Earth losing fresh water faster than thought, satellite data shows

Full text

Earth's continents are losing fresh water much faster than scientists thought, mostly because aquifer pumping in regions like northern India, southern Iran and parts of the U.S. is 45% more pronounced, on average, than previous estimates suggested, new research shows.

The study used satellite data to map freshwater trends across the globe between 2002 and 2025. Thanks to a high-resolution rendering method, researchers revealed regional gains and reductions in terrestrial water storage that were 33% more marked, on average, than in previous studies. The amount of fresh water lost was much higher than the gains overall, highlighting the urgent need for better water management to stave off a global water crisis.

"It has become very dire in a lot of places," Hrishikesh Chandanpurkar , a hydrologist and the co-founder of the nonprofit Evergreen Resilience Institute, who was not involved in the study, told Live Science.

Researchers created a map of terrestrial freshwater trends from 2002 to 2025. They then attributed these trends to human and climate drivers. (Image credit: O'Neill, Rodell and Loomis, 2026. PNAS . (CC BY 4.0) .) The study also attributed the changes in freshwater storage to different human and climate drivers, said first author Mary Michael Forrester O'Neill , a researcher at the University of Maryland's Earth System Science Interdisciplinary Center and the Hydrological Sciences Laboratory at NASA's Goddard Space Flight Center.

"Earlier studies could see broad patterns of long-term freshwater loss or gain but couldn't reliably tell how much was due to human activity versus climate swings, in part due to resolution limitations," Forrester O'Neill told Live Science in an email. "This study sharpened the picture and used geostatistical methods to rank the impacts of human causes, like irrigation and dam building, against climate factors, like rainfall and drought."

The researchers found 94 regions with above-background shifts in freshwater storage, 40 of which were influenced mainly by human land or water use, rather than climate variability and global warming . The team then focused on these 40 hotspots to pinpoint exactly where people can act to conserve water.

"Water managers and downstream communities need to know which aquifer, reservoir, or irrigation district is responsible for a given trend in large-scale freshwater storage, not just that 'somewhere in this country, water is being lost,'" Forrester O'Neill explained.

Freshwater gains in 22 regions were due to rain-fed agriculture, irrigation with surface water, reservoir impoundment and deforestation, which decreases evapotranspiration and plants' groundwater consumption, according to the study. In contrast, freshwater losses in the remaining 18 hotspots were caused by irrigation with groundwater and the construction of canals, diversions and other water infrastructure.

The findings, published Sept. 14 in the journal PNAS , show that human water use is a significant driver of water storage changes on every continent except Australia and Antarctica , where climate factors are more influential, O'Neill said. Local freshwater gains and losses may cut freshwater availability downstream or in the surrounding region, she and her colleagues noted in the study.

One example of a region that gained fresh water in the study is the area of central Russia that hosts the Boguchany Reservoir, a giant artificial lake that was filled in 2012 after the construction of a hydroelectric dam. And three stark examples of places that lost fresh water are northern India, southern Iran and the southern High Plains (a region at the intersection of Colorado, Kansas, New Mexico, Oklahoma and Texas), because these regions host major breadbaskets that tap aquifers for irrigation, O'Neill said.

The results are concerning because the places with the worst groundwater depletion overlap with areas that irrigate crops unsustainably, O'Neill said. Previous research shows that critical food-producing regions ‪—‬ such as the U.S. High Plains, California's Central Valley, northern India, Pakistan, North China and the Fertile Crescent ‪—‬ consistently overuse groundwater. "Within these regions, we found the pace of depletion is considerably more severe than earlier estimates suggested, reinforcing food security concerns," she said.

One of the novelties of the paper is its very high resolution, said Chandanpurkar, who last year led a study describing "mega-drying" regions across the world and co-authored a report on continental drying . The new study used data from NASA's Gravity Recovery and Climate Experiment (GRACE) mission and its Follow-On successor (GRACE-FO), which have measured subtle redistributions of mass on Earth. But unlike previous studies, the new research resolved the data in high resolution, he said, like correcting blurry vision with glasses.

"This is not a brand-new method, but it's the first time they're applying it to these regional hotspots and then trying to figure out what they mean," Chandanpurkar said.

The new study (A) produced a much sharper picture of freshwater trends than standard processing methods and previous studies did (B). (Image credit: O'Neill, Rodell and Loomis, 2026. PNAS . (CC BY 4.0) .) Scientists already knew that groundwater depletion and other human water uses are "quite staggering," and this study is a reminder that continental drying is "a very big deal," Chandanpurkar said. "Water is really undervalued," he said. "Giving it its proper dues would be essential for this [freshwater loss] to really be acknowledged by leaders and economists."

It's impossible to say exactly how many years or decades humans can go on consuming this much fresh water before it runs out , partly because the climate can be unpredictable and partly because it's hard to tell how much water remains in Earth's reservoirs, Forrester O'Neill said.

"GRACE satellites measure the rate of change in water mass, not the total volume stored in an aquifer, so we have no independent basis for estimating time to depletion or a timeline for systems to collapse," she said.

RELATED STORIES

'An increasing attack on water resources from multiple fronts': Scientists warn 'day zero droughts' could hit before 2030

Kabul could become the first modern capital to run out of water — here's why

'If this is the new normal, then we are in very deep trouble': Brazil's drying aquifers could have ripple effects across the world

However, there will come a point when the cost of pumping groundwater, for example, will be too high to pay in some regions. Water table declines of just a dozen feet (a few meters) or so can cause a substantial share of wells to run dry, research shows , and the expense of drilling deeper will affect at-risk communities most of all, Forrester O'Neill said.

"We are most concerned about the rising cost and inequity of continuing to reach water that keeps retreating deeper underground, paired with strain on existing surface water resources," she said. "Of course, water-scarce regions exhibiting acute groundwater depletion are at risk, like Iran , Saudi Arabia, Syria, Türkiye, Iraq, and Libya. Another one to keep an eye on would be Mexico , simply because of the documented, progressively increasing well depth paired with the observed terrestrial water storage loss."