Topic: How to Read River Forecasts Posted: 12 Apr 2010 at 11:22am
The only posts I could find on forecasts are side comments on other topics, and mostly warn against using them for anything more than seeing what the "trend" is.
Partly, to make sure I know how to read these, and partly for posterity (as though new comers to any forum search before posting), I wanted to detail what I get from these graphs and ask what I'm missing.
So, first off, how are these to be read?
Here is NOAA's MF Snoqualmie "near Tanner" chart:
http://www.nwrfc.noaa.gov/river/station/flowplot/png/TANW1.HG.0.0.1.0.png
I assume everything but the line to be the historical levels recorded on this day of the year, not the levels we could expect this year for these dates.
The "observed" is clearly what has been recorded this year.
The "forecast" is what we could expect for this year, but how frequently is this updated? Is it an automatically derived statistical forecast, or are there hydrologists, analogous to the meteorologists on the evening news, who tweak the statistical models based on intuition and whether Venus is in parallax or whatnot.
I would nearly always expect the "mean" to be within the limits of the trend, except in outlier cases as you see on 4/20 where there was some very high max, or for some date that had a very low low.
So, am I missing anything here? Are there other resources I can better use to figure out whether I should plan to paddle a certain stretch on a future date? Of course I'll always have a backup plan, even if that just means the buffet at the casino. Still, I'd like to make sure I'm using the available tools to best extent I can.
Is the only difference between the "NWRFC" and the "FC Page" the amount of accompanying text at the NOAA page?
What is the faint gray line at the bottom of the chart with two peaks and two troughs for each day. My guess is high tide and low tide, but I'm not sure.
Note that there is only one "peak" per day on the x axis; the labels are every two days. As such, I always assumed the bottom gray squiggle just showed real days, not just labeled days. On the graph that you posted, river discharge is generally expect to climb through 4-13, and then start declining on 4-14.
And I also assumed that the forecast is based on a hydrology model that computes flows based on anticipated precip, freezing levels, vegetation, historical run off rates, etc, and not hydrologists sitting around a table.
Yeah, not sure if I interpreted your post correctly, but I can definitely clarify the charts.
The historic min, median, and max DAILY flows represent the average flow over a 24-hour period, thus the graph describes the data as discrete points.
The line represents the observed INSTANTANEOUS discharge, which is why the graph represents the data as a continuous line.
What does this mean? Instantaneous flows during peak events will easily outstrip the 24-hour historic average max discharge for that day. This is why the line is often greater than the points.
Not sure how they model the forecast, but it is a model. The fact that it's a model is the reason to not totally trust it, not because it doesn't represent the condition in which we as paddlers are interested.
In the end, any predictive model is uncertain. I think it would be useful if the forecasted flows also had confidence intervals to illustrate what level of uncertainty is present. Without a clarification of the uncertainty associated with the model, it's pretty useless.
sitting all alone on a mountain by a river that has no end
The only way to forecast river levels (in my opinion, which is, obviously, opinionated -- and keep in mind, I'm not a scientist, but only someone who's been paddling around here for the last 17 years) is to look at what the river's doing now, and whether it's going up or down, and to look at the weather forecast and the calendar. (The snowpack report is optional, but if you're a skier you kind of know what's going on there).
If there's rain in the forecast and the river basin is low or if it's the warmer time of year, the river will most likely be going up. It helps to know a bit about the river basin itself, and whether or not snowmelt will affect it. For example, the Skykomish relies heavily on snowmelt, but the Tilton is pretty much all rain-driven. The freezing level for the time of the rainfall is important, too. If the freezing level is 2000 feet and rain is forecast for the Sky basin, the river won't be going up much, if at all, since most of the precip will be falling as snow. If it's 10,000 feet and there's still snow in the mountains, expect it to go up, possibly quite quickly. If there's a lot of snow in the mtns. and there isn't any rain but the FL goes that high, the river will go up, but mostly at night. Also, certain rivers go up and down more quickly. The Snoqualmie will go up and down very quickly, the Sky relatively quickly and the Wenatchee not quickly at all. And then there are the dammed rivers. All I can say about the Green is that if it's raining heavily they're not likely to release since they don't want to flood Renton.
IMO, the river forecast sites are about as useful as the 2-week weather forecasts they give on TV.
Great Topic, Duane. I'm surprised it hasn't come up as such an articulate question in the past. Or maybe it has and I missed it.
Sooner or later James will come along and I'm sure He'll have some insights to contribute. He's a sharp cookie, even if "he no longer boats"
I just use the grey jagged peaks at the bottom to more closely read the time frames within the two day period I'm interested in, relative to where the red line (present instantaneous) is.
As Chipper said, it's a dynamic model. For example, currently the forecast graph for Robe indicates it will be in tomorrow and/or wednesday. I'm taking that with a grain of salt becasue it's not raining much at all at the moment, and the 3-7 day forecasts don't seem to reflect snowmelt conditions (like- sunny and in the 70's).
The PNW is an incredibly dynamic region for weather. It's a moving target anytime the weather is already dynamic enough to make a difference. I try to integrate all of the following resources into my peripery when I am gearing up for a few days of paddling:
-Multiple weather forecasts from multipe sources. (the internet is great, but sometimes the animated dopler radar images on the evening news are really helpful too, and different websites are better to interact with than others)
-The river forecasts you're talking about-- while sometimes inaccurate, sometimes they are spot on. The more you know the micro climates you are looking at, the more you can personally know these inaccuracies. sometimes to the point of being able to look up towards the region you're interested in (like looking up towards the Cascades from the Aurora Bridge, for example)
Here's the most overlooked indicater most paddlers today seem to be missing:
Your boats are tied down and you are rolling towards the river- either hellbent on running something specific, or perhaps up in ther air as to what to do. Maybe you aren't sure if what you want to run will be low enough to be safe. Maybe you want to BE SURE that what you're after is PUMPING solid enough to warrent an extra 30 - 60 minutes up some slow rocky forest service road.
As paddlers drive to "The River" they are constantly crossing litteraly hundereds of visual gauges: creeks and rivers big and small, from Downtown Seattle all the way out to Early Winters or where ever you are goin. As you get closer to your destination you should pay closer attention to every little creek you cross. These relative guages will often inform you greatly. If no one is tailgating you, slow down as you cross the bridge and look at these side creeks.
This sounds all obvious and simple. But here's the catch: for this sort of practice to work for you, it has to be a habit. It's not something you do once in awhile. it's a keen awareness that, if practiced habitually, will dial you into your local region (and from Squamish to Hood River and everywhwere inbetween that's one big region!!). Over time the information will collect somewhere in your brain, and you'll have an internal resource that opperates on a mostly subconscious level. Now when you combine that with all the other info and resources at your disposal, you'll reallly be able to intuit what is going on. It's a relationship that takes time.
Just something to think about
I know I'm digressing form your original topic, but the other thing is this: pay attention to the other localized guages near the ones you are interested in. Sometimes you have no other choice because the creek you want to run doesn't have a guage at all. Try to see the larger tapestry that composes the watershed you are interested in and it's nieghbors. I find it really fascinating to compare the forecast graphs of say, the S. Frk Stilly, the Sky, and the Middle Middle. The shape of their graphs tend to have some predictably familiar signatures to them, like personalities that assert themselves again and again. Compared side by side? It's almost like you could show me all three graphs, without labeling which river it was, and I might be able to tell you which graph represented which river. almost .
This was a great thread, and I appreciate everyone's input. It was great to get both an explanation of the graphs, but also more practical advice on using the streams and weather reports as a tool.
Now I'm trying to guess when either the Yakima or the SF Snoqualmie may come in this year. I'm still looking for some class II water to raft with my daughter.
On the charts, I see the mean and median flows reported over the past 40+ years, but they only show information for dates up to the current date. I can see the past week, including today, or even the past 120 days, but I'd like to see the median or mean flows for the year, or even just the next 120 days. The data must exist or it wouldn't be available for when next Wednesday is the current day, and such.
Does anyone know where to find the mean or median flows across the year, or even for the coming few months?
Keep in mind that the Nisqually, section McKenna to Yelm, is not a bad beginner run. Just mind the fact that there are a couple logs blocking the left side of the class 3 rapid Kahuna. One was cut and the lower one is above water but still be careful. I run it on the right. That rapid is about 4 miles into the trip. It appears after the housing goes away and a big dirt cliff forms on the right. Just below that is another rapid where a car is stuck in the left channel. Obviously stay right. 60 feet below the car is a tree, also from the left side. It is a good easy run, other than the fact that there are a couple hazards. I take new folks down it all the time. Old hands will have half a dozen surf waves that are of good quality, with fat eddies for the new folks to hang out in.
actually, I forgot about that run- it's not bad. I'd think it's pretty darn low by now though. Still- There's the whole Nisqually below there which I haven't explored and would like to. Maybe better suited to canoes or sea kayaks, but the lower reaches of the nisqually are supposed to nice.
Yesterday's rain event is a great example of how notoriously off these predictions on the graphs can be...
The Sky was predicted to go roughly upto 1200 or so, but well below the 1600 mark... ha ha ha, that's funny stuff. Glad it's all wrong.
So this illustrates a good reason why you shouldn't go planning a rio trip based on graph predictions. They are rarely accurate and more often than not well of base.
2 more dirty cents added to this pile of coins....
I think they are usually closer to the mark during a rain event in the rainy season. I think this being summer through them off more than normal. But still, I agree with ya Harms, planning trip based on these predictions is something to be taken with many grains of salt.
Lots of good points here. At best it's a guess. There are tools to make it an "educated guess", but a guess nonetheless. Predictors all come from computer generated trends based on what's happening and what forecasters knwo about a particular region. As the "real-time" updates so do the prediction trends. JP brings up a good point of "studying" your drainages, the correlation between basins will help you figure out where the rain is falling (whether your run is in or not) and you will learn how to incorporate other things, like "back-up" rivers in the area. As you drive to the river, like JP pointed out, look at the streams you cross and "study" them. They are pieces of the puzzle.
This is what I did last year. First, go to www.nwrfc.noaa.gov and find your region. Then look at the prediction pages for the other runs. Sometimes I got help from others who really knew the area, but I still did the leg work of studying the gauges, maybe even looking at them every night throughout the week. Looking at the sister drainages. Sometimes I would look and see what other groups were paddling to see if I was seeing something they weren't or vice versa. I did a fair amount of paddling on the OP last winter and I learned a lot about gauges, predictions and all that. The other thing to consider is the freezing level. We did the Upper Upper Sitkum one day and Hoov didn't even come out, said we were crazy and the river was "bloating". When we got to the put in there was about 3 or 4 inches of snow and the river was butt low.
Learning the gauges (even though it is still a guess), watching freezing levels (also how high above or below the gauge is to the run. the gauge could be way below like on FITW and will read higher than what's in the riverbed way up where you'll be kayaking), having a good source if you don't know the area, and watching the basic trend of the forecasts. Last year I watched some weeks things get predicted high, consistently, and others similar trends. One other "tool" I developed for winter boating with predictions is having a list of back-up rivers in the area. Maybe on the way in you pass a couple and notice they are low, but good, and your buddy tells you that "if these are low, they upper is gonna be the bone zone", then the choice becomes clear. Cuz it's all about gittin on the river for a good day of boatin anyway, right?
Or I guess there is one last option... for a small beverage contribution, I could be persuaded to do thework and tell you when and where ;) But that kinda takes the fun out of it!! Hope this helps, have fun!!
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum
Register
Login
Active Topics
Memberlist
Search
IP Logged
