What happens when water you built everything around turns out to be temporary?
The American dry side holds two versions of this question. Both involve communities that treated a water supply as permanent, a specific documented moment when the supply's limits entered the public record, and people living right now with what followed.
The mechanisms split. In southwest Kansas, farmers tapped the Ogallala with center-pivot irrigation starting in the 1950s and pumped it down while federal hydrologists published the drawdown numbers. The water didn't leave. It was extracted. The community that used it is the community that depleted it, and the reckoning belongs to the same hands that turned the valves.
At California's Salton Sea, the water was never supposed to be there. Agricultural runoff created the lake and sustained it for a century. When the 2003 Quantification Settlement Agreement began redirecting that runoff to coastal cities, the lake started shrinking, and the communities on its shore inherited the consequences of a landscape being abandoned by the political economy that accidentally made it.
Depletion and abandonment. Two forms of the same loss, separated by who holds the pump handle.
The Aquifer Curve
In January 1946, southwest Kansas had about 420 irrigation wells. By 1975, roughly 7,000. The machine that made that possible was specific enough to have a patent number.
Frank Zybach, a Nebraska farmer, filed for a patent on a self-propelled sprinkling apparatus in 1949 and received it in 1952. Robert Daugherty's Valley Manufacturing bought the rights by 1954 and started selling center-pivot systems that could irrigate terrain no flood setup could reach. The pivots spread across the High Plains from South Dakota to the Texas Panhandle through the 1960s. In western Kansas they changed the fundamental arithmetic of what the ground could produce. Corn became reliable. Corn supported feedlots. Feedlots supported packing plants. Packing plants supported towns. A regional economy assembled itself around pumped groundwater with the confidence of something that had never checked the bottom of the tank.
Haskell County sits in the thick of that transformation. The county reported $1.636 billion in crop and livestock sales in the 2022 USDA Census. Beef, feedlots, grain, oilseed. Every dollar downstream of the pivot. Every dollar dependent on water that was already declining before the irrigation boom turned twenty.
The record is specific, and it was never hidden. In 1974, USGS hydrologists Edwin Gutentag and Lloyd Stullken published Hydrologic Atlas 515, a county-level groundwater study of Haskell County mapping water-level declines from predevelopment through 1970. Average drop: about 15 feet. Some areas down more than 30. Groundwater appropriation applications had averaged 400 per year from 1953 to 1969. The development surge was barely two decades old, and the drawdown was already measured, mapped, and sitting in a government publication with the authors' names on it.
Seven years later, a joint KGS-USGS synthesis made the ratio explicit for all of southwest Kansas. Annual recharge to the main irrigable aquifer: approximately 210,000 acre-feet. Annual irrigation pumpage: 2.1 to 2.8 million acre-feet. Net withdrawal after return flow: 1.7 to 2.2 million.
For every acre-foot of water nature returned to the Ogallala, farmers in southwest Kansas pulled ten.
By 1975, about 14 million acre-feet had been removed from storage, with declines reaching 135 feet in the hardest-hit areas. Haskell, Finney, and Stanton counties had already dropped more than 50 feet.
The turning point is difficult to sit with precisely because it wasn't hidden. The 1981 report didn't reveal a secret. It quantified what anyone drilling wells in Haskell County could already feel in their pump pressure and their electricity bills. The aquifer curve was on paper, with maps, in a document anyone could request. But knowing the curve and acting on the curve are separated by everything that makes a life: the corn is growing. The feedlots are buying. The packing plant is hiring. The note on the combine comes due in November regardless of what the saturated thickness was last spring. Every individual decision made sense. The aggregate consequence was someone else's problem until it wasn't.
What people did, mostly, was keep pumping. Anyone who has written a report that got filed and forgotten understands the logic. The information existed. The institutional will to act on it did not, because acting on it meant shrinking the economy the information described.
The KGS index well in Haskell County now records seasonal drawdowns exceeding 120 feet during irrigation months. Smaller fluctuations since 2014 reflect court-enforced pumping reductions, not recovery. Across Groundwater Management District 3, which covers southwest Kansas, the average decline since predevelopment is 103 feet. Average remaining saturated thickness: 151 feet. Between 1996 and 2016, the district lost 1.69 feet per year. The farmers in GMD3 can do that math themselves. Some of them grew up doing it at the kitchen table with fathers who ran the wells.
Travis Leonard, a Haskell County farmer, shut down a well that had irrigated his family's grain field for more than sixty years. When he took over the farm sixteen years earlier, it had more than a dozen irrigation wells running. By 2023, he was down to three.
"We didn't have any idea when it was going to end, but that day is coming."
In December 2024, more than a hundred farmers gathered in Sublette, Haskell County's seat, to argue about water-resource futures. The state had required groundwater districts to submit action plans by July 2026, and the room was full of people trying to negotiate the terms of their own contraction. Farmer Bret Rooney told Kansas News Service that southwest Kansas "boomed through the development of the aquifer." The boom was the problem and the livelihood simultaneously. Nobody in that room could separate the two.
By February 2026, GMD3 had proposed stabilizing the aquifer over twenty years through a 27.7 percent reduction in use, cutting pumping 5 percent annually with an initial 10 percent cap on individual wells to limit economic hardship. The KGS estimates a 31 percent reduction would be needed just to achieve zero net water-level change. Not recovery. Not recharge. Just holding the line at 103 feet below where it started.
What Haskell County is negotiating now is not a fix. It is managed retreat from an economy built on a resource everyone involved knew was being mined. The 1974 atlas is still in the USGS archive. The 1981 synthesis is still on the KGS website. The curve was never a surprise. It was a trajectory that made sense at every individual decision point along the way, until the aggregate consequence arrived and the wells stopped producing enough water to justify the electricity it took to run them. The farmers are still there. The aquifer is still declining. The action plan is due this month.