In 2001, the Bureau of Reclamation (USBR), a federal agency that manages the agricultural water supply within the basin, began paying farmers to pump large amounts of groundwater out of the aquifers below. The idea was supposed to temporarily help producers access underground reserves whenever there was a run on surface water, referring to lakes, rivers, and other above-ground sources. Surface water has historically been the primary water source for many farmers in the region. However, as droughts persisted year after year, the pay-to-pump program gradually became the norm rather than an exception. Since 2001, such programs have now been activated in more years than not—even as the agency’s own scientists warned that water was getting pulled out of the earth faster than it could be replenished. 

“The prevailing policy for the last 20 years has been to just try to shift surface water users temporarily, inexpensively to groundwater,” said Jim McCarthy, southern Oregon program director for Oregon WaterWatch, an organization that advocates for sustainable management of water resources. “The consequences of this policy have borne some pretty bitter fruit.”

For most of the 20th century, USBR’s primary obligation within the Klamath Project was to provide surface water to farmers. The agency used Upper Klamath Lake in the northern part of the basin as the Project’s primary reservoir. It fed into the Klamath River, from which a system of canals then diverted water into the deep, interconnected irrigation ditches that run lateral to almost every parcel of farmland, flowing from acre to acre and quenching the thirst of the alfalfa, potatoes, and cattle that have long comprised some the area’s biggest agricultural exports.

Then came new conservation regulations and litigation that complicated USBR’s water obligations. In the 1980s and ‘90s, three salmon and suckerfish species that call the Klamath Basin home were listed under the Endangered Species Act, a move that required USBR to limit surface water for farming during droughts to protect the fishes’ habitats. Around the same time, the Klamath Tribes regained federal recognition and a series of court decisions reaffirmed their water rights in the region.

Competing interests over the use of surface water clashed during a particularly severe drought in 2001, when USBR for the first time curtailed the release of almost all water for irrigation to preserve adequate water levels for fish in the Upper Klamath Lake and Klamath River. The decision sparked outrage among members of Project communities and set off a series of high-profile protests, including threats to force open irrigation canal gates.

In an attempt to prevent future conflicts, USBR that year began experimenting with what it called a “water bank,” which it hoped would reduce demand for surface water during dry years. The idea was two-pronged: First, the water bank would pay some farmers not to farm, or to farm without water, through a land-idling program. And second, it would pay other farmers to pump groundwater and use that to irrigate instead. The water bank would effectively increase surface water supplies everywhere else, the thinking went.

The initiative was only supposed to be a short-term solution. Even in the water bank’s early years, USBR was aware that paying farmers to pump too much groundwater could be ecologically damaging in the long-run.

“The impact of groundwater pumping on basin aquifers during ongoing drought conditions is largely unknown and continued reliance may not be sustainable,” cautioned a report on the water bank published by the Government Accountability Office (GAO) in 2005.

A decade later, USBR’s own scientists sounded the alarm on the risks that the pay-to-pump program posed to the upper Klamath Basin, including to the very fisheries that the agency was mandated to protect.

“The unsustainable groundwater pumping that has occurred under the [water bank] reduced the total amount of water in the Klamath Basin, to the detriment of fish and wildlife, including endangered species,” wrote USBR scientists in a 2015 whistleblower complaint, in which they alleged that the water bank had misspent taxpayer funds. (USBR and its parent agency, the Department of the Interior, did not respond to a detailed list of questions from The Counter. In a general emailed statement, a DOI spokesperson wrote: “In response to 20 years of water supply challenges in the Klamath Basin, Congress authorized Reclamation to fund drought response measures in the Klamath Project area.”)

The researchers’ concerns about sustainability were backed by growing scientific evidence: In 2012, federal and state researchers established that groundwater and surface water were interconnected in the Klamath Basin. Specifically, they found that when too much groundwater is pumped in certain areas, the resulting overdraft can impact springs and streams, indirectly reducing the amount of surface water available. In other words, overreliance on groundwater could potentially contradict the water bank’s stated aim: conserving surface water.

If USBR was attuned to these warnings, it wasn’t reflected in policy. As one dry year rolled into the next—Oregon governors have issued 13 drought declarations for Klamath County in the past 20 years—the water bank quickly evolved from an interim measure to the new status quo. Since 2001, it has paid farmers to pump groundwater approximately two out of every three years. In multiple instances, the program incentivized the extraction of volumes much higher than the average pumping amounts recorded prior to its implementation.

For example, during a drought in 2010, the water bank paid farmers to pump approximately 101,000 acre-feet of groundwater, according to a summary report on the program. That equals a more than three-fold increase relative to the average pumping volume in the Project before the water bank was established, which the U.S. Geological Survey estimated to equal roughly 28,600 acre-feet per year. (An acre-foot is the amount of water it takes to cover an acre of land with a foot of water.) Subsequently, multiple Project towns saw water levels sink to their lowest levels on record at the time.

Still, USBR continued to fund groundwater pumping. In 2020, the program paid farmers to pump an estimated 96,920 acre-feet of water—again tripling the area’s historical pumping average and pushing water levels even further away from the surface in many Project areas.

Exactly how much the water bank compensated farmers for groundwater pumping varied widely from year to year, in part because there has been little consistency in who is actually running the show. ​​In its first few years, USBR spearheaded the water bank itself; then, in 2010, it handed the reins over to an organization run by Project farmers that later dissolved following the aforementioned whistleblower complaint. Since 2018, a new group with many of the same members and staff called the Klamath Project Drought Response Agency (KPDRA), has been administering the water bank’s land idling and groundwater pumping programs, subject to USBR funding and approval.

The general idea behind the groundwater pumping program is that the federal government, rather than farmers, bear the costs associated with extraction—and then some. Most years, the program has provided not only a full reimbursement of irrigators’ power bills, but also an additional stipend that increased in value the more that one pumped. A KPDRA attorney said that the stipend was supposed to cover “operating costs, maintenance and depreciation,” but acknowledged that it did not verify whether or not the stipend was actually spent on such expenses.

The exact formula for calculating the stipend’s value was also inconsistent. In some years, for example, the water bank paid farmers an extra $10 to $20 for every acre-foot pumped. By contrast, in 2018, the water bank reimbursed farmers 160 percent of the cost of their pumping power bills—meaning that more than a third of the money expended through the groundwater program that year was paid out as no-strings-attached stipends.

This year, the water bank chose not to fund a dedicated groundwater program. KPDRA’s attorney declined to elaborate on why, but the group’s meeting minutes suggest that it ran into difficulty meeting new groundwater monitoring requirements. Nonetheless, it’s still reimbursing pumping costs in a roundabout way: With $15 million in CARES Act Covid-19 relief funds, KPDRA is distributing a flat per-acre payment to all Klamath Project farmers, which it said in its program description could help shoulder groundwater-related expenses. KPDRA’s attorney for the water bank said that the exact per-acre payment amount has not yet been determined.

To date, the federal government has spent at least $105 million on Klamath water bank programs, The Counter found. USBR has not responded to requests for public records detailing how much of that was spent on groundwater pumping specifically.

Even those wary of the groundwater pumping program acknowledge that it may have been a helpful means of managing limited water resources on a temporary basis. But the program turned out to be far from temporary in nature.

“A water bank might be helpful in the short term,” said Keith Schultz, a former USBR fisheries biologist and one of the whistleblowers who warned the agency about the unsustainability of the program. “But year after year it’s self-defeating.”

A key part of the issue, say hydrologists, is that USBR’s water bank never actually considered how it might help restore aquifers in non-pumping years—the way a person might consider putting money back into a bank account after making a large withdrawal. Instead, for the past two decades, the federal government has treated groundwater as a limitless resource that it financially incentivized farmers to drain with little apparent thought to long-term consequences.

“In my opinion, the use of the term ‘water bank’ for Reclamation’s groundwater pumping program is a misnomer,” wrote Marshall Gannett, a retired hydrologist who studied the Klamath Basin, in an email. “The term ‘water bank’ when used for groundwater implies that the water pumped was previously added to the aquifer system (for example through artificial recharge) or that it will be replenished at a later time by either artificial recharge or reduction in normal pumping in wet years. None of this was done or, as far as I know, contemplated. Simply pumping groundwater with no provision for augmenting the aquifer’s recovery is not really a water bank.”

Documenting the disappearance of groundwater

Around 10 a.m. on a sunny morning this November, Michael Thoma, a hydrogeologist for the Oregon Water Resources Department (OWRD), was unspooling a black wire into the ground, via the mouth of a six-foot tall metal cylinder in the middle of an empty parking lot. We were at Bonanza Big Springs Park in Bonanza, a town located near the southern end of the county. The park’s namesake springs were nowhere to be seen. Instead, the ditch where groundwater would typically seep out of the earth—and then eventually flow into the Lost River a few steps away—was dry and overrun with grasses and weeds.

The state of Oregon monitors the cyclical rise and fall of groundwater using what is called a hydrograph—a chart that tracks the water level over time, as measured by its distance from the surface of the ground. In a healthy aquifer, springtime water levels return to the same peak year after year. That’s not what’s happening in many parts of Klamath County.

Instead, hydrographs taken from observation wells across the region exhibit what Thoma described as a “stair-step” pattern: Groundwater levels may recover between fall and spring, but they trend downward on a year-to-year basis.

When pumping is kept to a minimum, water levels return to their usual highs every year, Thoma said. “But if you pump a little too much, then the water level is not quite going to come back to where it was the previous year. If you do that again the following year and the following year—in any one year, you’re not really taking out that much more water, but over the course of several years, in this case, since 2001, you’re seeing a significant stair-stepping in the high-pumping, or drought years, and barely enough recovery in the low-pumping, or non-drought, years.”

There’s a noticeable correlation between periods of water bank-supported pumping and steep water level declines where pumping occurred. In 2010—a year when the water bank funded heavy pumping—water levels in Klamath Valley wells fell by an average of 20 feet between spring and fall, according to a U.S. Geological Survey report. By contrast, the next year, when there was no pay-to-pump program, water levels actually rose an average of three feet within the same time frame. Water levels then dipped again in 2012, when the water bank was reactivated.

While the federal water bank incentivized overpumping of the aquifer, the responsibility of protecting groundwater resources actually falls on the state. In past years, OWRD has urged Klamath Project farmers to reduce the amount of groundwater they were pumping through the water bank program, though the agency has not issued official orders limiting use. In drought years especially, the state agency has had to walk a tightrope: When the federal government restricts surface water use, farmers find themselves particularly dependent on groundwater to stay afloat. But without limits to pumping, and with USBR incentivizing the practice, the resource could become more scarce in years to come. 

“Currently, there are no mandated curtailments of groundwater use within the Klamath Project area,” wrote Bryn Hudson, water policy analyst for OWRD, in an email to The Counter. “However, the Department has communicated to groundwater users since 2010 that the supplemental irrigation pumping from the aquifer system, when surface water supplies are short, is not sustainable.”

A key part of the issue here is that once a business already has a groundwater pumping permit in hand—a permit that may have been issued years or even decades before the current, ongoing drought—there’s little that the state can do to retroactively limit the amount used. In order to actually mandate curtailments, OWRD has to first designate a region as a “critical groundwater area.” And the reason that hasn’t happened yet is mundane: Hudson wrote that while the agency intends to do so at some point in the future, “no timeframe has been established for the designation due to other workload priorities.” 

It’s important to note that water levels are not solely influenced by groundwater pumping: Hotter weather, a consequence of our ongoing climate crisis, can increase the amount of water lost to evaporation, for example. Altitude, relative proximity to lakes and rivers, and the type of sediment below ground can also affect how and where exactly water moves during times of scarcity.

It also bears mentioning that what happens at any single observation well is not indicative of what happens across the region as a whole. For example, near Malin, Oregon, a Klamath Project town close to the state’s southern border, the disappearance of groundwater is particularly stark: The water level in one observation well fell to a depth of 217 feet below the surface this September, the lowest point ever recorded. That depth is 20 feet lower than the level recorded the same time last year, and 65 feet lower than levels recorded in 2000, the first year that OWRD began tracking levels at this well.

Thoma suspects that local groundwater pumping is not the sole reason for this decline. A stone’s throw across State Line Road separating Oregon from California are a series of deep, powerful irrigation wells that reach up to 2,380 feet into the ground. (Infamously, California does not currently regulate groundwater.) These high-volume pumps may contribute to a phenomenon known as “interference,” where water levels in one well drop significantly because of heavy extraction nearby.

On the other hand, wells located near irrigation ditches may actually benefit from leakage from those sources. As residential wells went dry this summer, agricultural groups in the Klamath Project have argued that diverting surface water to the Klamath Project could have dampened the severity of the crisis.

“The practice of retaining all this water in Upper Klamath Lake and […] Klamath River is really the cause of the groundwater challenges in the Klamath project area,” said Paul Simmons, executive director of the Klamath Water Users Association (KWUA), an organization representing Project farmers. “It’s not unusual for wells to be located adjacent to structures like canals that carry water normally in the summer, so they’re a sort of source of supply to groundwater.” It’s possible that this can, in turn, pose a health risk: Irrigation water in general is not recommended for human consumption, as it is often heavily polluted with bacteria and agricultural runoff.

Thoma acknowledged that irrigation canals can help replenish groundwater reserves nearby, but such sources alone may not have completely prevented this year’s emergency. 

“There are certainly shallow wells that are near canals that are benefiting from water from irrigation canals,” Thoma said. “But there’s also wells that we’ve measured this year that are 100 feet up the side of the hill, well above where the canals [are]—they’re dry. There’s some that are not really near any canals, they’re among a bunch of irrigation wells that are pumping.”

In other words, certain wells might get some benefit from leaky canals based on their location, but we probably can’t expect surface water diversions alone to help the aquifer recover—even over an extended period of time. After all, a large portion of the water used for farming doesn’t actually stay in the ground that it’s applied to: It’s consumed by the cattle that are then sent to meatpackers in California. It’s taken up by the alfalfa that is then cut and shipped in bales to ranches across the country or overseas. It’s absorbed by the onions and potatoes that are then harvested from the ground, shipped out of the Project, and distributed through our food system.

How groundwater shortages squeeze Klamath Basin communities

Worsening water scarcity in the Klamath Basin has put its communities in a challenging bind.

“Since [2001] you’ve had some kind of water shortage on average about every other year that’s gotten seismically worse in the last couple years,” said Ty Kliewer, a cattle rancher in Klamath Falls.

Unlike some Project farmers, Kliewer doesn’t have an irrigation well, meaning that when surface water is curtailed, his domestic well is the only source of water for his cows and his home, as well as the small beer company, Skyline Brewery, that he started onsite. The sharp year-to-year swings in water availability have taken a toll over the past two decades.

“A lot of the people here right now feel like their place and their purpose has totally been ripped away from them,” he said.

Kliewer said that the Klamath water bank program has been a lifeline keeping his business in the black during dry years. Under non-drought circumstances, for example, he would typically grow his own hay, and sell a good portion of it. This year, he had to pay out of pocket to buy 400 tons. When I visited his farm in November, he was mixing spent grains from the brewery into the hay he’d purchased to feed his cows. He’s participating in the water bank’s land idling program, and expects that it will help him to recover two-thirds of his hay costs.

For Klamath Project towns, water scarcity also poses an existential threat. Take the city of Malin, for instance. According to public works director Rob Grounds, the primary well serving the city’s residents fell by 70 feet in the summer. Malin had to set up an emergency water distribution center for families living on the outskirts of its boundaries, for whom a record number of wells went dry. Grounds is now exploring the idea of deepening the city’s main well by 200 feet in order to keep the resource dependable in the years to come. Right now, he estimates that deepening the primary well may cost around $500,000. That’s if he can even get a driller on the phone—waitlists for well services are notoriously long in the region right now. Nonetheless, Grounds sees groundwater pumping for agriculture as a necessary cost of doing business, one that keeps the community at large afloat.

“If they stop pumping ag wells, little towns like Malin won’t survive,” said Grounds. “Ninety percent of our people work in agriculture. If agriculture ceases to exist, so does Malin.”

But pumping also comes with costs that can’t be measured in dollars. How do you appraise the value of keeping water stored in the aquifer and flowing through the region’s tributaries to ensure the survival of threatened species, for example? Tribes and environmental groups alike have argued that limits on groundwater pumping should be a key element in a broader effort to protect natural resources across the Basin.

“Groundwater mining is having an impact on the whole system,” said Don Gentry, chairman of the Klamath Tribes. “There needs to be a real look at what’s changed and what’s sustainable.”

The Klamath Tribes’ former reservation boundaries fall a few miles north of the Klamath Project, but within the bounds of the upper Klamath Basin as a whole. Gentry expressed empathy for farmers and ranchers who were struggling with water scarcity, noting that many members of the Tribal community are involved with agriculture, as well.

“There’s a mismatch between federal and state when it comes to water rights and protections,” said Brad Parrish, water rights specialist for the Klamath Tribes. “We have federal water rights that aren’t being protected properly under state law.”

It’s a regulatory gap that has ramifications for the entire region. Broadly speaking, if the streams that feed into Upper Klamath Lake are diminished because of groundwater pumping, the volume of water available for everyone else downstream shrinks even further. And if lake levels fall beyond a certain threshold in the spring—as was the case this year—surface water for Project farmers may get curtailed again, further exacerbating the Project’s reliance on groundwater, and setting off yet another cycle of pumping that risks pushing water levels further out of reach.

What does a solution look like?

Across the West, communities are taking drastic measures to secure water resources amid dwindling supplies.

In Arizona, one suburban development is trying to purchase 2,000 acre-feet of water from a farm 260 miles away, a move that has sparked concerns over the precedent it could set. In southern California, one water utility has confronted falling water levels in the Colorado River by purchasing tens of thousands of acres of farmland in the nearby Palo Verde Irrigation District and letting a portion of it fallow every year. In most Western states, including in Oregon, rights to water are determined on a seniority basis. The utility district’s strategy specifically targets land with higher-priority water rights. In doing so, it can add up to 110,000 acre-feet of water to its own supply annually, according to a September board meeting bulletin.

In the Klamath Basin, the Bureau of Reclamation has recognized that the gap between agricultural demand for water and its availability is likely to widen with time.

“Agricultural irrigation uses about 98 percent of the total human consumptive water use in the Klamath River Basin,” USBR wrote in a 2016 technical assessment of the region, which includes both Project and non-Project farming. Under the Basin’s average projected climate change scenario, the agency estimated that demand for water will likely increase by 18 percent in the next five decades. In the same timeframe, it expects spring snowpack—a crucial source of both surface and groundwater—to fall sharply, by around 62 percent.

But farmers, an important political constituency, bristle at the idea of retiring farmland to reduce demand for water.

“Agriculture is a major economic engine in our basin,” wrote Tricia Hill, a Project farmer and former president of KWUA in a recent post on the organization’s website. (Hill declined a request for an on-the-record interview.) “Money that a farmer receives for a crop is passed through to employees, seed growers, equipment dealers, and many others. All of our employees buy cars, clothes and groceries. They go to restaurants. Hire plumbers. Pay the neighbor’s kid to mow their lawns.”

Environmental experts acknowledge that cutting back on water use will likely be painful. However, they say, it’s also unavoidable given what’s happening on the ground.

“It’s this elephant in the living room that we’re all kind of walking around pretending not to notice,” said Steve Pedery, conservation director for Oregon Wild, an organization that advocates for wildland protection. 

For the past two decades, Klamath water bank programs have helped Project farmers respond to droughts on a year-to-year basis. But funding land idling and groundwater pumping ad-hoc only gets more expensive every year, Pedery said. It also represents years of missed opportunities to invest in permanent reductions in water demand.

“It’s been 20 years since the original water crisis and we’re still doing it,” he said. “It certainly feels like that money could be better spent on solutions that are actually enduring.”

Two decades ago, Congress came close to funding one such solution. In discussions for the 2002 Farm Bill, the Senate passed an amendment that would have created a Klamath Basin Task Force, funded to the tune of $175 million, to explore long-term water conservation. This would have opened the door to voluntary agreements, in which Project farmers were paid to permanently scale back water use. KWUA, the group representing Project farmers, lobbied then-Representative Greg Walden to kill the amendment, much to the dismay of conservationists and even a handful of Project farmers who’d supported it. One Project farmer reported receiving death threats from fellow irrigators after disagreeing with KWUA’s position. (In response to a request for comment, Walden wrote in an email: “While I supported selected land acquisition and water banking and conservation efforts, I did not support shutting down the farm economy in the basin through massive federal takeover of farms and ranches, as I and my constituents believed the Senate proposal would have done.”)

That hasn’t stopped some from forging ahead on efforts to reduce agricultural demand for water in the region.

The other option, pumping for groundwater when surface water is restricted, no longer seems as viable as it did in 2001.

“Every time you pump more from your well in a basin where the groundwater and surface water are connected […] you’re making the next drought worse,” said McCarthy, of Oregon WaterWatch. “So you’re digging yourself into a bigger and bigger hole.”

Just look at the plight of hundreds of residents of Klamath County, who continue to bear the brunt of disappearing groundwater. In an October letter to the Oregon Emergency Board, one resident wrote to legislators complaining that their household well had yet to recover. And because their home was heated with a water-based heat pipe, the resident said that they actually had no heat going into the winter, an issue temporarily addressed through the use of a fireplace insert.

Whether the federal government will fund future groundwater pumping programs, or whether the state of Oregon will put limits on the practice remain to be seen—but the fate of the rest of the community might depend on it.

“If you know people don’t have water, every day is misery,” said Goettl, the regional emergency coordinator for the state. “It’s in everybody’s interest to be able to better stabilize the water system in general.”

Goettl is hoping that this summer’s emergency doesn’t play out again next year, but he’s not holding his breath. He’s got an additional 118 emergency water tanks in storage just in case.

This piece is part of a collaboration that includes the Institute for Nonprofit News (INN), California Health Report, Center for Collaborative Investigative Journalism, Circle of Blue, Colorado Public Radio, Columbia Insight, The Counter, High Country News, New Mexico In Depth and SJV Water. The project was made possible by a grant from the Water Foundation with additional support from INN. For earlier stories in the Tapped Out series, click here.

Kimberley Hasselbrink contributed photography for this piece.