67 places where aquifers stopped shrinking - and what they did right

More than a third of the world's aquifers are declining. Groundwater supplies drinking water to half the global population and feeds 40% of irrigated agriculture. The bad-news cases far outnumber the good ones.

But there are good ones. And Scott Jasechko, a professor at UC Santa Barbara's Bren School of Environmental Science and Management, wanted to know what they have in common. His study, published March 20 in Science, examines 67 cases where depleted aquifers actually recovered - and finds that the successful places almost never relied on a single fix.

Beijing's aquifer: from 20-meter drops to rising springs

Beijing offers one of the most dramatic examples. Between 1950 and 2000, groundwater pumping caused water tables around the Chinese capital to plummet by more than 20 meters in some areas. A megacity was draining its own foundation.

The turnaround involved multiple interventions layered on top of each other. Starting in 2003, the government built canals and pumping stations to deliver water from wetter regions in the south. The city simultaneously expanded its use of reclaimed water for environmental purposes - watering trees and grasslands, replenishing lakes and rivers. And after the southern water deliveries began, authorities banned pumping from deep confined aquifers for industrial use.

Both the shallow and deep aquifers began recovering. Land subsidence rates slowed. Springs that had dried up started flowing again. Irrigated agriculture remained productive, and its sustainability was no longer threatened by falling water tables.

Beijing used all three of the broad strategy categories that Jasechko identified in his analysis: alternative water sources, policy interventions, and artificial aquifer recharge. That combination was not unusual. Two-thirds of the 67 success cases involved interventions from multiple categories.

Green Bay's cautionary pipeline

But recovery is not permanent. Green Bay, Wisconsin, provides the counterpoint. In 1957, the city built a 43-kilometer pipeline to supplement groundwater with water from Lake Michigan. The stressed aquifer recovered - for a while. Then additional demand sent it falling again for decades. In 2006, Green Bay built a second pipeline, this one 100 kilometers long, to bring more Great Lakes water. The aquifer is recovering again.

The Green Bay story illustrates one of Jasechko's key themes: improvements are not guaranteed to last. Conditions change. Demand grows. Climate shifts. An intervention that was sufficient in 1957 may not be sufficient in 2006, and what works in 2006 may need reinforcement by 2030.

The menu of strategies

Jasechko sorted the interventions across all 67 cases into three broad categories, then examined their frequency and combinations.

Alternative water sources appeared in 81% of cases - the most common strategy by a wide margin. The appeal is straightforward: if another water source is available, tapping it reduces groundwater demand without requiring anyone to use less water overall. But Jasechko noted the drawbacks. Alternative sources are often expensive to develop, and they can simply move the depletion problem to another location. Importing water from a distant river basin helps the receiving aquifer but may stress the source.

Policy and environmental markets benefit from low costs and directly target the behaviors that caused depletion. Pumping restrictions, water pricing reforms, and trading schemes can reduce demand without new infrastructure. The tradeoff is economic disruption. Communities that built their economies around cheap groundwater face significant adjustment costs when that access is curtailed.

Artificial recharge - deliberately channeling water into aquifers through injection wells or infiltration basins - can avoid the need to reduce pumping. This is attractive for communities whose economies depend on groundwater, but the water has to come from somewhere, and getting it underground requires energy and infrastructure.

Ten patterns across 67 recoveries

Jasechko distilled his analysis into ten recurring themes. Multi-pronged strategies outperformed single interventions. Reducing pumping consistently helped. Good policy design meant little without enforcement. Recovery timelines varied dramatically - some aquifers bounced back in years, others needed decades. Where recovery was slow, gradual policy phase-ins were more politically sustainable than abrupt restrictions.

Recovery was also uneven geographically. Within a single region, some areas might recover while others continued to decline, depending on local geology, pumping patterns, and proximity to recharge sources. This spatial variability complicates assessment: declaring an aquifer 'recovered' based on monitoring wells in one location can mask ongoing depletion nearby.

Two of Jasechko's themes carry particular weight. The eighth - that improvements can easily reverse - argues against complacency. And the ninth - that groundwater quality matters alongside quantity - warns against a narrow focus on water levels. An aquifer that recovers its volume but becomes contaminated by agricultural runoff or saltwater intrusion has not truly recovered.

What the analysis does not prove

Jasechko was forthright about the study's boundaries. This is a review of published case studies, not a controlled experiment. The 67 cases were drawn from scientific literature, which likely overrepresents places where English-language research has been published. Successful recoveries in regions without robust groundwater monitoring or academic research traditions are probably underrepresented.

The study does not trace interventions forward from implementation to outcomes. It identifies places where aquifers recovered and catalogs what was done, but it cannot isolate which specific intervention was responsible for how much of the recovery. Many cases involved simultaneous actions, making attribution difficult.

There is no guarantee that strategies that worked in one place will work in another. Geology, climate, governance structures, economic conditions, and water law all vary enormously between regions. The case studies provide a menu of options, not a recipe.

Jasechko is working with UCSB colleague Debra Perrone toward a more systematic database that would catalog all places where interventions were implemented and track groundwater levels before and after. That kind of dataset would allow correlative analysis - connecting specific actions to measurable outcomes - which the current review cannot do.

A menu longer than expected

In 2024, Jasechko and colleagues compiled the largest global assessment of groundwater levels, spanning nearly 1,700 aquifers. That study presented a daunting picture: accelerating declines across much of the world. The current paper is, in a sense, the constructive follow-up - a survey of what has actually worked.

The findings are cautiously encouraging. Recovery is possible. It has happened in 67 documented cases across different climates, governance systems, and economic contexts. But it requires sustained effort, usually multiple types of intervention, and ongoing monitoring to ensure that gains are not erased by changing conditions.

For groundwater managers and policymakers facing depleted aquifers, the practical value is in the breadth of examples. The menu of strategies, as Jasechko describes it, is longer than he originally anticipated. That does not make the problem easy. But it makes the claim that nothing can be done harder to sustain.