The question of reservoir evaporation – How much water are the Lower Colorado River Basin states really using?

Accounting for Lower Colorado River Basin water use, with and without evaporation

The conventional simplification of the Colorado River Compact’s water allocation scheme is that it set aside 7.5 million acre feet of water use for the “exclusive beneficial consumptive use” of the states of the Lower Basin – Nevada, Arizona, and California.

In the 21st century, the official accounting shows the Lower Basin states using an average of 7.4 million acre feet per year (the blue line in the graph above), so we’re good, right?

Well, actually….

When the states of the Upper Basin gathered in the late 1940s to negotiate the Upper Basin Compact, they recognized that if you’re going to build a reservoir to enable the consumptive use of water, you need to account for the evaporative losses from that reservoir.  Article V of the Upper Basin Compact goes into some detail about how to account for “all losses of water occurring from or as the result of the storage of water in reservoirs” built by the states and the federal government.

But, as Eric Kuhn reminded us in a post yesterday, we don’t have a corresponding Lower Basin Compact. Instead, we have the 1963 Supreme Court decision in Arizona v. California, and the subsequent implementing decrees, to set the parameters for the water’s allocation. The court didn’t require an accounting of evaporation, only the water actually released from the reservoir.  Here’s Eric:

I have no doubt that the 1922 compact negotiators considered this evaporation a man-made beneficial consumptive use to be covered by the (water) apportioned to the Lower Basin, but it is not covered in the Supreme Court’s decision.

Not everyone agrees about the compact negotiators’ intent. See for example Jason Robison and Larry MacDonnell in their review of AZ v. CA. But Robison and MacDonnell also note that the subsequent failure to fully wrestle with the question of reservoir evaporation and related questions like Lower Basin tributary use – the sort of things that negotiating a Lower Basin Compact might have adressed – has left a bit of a mess.

Legal nuances aside, the reality is that in the absence of a Lower Basin Compact to come to a negotiated agreement on questions like this, we’re left with significant uncertainties on this and other questions. And by my math, if you count reservoir evaporation, the Lower Basin states have been using an average of 8.4 million acre feet a year in the 21st century (the pinky-orangey line above). On this point Kuhn, Robison, and MacDonnell agree – that’s a problem.

The Lake Powell Pipeline and the problems posed by the lack of a Lower Colorado River Basin Compact

By Eric Kuhn

The Lake Powell Pipeline, moving Upper Colorado River Basin water to the Lower Basin

As the Colorado River Basin’s managers wrestle with thorny questions around the proposed Lake Powell Pipeline, a colleague who works for a Lower Colorado River Basin water agency recently asked a question that goes to the heart of the future of river management: With land in the Lower Colorado River Basin, why doesn’t Utah have a Lower Basin allocation?

The answer, arising from deep in the history of the Law of the River, goes far beyond the possible use of some of Utah’s Upper Basin water in the Lower Colorado River Basin’s Virgin River Valley. It strikes at the heart of important and as yet unresolved questions in the river’s future – about accounting for reservoir evaporation, and who bears the responsibility for, and benefits from, water flowing down the Lower Basin’s tributaries.

The lack of a “Lower Basin Compact”

The simple answer to my colleague’s question is that Utah has no apportionment of Lower Basin water because there is no Lower Colorado River Basin Compact.

The negotiators of the 1922 Colorado River Compact allocated water to the Upper and Lower Colorado River Basins with the clear intention that the states of each basin would return to the negotiating table to work out the details of how their share of the water would be allocated and accounted for. The states with Upper Basin interests completed the task in 1948, negotiating the Upper Colorado River Basin Compact. Notably Arizona, which is not an “Upper Division State”, but which has land in the Upper Basin, participated, and was allocated a share of the Upper Basin’s apportionment.

In the 1920s and ’30s, there were many attempts, all unsuccessful, to negotiate a Lower Basin compact.  They came close in 1927 when the basin governors met in Denver for the specific purpose of resolving the differences between Arizona and California.  The Senate used the deliberations from 1927 as input to the mainstem apportionments to Arizona, California, and Nevada provided for in the 1928 Boulder Canyon Project Act. But the language of the Act clearly suggests Congress expected a Lower Basin Compact to be negotiated to flesh out the details, much as later happened in the Upper Basin.

That never happened, though, and in the absence of a Compact, the present-day apportionments we all know so well – 4.4 million acre feet for California, 2.8maf for Arizona, and 300,000 acre feet for Nevada – are based on the Supreme Court’s interpretation of the intent of Congress when it debated and passed the Boulder Canyon Project Act in 1928. But in the absence of the a careful process of negotiation to cover the full range of issues that needed to be considered (as happened in the Upper Basin), the Supreme Court’s decision was narrow. It only covered under contracts between the federal government and water users downstream of Hoover Dam – not for water consumed on the Lower Basin tributaries or on the Lower Basin’s portion of the mainstem above Lake Mead.

In the 1930s, William Donovan came close

In 1930 President Hoover appointed Colonel William Donovan as special mediator to negotiate a Lower Basin compact.  Donovan came close, but Arizona and California just could not make one final compromise. In both the 1927 and 1930 near misses Utah and New Mexico would have received from the 8.5 million acre-feet apportioned by the 1922 compact to the Lower Basin “all water necessary for use on areas of those States lying within the lower basin” – (see Science Be Dammed, chapter 8).

In 1952 when Arizona filed its Supreme Court case against California, one of its initial goals was a Court ruling that would be functionally equivalent to a Lower Basin compact. Among its claims for relief was that the court find that Arizona had the right to use all the one million acre-feet apportioned to the Lower Basin under Article III(b) subject only to the “rights of New Mexico on the Gila River and Utah on the Virgin River.”  As the case proceeded, Arizona changed its tactics.  It amended its original claims for relief arguing that the case was no longer about the compact.  Instead, it was only about the intent of Congress to apportion water under the 1928 act.  The Special Master and Supreme Court agreed and ruled that the Colorado River Compact need not be interpreted to decide the case. Since the court’s decision was limited to the water used in and below Lake Mead and avoided the compact, it falls well short of a Lower Basin compact.

Would we be better off with a Lower Basin Compact?

The entire basin would be much better off with a functioning Lower Basin compact. The problem is that to get there, the five states with Lower Basin interests – Arizona, California, and Nevada (the “states of the Lower Division”), plus Utah and New Mexico, which also have watersheds falling within the Lower Basin’s boundaries – would have to negotiate through several difficult issues that have never been resolved and for the moment are conveniently tucked away.  These issues include dividing up the one million acre-feet of III(b) water, the “bonus water” provision added to the compact late in negotiations to sweeten the deal for Arizona. In addition to addressing the Lower Basin needs of Utah and New Mexico, Nevada also needs a piece of III(b) water to cover its uses on the Virgin and Muddy Rivers.  Water from the dry-up of previously irrigated lands in these two drainages supplies the Southern Nevada Water Authority.

A second issue is dividing up the substantial evaporation use on the Lower Basin mainstem reservoirs.  I have no doubt that the 1922 compact negotiators considered this evaporation a man-made beneficial consumptive use to be covered by the 8.5 million acre-feet apportioned to the Lower Basin, but it is not covered in the Supreme Court’s decision.

Third, there are difficult Lower Basin accounting issues to resolve, including how compact apportionments are to be measured. Yes, a century after the compact was signed, this fundamental issue has never been resolved (see Science Be Dammed, chapter 12). Another unresolved accounting  question is how to address the depletion of groundwater hydrologically connected to the Colorado River.

These accounting issues point to a bigger issue. The primary purpose of a Lower Basin compact would be to allocate among five states 8.5 million acre-feet of beneficial consumptive use – the amount apportioned to the Lower Basin by the 1922 compact. How would a compact address the reality that when tributary uses, reservoir evaporation, and hydrologically connected groundwater are considered, the Lower Basin is currently using more than 8.5 million acre-feet?

These major unresolved issues are why in Science Be Dammed John Fleck and I conclude that there is little incentive and major downsides for important playsers to a Lower Basin compact, thus it is unlikely to happen. The problem from my perspective as a former manager of an Upper Basin water agency is that through the 1948 Upper Basin compact, reservoir evaporation, hydrologically connected groundwater, and overuse are all addressed. Without a Lower Basin compact, important elements of the equity between the basins are missing.

The Gage Selfie Collection: Rio Grande at Albuquerque

One of our recent University of New Mexico Water Resources Program graduates suggested an extra credit assignment for this year’s students: stream gage scavenger hunt, with selfies.

Here’s the measurement point for USGS 08330000, Rio Grande at Albuquerque, NM. Flow at the time I took it yesterday morning measured 111 cubic feet per second.

Am I eligible for the extra points?

Record low flows on New Mexico’s Rio Grande

Record low flows on the Rio Grande at Embudo

Flow on the Rio Grande at Embudo in north-central New Mexico as I write this is measured at 156 cubic feet per second, which is the lowest for this date in history. And its quite a history – Embudo was the first gage installed by the USGS, back in the winter of 1894-95.

Note that I’ve switched to log scale for the graphs, which better distinguish the really flows. Of which 2020 is one.

Happy New Water Year, where’d all that Colorado River water go?

End of Water Year 2020

Shrouded in pandemic fog, I’m only now getting to my sorta annual “Happy New Water Year!” post, where I traditionally look on in alarm at dropping Colorado River Basin reservoir levels and make fun of the Lower Basin for using too much water.

The alarm remains – after a crappy runoff, combined storage in the two big reservoirs behind Lake Powell and Lake Mead is down 1.7 million acre feet from a year ago. For the ~40 million people depending on the Colorado, this is cause for concern. But the water use trends continue to bend in a way that has taken some of the steam out of my old jokes about users wasting all their bonus water on “hookers and blow”.

Reviewing the ’07 Guidelines

The start of water year 2020-21 offers a good moment to look back beyond a single year and see how the basin’s been doing.

Doing trend line analysis is always a rigged game, because so much of what you see depends on the arbitrary starting point you use.

But with the Bureau of Reclamation in the midst of its review of the basin’s 2007 Interim Guidelines, we have a non-arbitrary starting point at which to look at trends. 2007 is when the basin first adopted what we might think of as “drought operation rules”, aimed at reducing the decline of the two big reservoirs.

So, since 2007, how has the basin done?

Shrinking supply

From the supply side, things look bad. The river’s flows since 2007 have been ~1.2 million acre feet below the long term averages fed into the model projections done in support of the development of the 2007 guidelines, an 8 percent reduction in flow. The climate change shark is, as Brian Richter noted recently, circling nearby.

But despite that reduction in flow, total storage behind Glen Canyon and Hoover dams, the two primary system reservoirs, has dropped only 2.6 million acre feet. That is far less than you’d expect from 12 years of 1.2 maf per year flow reductions alone.

That kind of a flow reduction should have been enough to nearly empty the reservoirs. Why hasn’t that happened?

Shrinking demand

Because we also have been using less water.

Municipal and agricultural use of Colorado River water since 2007 has averaged 1 million acre feet per year less than the projections used to underpin the analysis done in support of the 2007 interim guidelines. With growing water conservation efforts, 2020 water use is on track to be 1.5 million acre feet below the demand projections used in the 2007 guidelines’ analysis.

Across the Colorado River Basin, among both Upper Basin users and Lower Basin users, folks are using less than their legal entitlement under the Law of the River. This is not enforced conservation. This is folks realizing they’ve got less water, so they’re using less.

As regular readers of this blog know, I think a better understanding of this demand side change is crucial to mapping out our policy options for keeping the reservoirs from collapsing. Here is how a group of academics (myself, Anne Castle, Jack Schmidt, and Doug Kenney) put it in a May 1 letter in response to the Bureau’s request for comments on the scope of its current review of the Interim Guidelines (full letter here):

The bottom line

In fact the modeling done for the ’07 guidelines, when you look at the projections for reservoir levels for the last 12 years, looks pretty good. The levels of both Powell and Mead are well within the range of likely scenarios the Bureau modeled for the Interim Guidelines EIS – a bit below the median, but well above the “OMG SCARY” scenarios the Bureau considered as unlikely but plausible. Our misunderstandings of both supply and demand were, to first order, offsetting errors of roughly the same magnitude. But as we look back in preparation for looking ahead to the next round of negotiations, it’s important to to dig into the details of what we got right and wrong in that remarkable ’07 planning exercise, and why.

nota bene: A huge thanks to Jack Schmidt and Jian Wang at Utah State University for their analysis of Upper Basin demand projections, and especially to Jian for open-sourcing the underlying data.

The “Colorado River Simulation System” and Elinor Ostrom’s “authoritative image of the problem”

Ostrom 1990, Governing the Commons

Burnishing my notes for UNM Water Resources class this afternoon to talk about Elinor Ostrom, I spent a bit of my morning going back through the underlined bits in my copy of her seminal book Governing the Commons. I first read it in the fall of 2009, when she won the Swedish prize. My initial scratchings in the book – Wait, what? They imposed constraints on themselves? – are instructive.

My rereading of Ostrom comes as I’m in the midst of a little side project digging into the history of something called the “Colorado River Simulation System”, a computer model that is central to 21st century Colorado River management. CRSS, as it’s called, brings together the best hyrdologic understanding of how the basin operates with policy rules, serving as a sort of common language for talking about the river and making decisions about its future.

It is impossible to overstate the impact of the intellectual light bulb that fall when I first read Ostrom. I’d recently finished my first book and was eyeing taking a stab at writing something about the Colorado River. Lake Mead had been steadily dropping and, fully steeped in the “tragedy of the commons” narrative, I imagined that my task was to chronicle its collapse. The federal government did not seem up to the task of imposing a centralized authority on the system, and the “tragedy of the commons” conventional wisdom suggested water users would be unwilling to impose constraints on themselves?

Yet here was Ostrom, as characterized by the Nobel committee’s press release:

Elinor Ostrom has challenged the conventional wisdom that common property is poorly managed and should be either regulated by central authorities or privatized. Based on numerous studies of user-man-aged fish stocks, pastures, woods, lakes, and ground-water basins, Ostrom concludes that the outcomes are, more often than not, better than predicted by standard theories. She observes that resource users frequently develop sophisticated mechanisms for decision-making and rule enforcement to handle conflicts of interest, and she characterizes the rules that promote successful outcomes.

Crucially, Ostrom argued that such successful problem solving was not a given. But she offered an intellectual toolkit for understanding what to look for.

Central to that is what, in Governing the Commons, she described (in her case study of Southern California’s Raymond Basin) as “a single, authoritative ‘image’ of the problem.” This amounts to a common, data-based understanding of the system on which decisions can be based, scenarios played out.

In the years since, I have come to understand the way CRSS plays that role. A sophisticated model built on the Riverware platform, it is used by the Bureau of Reclamation, major water agencies across the basin, environmental groups, and university researchers to study the implications of drought, climate change, and water policy options as we navigate the Colorado River Basin’s future. It is, to borrow a phrase from one of the people I spoke with while working on my little dive into CRSS’s history, the common language we speak in the basin.

Want to know what the risk of Lake Powell dropping to levels at which Glen Canyon Dam can no longer generate electricity? The Colorado River District’s Risk Study used CRSS to do that. Want to know the risk of a continuation of the dry conditions of the 21st century continuing, or worsening? We have the “Stress Test” and the “Super Stress Test”, simulations written in the language of CRSS, a language that basin managers can understand, to help us. Want to evaluate climate change scenarios against paleo droughts? Homa Salehabadi and colleagues at Utah State have done that, offering their answers in CRSS, a language decision makers can understand.

It is not without its limitations. It constrains the questions you can ask, and I watch people struggle now to bend it to the task of sorting out things like environmental flows through the Grand Canyon, or better understanding the Upper Colorado River Basin’s consumptive uses, past and future. Some of the constraints are technical. Some are political. But for the basic task of helping the basin’s water management community understand its choices, providing an “authoritative image”, it continues to perform admirably.


Ghost of Water: The Inauspicious end of the Alameda Lateral

Alameda Lateral, Albuquerque, New Mexico, 2020, by John Fleck

I’ve driven by the spot in the picture a jillion times in the 30 years I’ve lived in Albuquerque, and never noticed the ditch squeezed between #1 Plumbing and Air and the Chevron station on the corner of Edith and Candelaria.

My latest pandemic bike riding project involves scouring GIS data from the Middle Rio Grande Conservancy District, the irrigation district in our part of the Rio Grande Valley, in search of abandoned ditches. The last few weekends my friend Scot and I have been piecing together the remnants of the Barelas Ditch south of downtown. This weekend we turned our attention to the Alameda Lateral, which used to irrigate the east side of what we call Albuquerque’s “North Valley”.

The air was smoky today, so it was a perfect day for the sort of lazy stop-and-go riding needed to find traces of old ditches – folded up paper maps with scribbles in my pocket along with my oatmeal cookies.

MRGCD Alameda Wasteway control structure

At the spot where I took the picture, the Alameda is just a wasteway, no irrigation water because no irrigated land “downstream” of the trash rack in the picture. A block or so downstream, it’s completely underground, but there’s an incongruous irrigation system control structure flanked by busy street, apartment complex, and construction company equipment lot. You can’t just plug an irrigation canal, you have to have a way to let the water drain out the bottom end. So whenever the Alameda downstream from here was formally abandoned, they buried a wasteway conduit to carry what was left over to the big open air “Alameda Interior Drain”, one of the most visible water features in this part of the North Valley.

Scot and I spent a bunch of time trying to find the upstream bits of the canal (our Strava GPS from the ride is hilariously squiggly), riding up and down the neighborhood streets. A lot of it is underground, passing beneath the light industrial stuff that long ago displaced the marginal agriculture that used to scratch by in this valley.

My GIS skills aren’t great, but here’s my attempt at showing this segment of the ditch:



Six Colorado River Basin States to Interior: Don’t Allow Utah to Blow up Basin Collaboration

The six Colorado River Basin states that do not have the letters “U-T-A-H” in their names just sent a remarkable letter to Secretary of the Interior David Bernhardt with a plea – don’t let the rush toward federal approval of Utah’s proposed Lake Powell Pipeline blow up the Colorado River Basin’s framework of collaborative rather than confrontational problem solving:


The six-state letter, the product of intense discussions in recent weeks among the states (including the one with “U-T-A-H” in its name) takes great pains to point to an important historical norm in Colorado Basin governance – states don’t mess in other states’ internal water use decisions. But in asking to move Upper Basin water to a Lower Basin community, Utah has crossed a line that the other states simply couldn’t let pass.

The full letter is worth a read.

How municipal water conservation is keeping the Rio Grande through Albuquerque from going dry

Rio Grande at Albuquerque’s Central Avenue Bridge, Sept. 6, 2020

One of the traditional “tragedy narratives” of western water is the idea that thirsty cities are draining our rivers. But in two of the last three years, precisely the opposite has happened here in Albuquerque.

We’ve been limping along on a very bad year on the Rio Grande, with some of the lowest flows through Albuquerque that we’ve seen in a while. And the limping will continue. But with irrigation water in storage just about gone, an agreement is taking shape that will use an unused chunk of Albuquerque’s imported Colorado River water to keep the Rio Grande from drying through Albuquerque in coming months.

This is possible because Albuquerque’s water conservation success has left it with more water rights than it currently needs, including water we import through the San Juan-Chama project, a transbasin diversion that brings Colorado River water through tunnels beneath the Continental Divide. Some of that, now sitting in storage in reservoirs up on the Chama, will be released in coming weeks to maintain flows in the river here in town.

A similar deal in the very dry summer of 2018 also used some of Albuquerque’s unused Colorado River apportionment to keep the Rio Grande wet.

To be clear, this isn’t a charitable contribution on Albuquerque’s part. As I understand the deal, three government agencies with a shared interest in keeping the river wet – the Middle Rio Grande Conservancy District, the New Mexico Interstate Stream Commission, and the U.S. Bureau of Reclamation – are paying the Albuquerque Bernalillo County Water Utility Authority for the water.

Also to be clear – the environment and our community’s cultural values around a flowing river in our midst will not be the only beneficiary. After what water managers call “system losses” – the trees and the shallow aquifer using a share of the water as it wanders downstream – some portion of the water will arrive at an MRGCD dam south of Albuquerque, where it can be diverted for use by farmers south of town.

Also to be clear – absent upstream dams, which held back a bit of the high spring flows for use over the summer, the Rio Grande through Albuquerque would have this summer long since been dry. This is such an altered system that in a year like this, it’s only human water management that is keeping anything flowing here at all.

Also to be clear – the only reason we’ve got this extra water at all is because we divert large quantities away from the Colorado River, which has problems of its own.

But it’s intriguing to see the traditional narrative turned on its head – water available for the environment because a city has more than it needs.


San Diego’s puzzling pursuit of a big new pipeline to the Colorado River

I’m puzzling over the San Diego County Water Authority’s pursuit of a new Colorado River pipeline.

I’ve been puzzling for a while, given a that it would be really expensive and that a really big pipe (aqueduct) already exists to carry the water to San Diego. My puzzlement was goosed by a report that surfaced last week at a board meeting of one of its member agencies suggesting that the general managers of agencies representing the majority of the Water Authority’s actual water-using member agencies don’t seem to want it.

The puzzle

In 2018, the San Diego County Water Authority did something striking, given the water management community’s tendency toward inflated future demand projections. It calculated something it called a “demand reset”, involving a fresh look at demand projections published just two years earlier.

rethinking San Diego’s water needs

Demand calculations published in the agency’s state-mandated Urban Water Management Plan, the “demand reset” analysis found, had inflated near-term water demands (water needed by 2020) by 9 percent.

So how has the new “demand reset” analysis held up? Actual water use in 2019 was 14 percent lower than the projection made in 2018, just one year earlier.

To be clear – the new study, done out of fears that the last study had overestimated demand, itself overestimated demand.

This has been going on for a while. Since its 2002 peak, total San Diego County Water Authority use has declined by 27 percent, even as the county’s population has risen by 15 percent. See decoupling.

I was grinding through these numbers as I puzzled over the San Diego County Water Authority’s scheme to build a new pipeline across the desert to carry its supplies of Colorado River water – ~280,000 acre feet a year of water it got through various agricultural efficiency projects in Imperial Valley.

That water currently is delivered via exchange agreement through the existing Metropolitan Water District Colorado River Aqueduct. So San Diego already has a way to get the water. But an arcane legal and financial and cultural feud (c’mon L.A., know you hate the Padres) between the San Diego County Water Authority and the Los Angeles-based Metropolitan Water District of Southern California has led the San Diego County Water Authority to pursue a pipeline of its own, so that it would no longer be beholden to Met.

I’ve been careful in my language here to talk about “the San Diego County Water Authority”, rather than “San Diego” writ large. The Water Authority is a creation of state law that serves as a water wholesaler and system integrator across territory served by 24 member agencies, large and small. As is often the case with regional water authorities of this sort, its board of directors is made up of representatives of each of the member agencies.

Conceptual map for possible routes for a new San Diego pipeline to carry Colorado River water

In the summer of 2019, on a trip to Imperial Valley to gather stuff for a new book I’m writing hoping to start working on again when the pandemic fog clears, I heard a fascinating presentation about the project from San Diego County Water Authority Assistant General Manager Dan Denham to the Imperial Irrigation District board. Denham laid out the Water Authority’s basic argument – that with anticipated increases in the cost of getting its water through the existing Metropolitan system, it makes sense for the Water Authority to consider building its own pipe. In a net present value calculation, the Water Authority has been arguing, it would be cheaper. (The minutes are here if you’re curious, the questions were interesting.)

Here’s the fresh part of puzzle.

The idea that this is cost effective in the long run, being pushed hard by San Diego County Water Authority management, does not seem to be shared by the authority’s member agencies. 18 of the 24 member agencies, representing by my calculation 89 percent of the Authority’s water use, went out and hired their own consultants to take an independent look at the numbers. Their conclusion (which surfaced last week at a meeting of the Vallecitos Water District board, the report starts on pdf page 82):

The Draft Study’s finding that the project is economically competitive with other supply and transportation options is not reasonable. We find the project to be substantially more costly than other options.

Dueling consultants’ reports followed, and I’m not going to wade into the question is which team is right. I will note, rather, that the managers of the 18 water agencies (including the City of San Diego, by far the region’s largest water user) were sufficiently concerned about the numbers coming out of the Water Authority that they went out and hired their own consultant.

San Diego’s dream of its own Colorado River pipeline has been around since the beginning of water development in the region. In its current incarnation, the project has been at least somewhat actively pursued since 1998. So it’s probably never going to go away. But I’m adding this to my “yeah, but it also probably will never be built” list.

(H/t to Voice of San Diego’s Ry Rivard for his work on this topic over a number of years.)