Day: October 4, 2024

Delayed farm bill punted until after election with Congress stuck on how to pay for it — Source #NewMexico

A farmer stores grain near Eldridge, Iowa, on Sept. 28, 2024. (Photo by Kathie Obradovich / Iowa Capital Dispatch)

Click the link to read the article on the Source New Mexico website (Allison Winter):

October 3, 2024

Sweeping legislation that would set food and farm policy for the next five years is in limbo, waiting for lawmakers to decide its fate after the election.

The latest deadline for the farm bill passed unceremoniously at midnight on Sept. 30, without a push from lawmakers to pass a new farm bill or an extension.

Congress will have to scramble in the lame-duck session set to begin Nov. 12 to come up with some agreement on the farm bill before benefits run out at the end of the year — which if allowed to happen eventually would have major consequences.

The law began 90 years ago with various payments to support farmers but now has an impact far beyond the farm, with programs to create wildlife habitat, address climate change and provide the nation’s largest federal nutrition program.

Ag coalition in disarray

The omnibus farm bill is more than a year behind schedule, as the bipartisan congressional coalition that has advanced farm bills for the last half century has been teetering on the edge of collapse.

Congress must approve a new federal farm bill every five years. The previous farm bill from 2018 expired a year ago. With no agreement in sight at the time, lawmakers extended the law to Sept. 30, 2024.

The delay creates further uncertainty for farmers, who are facing declining prices for many crops and rising costs for fertilizer and other inputs.

Lawmakers have some buffer before Americans feel the consequences of the expiration.

Most of the key programs have funding through the end of the calendar year, but once a new crop year comes into place in January, they would revert to “permanent law,” sending crop supports back to policy from the 1938 and 1949 farm bills.

Those policies are inconsistent with modern farming practices and international trade agreements and could cost the federal government billions, according to a recent analysis from the non-partisan Congressional Research Service.

‘Groundhog Day’ cited by Vilsack

The stalemate between Democrats and Republicans over the farm bill has centered on how to pay for it and whether to place limits on nutrition and climate programs.

Agriculture Secretary Tom Vilsack told reporters in a press call on Saturday that the process “feels like Groundhog Day” — because he keeps having the same conversations about it. Vilsack said Republicans “just don’t have the votes” on the floor for legislation passed in the House Agriculture Committee, which is why it has sat dormant in the House for four months.

“If they want to pass the farm bill they’ve got to get practical, and they either have to lower their expectations or raise resources. And if they’re going to raise resources, they have to do it in a way where they don’t lose votes, where they actually gain votes,” Vilsack, a former Iowa governor, said.

The Republican-led committee approved its farm bill proposal largely on party lines at the end of May, amidst complaints from Democrats that the process had not been as bipartisan as in years past.

Partisan division is not uncommon in today’s Congress but is notable on the farm bill, which historically brought together lawmakers from both sides of the aisle. Bipartisan support can be necessary for final passage because the size of the $1.5 trillion farm bill means it inevitably loses some votes from fiscal conservatives and others.

Shutdown threat

Lawmakers are on borrowed time with both the farm bill and the appropriations bills that fund the federal government.

The House and Senate both approved stopgap spending bills at the end of September to avoid a partial government shutdown. The short-term funding bill, sometimes referred to as a continuing resolution, or CR, will keep the federal government running through Dec. 20.

Some agriculture leaders had asked for the continuing resolution to not extend the farm bill, to help push the deadline for them to work on it when they return.

The day after they approved the CR and left the Capitol, 140 Republican House members sent a letter to congressional leadership asking to make the farm bill a priority in the waning weeks of 2024.

“Farmers and ranchers do not have the luxury of waiting until next Congress for the enactment of an effective farm bill,” the letter states, noting rising production costs and falling commodity prices that have put farmers in a tight spot.

House Democrats also say they want to pass a new farm bill this year.

House Minority Leader Hakeem Jeffries, a New York Democrat, listed the farm bill as one of his top three priorities for the lame duck. Also on his list were appropriations and the National Defense Authorization Act, which sets policy for the Pentagon.

“It will be important to see if we can find a path forward and reauthorize the farm bill in order to make sure that we can meet the needs of farmers, meet the needs from a nutritional standpoint of everyday Americans and also continue the progress we have been able to make in terms of combating climate crisis,” Jeffries said in remarks to reporters Sept. 25.

Nearly 300 members of the National Farmers Union visited lawmakers in September to ask for passage of a new five-year farm bill before the end of 2024.

“Family farmers and ranchers can’t wait – they need the certainty of a new farm bill this year,” National Farmers Union President Rob Larew said in a statement after the meetings. “With net farm income projected at historic lows, growing concentration in the agriculture sector, high input costs and interest rates, and more frequent and devastating natural disasters, Congress can’t miss this opportunity to pass a five-year farm bill.”

Disagreements over SNAP formula

The key dispute for Democrats this year is a funding calculation that would place limits on the “Thrifty Food Plan” formula that calculates benefits for the Supplemental Nutrition Assistance Program, SNAP.

It would keep SNAP payments at current levels but place a permanent freeze on the ability of future presidents to raise levels of food support. Democrats have characterized it as a sneaky cut to vital support for hungry Americans that makes the bill dead on arrival.

Republicans are using the limits as part of a funding calculation to offset other spending in the bill. The bill would raise price supports for some crops like cotton, peanuts and rice.

“They have to do one of two things,” Vilsack said of lawmakers. “They either have to recognize that they can’t afford all the things that they would like to be able to afford, if they want to stay within the resources that are in fact available … Or another alternative would be to find more money.”

Vilsack recommended finding other sources of funding outside the farm bill, like changes to the tax code.

“You close a loophole here or there in terms of the taxes or whatever, and you generate more revenue, and you have that revenue directly offset the increase in the farm bill. … That’s the correct way to do it. And that’s, frankly, the way Senator Stabenow is approaching the farm bill,” Vilsack said, referring to Senate Agriculture Committee Chairwoman Debbie Stabenow, D-Mich.

The Senate Agriculture Committee has had no public markup or formal introduction of a bill. But leaders say committee staff have been meeting weekly to discuss a path forward. Stabenow has not publicly disclosed the offsets for the money she says is available to be moved into the bill.

A powerful sprinkler capable of pumping more than 2,500 gallons of water per minute irrigates a farm field in the San Luis Valley June 6, 2019. Credit: Jerd Smith via Water Education Colorado

When it comes to probabilities, don’t trust your intuition. Use a decision support system instead! — NOAA

Click the link to read the article on the NOAA website (Brian Zimmerman):

October 3, 2024

This is a guest post by Brian Zimmerman, a climate scientist at Salient Predictions. Salient is a startup that utilizes advances in machine learning and artificial intelligence to develop and provide accurate, reliable forecasts on the subseasonal-to-seasonal timescales. Additionally, Brian serves as a decision support specialist, helping clients to navigate uncertainty in the most effective manner possible.

The nature of uncertainty

There are not too many things more frustrating in this world than unmet expectations. Your favorite basketball team has an estimated 79% chance of beating the upcoming opposition; yet, they lose, and you’re out $10 to your work betting pool. There was only a 15% chance of rain in the forecast one day before your friend’s wedding, and it starts raining precisely 10 minutes before the outdoor ceremony is supposed to start. Or perhaps you’re in the ENSO betting market.

Regular readers of the blog recognize that climate predictions are uncertain, including for the El Niño-Southern Oscillation (ENSO), which are expressed as probabilities (70% chance of El Niño coming) instead of forecasting a single black-and-white outcome (El Niño is coming). To make decisions in the face of uncertainty, we sometimes consciously, often automatically, use our best guess at how likely the various outcomes are and what our tolerance for risk in the given situation is. For all but a few of us (nerd alert!), this is done in a qualitative, intuitive fashion. 

The trouble is, humans aren’t very good at it. Most of us find it difficult to intuitively understand probabilities except 0, 50, and 100%. In order to make sense of an uncertain world, we often craft narratives that enable us to make decisions and move forward. Unfortunately, these narratives tend to be compromised by two traits: our penchant for overconfidence and our aversion to risk (see footnote #1). This generally leads us to sorely misestimate the likelihood of almost all events facing us as we move through our lives. An even more frustrating truth to face is that even if we correctly act on probabilistic information (know the true odds of a given event and make the best decision possible) the outcome may be opposite of what we favored.

So, what to do? This blog post aims to offer some tips and tricks that are likely to lead to greater fulfillment of all your most accurately estimated dreams! Let’s learn from a simple example. I’ll use a sports betting analogy to highlight the benefits of using calibrated probabilistic forecast models (like the CPC ENSO forecast!) for real-world decisions.

Betting on the Bulls: building a decision support system

Say you’re a Bulls fan, and there’s a betting pool at work. Given no information about them or their competition, you would assume that the odds that they win any given game is 50-50. In reality, not all teams are equally matched, so the true chances the Bulls will win a given game will vary. To maximize your odds of large profits over the season, you need a system for guiding how much to bet based on a range of probabilities. Good probabilistic decision making requires two key items (see footnote #2 for additional commentary): 

1) A clearly defined event with a yes-no outcome:

The Bulls will win their game tomorrow night.

2) A set of actions you’ll take at specific thresholds:

This table shows what you will gamble based on what you think the probability is of the Bulls winning against whatever team they are playing. The maximum you can put into the pool is $50. 

Your decision support system (DSS) for determining how much to bet based on your estimate of the percent chances of the Bulls beating their opponent. Credit: NOAA

If the odds are less than 50%, you don’t risk your money. The higher the odds above 50%, the more you bet. (Note: this table above is your decision support system or DSS.)

Now that we have our DSS, we can mock up how this system would work out in 3 situations: 

  1. a scenario where you know absolutely nothing about basketball, and don’t try to learn. You just bet randomly based on what you had for lunch that day. This is akin to using the Farmer’s Almanac to forecast ENSO. 
  2. a scenario where you are an experienced basketball enthusiast with a discerning eye. Your estimation of the true probabilities of the Bulls winning against any team they play is perfect. This would be equivalent to climate scientists having a model that perfectly predicts the chances of El Niño in a given season: for instance, when the model predicts there is a 70% chance of El Niño coming, El Niño actually happens 70% of the time.
  3. a scenario where you’re enthusiastic and like sports, and your estimation of the true probabilities of the Bulls winning against any team they play is good, but not perfect. This would be equivalent to climate scientists having a model that does not perfectly predict the chances of El Niño, but is pretty good! For instance, when the model predicts an 80% chance of El Niño coming, El Niño actually happens 60% of the time. 

How would you wind up in each of these situations? I worked up some code (see footnote #3) so we can explore the outcomes!

Outcomes!

Each experiment simulates 1,000 seasons of 82 games each. The tables below show excerpts from one season of the second (perfect estimates of the chances of winning) and third experiments (estimates of the odds of a Bull’s victory are good, but not perfect.)

(top) A simulated season in Experiment 2, where if you predict the Bulls have a 90 percent chance of winning a game against the 76ers, they do, in fact, win against them 90% of the time (line 2). (bottom) A season from Experiment 3, where your estimates of the chances are good, but not perfect. You estimate the Bulls have a 60% percent chance of beating the Wizards over time, but they only beat them 50 % of the time (Line 81). NOAA Climate.gov graphic, based on data from Brian Zimmerman.
Each bar chart shows the cumulative profit or loss from betting after 1,000 simulated basketball seasons. The amount of profit or loss is shown across the bottom of the chart, and the height of the bar indicates how many seasons had that total. When you bet randomly (top), you lose as often as you win, and your average profit is close to zero. When you bet following a decision support system that risks more when you think the Bulls’ odds of winning are higher, you make more than $500 on average even if you can’t predict their odds of winning perfectly (bottom). NOAA Climate.gov image, based on data from Brian Zimmerman.

The first thing to notice is that by simply utilizing a tailored decision support system (DSS), you can come out ahead over time even if all you know is the probability of your team winning a game. The figure above makes this obvious; randomly betting performs much worse overall! Yes, sometimes the Bulls lose when the chance of winning is high, but if you stick to the program (your DSS), you will end up ahead at the end of the season (in this simulation, see footnote #4).

Now, you might be saying to yourself, “Okay, sure, you can come out ahead if you know the probabilities perfectly, but no one actually knows the true odds of their team winning any game.” Not to fear! The bottom row of the graphic shows that, over the long haul, the consistent implementation of decisions derived from your DSS is almost always lucrative even when your forecasts are not perfectly reliable! It’s the DSS that is critical here.

Comparing the three experiments, we see something we’d expect—that using a decision support system based on perfect knowledge of the probabilities is the MOST lucrative. But we can also see something that is perhaps not intuitive: the figure below shows that over shorter periods of time, our imperfect estimates (or even random bets!) will sometimes put us further ahead than using perfect estimates of the odds.

Out of a thousand seasons, there will be some where random betting (top) or imperfect estimates of the chances of winning (middle) will make as much money as betting with perfect knowledge of the probabilities (bottom). Inevitable negative outcomes can be one of the hardest things for people to accept about forecasts that use probabilities. NOAA Climate.gov image, based on data provided by Brian Zimmerman.

That is because there is still some randomness in the actual outcomes of the game (see footnote #3 on “noise”). A 60% chance of victory means that you will lose your bets on occasion (60% does not equal 100%). It’s inevitable that there will be a single game, or even 5- or 10-game stretches (possibly even an entire season!), where you actually end up with higher profits using a DSS with imperfectly estimated probabilities or even just random bets.

This is shown in a different way in the line graph below. I cherry–picked a season that shows both Experiment 1(random guesses) and Experiment 3 (good, but not perfect estimates of the true probability) performing worse than Experiment 2 (perfect prediction of the probability)—but only for the 1st half of the season! In the end, if you follow your DSS, you end up much better off than randomly betting, even with imperfect estimation of the probabilities.

Game by game cumulative profits for season 796 from each of the three experiments. In this season of the experiments, the bettor scenario using a decision support system with true probabilities (blue) did worse than the other two strategies until the very end of the 82-game season! NOAA Climate.gov image, adapted from original by Brain Zimmerman.

Bringing it all together

This simple example shows how we can benefit from uncertain, probabilistic estimates of the chances a team will win a game. As frustrating as it may be, a single bad game or even one bad season doesn’t mean our strategy is flawed. Making decisions based on weather, climate, and ENSO forecasts is similar! In fact, any probabilistic forecast is similar.

In ENSO forecasting, a premium is put on using so-called “calibrated” forecast models, which are constructed using long hindcasts to create more reliable outlooks. In calibrating the model output, the goal is to make the forecast estimates closer to the true probabilities – which pushes us closer to the case outlined by Experiment #2. This allows for the highest chance of the best outcomes over the long haul, but also by no means guarantees them.

Having a solid DSS means we also have considered our tolerance for risk and how to act upon the estimated forecast probabilities. This should help us avoid disappointment and stay true to the course when we don’t get an outcome we hoped for. In an inherently uncertain world, it’s the best we can do.

Lead editors: Rebecca Lindsey and Michelle L’Heureux. 

Footnotes

  1. Some additional reading: 
    Kahneman, Daniel. (2011). Thinking, Fast and Slow. London: Penguin Books. 
    Here’s a related Guardian article/interview with Kahneman where he states that if he could wave a magic wand and eliminate one thing, it would be “overconfidence.” 
    Kahneman and Tversky (1979). Prospect Theory: An Analysis of Decision under Risk. Econometrica, 47 (2), 263-292. 
  2. Event Definition:
    An event can literally be anything! However, whatever you choose, it’s critical to be specific. It could be something like “the Bulls will win tomorrow’s game against the 76ers”, or “El Niño will develop between now and March 1st” – essentially, anything that can have a definitive outcome of either “yes it did happen” or “no it didn’t happen”. Thresholds of Actionability: Here is where it gets tricky. The goal for defining “thresholds of actionability” is to determine a set of actions you would take given the various probabilities of the defined event occurring. This can be complicated because it starts to incorporate all kinds of other concepts – and is highly prone to the influence of Narrative. It’s also highly personal because different people or organizations have different risk tolerances! Ideally, these thresholds of actionability would be defined and analyzed over some historical period in order to optimize the thresholds, but let’s leave that alone for now and just make up some that seem reasonable. We can go back to our sports example for some simple fun.
  3. I created a Betting Simulator to generate these experiments. You can run it yourself because I put the python code here: https://github.com/bgzimmerman/enso_blog. The results in this blog are generated from the output of 1000 synthetic seasons of betting in the pool. There are 82 games in a season – the true probability of the Bulls winning is created using a stochastic generator – your estimated probability (on which you cast a bet) is generated conditioned on the true probability. For simplicity, the profit on a win is equivalent to the bet placed. We explore two scenarios – one in which you are perfectly prescient (i.e. your estimated probability equals the true probability) and another in which we incorporate estimation error (a tunable parameter – it adds noise to the true probability to create your estimated probability). The addition of noise here is meant to emulate both the impact of having an imperfect model and how human Narrative can lead you astray from the true probabilities (i.e. “Oh they lost the last two games so they’re due for a win this time!”). Additional Notes: In this simple example, we’re not getting into issues of who pays out the bet, what are the house odds, etc. In real betting, a gambler allocates bets based on having an edge. Finally, reliable probabilities will give you a higher payout but only if the scoring is “proper” (see Brocker and Smith, 2007 and Gneiting et al., 2007).
  4.  Astute readers may note that in this simulation with the true probabilities there is not a single season of 82 games in Experiment 2 where you lose money. Even in Experiment 3 (estimated probabilities) you don’t often lose money over a full season. However, do not jump to the conclusion that as long as you know a little bit about the sport, betting for a full season will return some money! This outcome is a result of the assumptions in the Betting Simulator. The simulator could be easily adjusted so there are more opportunities to lose money. Download it and experiment yourself if interested!

Reclamation announces $9.2 million for Tribal water projects and emergency drought relief supported by the Investing in America agenda: Reclamation’s Native American Affairs Program is providing funding for technical assistance and drought mitigation for Tribes

Rio Grande. Photo credit: USBR

Click the link to read the article on the Reclamation website:

October 1, 2024

The Bureau of Reclamation today announced a $9.2 million investment supported by President Biden’s Investing in America agenda to support Tribal efforts to develop, manage and protect water and related resources, and mitigate drought impacts and the loss of Tribal trust resources.  The 25 projects selected through the Native American Affairs Technical Assistance Program, with funding from the Inflation Reduction Act and annual appropriations will benefit 18 federally recognized Tribes across 11 western states.

“Reclamation is committed to working with Tribal nations to prepare for and respond to the impacts of climate change across all western basins,” said Bureau of Reclamation Commissioner Camille Calimlim Touton. “The projects we’re funding today will improve water use efficiency and increase Tribal water supplies by upgrading infrastructure and programs and modernizing existing facilities. Reclamation is providing the resources necessary to ensure these sovereign nations have the modern water infrastructure crucial to the health and economic vitality of their communities.” 

Projects will assess and repair a water treatment plant and drinking water system, replace failing irrigation system equipment and lower pump elevations for river access, establish an on-site training and testing center for Tribal water system operators, and map a reservation water utility system to aid future improvement, expansion, and enhancements.

Examples of the projects selected for federal funding include: 

  • Hopi Tribe (Arizona) – $397,476 to establish an on-site training and testing center to provide specialized training, operator exams, and attainment of Tribal Utility Management certifications. This will alleviate the need for water system operators to travel for training. 
  • Chickasaw Nation (Oklahoma) – $400,000 to develop a project to protect and manage diminishing groundwater supplies, accomplish community water assessments, and devise a regional water management plan to safeguard critical community water supplies. The project is in partnership with the Southern Oklahoma Water Corporation, Arbuckle Master Conservancy District and the city of Ardmore, Oklahoma. 
  • Ute Mountain Ute (Colorado) – $278,434 to design the 1,000 acre-feet Red Arrow Regulating Reservoir to help stabilize irrigation water supply. The reservoir design will allow for banking water during wet years and capturing the operational spill at the end of the 39.9-mile Towaoc- Highline Canal. 
  • Fort Mojave Indian Tribe (California) – $400,000 to replace irrigation intake pumps and related equipment on the Tribe’s land along the Colorado River. The declining river level is impacting the Tribe’s ability to irrigate agricultural fields which members depend on for income. Declining water level due to extended drought conditions necessitates the revamping of pumping stations. Replacing the 1980s-era pumps and lowering their elevation will improve the supply of irrigation water.  
  • Ute Indian Tribe (Utah) – $400,000 to complete an assessment and repairs to its water system treatment plant to benefit the Tribes’ drinking water system. 

View a full list of projects on Bureau of Reclamation’s website. 

Section 80004 of the Inflation Reduction Act appropriates $12.5 million for Reclamation to provide near-term drought relief to Tribes that are impacted by the operation of a bureau water project. 

Reclamation’s Native American Affairs Program provides funding opportunities and technical assistance through cooperative working relationships and partnerships with Tribes. To learn more about these and other funding opportunities, visit www.usbr.gov/native. 

Native American Affairs Program

CPW introduces Trojan Male brook trout in a historic effort to protect native cutthroat trout in #Colorado

Aquatic Biologist Jon Ewert stocks Trojan Male brook trout into Bobtail Creek during a historic stocking event in the headwaters of the Colorado River basin. Photo credit: Colorado Parks & Wildlife

Click the link to read the release on the Colorado Parks & Wildlife website (Rachael Gonzales):

September 27, 2024

 On Tuesday, Sept. 17, in an effort to restore native cutthroat populations in the headwaters of the Williams Fork River, Colorado Parks and Wildlife stocked 480 Trojan male or YY brook trout into Bobtail and Steelman creeks.

“This is a pretty historic moment for Colorado and native cutthroat trout restoration across the state,” said CPW Aquatic Biologist Jon Ewert. “This is a combination of both the hard work and dedication of CPW biologists current and retired.” 

“This is yet another example of the groundbreaking work done by CPW biologists and researchers to preserve native species,” said George Schisler, CPW Aquatics Research Section Chief. “While Bobtail and Steelman creeks are the first to be stocked with YY brook trout, they will not be the last. This is just the first of many for Colorado.”

In 2010, an alarming number of non-native brook trout were discovered after completing a fish survey in the headwaters of the Williams Fork River. While it is unknown when brook trout invaded these creeks, it was evident the thriving brook trout had nearly decimated the native cutthroat population over time.

Cutthroat trout found within these two creeks are some of the highest-valued native cutthroat populations in the headwaters of the Colorado River basin. Considered a species of special concern in Colorado, this subspecies of trout is genetically pure and naturally reproducing. 

“In 2011 we found 123 cutthroat trout combined in both creeks. Today, after 13 years of hard work by dedicated biologists we are seeing a little more than 1,400 cutthroats in these creeks,” said Ewert. 

Trojan male brook trout are often called YY because they have two Y chromosomes, unlike wild males with an X and Y chromosome. These trout are stocked into wild brook trout populations and reproduce with the wild fish, producing only male offspring. Without a reproducing population (male and female fish), the brook trout will eventually die out, allowing for native cutthroat trout to be restored.

Colorado Parks and Wildlife will continue to stock both streams with YY brook trout over the next several years to sustain the number of Trojan males in the population, eliminating the production of female brook trout in the creeks. 

To learn more about Trojan male brook trout and cutthroat trout restoration project in the Upper Williams Fork drainage, read our latest Colorado Outdoors Online Magazine article. 

Cutthroat trout historic range via Western Trout

At last, juice from Taylor Park Dam: It took awhile to make this happen but it immediately is cheaper energy for Gunnison County Electric Assocation — Allen Best (@BigPivots)

Taylor Park Dam. Photo credit: Allen Best/Big Pivots

Click the link to read the article on the Big Pivots website (Allen Best):

September 25, 2024

When work was completed on Colorado’s Taylor Park Dam in 1937, at least some thought existed that it would eventually be modified to produce electricity.

In 2024, it is finally happening. The first commercial power production has or will very soon happen in the first days of autumn.

The new 500-kilowatt hydroelectric turbine and generator installed in the dam will operate at or near full capacity 24/7/365. It is projected to produce an average 3.8 million kilowatt-hours annually. That compares to a  2.5-megawatt fixed-til solar array.

The electricity will get used by Gunnison County Electric Association. Mike McBride, the manager, says the electricity delivered will immediately save the cooperative money compared to the power delivered by Tri-State Generation and Transmission.

Under its contract with Tri-State, Gunnison County Electric can generate up to 5% of its own power. This hydroelectric facility will get it to 3%. The association is working to gain the other 2% from local solar array developments, one near Crested Butte and the other near Gunnison.

Map of the Gunnison River drainage basin in Colorado, USA. Made using public domain USGS data. By Shannon1 – Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=69257550

The Taylor River originates on the west side of Cottonwood Pass in the Sawatch Range. The road across the pass connects Buena Vista and Crested Butte and Gunnison. After being impounded by the dam that creates Taylor Park Reservoir, the river descends to meet the East River, which originates near Crested Butte. Together they become the Gunnison River.

Grand opening of the Gunnison Tunnel in Colorado 1909. Photo credit USBR.

The 206-foot-high earthen dam is owned by the U.S. Bureau of Reclamation but operated by the Uncompahgre Valley Water Users Association, which delivers water to the Montrose and Delta area via [the Gunnison Tunnel].

In 2020, that water association joined with the Gunnison County Electric Association to form a legal entity to finance the $3.6 million project.

George Sibley, a historian of all things water in the Gunnison Basin (and beyond), said the dam was originally intended for storing water for July through September.

In the 1970s that changed in a collaboration of the Bureau, the Uncompahgre water district, and Upper Gunnison Regional Water Conservation District. That collaboration allowed them to store water from Taylor in Blue Mesa Reservoir. This allowed water to be released continuously through the year.

“That year-round flow potential made it more possible to think of the Taylor Dam as a possible year-round power source,” he says.

But the coal-burning units at Craig were delivering plenty of cheap power. Only in the last couple of decades have the electrical cooperative started getting pressure from some members and “other cultural entities” to reduce emissions associated with their electricity, he says.

A study was commissioned in 2009 and wrapped up in March 2010. Beyond were more complications — but now success.