The single most useful piece of information available to anyone who fishes, paddles or wades a river in this country is free, recorded every fifteen minutes and posted to the public about hourly, and almost nobody reads it correctly.
Stream gauges are scattered across the country reporting water level and flow in near real time. The data is excellent. The interpretation is where it goes wrong, and it usually goes wrong on the very first number.
Stage is not depth
The gauge page will show you a gauge height, sometimes called stage, in feet. It looks like a depth. It is not.
Stage is measured from a datum that was established locally at that gauge when it was built, for that gauge's own convenience. It is an arbitrary reference. A stage of 4.2 feet does not mean the river is 4.2 feet deep, it does not mean the water surface is 4.2 feet above sea level, and it does not mean anything at all in comparison to a stage of 4.2 feet at a different gauge on a different river.
This matters because people trade stage numbers constantly. Somebody says the river fishes well at three and a half feet, which is genuinely useful information about that gauge on that river, and it is worth nothing anywhere else. Two gauges twenty miles apart on the same river will report different stages for the same water.
Stage is still useful. It is a consistent measure at one location over time, so it tells you whether this river is higher or lower than it was yesterday and than it was on the day you did well last spring. Just never carry the number across a boundary it does not travel over.
Finding Streamflow Data on the USGS NWIS WebsiteDischarge is the number that travels
The other figure is discharge, reported in cubic feet per second, and it is a real physical quantity: the volume of water passing the gauge each second.
Because it is a physical measure rather than a local convention, it is comparable. Comparable between days, comparable between seasons, and, with judgment, comparable between rivers of similar size. When people talk about a river being at 400 cfs, that is a statement with content.
Discharge is not measured directly, which is worth knowing. The gauge measures stage continuously, and a relationship called a rating curve converts that stage into a discharge. The rating curve was built by hydrographers going out and physically measuring the flow at a range of levels, and it gets rechecked and revised.
Which leads to the caveat: after a big flood rearranges the channel, the old relationship between stage and flow may no longer hold, and the reported discharge can be off until the rating is updated. This is also why real-time data is labeled provisional and is subject to revision. Provisional data is good enough to plan a Saturday around. It is not the number to build an argument on.
The statistic that makes it all legible
Here is the feature most people never notice, and it is the one that converts a raw number into a decision.
The gauge page will plot, alongside today's reading, the long-term median for this calendar day across the gauge's period of record. That single line answers the question you actually have, which is not "what is the flow" but "is this a lot or a little for this river at this time of year."
Six hundred cubic feet per second means nothing on its own. Six hundred against a median of two hundred for early September means the river is three times its normal flow and something has happened. Six hundred against a median of two thousand means it is very low and the fish are stacked in the deep water and probably stressed.
Many gauges also publish percentiles, showing where today sits within the historical distribution. Same idea, more resolution. Learn to look at that before you look at anything else.
Read the shape, not the point
A single reading tells you where the river is. The hydrograph, which is the plot of readings over time, tells you what it is doing, and what it is doing usually matters more.
A rising limb is water arriving. Levels climb, the river picks up sediment and debris, visibility drops, and in most fisheries this is the worst time to be there. It is also the most dangerous, because the river is changing under you and a crossing that worked on the way out may not work on the way back.
The peak is brief and is mostly of interest to people getting off the water.
The falling limb is where most of the good comes back. As the river drops and clears, fish reposition and feed, and there is a window on the descent that many anglers regard as the best of the year. The classic advice is to fish the drop rather than the rise, and the hydrograph is how you see one coming.
The recession, where the line flattens toward baseflow, is stable, predictable water. Less dramatic, more reliable.
The other thing the shape shows you is rate of change. A river gaining a foot over three days is doing something different from a river gaining a foot in four hours, even though both end up in the same place. The steep one is a flash event with everything that implies.
Fly Fishing: Reading a USGS Stream Flow GaugeWhere the gauge is relative to you
A gauge is a point measurement on a long river and its usefulness depends on geography.
If the gauge is upstream of where you are going, it is a forecast. Water passing it now will reach you later, and how much later depends on distance and gradient. Watching an upstream gauge climb is the most useful early warning available in a river canyon.
If the gauge is downstream, it is a summary. It is reporting on your water plus everything that joined below you, which means a big tributary between you and the gauge can make the number tell you about a river you are not on.
If there is a dam between you and the gauge, the number describes the dam's schedule rather than the weather. Regulated rivers change flow because an operator changed a setting, sometimes on a daily cycle, and the useful document there is the release schedule rather than the hydrograph.
Check where the gauge actually sits before trusting it. On a map, that means finding the confluences between you and it, which is exactly the kind of reading a topographic map is for.
The other channels
Most gauges report more than water level, and the extras are frequently the most valuable part.
Water temperature, where it is available, is the number that decides a great deal in a trout fishery: whether fish are feeding, whether they will survive being caught and released, and in many western states whether the river is under a heat-related fishing restriction at all. If your river reports it, it belongs in your routine.
Turbidity, where available, is a direct measurement of the thing you were trying to infer from the flow. Clarity is what determines presentation, and a turbidity trace tells you when the river cleared without you having to drive there.
Precipitation, at gauges that have it, closes the loop between what fell and what arrived.
The annual shape, which is the other half of planning
Zoom the hydrograph out to a full year, or look at the median curve across all twelve months, and you get a picture that plans a season rather than a Saturday.
Snowmelt rivers, which is most of the mountain West, have a characteristic annual shape: low and stable through winter, a long steep rise through spring as the snowpack melts, a broad early-summer peak, and a long slow recession through late summer into fall. The fishable windows sit before the rise and on the back of the recession, and where exactly they fall in a given year depends on how much snow accumulated and how fast it left.
Rain-driven rivers, which is most of the East and the coastal Northwest, look completely different: a jagged line all year with sharp peaks after storms and fast recessions between them. There is no single season-defining event, and planning is a matter of watching the week rather than the year.
Regulated rivers look like neither, because they look like whatever the operating agency decided. A tailwater below a large dam can be the most stable water in a region, which is why those fisheries hold up in drought years when everything around them is unfishable.
Knowing which of the three you are dealing with tells you what kind of planning even makes sense. It also tells you where to look on a map: a snowmelt river's condition in July is really a question about elevation and aspect in its headwaters, which is a terrain question of exactly the sort reading a topographic map handles, and which basins feed it is a question the public land and ownership picture in public land layers explained often illuminates.
Building your own number
The point of all of this is to stop guessing and start recording.
Every time you have a good day, write down the discharge, the stage, the temperature, the date and the clarity. Do it for a season on one river. What you will have is a personal rating curve for fishability, and it will be better than any advice you can get from anyone, because it is calibrated to your water, your methods and your judgment.
The reason this works is that the number people are really looking for, the "it fishes well at" figure, is a genuinely local quantity. It varies by river, by reach, by season and by what you are trying to do. Nobody can hand it to you. The gauge can help you derive it in about ten trips.
Safety, without a formula
Moving water is the outdoor hazard people most consistently underestimate, and this is a place where a precise-sounding rule would be worse than none.
Various rules of thumb circulate for when moving water becomes unwadeable, usually combining depth and velocity. They disagree with each other, they take no account of bottom type, temperature, what you are wearing or what is downstream, and treating one as a threshold gives false precision about drowning.
The honest version: force rises steeply with velocity, so a modest increase in flow is a much larger increase in what the water is doing to your legs. Cold water shortens your capability fast. A river that is rising is more dangerous than a river at the same level that is falling, because it is still changing. And the most important variable is what is downstream of you if you go in.
We treat this properly in river and creek crossings. Read the gauge before you go and let it change your plan rather than confirm it.
The gear
The national real-time water data site, bookmarked to your specific gauges rather than searched for each time. Set up a page of the four or five that matter to you and check it as a habit.
An app or alert that will notify you at a threshold, so a rising river reaches you before you have already driven three hours.
A record of your own numbers, in whatever form you will actually maintain. A notes file beats a beautiful system you abandon in March.
A thermometer, if your gauges do not report temperature, because it is the number that most often determines whether you should be fishing at all.
The dam release schedule for any regulated river you fish, which is published by the operating agency and turns an unpredictable river into a scheduled one.
Your state agency's current restrictions, since temperature-driven closures are common in summer and are set by the state rather than signaled by the gauge.
And knowledge of where the gauge sits relative to your access point, which pairs with how to find public fishing access and with the question of what you are permitted to do once you are standing in the river, covered in wading versus floating rights.
The short version
Stage is local and arbitrary. Discharge is real and comparable. The median line is what makes either of them mean something. The shape of the curve matters more than the point on it. And the gauge's position relative to you determines whether you are reading a forecast, a summary or a dam's schedule.
Learn those five things and a free public dataset becomes the most reliable trip-planning tool you own.
It is also worth appreciating what this actually is. A national network of instruments, maintained continuously for over a century in some locations, recording every fifteen minutes and reporting to the public about hourly, for nothing. Nobody built it for anglers. It exists for flood warning, water supply and engineering, and we get to use it because the data was made public by default. Individual gauges do get discontinued when funding moves, so if there is one you rely on, it is worth knowing that its continued existence is not guaranteed and that local agencies and watershed groups are frequently the reason a gauge is still reporting at all.
