In the contentious talks over how states will split the shrinking Colorado River, negotiators are reaching consensus on one point: Just go with the “natural flow.”
The concept is a somewhat simple one. Instead of negotiating future cuts across the entire seven-state region, the process would rely on recent water records — the amount of water flowing from the Colorado River headwaters in the Upper Basin to a point in Arizona marking the boundary of the Lower Basin states.
Negotiators recently heralded the move as a potential breakthrough in the long-stalled talks: It could help end a stalemate over how to share the pain of future water reductions and at the same time respond to the impacts of climate change. But that belies a set of lingering questions.
For one, just determining the water in the river will require complex calculations relying on evolving research. Even more critically, there’s no indication negotiators are close on the particularly difficult issue of deciding how big a share of water each group of states can claim.
Still, observers say it could mark an important change.
“There’s the philosophical idea of a natural flow regime guiding the management of the river, and then there’s the technical detail of what that actually entails,” said Eric Balken, executive director of the Glen Canyon Institute. “It’s a logical but revolutionary concept.”
At a meeting last month where a key Trump administration official gave some Colorado River negotiators a November deadline to reach a deal, state officials explained they were coalescing around the concept known as supply-driven planning, or just “natural flow.”
The theory — the premise of sharing the river based on how much water would travel downstream without dams or diversions or other human interventions — is actually a complex mathematical problem, rife with potential pitfalls and technical issues.
The 1,450-mile Colorado River is divided among seven states under a 1922 compact, but the river system’s annual operations are guided by a series of term-limited deals among the states, as well as agreements with Mexico.
Under the 1922 compact — which serves as the cornerstone of the “law of the river,” or all the laws and regulations that govern the region’s water supply — Arizona, California and Nevada in the Lower Basin and Colorado, Utah, New Mexico and Wyoming in the Upper Basin each get 7.5 million acre-feet of flows.
But the temporary agreements that govern reductions and attempt to respond to changing conditions are set to expire next year, and a new long-term operating plan must be in place by Oct. 1, 2026, which marks the start of the 2027 water year.
Negotiations among the seven states for a new deal have gridlocked while the two basins failed to reach agreements on how to share the pain of future water cuts, as well as how to account for all the water in the sprawling river system.
Researchers have found that over the past two decades, flows in the Colorado River have shrunk by about 20 percent — there’s no longer enough water for seven states to split 15 million acre-feet. A top water researcher in the region predicts the river could even become as small as 10 million acre-feet in response to climate change.
“No matter what natural flow regime you decide on, it doesn’t change the net bottom line: We don’t have enough water,” Balken said.
Under a natural flow system, the states would rely on a supply-driven calculation of how much water would be in the Colorado River were it not diverted to farms, stored in massive reservoirs or otherwise used by humans. The river in its current state is engineered to channel water into reservoirs across both basins and into canals to far-away farms in Arizona and California, while also generating electricity through dams at Lake Powell and Lake Mead.
“The Basin States have been exploring an explicit supply-driven operational framework based on the natural flow of the river,” said Becky Mitchell, who serves as both Colorado’s Colorado River commissioner and acting chair of the Upper Colorado River Commission. “The concept under discussion is that Powell would release a certain percentage or volume of the average of the last few years of natural flow, as measured at Lee Ferry.”
‘Accuracy and timeliness’
The Bureau of Reclamation already studies the natural flow of the Colorado River, providing annual estimates drawn from data at 29 points on the waterway and its tributaries, combined with consumptive uses and loss, such as water used in growing crops or lost to evaporation.
“Natural flow is an estimated value of what would flow in the river in the absence of human development,” explained Jack Schmidt, program director of the Center for Colorado River Studies at Utah State University.
Essentially, Reclamation examines the flow of water in the Upper Basin’s river system while also adding back in any water diverted for use in cities and farms. That data is used to calculate how much water would be found in a hypothetically free-flowing river at Lees Ferry, Arizona, a point south of the Glen Canyon Dam that is the dividing line between the Upper and Lower basins.
But these calculations get tricky.
While it is relatively simple for Reclamation to add up how much water the Upper Basin uses for cities like Denver — the water is piped across the Front Range in tunnels and pumps — it is significantly more complex to determine how much is used for irrigated agriculture.
In the Upper Basin states that includes how much water is lost to “evapotranspiration,” or the evaporation of water into the atmosphere from the soil or from soil to the air via plants, Schmidt said. Water used for agriculture can account for 70 percent of the water used in the four Upper Basin states, and how much of that is consumed, rather than returned to the river system, is an important data point.
While methods for measuring that water usage have advanced dramatically — from standard formulas based on local temperatures to new calculations that combine satellite data with water vapor measurements and other climate information drawn from monitors near the fields — Schmidt said that a standard would need to be codified.
“Accuracy and timeliness are going to have to be incorporated into these agreements,” Schmidt said, noting that another significant challenge with adopting an agreement based on natural flow is the fact that current calculations can take years to finalize.
According to Reclamation, its calculations typically lag at least 1.5 years behind the water year, as data on consumptive uses and losses is collected.
“We have this evolving change leading to more accurate estimates of natural flow but, of course, all that takes time to process that data,” Schmidt said, noting that Reclamation has yet to finalize data as far back as 2021, instead publishing provisional estimates. The states are weighing an agreement that would rely on the average of the three most recent years of flows.
There would also need to be agreement on how to determine evaporation from reservoirs, including Lake Powell, the major federal site that is home to the Glen Canyon Dam that sits above Lees Ferry.
Cuts to Reclamation staffing could further slow that process, Schmidt warned. POLITICO’S E&E News reported in June that Reclamation has shed 1,180 of its 5,700 employees through early retirement and buyout offers, according to a Freedom of Information Act request.
“Reclamation absolutely is going to have to be appropriately staffed so that they can generate these numbers in a timely way,” Schmidt said.
Schmidt also advised state negotiators to continue to look at the water levels of both Lake Powell and Lake Mead. Current operating plans include a series of reductions in the amount of water going to Arizona and Nevada when water levels in Lake Mead drop below specific elevations.
“The amount of water in the reservoirs — which is the bank account — we know it precisely and accurately, and we know it in a timely fashion,” Schmidt said. “We still don’t know what the natural flow of the Colorado River in 2025 is, but I can tell you the duration of time that Lake Powell and Lake Mead, collectively, increased in storage only lasted for two weeks this year.”
‘The ultimate sticking point’
The major stumbling block — even with agreement on how much water is flowing in the river — will be how much each basin can use each year and what must remain in storage in the reservoirs, in large part to keep the dams generating electricity.
While the 1922 compact gives the Upper and Lower basins the same amount of water — each claiming 7.5 million acre-feet — the two regions do not use the same quantities. A 1944 treaty with Mexico also gives that nation 1.5 million acre-feet of water.
An acre-foot is equal to about 326,000 gallons of water, or enough to support two to three families each year. The Colorado River supplies water to 40 million individuals and 5.5 million acres of irrigated farmland.
Reclamation reports show that the Upper Basin consumed 4.7 million acre-feet in 2023, the most recent data available. That same year, as the Biden administration pursued aggressive conservation efforts to address short-term drought, the Lower Basin reported using 5.8 million acre-feet.
By comparison, in 2021, the Upper Basin used 3.9 million acre-feet while the Lower Basin used nearly 7.1 million acre-feet.
State negotiators have not publicly discussed how the two basins would divide the total amount of water or what percentage either region is angling to claim.
“You could imagine on one side of the ledger, it’s 50-50, that’s how we interpret the compact,” Schmidt said. “And on the other side, the Lower Basin could say, we use more and we use it productively.”
He added: “That is the ultimate sticking point. … What is the percentage?”
Balken — whose group has advocated for a bypass around Glen Canyon Dam due the facility’s “antique plumbing” that could slow the Colorado River to a trickle under low water conditions upstream — noted that how the water is split will also be key to determining how individual states are forced to share any reductions.
“The real question for me is, ‘What is the percentage of the natural flow that’s going to be sent downstream?’ That will by default imply what the curtailments are going to be for the Upper and Lower basins,” he said.
Correction: This story was corrected after publication to reflect that Jack Schmidt’s comment about increased storage lasting for two weeks included both Lake Powell and Lake Mead.
