 Poster: A snowHead
|
I started this last year in a bid to have one page with all the 'important' info on it, seems to work well for me.
I'd be happy to add in any other resorts if other folk think it useful, just let me know the resort name and wiggly URL and I can add it....
http://www.gregh.co.uk/php/gfsruns.php?select=stuben
Cheers,
Greg
|
|
|
|
|
|
Obviously A snowHead isn't a real person
Obviously A snowHead isn't a real person
|
|
|
|
brian
brian
Guest
|
Ahem, pay attention at the back there young kitten ...
"including the last dozen GFS forecasts and an annual average. The average data is shown by the red line, the current forecast by the green line, and the average of the last dozen forecasts by the grey line."
The GFS ensemble members shown on the graph are run at the same time using the same input data but each run slightly tweaked within the margin of error of the input data. It's not the last dozen forecasts, it's the current forecast done multiple times.
|
|
|
|
|
|
You need to Login to know who's really who.
You need to Login to know who's really who.
|
brian, go on then jedi master, what's wrong with that sentence??
Shakes his head at himself in disgust thinking that he had really understood this 
|
|
|
|
|
|
brian
brian
Guest
|
kitenski, I thought I explained it in my sentences following yours. 
|
|
|
|
|
|
You'll need to Register first of course.
You'll need to Register first of course.
|
|
sorry, just to quibble some more, isn't the red line a 30 year average? The phrase "annual average" is a bit confusing.
|
|
|
|
|
|
|
|
ah yes , how does this sound then?
The graph that is displayed shows precipitation and the 850mb temperature forecast including the current GFS forecast runs (each run is the same forecast but with tweaked input data) and an average of all those runs. The average historical data (from the past 30 years) is shown by the red line, the current forecast by the green line, and the average of the last dozen runs by the grey line. Each other line is the same forecast with the tweaked data.
|
|
|
|
|
|
brian
brian
Guest
|
|
|
|
You'll get to see more forums and be part of the best ski club on the net.
You'll get to see more forums and be part of the best ski club on the net.
|
|
brian, I've suddenly had a eureka moment here now with these charts, are they essential a probability fan chart, showing a range of probabilities and the further you get from the central prediction the less likely the probability?
|
|
|
|
|
|
|
|
|
stevew, well put! Sounds exactly how I see them too...
|
|
|
|
|
|
snowHeads are a friendly bunch.
snowHeads are a friendly bunch.
|
|
brian, kitenski, or you could say that 850mb exactly corresponds to 1500m, given an ISO standard atmosphere, with pressure of 1013mb at sea level.
|
|
|
|
|
|
And love to help out and answer questions and of course, read each other's snow reports.
And love to help out and answer questions and of course, read each other's snow reports.
|
|
kitenski, i thought there were 20 'tweaked runs' (haven't bothered counting though). you might also like to mention the thick blue line which is the current forecast, but run on older 'lower resolution' software.
|
|
|
|
|
|
|
|
*walks into room sees nothing interesting, walks straight back out again* 
|
|
|
|
|
|
You know it makes sense.
|
| rogg wrote: |
| kitenski, i thought there were 20 'tweaked runs' (haven't bothered counting though). you might also like to mention the thick blue line which is the current forecast, but run on older 'lower resolution' software. |
Where's the thick blue line?? I see red, green and grey???
|
|
|
|
|
|
Otherwise you'll just go on seeing the one name:
Otherwise you'll just go on seeing the one name:
|
kitenski, show below:
|
|
|
|
|
|
Poster: A snowHead
|
|
kitenski, Are you sure you are not colourblind t o blue? The broad blue line ends just above 10 on the right axis, just below the narrow red - if you don't see that one as blue more info here
|
|
|
|
|
|
Obviously A snowHead isn't a real person
Obviously A snowHead isn't a real person
|
| Quote: |
the further you get from the central prediction the less likely the probability?
|
is that really the case? I thought it was more that where there is a very big spread and confusion in the wigglies, the data were such that very small changes in input led to very big changes in output. Where that's the case, the "central prediction" is probably not much more probable than the rest? So when the lines are absolutely all over the place, best take very little notice of any of them. Is that right, jedi master?
And in terms of the fan over time, I suppose it's theoretically possible to envisage a reverse fan - where a big weather event a few days away is pretty certain, but that there's lots of instability and uncertainty in the very short term, until it hits?
the nature of the beast obviously means that that outcome would be unlikely - but it's not impossible?
|
|
|
|
|
|
Well, the person's real but it's just a made up name, see?
Well, the person's real but it's just a made up name, see?
|
|
Oh - another thing! When people embed GFS pictures, is there a way of locking them? The best way to learn is to be able to read an informed interpretation of the picture - but in the sticky forecasting thread, when carled says that a chart for April 2006 was a "belter", and explains why, it's not helpful if the chart in the same post is for 4 December 2009. If we could look at the words and the pictures together, I'm sure we would learn a lot quicker. I for one would really like to be able to read interpretations (from the people who know about them) alongside the pictures. The sticky thread is useful, but although I'm now used to the changing of the charts it caused me lots of head scratching when I first started trying to follow it, and couldn't make sense of what carled was saying. It took me a while to realise that the chart wasn't the one he was talking about, and I'm sure I'm not the only thickie round here.
|
|
|
|
|
|
You need to Login to know who's really who.
You need to Login to know who's really who.
|
cheers all, I saw the blue line just didn't think of it as 'thick' ( a bit like me .....)
pam w, you can only lock them by downloading the image and then uploading to the snowmediazone and linking to that, ie from 2007 looks like they had less 'runs'
|
|
|
|
|
|
brian
brian
Guest
|
Ignoring the green line for a second (referred to as the operational and run at a higher resolution, ie. over more locations with a finer granularity) then all the other ensemble members are equally likely. So if 19/20 ensemble members predict a big dump of snow a few days out then it's fairly likely.
However, the fairly large caveat to all of this is the reliability of the model in general. It's not unusual to have cast iron agreement of practically the whole ensemble for 5 or 6 days out but for the whole thing to flip in a couple of runs time.
My rules of thumb are -
- within 4 days, pretty likely but small variations still likely (and small variations can have big effects, eg. missing a localised snowstorm completely).
- 5-6 days, worth looking at but could be subject to major change
- beyond 6 days, pretty much for interest only, can sometimes pick up trends of interest
|
|
|
|
|
|
You'll need to Register first of course.
You'll need to Register first of course.
|
As a webby geek rather than a weather geek I thought this might be useful:
GFS graph grabber
How does it work?
1) Select the co-ordinates for the GFS graph you want (e.g. 45,7)
2) Click "Get GFS graph button"
3) A copy of that graph, for that moment in time is then stored on my server
4) You then just need to cut and paste the code from the box beneath the graph image
Any suggestions or comments as always appreciated.
e.g.
Note: To save space it will only allow one run per set of co-ordinates per hour (sorry!)
|
|
|
|
|
|
|
|
| Quote: |
It's not unusual to have cast iron agreement of practically the whole ensemble for 5 or 6 days out but for the whole thing to flip in a couple of runs time.
|
Sometimes, that's a really happy thought. 
|
|
|
|
|
|
|
|
|
|
|
You'll get to see more forums and be part of the best ski club on the net.
You'll get to see more forums and be part of the best ski club on the net.
|
|
You're a genius, and I'm not just saying that because its predicting snow and a drop in temps the day I arrive.
|
|
|
|
|
|
|
|
FAQ
