The Man Three Miles from the Table Who Knew the River Was Smaller

The following conversation never happened. Eldon Cray Lansford is not a real person, though nearly everything he describes is. His knowledge, his frustrations, and his precise professional situation are drawn from the documented experiences of USGS hydrologists, particularly E.C. LaRue, who held data the 1922 Colorado River Compact Commission chose not to use.¹ We've placed an imagined man in a real chair, in a real room, during a real November, and asked him what it was like to sit there with a number no one wanted to hear.

It's mid-November 1922. The Colorado River Commission has been meeting at Bishop's Lodge outside Santa Fe for a week. Eldon Cray Lansford, a field hydrologist with the U.S. Geological Survey, is not at Bishop's Lodge. He's at a boarding house in town, three miles south, where rooms cost considerably less and no one has asked him to limit occupancy to two per room.² He has a copy of USGS Water-Supply Paper 395 in his satchel and nothing on his calendar.

You've been in Santa Fe for nine days. What exactly are you doing here?

Lansford: Waiting, mostly. I came because Mr. LaRue thought it would be useful to have someone from the Survey nearby in case the Commission wanted to consult our flow records. He wrote to Secretary Hoover directly, offering our services.³ We received a very gracious reply.

What did it say?

Lansford: Thank you.

That's it?

Lansford: It was a well-composed thank you. Government stationery. I've kept it.

So you're here in case they call, but they haven't called.

Lansford: I would not describe myself as sitting by the telephone, no. The Commission has its own technical people. Director Davis has been presenting the Reclamation Service's figures. He spoke to them on the eleventh, I'm told.⁴ He knows the river. He's been studying it since 1895. He's a serious man. He's also Powell's nephew, which... well, Powell understood aridity better than anyone who ever lived, and I sometimes wonder what he'd make of the numbers his nephew is putting in front of those commissioners.

What numbers is Davis presenting?

Lansford: Sixteen and a half million acre-feet per year at Lee's Ferry. Roughly. Based on the gauge record at Laguna Dam, near Yuma.⁵

And your number?

Lansford: Fifteen million. We get there differently. We're building the flow up from upstream gauges: Green River, Fruita, several tributaries, synthesized to estimate what reaches Lee's Ferry.⁶ Many measurement points, assembled into a picture. They're reading it at one point near the bottom, after diversions, after losses, and then adjusting upward.

Adjusting how?

Lansford: The Fall-Davis report puts the average discharge from 1899 to 1920 at about 16.3 million acre-feet. But then it suggests that certain early years, the late 1890s, were unusually low and perhaps shouldn't weigh so heavily. Trim those off, focus on 1906 to 1920, and the average climbs to 17.3 million.⁷

I understand the impulse. You have a short record, you're uncertain about the early measurements, you want to be fair to the data. But you can also read that choice as selecting the wettest fifteen years in your short record and calling them normal.

And you believe the late 1890s data should be included?

Lansford: The river doesn't care what I believe. The gauge was running. The water was low. That happened. Mr. LaRue's 1916 report included those years. That's why our number is lower. It's not really a disagreement about methodology. It's a disagreement about which years count.

A million and a half acre-feet is a significant gap.

Lansford: It's the difference between a river that can support what seven states want to build and a river that can't. I want to be careful here. I'm not saying Director Davis is dishonest. I think he genuinely believes the river is generous. He's spent thirty years looking at it and seeing possibility. I've spent eight years measuring it and seeing what the current meter shows me. Those are different relationships with the same water.

You started the fieldwork in 1914, 1915?

Lansford: Fourteen, fifteen. Before the war. I was hiking tributaries Mr. LaRue couldn't get to himself, running measurements. You wade in, you hold the meter, you record the depth and velocity at intervals across the channel, you move upstream, you do it again. You do this for years. You get to know a river the way you know a neighbor. Not what they say about themselves at church. What you observe through the fence.

What does that kind of knowledge feel like when it's not wanted?

Lansford: [long pause] Like being a doctor at a banquet where everyone's drinking water you know is bad. Except the host has already told the guests it's fine, and you weren't invited anyway, and also the host is solving a real problem while he's doing it. California is going to take the whole river if these men don't reach an agreement.⁸ The prior appropriation doctrine means whoever uses the water first owns it, and California is using it fastest. A compact, any compact, prevents that. They absolutely need a deal. I just don't know why the deal needs a number that the river doesn't have.

Does it need that number?

One of the commissioners apparently acknowledged they had no idea how low future droughts might go.¹⁰

Lansford: They know. That's what people get wrong about this. They assume the commissioners are ignorant. They're not ignorant. They have a twenty-year record and they know twenty years is short. They've said so in the room. But knowing the record is short and acting on that knowledge are very different things. Acting on it means accepting that maybe there isn't enough water for everyone's plans. Nobody traveled to Santa Fe to accept that.

What happens when the compact is signed with the wrong number?

Lansford: I measure rivers. I don't predict the future. But water law is permanent in a way that water isn't. They're going to write a number into a legal document, and then they're going to build dams and canals sized to that number, and then they're going to invite people to move to a desert and grow things in it based on that number.

It's snowing outside.

Lansford: It's been snowing all week. Good for the snowpack. [half-smile] Every hydrologist's reflex. You see weather and you think about what it becomes when it melts. My wife finds it unbearable. She'll say, "It's a beautiful morning," and I'll say, "Twelve to fourteen inches of water equivalent," and she'll go talk to someone else.

Will you stay through the signing?

Lansford: I'll stay. I'm not entirely sure why. Nobody's asked for my opinion and I doubt anyone will. But I've spent eight years on that river. I want to be nearby when they decide what it is. Even if they're wrong about it. [pause] Someone should be here who knows what the gauges actually say. Whether or not it makes any difference at all.

The Colorado River Compact was signed on November 24, 1922, allocating water based on an estimated annual flow of roughly 16.4 million acre-feet.¹¹ Actual flows from 1906 to 2024 have averaged approximately 14.6 million acre-feet, and just 12.4 million since 2000.¹² LaRue's 15 million acre-feet estimate, dismissed in 1922, proved optimistic.

Footnotes

1. Eric Kuhn and John Fleck, Science Be Dammed: How Ignoring Inconvenient Science Drained the Colorado River (University of Arizona Press, 2019). The definitive account of how available hydrological data was sidelined during the Compact negotiations. ↩

2. Hoover ordered the Bishop's Lodge manager to limit occupancy to two per room and reduce the guest list. John Fleck, "A century ago in the Colorado River Compact: Converging on Santa Fe," Inkstain.net, November 9, 2022. https://www.inkstain.net/2022/11/a-century-ago-in-the-colorado-river-compact-converging-on-santa-fe/ ↩

3. LaRue wrote directly to Hoover offering to contribute to the Compact's development. He received a thank-you note. Ge et al., "Fixing the Flawed Colorado River Compact," Eos, June 2023. https://eos.org/features/fixing-the-flawed-colorado-river-compact ↩

4. Davis presented the Reclamation Service's flow data to the Commission on November 11, 1922. Utton Center, Colorado River Compact Centennial, Episode 8. https://uttoncenter.unm.edu/resources/colorado-river-compact-centennial/episode-8.html ↩

5. The Fall-Davis report's estimate was based primarily on measurements at Laguna Dam near Yuma. No gauging station existed at Lee's Ferry until 1921. Wayne's Words, "Was the 1922 Colorado River Compact Based on Bad Data?" March 2026. https://wayneswords.net/threads/was-the-1922-colorado-river-compact-based-on-bad-data-or-just-bad-timing.11158/ ↩

6. LaRue calculated average discharge at Lee Ferry at 15.0 million acre-feet per year using upstream gauges near Green River, Utah, and Fruita, Colorado, combined with tributary records. Ge et al., Eos, 2023. ↩

7. The Fall-Davis report noted average discharge of 16.3 MAF from 1899–1920 but suggested focusing on 1906–1920 (average: 17.3 MAF), dismissing earlier dry years. Wayne's Words, 2026. ↩

8. By April 1922, the greatest risk to upper basin states was that California would establish prior appropriation rights to the river's water before other states could develop their shares. Utton Center, Compact Centennial series. ↩

9. Delph Carpenter, Colorado's representative, proposed the two-basin division framework that broke the negotiating impasse. The division required sufficient water in both basins to satisfy all parties. Big Pivots, "Why not create the Colorado River Compact they wanted in 1922?" September 2025. https://bigpivots.com/why-not-create-the-colorado-river-compact-they-wanted-in-1922/ ↩

10. One commissioner acknowledged that with only a 20-year record, guaranteeing future deliveries was inherently risky. John Fleck, "A century ago: How much water to send past Lee's Ferry?" Inkstain.net, November 2022. https://www.inkstain.net/2022/11/16640/ ↩

11. Congressional Research Service, "Colorado River Compact of 1922," updated through May 2026. https://www.congress.gov/crs_external_products/IN/HTML/IN11982.html ↩

12. CRS, 2026. Actual annual flows from 1906–2024 averaged approximately 14.6 MAF; flows since 2000 have averaged 12.4 MAF. ↩