A NASA-led analysis using over two decades of gravity satellite data has identified regions where human activity has altered terrestrial water storage more significantly than climate forces alone. The study, published in the Proceedings of the National Academy of Sciences, leveraged measurements from the GRACE and GRACE-FO missions to detect subtle shifts in Earth's gravitational field caused by changes in water mass across continents. These shifts, tracked from April 2002 to November 2025, reveal patterns of water gain and loss tied to irrigation, reservoir construction, and land-use modifications.

The research highlights 40 areas where anthropogenic factors correlated more strongly with water storage trends than climatic variables, spanning all nonpolar continents except Australia. In California’s Central Valley and northern India, intensive groundwater pumping for agriculture has driven measurable declines in terrestrial water storage, with the latter showing a steeper downward trajectory than prior estimates. Conversely, in parts of sub-Saharan Africa and northern Brazil, expanded rainfed cropland and shifts from forest to soybean cultivation correlated with localized increases in water storage, potentially due to agricultural practices that enhance groundwater recharge.

Hydrological engineering projects also emerged as focal points. The Boguchany Reservoir in Russia’s Angara River basin, one of the largest reservoirs filled this century, exhibited the strongest storage trend in nonpolar northern Asia. Smaller-scale interventions, such as the Karakum Canal in Turkmenistan, demonstrated detectable impacts on desert water tables through seepage from unlined canals. These findings underscore the capacity of high-resolution satellite data to resolve localized hydrological changes previously obscured by coarse spatial coverage.

The study’s methodology combined statistical weighting of anthropogenic and climatic influences with enhanced spatial detail, enabling detection of human signatures at scales five to ten times finer than earlier analyses. While temporal resolution was limited, the approach allows for tracking of storage trends in response to events like droughts or policy shifts. Researchers aim to refine this technique further to disentangle local from non-local impacts of water management and to integrate high-resolution data with standard monthly products for improved monitoring. The work provides a quantitative foundation for assessing how land and water management practices interact with climate dynamics in a changing hydrological cycle.