research News
Satellite-based Study Unveils Rapid Groundwater Depletion in High Mountain Asia, Threatening Water Security
A recent satellite-based study published in Environmental Research Letters has revealed alarming declines in groundwater storage across High Mountain Asia (HMA), often called the "Asian Water Tower". This vital water source, which underpins agricultural irrigation, urban water supply, and ecological security for hundreds of millions of people in more than a dozen downstream countries, is depleting at an alarming rate of approximately 24.2 billion tonnes per year.
The research, led by Prof. WANG Shudong at the Aerospace Information Research Institute, Chinese Academy of Sciences (AIRCAS) addressed the challenges of data scarcity, and complex terrain in the region. The team developed a novel AI-powered assessment by integrating multi-source satellite observations, Earth system modeling, and explainable artificial intelligence. This allowed them to reconstruct two decades of groundwater storage (GWS) change, identified dominant drivers, and conducted forward-looking risk scenario analyses.
The study reveals that, from 2003 to 2020, roughly two-thirds of HMA experienced declines in groundwater storage. The most significant decreases occurred in densely populated, irrigation-intensive downstream basins such as the Ganges–Brahmaputra, Indus, and Amu Darya. Meanwhile, some high-elevation inland regions exhibited localized recovery. Climatic factors were responsible for nearly half of the observed GWS variability, with cryospheric playing a prominent role. However, human water withdrawals, especially in downstream irrigation zones, have increasingly contributed to the depletion of groundwater, a trend that accelerated after 2010.
Projections suggest that downward trend will continue under current water-use practices. While glacier melt may temporarily slow losses in some areas around the 2060s, this "buffer" is unsustainable, and a faster decline could follow. If water-use practices remain unchanged, a more rapid decline is anticipated, putting downstream agricultural zones at even greater risk.
The study used a knowledge-guided data-driven assessment framework that use multi-sensor satellite data to retrieve GWS changes over the past twenty years. The framework employs a lightweight Transformer architecture to capture hydrologic memory and lag effects in mountainous catchments. Explainable machine-learning techniques were also used to provide mechanistically consistent attribution of the inferred changes. The results were cross-validated against thousands of groundwater well measurements and independent datasets, enhancing the reliability of the conclusions.
Basin extent and hydrological characteristics across High Mountain Asia, and the lightweight deep-learning framework employed in this study. (Image by AIRCAS)
This innovative framework effectively integrates remote sensing with physical knowledge and interpretability, overcoming challenges posed by complex terrain and incomplete human water-use data. The study emphasizes the urgency of addressing the growing water-security crisis in HMA, calling for immediate actions such as water-saving measures, efficient water management, managed aquifer recharge, and ecosystem restoration to enhance the region's resilience to water shortages.
Groundwater storage changes and projected future trends in major basins across High Mountain Asia. (Image by AIRCAS)
The research was led by AIRCAS in collaboration with multiple domestic and international partners and was supported by the National Key R&D Program of China and the National Natural Science Foundation of China (NSFC) Key Program.