Guiding the Columbia River

Read the discharge column down the Columbia and the river appears to shed 87,600 cubic feet per second in its last hundred and fifty miles, roughly seven tenths of the total, over ground that is still adding 19,900 square miles of catchment. No diversion in North America is that size. The number is not describing a loss; it is describing a tide, and the moment a reading stops meaning what it meant upstream is the most useful thing on this page. Other large water gets the same treatment across the iconic fisheries writing.

The bottom three stations are measuring something else

Vancouver sits at 1.82 feet and Port Westward at minus one.

The station near Quincy publishes a datum elevation below sea level, which is the tell. In a tidal reach an instantaneous discharge figure is the net movement of water past the gauge at that instant, and a flooding tide subtracts from it.

Upstream, at Priest Rapids or the Canadian line, discharge means supply: how much water the catchment and the dams are delivering. Below Bonneville it means something closer to a balance between river and ocean at one moment.

Same units, same service, same column heading, different physical quantity. Nothing in the data announces the change, and the only clue in the published metadata is the station elevation.

Which makes this the clearest example in the series of a method reaching its limit. Every technique this writing has used, yield ratios, mass balance, upstream-downstream comparison, is valid above The Dalles and misleading below it.

Glacier Anglers, MT

An eight and a half foot tide at the mouth

Mean higher high water 10.84 feet, mean lower low water 2.23.

The tide station at Astoria publishes accepted datums in feet: mean higher high water at 10.84, mean high water at 10.17, mean tide level at 6.78, mean sea level at 6.74, mean low water at 3.4 and mean lower low water at 2.23.

The difference between the highest and lowest of those is 8.61 feet. That station is flagged as tidal and sits about twenty-eight miles downstream of Port Westward.

Its observed water level that evening was 3.705 feet, between mean low water and mean higher high water and much nearer the low end, with air temperature at 64 degrees Fahrenheit.

An 8.61-foot range moving in and out of a channel this wide displaces an enormous volume, and it is the volume that turns a discharge reading at Port Westward into a number seventy per cent below the one at The Dalles.

Where the discharge column is still doing its job

Above The Dalles the yield falls in a way that makes sense.

At the Canadian line the river drains 59,700 square miles and was carrying 103,000 cubic feet per second, a yield of 1.725. That is a genuinely high figure and it reflects a wet mountain catchment upstream.

Below Priest Rapids the drainage has grown to 96,000 square miles, an increase of sixty-one per cent, and the flow to 111,000, an increase of under eight. Yield falls to 1.156.

At The Dalles the drainage is 237,000 square miles, up another 147 per cent, and the flow 125,000, up thirteen. Yield falls again to 0.527.

That sequence is ordinary and legible: a wet headwater basin joined by progressively drier interior country, exactly the pattern this series has measured on smaller rivers across the west. It is the sequence below The Dalles that breaks.

The largest catchment in this series, by a factor of three

256,900 square miles at Port Westward.

Nothing else measured across fifty states comes close. The previous largest station used in this writing sat on 80,650 square miles in eastern Montana, and this one is more than three times that.

Scale changes which questions are worth asking. On a catchment that size the notion of a single water year, a single snowpack or a single management decision stops applying, because the basin spans two countries, several mountain ranges and a full continental rain shadow.

It also means the river is never really responding to local weather. Whatever a client sees on a forecast for the reach they are fishing is nearly irrelevant to the volume moving past them.

For an operator that is a genuine simplification. On most water the job includes reading conditions; here the mainstem's volume is effectively a constant on the timescale of a booking, and the variables that matter are different ones.

One thermometer on the whole river

21.9 degrees at The Dalles, and nothing anywhere else.

Of the six stations pulled, exactly one served a water temperature. The station below Bonneville served no discharge at all, only stage.

On a river this important that is a striking gap, and it repeats the pattern this series has found across the interior west: discharge and stage are maintained, temperature is patchy, and some stations quietly serve only one parameter.

The single reading that does exist is worth having. Twenty-one point nine degrees in late July at a station sitting at sea level, a hundred and ninety river miles from the ocean, is a real number about a real problem for cold-water species.

But one station cannot describe a river that crosses two countries, and an operator working any specific reach here is going to be relying on their own measurements. That is the same conclusion this writing reached on a river with no working thermometers at all, for the same reason.

What sea level at The Dalles actually means

The gauge datum is 0.0 feet, a hundred and ninety miles inland.

The Dalles, Bonneville, Vancouver and Port Westward all publish station elevations of essentially zero. From The Dalles down, this river is at sea level for its entire remaining length.

That is a consequence of the dams as much as the geography: a series of pools holds the water surface near a fixed elevation through the gorge, so the river descends in steps at the structures rather than continuously along its bed.

Above them the drop is real. The Canadian line station sits at 1,202.84 feet, so the river loses its entire elevation before The Dalles and none of it after.

Practically, that shapes everything about how the lower river behaves: slow water, deep pools, tidal influence reaching far inland, and none of the gradient-driven character that defines the tributaries feeding it.

A yield of 1.725 at the border, and what it says

Fifty-nine thousand square miles delivering 103,000 cubic feet per second.

The station at the Canadian line returned the highest yield on this page and one of the higher figures in the series for a catchment of any size. That is what a wet mountain basin at the top of a continental system produces.

The comparison that makes it concrete is downstream. By The Dalles the drainage has nearly quadrupled and the flow has risen by twenty-one per cent, so almost four fifths of the basin is contributing almost nothing to the total.

That is the Columbia Basin proper: a rain shadow behind the Cascades, farmed heavily, and dry through the summer. On any other river this writing has covered, that country would be the whole story. Here it is a footnote in the volume column.

The practical version is that the water in the lower river came overwhelmingly from a long way north and west of where a client will be standing, and did not respond to anything happening locally.

Stage without discharge is its own kind of gap

The Bonneville station served a gauge height and nothing else.

Ten point nine eight feet of stage, current to the minute, and no flow figure at all. That is the fourth distinct category of missing data this writing has run into, after dark sensors, no-data sentinels and absent drainage areas.

It matters because stage looks like a measurement of the same thing. A number in feet at a named dam reads as informative, and on its own it supports no comparison at all: stage relates to discharge through a rating curve specific to that channel, and without one the figure travels nowhere.

Anyone assembling a picture from several gauges has to notice which parameter each one actually served. Six stations went into this page and four discharge figures came out of it that mean the same thing as each other.

Counting is again the check that catches it. If the number of stations and the number of usable comparisons differ, the difference deserves a sentence rather than a silent omission.

Why the middle river runs at 22 degrees

Sea-level pools, a rain shadow, and 190 miles with no gradient.

The single temperature on this page, 21.9 degrees at The Dalles, is not an anomaly and it is not a bad year. It is what a wide, slow, sea-level river does in eastern Oregon and Washington in the last week of July.

Depth and surface area are both large, the water moves slowly through a chain of pools, and the country around it was recording air temperatures in the nineties elsewhere in this series' pulls for the same region.

The mechanism is the opposite of the spring-fed tributaries this writing has measured nearby, where water emerges with no exposure history and stays cold through a heatwave. Here the exposure history is essentially the whole summer.

For an operator that has direct consequences for species timing and for what can honestly be promised in August, and it argues for pushing clients toward the tributaries or the tidal reach in high summer rather than the middle river.

Maine Bass Fishing Guide Service Inc, ME

What a quarter of a million square miles rules out

Any explanation that depends on one place.

Most of the rivers in this series can be understood from a handful of local facts: a dam, a farming valley, a spring complex, a snowpack. This one cannot, and pretending otherwise is the commonest error in writing about it.

No single reservoir controls it, no single winter determines it, and no single agency governs it. It crosses an international boundary, two states, several tribal jurisdictions and a continental climate divide.

That means an operator's expertise here is necessarily narrow and deep rather than broad. Knowing one reach properly is achievable; knowing the river is not, and claiming to is a tell. What different species pay is where that narrowness starts earning.

The honest positioning follows from that. Being the person who knows sixty miles of it extremely well is both more credible and more useful to a client than being the person whose website says Columbia River.

Two forecast offices, one radar, and a river that ignores both

The Dalles answers to Pendleton, Port Westward to Portland, and KRTX covers both.

Query the forecast grid service for The Dalles and it returns Pendleton office territory, grid 53 by 111, zone ORZ041, radar KRTX. Ask the same service about Port Westward and it answers Portland office, grid 103 by 138, zone ORZ108, on that identical radar.

Two offices and one radar across the two ends of the measured reach. In practice the difference between them is the Cascade crest, which is one of the sharpest climate boundaries in the country.

What neither product describes is the tide, which is the dominant control on the lower river's behaviour and belongs to a completely separate service.

An operator working the tidal reach needs a tide table more than a forecast, and the two are published by different parts of the same agency with no obvious link between them. Knowing that is itself part of the job.

Three different businesses along one name

A cold headwater reach, a warm impounded middle, and a tidal estuary.

The reach near the Canadian line is high-yield mountain water. The middle river through the Columbia Basin is a chain of pools at sea level running near 22 degrees in summer. The lower river is a tidal estuary where the water moves both directions twice a day.

Those are not variations on a theme. They differ in species, in seasons, in the boat required, in the licensing picture, and in whether a tide table or a release schedule is the document you check.

Almost all the search traffic and almost all the marketing treats them as one river, which is why so much of it is vague. Naming the reach is more accurate and, as usual, easier to be found for.

The comparison worth making is with saltwater work in the same corner of the country, because the lower Columbia has more in common with that than with anything upstream of Bonneville.

Two states, two licences, one boat

The mainstem is the border for hundreds of miles.

Between the Washington and Oregon banks the river is a state line, and an operator working it is working in two jurisdictions simultaneously, often within a single drift.

That is a well-known complication locally and a completely invisible one to a client booking from out of state. It affects credentials, reporting and which set of rules governs a given fish.

The arrangements that cover it change, and they are not the sort of thing to run on memory or on what a neighbouring operator did three seasons ago. Check the current requirements with both state agencies before the season, every season.

For the paperwork floor, the state-by-state credential picture is the starting point; for what the work pays, the regional day-rate figures.

Scale is a marketing problem, not just a hydrological one

Two hundred and fifty-seven thousand square miles under one search term.

A client searching for this river is searching for something the size of a small country, and the results they get are unlikely to be near where they will actually be.

That is the single biggest structural disadvantage of guiding on very large named water, and it is invisible from inside the business. Operators assume they are competing with the outfit down the road when they are competing with a name.

The fix is the same one this writing keeps arriving at from different directions: be specific about the reach, the season and the species, and let the river name be context rather than the claim.

It is a smaller pond by definition, and being findable in it beats being invisible in the large one. The route into guiding covers none of this, which is part of why it stays available.

When to distrust your own method

When a number implies something physically impossible.

A seventy per cent loss over 150 miles of river that is simultaneously gaining 19,900 square miles of catchment is not a finding, it is a signal that the quantity being compared has changed underneath the comparison.

Earlier in this series the same reflex caught a release masquerading as a catchment yield, a diversion masquerading as a dry tributary, and a nine-year-old sensor reading masquerading as today's. Each time the check was the same: does this number imply something that cannot happen?

Here it does, and the resolution was one line of station metadata, an elevation of minus one foot, plus a tide table from a different agency.

That is the working habit worth taking away from this page more than any particular figure. A river where an impossible yield had a different explanation makes the same argument from the opposite direction.

What is actually stable here

Volume. Almost everything else varies more than it does.

For all the complexity above, the mainstem Columbia is among the most dependable large rivers an operator could work. A quarter of a million square miles of catchment does not have bad weeks.

What varies instead is temperature, species timing, tidal state and regulation, and those are the things a client actually experiences. An operation built around explaining them is offering something genuinely useful.

That is a different value proposition from the one most guiding businesses lead with, and it fits a river where nobody needs to be told the water will be there.

Where it turns into money is in trip design, particularly for clients travelling a long way. The mechanics sit in the writing on multi-day packages, and the calendar side in the booking-window data.