Water reaching a Colorado State University dorm has crossed the Continental Divide, passed through tunnels and treatment systems and entered a network built on a river that is shrinking with rules that assumed it wasn’t.
The science of the Colorado River’s decline is not new. It is, however, arriving faster than expected.
Since the early 2000s, flows in the Colorado River have dropped roughly 20% compared to 20th-century measurements. Average annual flows during the 20th century ran about 14.6 million acre-feet per year; since 2000, that number has fallen to roughly 12.4 million acre-feet.
That gap matters because the 1922 Colorado River Compact was drafted around an assumed flow of roughly 16.4 million acre-feet per year, a figure the river has never reliably produced.
Brad Udall, a senior water and climate research scientist at CSU’s Colorado Water Center, said the error wasn’t purely a measurement problem.
“The conventional wisdom is that they overestimated, and they did the best they could, and they just made a mistake,” Udall said. “But the compact commissioners in 1922 didn’t do their homework. There were plenty of warning signs, even back then, that there wasn’t enough water in the river. And they basically ignored the science.”
Udall said one driving force of the decline is that the upper Colorado River Basin has warmed approximately 3 degrees Fahrenheit since the 1970s.
The warming also shifts snowmelt timing. As Udall explained, a warmer atmosphere means spring arrives earlier and the snow melts faster, reducing the slow seasonal release of water that the river and its users have historically depended on.
A 2017 study in Water Resources Research, co-authored by Udall, found that roughly one-third of the flow decline between 2000 and 2014 was attributable to warming temperatures — and that continued warming under business-as-usual emissions could produce further reductions of 20% by midcentury and 35% by end of century.
This year has compounded those trends. As of April 2026, Lake Powell‘s water levels sit barely above 3,500 feet. Lake Mead is projected to decline by 20 feet over the year, which would decrease Hoover Dam’s hydropower output by 40%. Spring and early summer runoff into Lake Powell is forecast at roughly 36% of the average.
Udall said roughly half of the water used on Colorado’s Front Range is Colorado River water, moved across the Continental Divide through interbasin diversion projects.
Former Director of the Colorado Water Center John Tracy explained that for Fort Collins specifically, that means the Colorado-Big Thompson Project, a federal trans-mountain diversion operated by Northern Water, is a backbone of the local water supply.
State Engineer for the Colorado Division of Water Resources Jason Ullmann said the immediate effects of this year’s record-low snowpack are already hitting water users across the state.
“We’re going to be doing a lot of actual shutting people off,” Ullmann said. “I have many water commissioners that go out to people’s head gates, and they close them, and they lock them because they can’t take water because there’s not enough water supply.”
Ullmann said curtailments in 2026 are expected to exceed any prior year, driven by what state data shows is historically low snowpack. According to CSU’s Colorado Water Center, Colorado uses roughly 5.3 million acre-feet of Colorado River water annually on average, well below its compact allocation.
The management rules that have governed how Lake Powell and Lake Mead store and release water, the 2007 Interim Guidelines and the 2019 Drought Contingency Plans, expire at the end of 2026. The Bureau of Reclamation has set an Oct. 1 deadline to finalize new operating guidelines.
The seven basin states missed the November 2025 deadline to submit a preliminary consensus plan, then missed a second deadline Feb. 14. If the states cannot reach a consensus agreement, the Bureau of Reclamation will impose its own operating plan under existing legal authority.
The central dispute is over who absorbs cuts in dry years. The Lower Basin — California, Arizona and Nevada — has offered to conserve 1.5 million acre-feet annually but wants the Upper Basin to share mandatory reductions if further cuts are needed. The Upper Basin states have declined.
Chris Winter, executive director of the Getches-Wilkinson Center for Natural Resources at the University of Colorado Law School, said the federal government’s responsibilities in the basin are layered, as an infrastructure operator, as quasi-mediator between states and as trustee for the 30 tribes that hold federal reserved water rights.
One dimension of the crisis that experts say has received insufficient attention is the river’s environmental condition itself. The 1922 compact assigns no water to ecological uses, nothing for fish habitat in the Grand Canyon and nothing for the Colorado River Delta, which once supported wetlands and migratory bird populations before upstream development reduced flows to a trickle.
Winter said the gap is still largely absent from the current negotiations.
“It’s very difficult because in the 1922 compact and the law, the river really focuses on how we allocate water for consumptive uses, not how we keep water in the stream and rivers for environmental and recreational values,” Winter said.
Research modeling suggests that under existing policy and continued warming, both Lake Powell and Lake Mead face substantial risk at above 80% probability of reaching dead pool before 2060. Dead pool refers to the point at which a reservoir drops so low that water can no longer flow through the dam outlets via gravity, effectively ending both hydropower generation and downstream delivery.
For Udall, who has been warning about these dynamics for more than two decades, the lesson from 1922 applies directly to how decisions are made now.
“Trusting scientists and engineers and empowering and listening to the ones that are telling you stuff you don’t want to hear is really important,” Udall said.
Reach Maci Lesh at science@collegian.com or on social media @RMCollegian.
