| Tak |
| (Weather Watcher) |
| Tue Feb 14 2006 11:19 PM |
2005 Forecast model performance
|
Do any of the mets/moderators know of any analysis of tracking model accuracy for last year? Although I cant find any info, I believe that the GFDL and NOGAPS did the best from 120 to 24 hours across the board.
| Unregistered User |
| (Unregistered) |
| Wed Feb 15 2006 12:14 AM |
|
Re: 2005 Forecast model performance
|
Ah, a statistics question!
Best is relative:
1. Is the model that got 5 storms perfect but missed the remaining 20 storms by over 500nm to the south the best?
2. Is the model that never got one right but got half right by being off exactly 300nm north and the other half exactly 300nm south best?
The first one you had 5 right, but the average is skewed way south so you're likely to get bad predictions southward with the storm, thus it is inaccurate.
The 2nd one you know you are always in the envelope, and by manipulating statistics, you can even say that, "the model predicted the average path of every storm correctly," becuase (300N + 300S)/2 = (0N)/2 = 0nm off.
---
Now, on to the opinions
GFDL, when it was correct, was very correct. When it was wrong, it was very wrong. It was more often correct than wrong. GFS and NOGAPS were ok, not great, but not bad. BAM was often really bad, but occasionally correct (usually when GFDL was wrong). CMC was erratic - sometimes right sometimes wrong and no commonality when.
That's my observations. Anyone have anything official?
| Sarasota3G |
| (Registered User) |
| Thu Mar 02 2006 02:01 AM |
|
|
Re: 2005 Forecast model performance
|
For fun, go to http://www.wunderground.com/tropical/ scroll down to the "Hurricane Archive" section, look at the "Tracking maps and info for all Atlantic storms, 1886-2005" box. Select a year, select a Storm and look at the "Forecast Verification" link. They take a while to run and I won't vouch for validity, but it is interesting to look at. There are patterns,
but for just research they should provide insight. "PLEASE, PLEASE, do not count on this data for future references or predictions"!!! Let's listen to the Met's and other experts for that.
| Tak |
| (Weather Watcher) |
| Thu Mar 02 2006 10:20 PM |
|
Re: 2005 Forecast model performance
|
Thanks for the tip. I'm on wunderground as much as I am on here but I have never used that feature. I usually print out the model updates and then compare the coordinates from 72, 48, 24 hours.