University of Colorado researchers published a stark analysis this week warning that the Colorado River system is on track for a complete "system crash" as soon as the start of the 2028 water year, which is roughly 18 months away, if the basin experiences another dry year similar to 2025.
The paper, co-authored by Anne Castle, a former Assistant Secretary for Water and Science at the Interior Department who has spent her career studying the Colorado River, describes a situation in which Lake Mead and Lake Powell drop below thresholds where they can no longer function as the water storage systems that 40 million Americans in seven Western states depend on.
When that happens, Castle said, it is over in the specific sense that no amount of management can fix what the physics of the reservoirs no longer allow. "If the reservoirs drop to that level, we're not getting any benefit from them anymore. If that happens, it's like the Colorado River is running free. We get what we get, what nature provides, and we have no ability to buffer dry years."
The current situation is not theoretical. Lake Mead is sitting at approximately 1,049 feet above sea level as of early June.
The critical threshold below which the reservoir loses its ability to function as meaningful water storage is 975 feet.
There are 74 feet between where the water is today and where it cannot go without constituting a catastrophic failure of the system.
A Trump administration decision to reduce releases from Lake Powell to Lake Mead could cause Lake Mead to drop 28 feet by July 2027 alone, a single policy decision that would push the reservoir roughly a third of the way to the crisis threshold in one year, past any previous record low.
The Specific Numbers And What They Mean
The Colorado River provides water to seven states, California, Nevada, Arizona, Utah, Colorado, Wyoming and New Mexico, plus Mexico. The system works because two massive reservoirs, Lake Mead behind Hoover Dam and Lake Powell behind Glen Canyon Dam, store the river's water across seasons and years.
In wet years, the reservoirs fill. In dry years, the reservoirs drain to compensate. The stored water is the buffer that makes a desert civilization possible.
The system requires that the reservoirs maintain water above specific elevations to remain functional. Lake Powell's critical threshold is approximately 3,490 to 3,500 feet above sea level.
The researchers' analysis and Bureau of Reclamation engineering reports describe damage to Glen Canyon Dam's lower outlet works that would prevent water from being released downstream if the reservoir drops below that level. Lake Powell currently sits barely above 3,500 feet. The margin is not large and the direction is wrong.
At Lake Mead, the critical threshold is 975 feet. The reservoir is at 1,049 feet. That gap looks more comfortable but the trajectory is downward and the management tools available to halt the decline are limited.
The Bureau of Reclamation has been conducting what amounts to emergency operations to keep Lake Powell above its critical threshold, beginning emergency releases from the Flaming Gorge reservoir upstream of Lake Powell on April 23, a release program that will continue through April 2027, and reducing the volume of water flowing from Lake Powell to Lake Mead by 1.48 million acre-feet through September 2026.
The Bureau's own analysis is alarming about what comes next.
A Bureau official told the Salt Lake Tribune that in terms of tools to prop up Lake Powell between December 2026 and March 2027, "there really aren't too many new and different things. We're basically stuck with using the same two levers that we use this year." The levers are running out. The water is running out with them.
What 2026 Has Already Revealed
The 2026 water year is on track to be one of the lowest runoff years on record for the Colorado River Basin, a basin that covers portions of seven states and whose annual snowpack and spring runoff determine how much water flows into the reservoirs.
After the wet 2023 year that briefly raised reservoir levels and generated optimism about the system's long-term health, 2024 and 2025 were dry years that reversed much of that progress. 2026 is continuing the pattern.
The structural reality that Castle's paper documents is the one that water managers have been reluctant to state as plainly as she does: even in years when conservation efforts have reduced water use, the reductions have not been enough to match what the natural inflows provide.
Water use across the basin continues to outpace natural inflows every year. The reservoirs decline by some amount every year that inflows do not exceed use. The wet years provide temporary relief.
The dry years accelerate the crisis. The trend, across a decade of drought conditions driven by a warming climate, is downward.
Even the optimistic scenario in Castle's analysis, a very wet year that boosts reservoir levels significantly, provides only approximately two years of relief before the system returns to crisis conditions, because the underlying mismatch between consumption and supply has not been resolved. A wet year buys time. It does not solve the problem.
The Policy Failure That Is Making It Worse
The management rules that governed the Colorado River system were written in 2019. They expired this fall of 2026.
The Bureau of Reclamation released five proposed frameworks for post-2026 operations and conducted analysis of what each framework would do under various climate scenarios.
The finding that the Sierra Club and other water policy organizations have been circulating is alarming, all five proposed plans fail under current drought conditions.
The alternatives that water managers developed to replace the expired guidelines are not adequate for the reality the river is currently experiencing.
The seven basin states, California, Nevada, Arizona, Utah, Colorado, Wyoming and New Mexico, remain divided on how to distribute reductions in water use among themselves. California, which holds the most senior water rights under the Law of the River that has governed allocation since 1922, has been reluctant to accept cuts proportional to its usage.
The Upper Basin states, Utah, Colorado, Wyoming and New Mexico, and the Lower Basin states, California, Nevada and Arizona, have different perspectives on who should bear the burden of reduced allocations.
The Trump administration's decision to reduce releases from Lake Powell to Lake Mead — a decision that federal water officials said would "accelerate the downstream decline of Lake Mead" and could reduce Hoover Dam's hydropower generation capacity by up to 40 percent as early as this fall, has added a new layer of complexity to an already difficult negotiating environment.
Reducing releases from Powell protects the upstream reservoir at the cost of the downstream one. It is a tradeoff that benefits the Upper Basin states and complicates the situation for the Lower Basin states and the cities, farms and tribes that depend on Hoover Dam.
What System Crash Actually Looks Like
Castle described what the Colorado River would look like in a system crash scenario in terms that are worth sitting with.
It would operate in what the experts call "run-of-the-river" mode, water flowing straight through the system as it arrives, in real time, without being stored. The dams would no longer provide a buffer against drought years.
The cities and farms that have been built on the assumption that the river's annual variability can be smoothed out over time by reservoirs holding water in good years for use in bad ones would face a reality in which there is no smoothing.
Whatever flows this year is what there is this year. If it rains, you have water. If it does not, you do not.
Forty million Americans in seven states plus Mexico depend on the Colorado River. The largest city in that group is Las Vegas, which gets virtually all of its municipal water from Lake Mead. Phoenix and Tucson depend on Colorado River water delivered via the Central Arizona Project.
Los Angeles and San Diego's water systems are partly dependent on Colorado River allocations.
The agricultural regions of the Imperial Valley and Yuma, Arizona, which produce a significant fraction of the winter vegetables eaten in the United States, depend on Colorado River water for irrigation.
The system crash that Castle is warning about is not a natural disaster in the conventional sense. It is the endpoint of a management failure that has been visible and documented for decades arriving on a specific timeline because the gap between what the river provides and what users take from it has never been fully closed.