Guiding the San Juan River

Two rivers meet at Farmington. One of them came out of the bottom of a dam fifty miles earlier at 13.1 degrees Celsius and was still 22.2 when it arrived. The other started in the Colorado mountains, lost fifty-eight per cent of its water in about fifteen miles, and turned up at 29.6 degrees carrying almost nothing. A guide working this system is working the gap between those two facts, and the gap is 7.4 degrees wide. Similar water gets the same treatment throughout the iconic fisheries writing.

A river that loses more than half its water in fifteen miles

118 cubic feet per second below Aztec, 49.3 at Farmington.

The two Animas gauges sit about fifteen river miles apart and 279 feet of elevation apart. Between them the catchment grows by 59 square miles, a rise of four and a half per cent, and the river loses 68.7 cubic feet per second, a fall of fifty-eight per cent.

Nothing natural removes more than half a river's water across fifteen miles of valley floor while the drainage area is still growing. Diversion does, and this valley is farmed.

The yield figures make the size of it plain. Above Aztec the Animas was delivering 0.091 cubic feet per second per square mile. At Farmington the same river was delivering 0.036, a drop of sixty per cent per unit of ground, over a distance where the ground barely changed.

This is a mountain river, and it arrives at its confluence as a small warm one. Anything a guide reasons about the Animas from its Colorado headwaters is wrong by the time it reaches New Mexico.

Fire Drill, IN

The highest yield on the page comes out of a valve

0.266 below Navajo Dam, seven times the Animas figure.

The station near Archuleta was carrying 867 cubic feet per second off 3,260 square miles. That is a yield of 0.266, which in a desert basin at the end of July is not a number the landscape produces.

It is more than seven times what the Animas was delivering at Farmington and nearly three times the yield at Aztec. The two rivers drain adjacent country under the same sky.

Set the temperatures beside it and the mechanism is unambiguous. Thirteen point one degrees against 29.6 in the same basin on the same evening is a difference of 16.5 degrees, and no geographic variable available here explains it.

What explains it is a reservoir with a deep outlet and an obligation to release. This series has now measured that same signature on several rivers, and it always looks the same: high yield, low temperature, and a catchment that could not produce either.

What the confluence actually contributes

Five per cent of the flow, at 7.4 degrees hotter.

Where the Animas joins, the San Juan mainstem was carrying 935 cubic feet per second at 22.2 degrees. The Animas brought 49.3 cubic feet per second at 29.6.

That is about five point three per cent of the combined volume arriving substantially warmer than the river it joins. Mixed proportionally, the effect on the mainstem is small, on the order of a few tenths of a degree.

Which is worth knowing precisely because the visual impression is the opposite. A confluence looks like a major event and this one, in volume terms, is not.

The useful reading is the other direction. The Animas is not changing the San Juan; the San Juan's own history is. By Farmington the mainstem had already climbed 9.1 degrees from where it was released, and it did that without help.

Nine degrees before anything joins it

13.1 at Archuleta, 22.2 at Farmington, on 7.8 per cent more water.

Between those two mainstem stations the drainage area grows from 3,260 to 7,240 square miles, an increase of 122 per cent, and the flow rises from 867 to 935 cubic feet per second, an increase of 7.8 per cent.

So the river more than doubled the ground behind it and gained under eight per cent more water, while warming 9.1 degrees. The additional catchment contributed heat and very little else.

That is the same pattern this series measured on the Green in Utah, where the cold released from a reservoir was spent by desert before the river had run its length. Here the spending happens faster and over a shorter distance.

For an operator the implication is identical and worth stating in the same terms: the sellable coldwater reach is measured in miles below the dam, not in the length of the river on a map.

Below the confluence the river keeps warming

25.9 degrees at Shiprock, 12.8 above where it started.

At Shiprock the San Juan drains 12,900 square miles at 4,890 feet and was carrying 1,170 cubic feet per second at 25.9 degrees.

From Farmington that is a gain of 235 cubic feet per second on 5,660 additional square miles, a yield on the increment of about 0.042, and another 3.7 degrees of warming.

The total from the dam is 12.8 degrees across 768 feet of descent. Elevation is not doing that work; the elevation change here is small and the temperature change is not.

Note also what the yield column does. It runs 0.266, then 0.129, then 0.091 down the mainstem, falling steadily as the river collects more of a landscape that has nothing to give it.

The state park at the top of it

Park elevation 5,600 to 6,600 feet, and the gauge sits at the bottom of that range.

The state park service describes Navajo Lake as the second largest lake in the state and gives the park elevation as 5,600 to 6,600 feet. It lists flyfishing the San Juan River among the park's activities and describes the river in the strongest terms the agency uses for any water it manages.

The gauge below the dam sits at 5,657.65 feet, near the floor of that range, which is a neat illustration of where the fishery lives relative to the storage above it.

The park also records 244 developed campsites across seven campgrounds, two marinas, 4.4 miles of trails, and accessible piers. That is the visitor infrastructure a guiding operation sits inside, and it shapes the client mix more than most operators account for.

Facilities, campsite counts and access arrangements change; confirm the current position with the park before building a season around any of it.

Camping infrastructure is a client-mix signal

244 developed campsites and two marinas set who shows up.

A destination with several hundred campsites at the top of it draws a different visitor from one reached only by lodge or by hotel. Self-supported anglers on multi-day stays behave differently from clients flying in for two guided days.

Both are real markets and they want different things. The camping visitor is price-sensitive, repeat-prone and often books late; the destination client books early, spends more per day and expects the trip organised end to end.

An operation that tries to serve both without deciding tends to price for one and market to the other. Deciding first, then building the offer, is the cheaper order.

The rate side of that decision is well documented across the trade. Day rates by state gives the regional floor, and booking lead times shows how differently those two groups behave in the calendar.

Where the 235 cubic feet per second comes from

Five thousand six hundred square miles delivering a yield of 0.042.

Between Farmington and Shiprock the river picks up 235 cubic feet per second. The Animas accounts for only 49.3 of that, so roughly 186 arrives from everything else across 5,660 additional square miles.

That works out at about 0.033 for the remainder, which places it alongside the thinnest catchments this series has measured anywhere in the country. Most of the tributary channels in that stretch are dry washes for much of the year.

Some part of the gain is also return flow: water taken out upstream for irrigation, used, and finding its way back to the channel warmer than it left. That is consistent with the temperature rising 3.7 degrees over the same reach.

The distinction matters commercially because return flow and snowmelt are not the same input. One arrives on an agricultural calendar and the other on a seasonal one, and only the first can change because of a decision made in an office.

Adjacent country, opposite readings

Two rivers, twenty miles apart, 16.5 degrees apart.

The Archuleta and Farmington Animas stations sit about thirty miles apart in the same basin, at 5,658 and 5,273 feet. Their temperatures were 13.1 and 29.6.

Nothing in climate, latitude, altitude or season separates them. They were under the same sky on the same afternoon and the difference between them is larger than the difference this series has measured between Florida and Montana on comparable dates.

That is the strongest possible demonstration of a point this writing keeps arriving at: on regulated water, geography stops being predictive and the release becomes the only variable worth reading.

It also explains why local knowledge is worth paying for in this basin specifically. A visitor with a map and a weather forecast cannot get anywhere near the right answer, and the two rivers look equally plausible from a distance.

Optimum Charters, WI

What a fifty-eight per cent loss does to a guiding business

It removes the tributary as an option, not just as a backup.

On most systems a warm mainstem sends an operator to a tributary. The arithmetic here closes that door: the Animas at Farmington was carrying 49.3 cubic feet per second at 29.6 degrees, which is neither cold enough nor big enough to hold a guided day.

That concentrates every operator in the basin onto the same reach below the dam, which is the reason that reach is as busy as it is. The crowding is not a fashion; it is arithmetic.

An operation planning for growth has to solve that structurally rather than hope for it. More water is not available in this basin in late summer, so the growth has to come from trip design, season length, or client value rather than from finding new stretches.

That is a harder business problem than most guides are set up to think about, and it is the actual constraint on the trade here. Everything else is downstream of the fact that there is one good reach and everyone knows it.

Reading this basin in the right order

Storage first, diversion second, weather last.

The order that works here is the reverse of the one most people use. Weather is the least informative input on this river and the one clients arrive with; storage and release are the most informative and are almost never discussed.

In between sits the irrigation calendar, which explains what happens on the Animas and part of what happens below Farmington, and which is public information that almost nobody in this trade reads.

An operator who works down that list in order will be right more often than one who watches a forecast, and will be able to say why, which is the part clients actually respond to.

It is also a durable advantage. Forecasts are free and universal; understanding how water moves through this particular basin takes a season to learn and cannot be looked up in five minutes. The same is true on other heavily managed water, where the release schedule outranks the snowpack.

A tributary whose thermometer stopped in 2023

The Animas below Aztec serves a temperature stamped October 2023.

The station's flow and gauge height were current to the minute on that evening. Its water temperature was last reported on 4 October 2023, and it is not quoted anywhere on this page.

That is the third dark or stale temperature sensor this series has hit in a handful of western rivers, and the pattern is consistent: discharge and stage are maintained, temperature quietly is not.

It matters here more than usual, because a live reading at Aztec would have shown exactly where on that fifteen-mile stretch the Animas gives up its cold as well as its volume. That measurement is simply not available, and no amount of reasoning substitutes for it.

The honest version is to say so. An operator quoting numbers to clients builds more credibility by naming the gap than by filling it with an estimate.

What the forecast covers and what it does not

Grid 68 by 201, zone NMZ203, radar KABX at Albuquerque.

The national forecast service places the Archuleta gauge in Albuquerque office territory, forecast zone NMZ203, with KABX as the nearest radar and Navajo Dam as the reference community.

The reference community being the dam itself is a fair summary of the region. The named place at this location is a piece of infrastructure.

What the forecast describes is air over high desert in late July. What the river was doing at the same moment was 13.1 degrees, and the two figures are close to unrelated for the reach that matters.

That disconnect is the single most useful thing an operator here can explain to a client, and it is worth a sentence on a website rather than a conversation at the put-in.

Two different kinds of scarcity in one basin

The mainstem is short of cold; the tributary is short of water.

These are not the same problem and they do not have the same solution. The San Juan below the dam has abundant flow and a finite thermal budget that is spent going downstream. The Animas has adequate temperature at altitude and loses its volume to use before it arrives.

An operator working the mainstem manages distance from the release. An operator working the tributary manages a water-rights and irrigation calendar that has nothing to do with fishing.

Most of the commercial activity sits on the first, for obvious reasons, and the numbers on this page support that choice rather than arguing with it.

But it is worth being explicit that the choice was made by the arithmetic. A yield of 0.036 and a temperature of 29.6 is not a coldwater guiding business, however good the water upstream in Colorado looks.

What this river rewards

Precision about reach, and honesty about distance.

The reach below the dam is one of the most heavily fished pieces of water in the southwest, and its reputation is national. That means an operator there is not competing on discovery; they are competing on execution and on being findable.

The specific, checkable claim available here is strong: 13.1 degrees on a July evening in New Mexico, with the nearest comparable water hours away. Almost nobody states it that way.

Below that reach, the honest position is that the product changes, and saying so builds more trust than blurring it. A client who is told plainly where the coldwater fishing ends is a client who believes the rest of what you say.

Operators on other engineered tailwaters face the same communication problem, and the ones who solve it tend to name reaches rather than rivers. The route into fly-fishing guiding and licence requirements by state cover the entry work behind that.

The whole basin in one column

Yields of 0.266, 0.129, 0.091, 0.091 and 0.036 in one drainage.

That is a sevenfold range inside a single river system on a single evening, and the highest figure belongs to the station immediately below a dam while the lowest belongs to a mountain river fifteen miles after a farming valley.

Neither extreme is a natural property of the ground. Both are the result of decisions about where water goes, and the temperature column follows the same logic.

Anyone building here should read the basin that way rather than as geography. Where the water is stored, where it is taken out, and how far a client is from the release are the three variables, and they explain almost everything in the table.

Longer trips are usually where that understanding pays, because they are the ones that cross reaches. Packaging longer trips is worth reading alongside what different species pay before committing to a single stretch of this river.