By Eric Kuhn and John Fleck

The Bureau of Reclamation’s January 2024 “Most Probable” 24-month study forecasts that annual releases from Glen Canyon Dam for both Water Years 2025 and 2026 will be 7.48 million acre-feet per year (maf). If this happens, the ten-year total flow at Lee Ferry for the 2017-2026 period will drop to about 83.0 maf, only about 500,000 acre-feet above 82.5 million acre-feet, the first 1922 Compact hydrology “tripwire.”

That line – 82.5 maf feet of Lee Ferry deliveries over a ten year period – has become a dividing line between two contending interpretations of the most important unresolved question in the century-old Colorado River Compact: How much water must the Upper Basin deliver to the Lower Basin? What happens if it doesn’t?

The consequences of triggering the tripwire, which might happen in 2027, are significant. In the worst case scenario, it could plunge the basin into Supreme Court litigation over the interpretation of the 1922 Compact, which could result in a forced curtailment of post-compact water uses in the Upper Basin. Or, alternately, if the Basin States are willing to settle their long-term disputed issues or implement basic changes to the Law of the River via the renegotiations of the post-2026 operating rules, it could be a “non-event.”  This is one of the fundamental issues facing the states as they meet to develop their basin-state alternative.

The 82.5 maf tripwire is based on the 1922 Compact’s two flow-related requirements at Lee Ferry; Article III(d) requires the four Upper Division States to not cause the progressive ten-year flow at Lee Ferry to be depleted below 75 maf. Additionally, Article III(c) provides that if there is not sufficient surplus water to meet the annual water delivery requirements of the 1944 Mexican Treaty, normally 1.5 maf, then each basin must provide half of the deficiency (the required annual delivery minus the available surplus). The Upper Division States must deliver their share of the deficiency at Lee Ferry in addition to their obligations under Article III(d).

The Upper Division and Lower Division States have never agreed on the meaning and interpretation of Article III(c). There have been numerous papers on the disputed issues by both compact scholars and practitioners and Article III(c) has never been interpreted by the U.S. Supreme Court. Suffice it to say that the Lower Division States believe that the Upper Division States must deliver at Lee Ferry a total of 82.5 maf every ten years (75 maf + 10 x 750,000), but the Upper Division States believe that they currently have no obligation to Mexico, so the number is at most 75 million. There are of course, nuances. The Lower Division States have suggested that the Upper Division States might also need to cover transit losses between Lee Ferry and Mexico and the Upper Division States have most recently suggested that if climate change, not Upper Basin depletions, is causing the ten-year flow to fall below 75 million, then their article III(d) non-depletion obligation must be appropriately adjusted. Further, if pursuant to either the extraordinary drought provision or a treaty minute, the required annual delivery to Mexico is less than 1.5 maf, then, even with no surplus, the Upper Division’s 50% share would be less than 750,000 acre-feet.

One should recognize that the annual releases from the Glen Canyon Dam and the annual flow at Lee Ferry (the compact point) are not the same.  Between the dam and Lee Ferry, the river gains flow from groundwater accretions (in part due to leakage around the dam) and from the Paria River. These gains can vary from about 30,000 acre-feet to over 300,000 acre-feet annually.  The ten-year Lee Ferry flow for 2014-2023 was approximately 86.1 maf.

That amount – 86.1 maf – might seem like a safe cushion. But because it is a ten-year moving total, we are about to drop out years with big releases (9 million acre feet) and replace them with years with just 7.48 maf. At the end of 2024, because both 2014 (the year that drops out) and 2024 (the year that is added in) are 7.48 maf years, the ten-year flow will stay about the same.  The way the ten-year flow calculation works is next year, 2015 will drop out and 2025 will be added in, and so on, but here is the problem; From 2015 through 2019, the 2007 Interim Guidelines dictated an annual release of 9 maf per year. With accretions, flows at Lee Ferry averaged about 9.18 maf per year (source: UCRC 74^th Annual Report). Thus, if 2025 and 2026 are 7.48 maf years, the ten-year flow will lose about 1.5 maf/year making the total about 83 maf for the 2017-2026 period. Because the 2007 Interim Guidelines expire, we don’t know what the annual release will be in 2027, but if it’s less than about 8.5 maf, because 2017 was a 9 maf year, the ten-year Lee Ferry flow could drop below the tripwire – 82.5 maf (with two more 9 maf years, 2018 and 2019, in the pipeline).

We recognize that the 24-month studies don’t predict the future. They are a planning and management tool. It’s plausible that by 2027, a series of wet years could result in a ten-year flow that is much higher than 82.5 maf, something that is, in our view, unlikely. But as we sit here in January 2024, the 24-month study is the only planning tool we’ve got. It would behoove us to pay attention to what it is telling us.

If a future 24-month study projects that ten-year flows will fall below 82.5 maf, it will be a big deal for the basin. What provision of the Law of the River will control annual releases from Glen Canyon Dam – the post-2026 Operating Guidelines, the Lower Division’s interpretation of the 1922 Compact (82.5 maf), or the Upper Division’s interpretation of the 1922 Compact (75 maf or less)?  If the Lower Division States agree to a ten-year flow target of less than 82.5 maf, are they effectively surrendering to the Upper Division States? If the Upper Division States agree to a flow target of 82.5 maf, are they effectively surrendering to the Lower Division States? If the Upper Basin states agree to either 82.5 maf or some smaller compromise delivery target and the hydrology remains bad, how will Wyoming, Colorado, New Mexico, and Utah – our states – approach the required water use reductions? What if there is not enough water in storage in Lake Powell (and the other CRSP reservoirs) to release sufficient water to bring the ten-year flows to 82.5 maf? Will the Lower Division insist that that the UCRC implement a curtailment of post-compact uses in the Upper Basin. If the UCRC refuses to do so, will that plunge the Basin into litigation?

At the recent 2023 CRWUA meeting, representatives of the Lower Division States stepped up and made it clear they own the “structural deficit,” and the conference was buzzing with talk of the innovative system approach the Lower Division has put on the table. This is great news, but as Colorado’s Royce Tipton concluded sixty years ago, the “structural deficit” is not a single number. It’s a range that depends on the interpretation of the Lee Ferry obligations of the Upper Division States under the 1922 Compact. If the average annual flow requirement is 8.25 maf/year, (which Tipton referred to as “fictional”) he calculated the deficit to be about 1.2 maf/year. Today we believe it’s about 1.4 -1.5 maf/year (Tipton’s assumptions about system losses were probably too low and perhaps his Lee Ferry to Lake Mead inflow assumptions too high). If, however, the average annual flow requirement is 7.5 maf/year, he calculated the deficit to be about 2 maf/year, today maybe 2.2 maf/year. This difference is huge, especially for the Central Arizona Project, the junior user on the Lower basin mainstem.

In the past decades, water managers in the Colorado River Basin have made accomplishments that even two decades ago were considered out of reach: The 2019 Drought Contingency Plans, the shortage sharing agreements with Mexico, the increased recognition of the rights of the Basin’s Native American communities are just a few. The Lower Division’s recent pronouncement that they own the structural deficit is another major step forward, but it has a fundamental flaw. Representatives of all seven Basin states continue to stubbornly insist that the Law of the River, and specifically, the 1922 Colorado River Compact, will serve as the “foundation” of the post-2026 operating guidelines. The flaw is that is that we don’t know if this foundation is based on a ten-year Lee Ferry flow of 82.5 maf, or 75 maf, or something different.

Put simply, the states agree that the Law of the River has to be the basis of what we do, but don’t agree on what the Law of the River actually says. Without an agreement on this fundamental issue, calling the 1922 Compact a foundation is nothing more than self-delusional wishful thinking. Now that the Lower Division States have agreed to own the structural deficit, is the next step for all seven states and the federal government to openly acknowledge that given the impacts of climate change on the river, the 1922 Compact’s overallocation of water and its disputed Lee Ferry flow provisions are core problems for the basin, not the foundation? Finding a sustainable future, without litigation, will require accepting and acknowledging the basic problems we face, not avoiding them.

By Jack Schmidt | January 9, 2024

This winter’s snowfall has been meager so far, and Reclamation’s January 1 prediction of unregulated inflow to Lake Powell is only 66% of average. Water managers are beginning to anticipate another summer of water-use restrictions. No one should think that last year’s runoff eliminated the ongoing crisis.

Figure 1. Graph showing active water storage in 42 reservoirs in different parts of the Colorado River basin. Conditions at the end of December 2023 are comparable to conditions in early May 2021, indicated by the black arrows. Data downloaded at https://www.usbr.gov/uc/water/hydrodata/reservoir_data/site_map.html.

2. Most of the basin’s water storage is in Lake Mead and Lake Powell (Fig. 2). Releases from Lake Powell and reductions in Lower Basin water use were sufficiently large that there was significant recovery of storage in Lake Mead. At the end of December, storage in Lake Mead (9.05 million acre feet) exceeded storage in Lake Powell (8.44 million acre feet) by approximately 600,000 acre feet. The difference in storage between the two reservoirs is much less than during the previous two years when more water was stored in Lake Mead.

Figure 2. Graph showing water storage since January 2021. Note that storage in Lake Mead was significantly greater than in Lake Powell in 2021 and 2022. Large spring runoff in 2023 was captured in Lake Powell, and some of that accumulated inflow was subsequently released to Lake Mead. The rate of reduction in storage in reservoirs upstream from Lake Powell significantly slowed after mid-fall 2023. The category “other Upper Basin reservoirs” includes Strawberry, Granby, McPhee, Dillon, Starvation, Nighthorse, and smaller reservoirs. Water storage in Lake Mohave and Lake Havasu remains nearly constant. Note that the vertical axis is an arithmetic scale that has a break. Data downloaded at https://www.usbr.gov/uc/water/hydrodata/reservoir_data/site_map.html.

3. The rate of loss in reservoir storage this year remains low relative to the rate of loss in previous years (Fig. 3), especially the rate of decline of the combined storage in Lake Mead and Lake Powell. (Fig. 4) The basin’s water managers are doing a good job of reducing use and conserving water in reservoirs. Reclamation’s estimate of probable consumptive water use in the Lower Basin in 2023, issued 31 December 2023, is 5.78 million acre feet, nearly 900,000 acre feet less than Lower Basin consumptive use in 2022. Will that degree of water conservation be enough? That depends on how much snowmelt occurs this spring.

Figure 3. Graph showing the rate of reduction in basin-wide reservoir storage in each of the past ten years. The reduction in storage has been at a much slower rate than in other years. Each year that plots lower than 2023 on this graph reflects a higher rate of loss in storage than this year.

Figure 4. Graph showing the rate of reduction in the combined storage in Lake Mead and Lake Powell in each of the past ten years. The reduction in storage has been slower than in any other recent year. Each year plotting lower than 2023 on this graph reflects a higher rate of loss in storage than in this year.

Acknowledgement: Eric Kuhn and John Fleck provided helpful suggestions that improved this posting.

Peering out across the Colorado River landscape for 2024, a few things I’ve been reading to help catch y’all up:

Paying to conserve

Annie Snider had a terrific story in late November about the role of federal cash in the short term solutions to the Colorado River’s challenges that needs to echo outside the usual Politico-reading crowd.

Big $$$ flowing from the federal government have been crucial to shoring up the reservoirs by paying people to leave water in them, but what does that do to the prospects of long run solutions that require permanent water use reductions:

The gusher of federal money is likely to make a broader, long-term deal to save the West’s most important river more expensive.

“It’s all a grand experiment,” said Kathryn Sorensen, a former head of Phoenix’s water department, who noted that Democrats’ Inflation Reduction Act funding is effectively creating a new market for water, with a new, higher price. “This market, especially one with a premium [price], might create some perverse incentives.”

Using hydrology, rather than reservoir levels, to determine how much water people use

Luke Runyon, now at the Water Desk (wait, what?) included this in his “what to watch for this year on the river” piece following last month’s Colorado River Water Users Association:

One more idea from the Las Vegas conference that’s still largely conceptual, but is gaining some interest from those in power, is to use annual measures of basic hydrology — like snowpack levels and streamflows — to determine how much water ends up being delivered to the basin’s varied users. It sounds simple: only use what nature provides.

But that idea flies in the face of the river’s foundational governing document, the Colorado River Compact, which put fixed volumes of water use on paper, regardless of whether it was a dry or wet year. For now, the idea seems to be more of a talking point than a specific policy proposal, and we will see if proponents can turn it into something Lower Basin users can get behind.

The current management regime is based on reservoir levels, rather than the underlying system hydrology. My collaborators Jack Schmidt, Eric Kuhn, and I argued in our 2022 NEPA comments for the hydrology approach rather than the reservoir level approach. We are not alone. It makes a lot of sense, though, as Luke notes, no one’s make a specific policy proposal (we just wave our arms in our comments and basically say “This’d be great, someone should figure out how to do it!”).

Which leads me to….

Modeling and a shared understanding

Kevin Wheeler, Terry Fulp, and Roberto Salmon have a super-important new paper on the use of modeling to create a shared understanding of the resources of the Colorado River basin:

As participants directly involved in recent agreements seeking to address this challenge, we describe how a critical basin-wide lens has been developed through information transparency, including the evolution and voluntary exchange of computational models. To achieve greater sustainability, further resolve and commitment from water managers in the United States and Mexico is necessary to ensure the required enhancement of the knowledge systems that will be used to inform extremely difficult and critical decisions.

This touches on a crucial point in negotiating these big, complex, shared resource management agreements – the importance of having a shared understanding of the resource. This is one of the key threads through my book Water is For Fighting Over, which drew for this point on the work of Elinor Ostrom.

When I wrote the following passage eight years ago, I was very specifically writing about the work that Wheeler, Fulp, Salmon, and others had done in negotiating a landmark U.S.-Mexico water management agreement known as Minute 319.

Much of the negotiations that followed hinged on the development of a shared understanding of a seemingly straightforward problem: how to account for the water moving through the Colorado River System from one nation to the other. But to deal with the issue required doing something that had previously been beyond reach: extending “the network” of US water managers into Mexico. One of the network’s key tools is a sophisticated computer model of the basin known as the Colorado River Simulation System. The Bureau of Reclamation supported an entire research team based at the University of Colorado in Boulder to develop and maintain it, and its calculations are used to make decisions about how much, where, and when, to release water.

Now Wheeler, Fulp, and Salmon have written for us a guide. I hope to return to this for a longer piece (see below, re business model), but in the meantime just read what they wrote, please, that’s your homework.

Busking as a business model

I’ve got no business model here, just an extremely part time University gig that ensures me library privileges (important!) and a semi-formal institutional structure to provide routine access to a bunch of really smart people (very important!).

But it occurs to me that I’m kinda like a busker, somewhere between my failed model of performing journalism in bars (this did not work out) and putting out a guitar case seeded with a few fives. Thanks to my faithful readers who provided the fives, here’s the guitar case.