Large language models, Homer (Simpson) and my writer’s brain

Posted by on

a Homeric odyssey

Like everyone else of a certain sort (on line and writerly), I’ve been playing with ChatGPT.

It’s a “large language model”, trained on a monstrous corpus of text, that basically projects (based on stuff people have said before) what might come next given a sequence of text. It doesn’t really “understand” it in a way that we understand “understanding”, but it’s pretty good at this point in saying stuff that sounds plausible and is sometimes useful.

As a writer, I’m fascinated by the mystery of “what comes next” out of my own writer’s brain.

Me:

What is the connection between Homer’s Odyssey and Homer Simpson?

ChatGPT:

Homer Simpson, the main character in the popular American animated sitcom “The Simpsons,” is named after the ancient Greek poet Homer, who wrote the epic poems “Iliad” and “Odyssey.” The name “Homer” is a reference to the epic poet, and the character’s traits and experiences often reflect elements from the stories in the “Odyssey.”

That’s pretty good stuff! I have no idea if it’s true or not – sounds like bullshit to me. One of the things we’ve learned in playing with ChatGPT is that large language models are superb bullshitters. I suppose I could Google it.

The fun thing for me as a writer is the “Holy moly, where did that come from?” feeling I sometimes get when I’m sitting at a keyboard and my fingers are flying.

When the work is flowing (which is often, I’m not one of those suffering “Woe writing is hard!” writers), my “writer’s brain” is like a playful, faithful golden retriever finding great sticks for me. I smile and pat it on the head and type up the stick, not much caring how the dog found it or where the next one is coming from.

Sometimes the sticks are weird and misshapen, but I don’t care because my dog brought them to me and they are therefore lovely by default!

Wikipedia and the large language model golden retriever that is my writer’s brain

The weird thing about this is that the my brain/golden retriever distinction is really sometimes how it feels. Like, “Where did that come from?” It’s not some muse or spiritual force. It’s like a pet dog! Which is why playing with large language models is so intriguing. I don’t usually have access to someone else’s golden retriever.

I have no idea if this is right, but as I’m in the business of trafficking in metaphor, I’ll cling to it for at least a moment. Let’s go through an example.

Consider Monday’s post on Homer Simpson, the Odyssey, and the difficulty of solving the Colorado River’s problems.

I’d planned to spend the morning working on my new Rio Grande book, but my golden retriever writer’s brain was feeling playful, and I agreed to take it for a walk.

I wanted to do a quick primer for readers about what’s up on the River. The premise was that there’s a bunch of stuff going on simultaneously, which makes solving the Colorado River’s problem a capacity challenge, and the first stick the dog brought me was the “walk and chew gum at the same time” cliche. That’s a nearby stick, I probably could have found it by myself. But I wanted another stick – I love to take a cliche and twist it in a way that makes you notice it. The golden retriever without evening stopping for a pat on the head and a “good dog” zoomed out and returned with Homer’s Odyssey – a long story, recited in an oral tradition, which is hard to do while you’re chewing gum.

Good dog!

This is the fun part. I have no idea where stuff like that comes from! I just smile and type. But the large language model probability thing gives me a nice mental model for what my writer’s brain is up to, sorting through its neural network for “the sort of thing that might come next.”

When my dog brings me a stick, I like to check it out, and my understanding of stuff like ancient Greece, like much of the material I pretend to know, is modest. So I went to the Google. Which gave me, right out of the box, the “Homer’s Odyssey” episode of the Simpsons! (Google’s a good dog – not up to the standards of my golden retriever, you really have to tell it where to look for sticks, but a good dog.)

The “truth” part is the sort of thing ChatGPT isn’t very good at yet – excellent at stringing together words that might reasonably follow, that sound plausible, but not so good about knowing if the words are right. The old journalist’s cliche, “If your mother tells you she loves you, check it out” applies to ChatGPT and also my dog’s sticks. The stodgy, plodding Google search is great to have at moments like this.

The meat of the post was straightforward – short term EIS, longer term EIS, etc. – but the real fun was yet to come.

In search of a Homer Simpson kicker, my golden retriever writer brain somehow rummaged up the old 2013 Eric Kuhn/Dave Kanzer* Homer Simpson Colorado River Stress Test CRWUA slide! Again, I’m in admiration of my dog’s fetching skills, like where the hell did you find that?

* I tried calling Eric to confirm, couldn’t reach him, but one of the sticks my retriever fetched was about Dave having something to do with that slide. (Some stuff I can’t Google and have to call Eric.)

** Eric called back. DK was, in fact, instrumental in the Homer Simpson Stress Test Slide, one of the great contributions to the 21st century Colorado River canon.

Deadpool Diaries: The numbers in the states’ two proposals

Posted by on

states’ proposal for Colorado River cuts. Lake Mead elevation along the x axis, millions of acre feet of cuts along the y axis

Getting ready for an interview this morning with Mark Brodie at KJZZ (waving at my Phoenix friends!) I put together a table to make it easier to compare the six-state proposal submitted Monday to reduce Lower Colorado River Basin water use, and the California proposal submitted yesterday (Tues. 1/31/23).

Perhaps worth sharing here? “Elevation” is Lake Mead elevation, the numbers are million acre feet of total cuts.

Two keys to note.

First, despite big disagreements about how to approach this, we have unanimity among all seven states that very deep cuts in Lower Basin water use are needed. At the lowest Lake Mead elevations, the numbers are similar.

The difference is in timing. California’s cuts don’t kick in until later – essentially a gamble on good hydrology once again helping us avoid conflict by letting us use more water in the short term.

The six-state proposal says “go big” any time Mead drops below 1,050. The California proposal doesn’t start “going big” until 1,025.

The six-state proposal yanks the bandaid off now.

Under the current “most probable forecast” for the coming year, we’d end up in 2024 with:

  • Six state proposal: 3,168 million acre feet in cuts
  • California proposal: 2,188 million acre feet in cuts
Tier Elevation 6-state California
Tier 0 1090 1,784 1,241
Tier 1 1075 2,156 1,613
Tier 2a 1,050 2,918 1,721
Tier 2b 1045 2,918 2,013
Tier 2c 1040 2,918 2,071
Tier 2d 1035 2,918 2,129
Tier 2e 1030 3,168 2,188
Tier 3a 1025 3,168 2,525
Tier 3b 1020 3,368 2,675
Tier 3c 1015 3,368 2,875
1,010 3,368 3,125
1,005 3,368 3,325

There are other differences too – huge disagreements on how to approach the allocation of the cuts! No time for that this morning, I’ve a book to write, but I hope to get back to that in the next few days, stay tuned.

As always, if you find this content useful, be sure to hit the “like” button, it really helps the channel! (There’s not really a like button, that’s just what YouTubers say, it’s a family joke after I started teaching my classes over Zoom back in the pandemic.)

Deadpool Diaries: Can the Colorado River community walk, chew gum, and recite Homer’s Odyssey at the same time?

Posted by on

The Colorado River Stress test, a Homeric odyssey

While we eagerly await whatever it is that might happen this week as the Colorado River basin states struggle to come up with a short term plan to use less water, the Bureau of Reclamation is inviting y’all to a webinar this afternoon (Monday Jan. 30 2023, details here) to begin thinking about a long term plan to use less water.

It’s a crazy time, and I worry about our collective capacity, but the river can’t wait, so buckle up!

A brief refresher is perhaps in order

The Supplemental EIS

I emerged from the writing cave (new book underway about the Rio Grande, which is a mostly a different river entirely) to share my thoughts about this week’s “deadline” (which as I explained isn’t really a “deadline”) for the seven Colorado River basin states to come up with a plan for managing the river for the next several years. This is a short-term effort, an attempt to limp through the 2025-26 time frame without breaking things. It requires temporary rules to reduce water use as needed in the Lower Basin, maybe some water use reductions in the Upper Basin, and tweaks to the reservoir operations rules to keep from breaking Glen Canyon Dam.

“EIS” here stands for “Environmental Impact Statement”, the process by which Reclamation will analyze our choices before picking one.

The key words here are short term.

The Real EIS

Post-2026, we need a much more robust and long-lasting framework for using less water and not breaking the dams and trying to respect tribal sovereignty and our evolving societal values around respect for the environment in the face of climate change stealing a bunch of our water.

In that regard, Reclamation has launched an expansive effort to help us collectively, as a society, think through these options.

A bunch of us wrote them letters last year telling them what we thought they should think about. They’ve summarized them nicely (pdf here). My favorite part is the people from Costa Rica and the UK who weighed in. This is a far-reaching issue.

The connection between the Supplemental EIS and the Real EIS

One of the difficulties in sorting out the near-term plans is that everyone’s angling for the high ground in the long-term plans. There’s a fear among water managers that if in the short term they demonstrate that they’re able to get by with less water, they’ll get screwed long term. A lot of what we figure out in the short term will echo into the long term.

Homer’s Odyssey

In season one of the Simpson’s, there’s a great episode called “Homer’s Odyssey” where Homer Simpson gets fired from the nuclear power plant and then becomes a citizen safety advocate who gets speed bumps and stop signs installed in Springfield, and Homer becomes a revered community leader, and Mr. Burns hires him back to become the chief safety officer at the nuclear power plant.

You didn’t think I meant reciting the entire Homeric epic, did you? I fear one episode of the Simpsons is the most we can hope for right now.

Picture courtesy Eric Kuhn’s 2013 presentation at the Colorado River Water Users Association.

Deadpool Diaries: Trapped, again, in a world we never made

Posted by on
abandoned boat at Lake Mead

Trapped in a world we never made.

As we get spun up for the second time in six months about a capricious notion of a “deadline” to fix the Colorado River, I’m reminded of the tagline from my favorite teenage superhero comic, Marvel’s Howard the Duck:

Trapped in a world he never made.

Once again, we are trapped in a narrative driven by a somewhat arbitrary “deadline” that misunderstands the nature of the ongoing processes as the collective community struggles to come up with a plan to save the Colorado River.

About that “deadline”

HTD#1, 1976

I’ve been absorbed in recent weeks writing about 1920s Albuquerque and the Rio Grande, so I’ve not had the time to do my usual “keep up with Colorado River stuff” that is typically a prominent feature of this blog, requiring me to read stuff and call a bunch of people, kinda like a journalist but as a hobby rather than a job with an editor glaring at me from across the room waiting for me to file.

But it’s been hard to avoid the spun up media attention to next week’s “deadline” for the Basin States to come up with a plan to…. Well, I’m not sure exactly what it needs to do, that’s part of the challenge.

Media coverage demands what we call “news pegs”. A Jan. 31 “deadline” for the seven Colorado River basin states to deliver what we all hoped would be a “consensus proposal” to reduce Lower Basin water use, reduce releases from Glen Canyon Dam to protect the infrastructure there, and save the West, and if the states don’t save themselves the feds will have to step in an impose their authority, provides such a news peg, a thing for reporters to pitch as a news peg to editors. And pitch it they have, with good reason! This is a really important story!

Thus we have Colton Lochhead last Thursday in the Las Vegas Review Journal: “Colorado River water managers optimistic about drought plan as deadline looms.”

Or Christopher Flavelle a day later in the New York Times: “The seven states that rely on the river for water are not expected to reach a deal on cuts. It appears the Biden administration will have to impose reductions.”

Wait, what?

A close read shows that the two stories, despite apparently contradictory headlines, are not inconsistent.

Flavelle’s pointing to the unlikelihood of a seven-state deal, which seems right. The AP’s Kathleen Ronayne did a great job with this part in a piece that ran this weekend:

California says it’s a partner willing to sacrifice, but other states see it as a reluctant participant clinging to a water priority system where it ranks near the top. Arizona and Nevada have long felt they’re unfairly forced to bear the brunt of cuts because of a water rights system developed long ago, a simmering frustration that reared its head during talks.

Thus in all the palavering over the last six months, California has been the odd state out, making it unlikely that you’ll see a seven-state agreement on how to reduce Lower Basin use and juggle releases from Lake Powell to reduce the risk of breaking Glen Canyon Dam.

Lochhead nicely captures this nuance in his story, because there’s nothing magic about a deal all seven states can sign:

Chuck Cullom, executive director of the Upper Colorado River Commission, which is made up of representatives from Wyoming, Colorado, Utah and New Mexico, said he expects that the states will reach a consensus for submitting a proposal before the Feb. 1 deadline, but whether every state will be able to endorse the approach remains less clear.

Thus we see the foundation for a hilarious conversation I had a few weeks back with one of the players in all of this about the meaning of the word “consensus”. What if something emerges that most of the states like, but not all seven?

Howard the Duck and Interior’s existential dilemma

For the second time in six months, we find our heroes at the Bureau of Reclamation and the Department of the Interior existentially trapped, like Howard, in a world they never made.

By “world they never made” here, I mean the rising frenzy over a deadline next Tuesday after which….

We’ll, I’m not sure what happens after Tuesday, let’s explore….

Why are we all (I’ve been doing this too!) reporting on a looming “deadline”, and concluding that if it is not “met”, as Flavelle reported in the grey lady, the federal government will have to step in and impose cuts? It’s really just a vague marker in a complex timeline needed to get a draft Environmental Impact Statement out by early April, and a final decision document done by summer.

There’s nothing firm about Jan. 31, a magical coach that will turn into a pumpkin if the states don’t agree on a plan by midnight. (One plus of the blog format is that there is no editor across the room glaring at me for mixing Howard the Duck and Cinderella.)

Last summer, the last time we played this game, the idea of a “federal deadline” took on a life of its own, setting off both a flood of stories about a “missed deadline” and a really productive process through the fall and winter that has moved us closer to clarity about what sort of cuts will be needed and how they might be apportioned. The Southern Nevada Water Authority has very publicly presented a proposal worth modeling, which has drawn interest from other states as a foundation for working out the issues. Other states (though not publicly) have offered up proposals of their own.

These exist! Reclamation can model them and tell us how they work under various hydrologic scenarios!

All of these can now be rolled into the modeling efforts being done by Reclamation in the next few months to help us all understand the “if this then that” questions about how we manage the system over the next few years to protect public health and safety and the dams.

Given the current frenzy over a looming “deadline” and the risk of failing to meet it, it’s worth returning to the original language in the Department of Interior’s federal register notice launching this process.

It laid out three scenarios the agency hoped to evaluate.

The first is a “no action” alternative which basically says fuck it, let’s just crash the system.

The second is the one we’re currently all spun up about, so it’s worth reading the actual language of the notice:

Framework Agreement Alternative: This alternative would be developed as an additional consensus-based set of actions that would build on the existing framework for Colorado River Operations. This Alternative would likely build on commitments and obligations developed by the Basin States, Basin Tribes, and non-governmental organizations that were included in the 2019 DCP.

No deadline here, no requirement for all seven states to sign on. Suitably vague, to allow the players the room to move as they try to come up with something workable. Which is what is now happening.

And finally….

Reservoir Operations Modification Alternative—This alternative would be developed by Reclamation as a set of actions and measures adopted pursuant to Secretarial authority under applicable federal law. This alternative would likely be developed based on the Secretary’s authority under federal law to manage Colorado River infrastructure, as necessary, and would consider any inadequacies or limitations of the consensus-based framework considered in the above alternative. This alternative would consider how the Secretary’s authority could complement a consensus-based alternative that may not sufficiently mitigate current and projected risks to the Colorado River System reservoirs.

Look, I’d be delighted if the seven states had settled on something on which they could all agree and handed it off Tuesday to Interior to model.

Following CRWUA, I was optimistic about the generally favorable response to Southern Nevada’s proposal.

I am discouraged that the continuing tension highlighted in a number of the stories between California’s stubborn insistence on its interpretation of its seniority under the Law of the River and the more adaptable interpretations being suggested by others in response to impacts of climate change unforeseen when the laws were written 60, 80, 100 years ago has made it impossible to get something done by next week. This may be the disagreement that leads us toward litigation and the “no action” alternative, see bad words before about crashing the system.

But it’s clearly not the case that it’s game over for a collaborative solution come Wednesday, that we’ll just leave it to the executive branch of government to save us (the water users across the basin) from ourselves.

The notion of a “deadline” has taken on a life of its own that is clearly unmoored from the reality of what actually needs to be done.

Support Inkstain

I drink coffee. You can help.

Ribbons of Green: The Hubbell Oxbow

Posted by on

The Hubbell Oxbow, a “lake” in Albuquerque’s South Valley turned flood control channel (with bonus cottonwood grove)

The big farms we have left in Albuquerque’s South Valley are weird.

Riding the old Hubbell “Lake” and Anderson Farms – red is the Sunday bike ride

I spent a good fraction of my weekend staring at maps of them, or riding my bike around them, or both.

My co-author Bob Berrens and I have zeroed on in this area for a key part of the storytelling in our new book, Ribbons of Green. It has it all:

  • the old Camino Real running up the river’s edge
  • an enduring indigenous Pueblo to the south
  • two enduring (?) Spanish-era villages
  • 21st century farm land that is crucial to our narrative precisely because it is not enduring

It’s this last bullet that was the focus of my latest squiggly Sunday outing.

Gardens

The book will be published (once we write it!) as part of the University of New Mexico Press’s New Century Gardens and Landscapes of the Southwest series.

The notion of “gardens” came to us after the notion of the book, which is an effort to tell Albuquerque’s story in terms of our community’s relationship with its river. By our “relationship” with the river, we do not mean how we feel about the river today, how we interact with it, in its modern form. Rather, we’re trying to tell the story of how we came to build our city in the bed of the river itself. Because the Rio Grande of today, pinned between levees, bears little relationship to the river the Puebloans, the Spanish, and the early Americans confronted when they set about to build a city here.

Here’s a squib from the book’s opening chapter:

Through its history, the English language word “garden” has done yeoman’s work, traveling with us as we made a modern world. At its simplest, it is a noun describing a place out behind the house where we grow flowers, and vegetables, and perhaps a few fruit trees. At its most expansive, it is “a region of great fertility.” Kent was “the Garden of England”, “known for its abundance of fruit and crops”. The province of Touraine was “the Garden of France.” But it is the noun’s interplay with garden’s verb form that does the word’s linguistic heavy lifting – “to bring a landscape into a particular state” – to change our world.[ “garden, n.” OED Online. Oxford University Press, December 2021. Web. 27 February 2022; “garden, v.” OED Online. Oxford University Press, December 2021. Web. 27 February 2022.]

It is this notion of “bringing a landscape into a particular state” that has brought me, again and again, to the neighborhood around the Walmart at the corner of Coors and Rio Bravo in Albuquerque’s South Valley.

Hubbell Lake

Hubbell Lake, 1920’s Plane Table Maps, sheet F6P113, courtesy MRGCD

Bob and I have been working with a bunch of old maps, trying to understand the evolution of that landscape as the community that was Albuquerque-in-becoming developed the institutions that eventually made it possible to put a Walmart in a spot that, as recently as the 1920s, had been labeled “lake” on the maps of the day.

The largest farms left in the Albuquerque reach of New Mexico’s Middle Rio Grande are here – much of the land in public ownership, leased to alfalfa farmers selling their crops to local equestrians, an effort by the community to preserve an agricultural heritage that looked nothing like this.

And not here. The agricultural heritage of this landscape is a recent creation, only made possible when the community added what looks kinda like Dutch technology – dykes to keep the river in its place, and drainage to lower the water table from the land beyond.

This was a flood path, a sweeping bend in a secondary river channel, inundated in high flows, that was used in the time before as pasture and a duck hunting club.

Thus does “lake” become farmland and a Walmart. It’s a very 20th century innovation. Before we did all this stuff, this stretch of the valley was salt grass marsh and “lake”.

This is the project of our book – to explain the processes of collective action that brought this particular landscape into this particular state – “gardening”.

A Lovely Grove of Cottonwoods

On our Sunday bike ride, my friend Scot and I rode into the Hubbell Oxbow from the west. The land is outside Albuquerque’s city limits, but is owned by the city’s Open Space program. It’s not really open to public access, but it’s easy to get in via the Middle Rio Grande Conservancy District’s Gun Club Lateral, the irrigation canal built in the 1930s (? – further research needed, but we think this timing is correct ?) and/or the Albuquerque Metropolitan Arroyo Flood Control Authority’s Hubbell Channel, built in 1978. (I’m more confident about that date, which I got from the AMAFCA GIS database we’re using.)

Scot’s an incredibly important unindicted co-conspirator on the project, assisting in the crucial research task of aimlessly riding our bikes around the valley floor looking at stuff.

Our current work in the South Valley is thorough.

Looking at stuff is key.

The stuff we found Sunday, slipping into the Hubbell Oxbow from the back, was lovely. Wrapping around the back of the farm, protecting the farmland itself with a levee, is a flood control channel and basin that have, in fine AMAFCA tradition, become a greenspace. I was about to type “community greenspace”, but that doesn’t seem quite right if, by “community”, we mean a place where the locals walk their dogs and ride their horses and bikes. Despite the easy access, it doesn’t seem to get much of that sort of use. To quote the city, “The property does not offer formal access.” (emphasis added)

But if by “community” we are comfortable attaching what the economists might call an “existence value” – a thing that we value simply in the knowledge that it exists, rather than a “use value”, which we value via horse and bike and dog walk, then yes: “community greenspace”, I guess.

This spot is a classic type section of the “ribbons of green” of our book’s title.

 

the rise and fall of “the flood menace”

Posted by on

via Google Ngrams, the rise and fall of the flood menace

Doing reading for the new book on early 1920s Albuquerque, as business leaders pursued what would become the Middle Rio Grande Conservancy District, I see the regular return of a phrase I’d come to see frequently my reading of Colorado River history in the same time period:

the flood menace

Here’s the Albuquerque Journal, reporting on a June 1921 meeting of the Albuquerque Chamber of Commerce, where leaders were gathering support for the District’s creation:

It was further made clear that unless this project is carried through to completion there will be a depreciation of production and land values and ever increasing menace from flood waters.

I love Google Ngrams for this stuff. It allows you to track the rise and fall of a word or phrase in the corpus of English-language writing.

Despite having built a city on the Rio Grande flood plain, we don’t much worry about the “flood menace” these days. But boy, howdy, was it on people’s minds here (and elsewhere) back in the 1920s!

 

Lukas: “really low #ColoradoRiver flows are off the table. “

Posted by on

Jeff Lukas:

 

Dead Pool Diaries: Climate change, the doctrine of prior appropriation, and the Colorado River crisis

Posted by on

A desert landscape. Corrales, New Mexico, January 2023. Photo by John Fleck

Writing in 2018 in the Seattle Journal of Environmental Law, Kait Schilling argued that the doctrine of prior appropriation – the notion that those who first put water to use hold priority over those who came later – was no longer compatible with a climate-changed world.

Climate change is diminishing water rights equally regardless of date of appropriation. Such a phenomenon makes the “first in time, first in
right” rule difficult to grapple with because right holders will be unable to access their water to its fullest extent. Because every human has a right to fresh water, the first in time, first in right mentality can no longer be sustained with the current state of climate change and population growth.

The two sides of the argument:

  1. equity requires sharing the pain across all water users – seniors (mostly farms but also Native American communities) and juniors (mostly cities)
  2. to respond to climate change-induced shortages, we need to cut off juniors, or make them compensate seniors for the water they need

This debate is the narrow eye of the needle we’re trying to pass through right now in the rapid-fire negotiations underway to deal with this Colorado River crisis.

The sticking point in the current Colorado River negotiations

In a letter submitted to the U.S. Department of the Interior in December, Arizona’s water leadership – the Department of Water Resources and Central Arizona Water Conservation District – argued for “1”. (Huge thanks to Daniel Rothberg at the Nevada Independent for collecting and posting the letters and also writing some smart stuff about what’s in them – journalism as a public good. The context here is Interior’s request for scoping comments on the agency’s crisis management Supplemental Environmental Impact Statement – SEIS):

All water users share risk from these conditions and the SEIS should ensure that the burdens associated with managing that risk are shared across all sectors and by all water users.

In their letter to Interior, the staff of the Colorado River Board of California argued for “2”:

Finally, some across the Basin have advocated for Lower Basin water users to be individually assessed for reservoir evaporation, seepage, and other system losses. The Board recommends that these losses continue to be treated as a diminution of available annual supply, which can then be met through application of the Law of the River as supplemented by voluntary agreements.

I plead guilty to a misleading oversimplification here, because in their arguments both Arizona and California are making a broader argument about equitable sharing of both the impacts of climate change but also the underlying problem of the river’s overallocation. In defense of my oversimplification, I’ll simply assert that it is the “hot” part of our “hot drought” (see Udall and Overpeck 2017) that makes the difference between the successful gradual process of negotiation we’ve using for more than two decades (see my book Water is For Fighting Over) and the crisis management of 2023.

Could a decent snowpack widen the eye of the needle we must thread?

An improving forecast. Source: Colorado Basin River Forecast Center

Curled up with my morning coffee this morning  (with a huge thanks to the supporters of this blog who bought it for me), I was happy to note that the snowpack is decent right now, and therefore the Colorado River runoff forecast, is up 650,000 acre feet from Jan. 1. That’s an extra ~10 feet of elevation in Lake Powell, which is ~10 feet farther from the white hot fire of crisis management at Glen Canyon Dam.

One argument here is that even a decent year, by lifting the pressure on Glen Canyon Dam -> more water to Lake Mead -> less pressure for really deep cuts now -> less risk of litigation. (I have my own views on this argument – I disagree with it, which I’ll explain below – but I’m trying to do the “view from nowhere journalist” thing here, and I want to give the people I disagree with their best shot, because the argument is not unreasonable.)

The litigation risk

In its EIS comments (see Daniel’s excellent work, did I mention its useful excellence? for the link), the Southern Nevada Water Authority laid out a plan calling for extremely deep cuts regardless of what the near term snowpack and runoff looks like – ~2.6 million acre feet of Lower Basin cuts from the states’ baseline allocations, essentially now. I’m torn between two similarly useful metaphors for what Southern Nevada says is needed – “ripping the bandaid off” and “a tourniquet, not a bandaid”.

In laying out its argument, Southern Nevada does version “1” above, much like the language of Arizona’s letter to Interior, with a call for distributing a portion of the cuts (those allocated to evaporation and system losses) sorta evenly across all water users.

I made much the same argument in my December letter to Interior, so I’m sympathetic to distributing the evaporation and system losses across all users before we think about allocation of the next level of cuts needed.

But here’s the thing that’s wrong with my argument.

To do that, you have to step outside the doctrine of prior appropriation. And the seniors – everyone mentions Imperial Irrigation District at this point, but they’re not the only senior with smart lawyers being asked to take cuts in this scheme, most notably Native American communities, who have deep legal and moral standing – will sue.

Flip the script, though, and try to take tourniquet-level cuts without spreading them broadly and you probably have to dry up the Central Arizona Project canal. Cue the “they’ve got smart lawyers and will sue” song. (To further complicate things, that would jeopardize the rights of Native American communities in Central Arizona that get their water through the CAP!)

This is the argument my smart friends who disagree with me make: A decent runoff this year would allow us to make more modest cuts (still substantial, but not nearly as deep) while avoiding tangling up the whole mess in the courts. I disagree, as I’ll explain below, but it’s not an unreasonable argument.

On bandaids and tourniquets

abandoned boat at Lake Mead

Obligatory “dancing with dead pool” visual reminder

In my comments to Interior, written in a Covid fever in December, I made an argument strikingly similar to Southern Nevada’s (“There was no collusion,” Fleck pleaded. “It’s just arithmetic!”) – that we need deep cuts now and forever. This is the tourniquet argument, and why I disagree with the “take advantage of a decent snowpack to make more modest cuts and avoid litigation” argument. The “now” part is because we’re staring down dead pool, and the “forever” part is that the river was always overallocated, and with climate change it’s now really honest truly overallocated for sure.

Even if we have a decent snowpack, I believe it is imperative that we use that water to refill reservoirs rather than irrigate alfalfa and lawns.

I’m genuinely concerned about throwing this whole thing into the courts. As we learned with Arizona V California six decades ago, litigation is a terrible way to manage a river. Much of our dilemma today is a result of the Supreme Court ignoring arithmetic and allowing us to overuse the river’s water.

My friends who argue for taking advantage of a bit of extra water, if we’ve got it, to try to stay out of court are not making an unreasonable argument.

But we can’t entirely blame the Supreme Court for our troubles. Part of today’s dilemma is the result of our failure to do the hard work of grappling with the court’s mistakes and sufficiently reduce our use of water. We’ve been avoiding litigation for too long by emptying the reservoirs and letting people use the water to irrigate alfalfa and lawns.

 

 

Dead Pool Diaries: Jack Schmidt on the hydrologic dance of operating Glen Canyon Dam at extremely low levels

Posted by on

An exchange on Twitter about the definition of “dead pool” sent me back to Jack Schmidt et al’s extremely useful (and now extremely relevant) 2016 analysis of what would be required to empty Lake Powell and move all the water down to Lake Mead.

It’s the thing that disabused me of my simplistic notion that you could operate an empty Glen Canyon Dam as if there wasn’t any dam there at all, just passing the “natural” hydrograph on downstream.

Here’s Schmidt et al:

The relatively small capacity of the river outlets would make it impossible for the flow regime of the Colorado River in the Grand Canyon ecosystem to resemble the pre-dam, natural regime. Instead, stream flow in the Grand Canyon ecosystem would be 15,000 ft3/s or less and would be steady throughout the day. During the spring snowmelt season, however, inflows would greatly exceed the capacity to release water downstream, and the elevation of Lake Powell would increase greatly.

Eventually, the reservoir would get high enough that you could use the old power plant outlets, meaning you could suddenly release a lot more water! Flows at that point jump dramatically. Here’s their graph, with flows on the top in brownish red, and reservoir elevation on the bottom:

Schmidt et al 2016 – operations of Glen Canyon Dam at low reservoir elevations.

 

The whole paper is worth reading for many reasons, not the least of which is the group’s analysis of evaporation savings of sending all the Powell water down to Mead for storage. (Spoiler alert: Evaporation goes down in one place and up in the other. No free lunch.)

Inkstain tip jar.