Topic Options
#263102 - 04/10/2005 13:32 Model Analysis
Daryl Offline

Registered: 29/03/2001
Posts: 3553
Loc: Mt Nelson , Hobart, Tasmania
I have always assumed that the models use soundings as their primary source of atmospheric profiling. So I was very surprised to see a recent LAPS 12Z Anal chart with the 5280 thickness line down midway between 45S and 50S below Hobart when the 12Z Hobart sounding had the thickness at 5280.

Does anybody have any idea how this could happen?

#263103 - 05/10/2005 08:57 Re: Model Analysis
Craig Arthur Offline
Wind hazard researcher

Registered: 08/05/2001
Posts: 3549
Loc: 149.152009°E 35.187056°...
The models use the soundings and their previous run to initialise the upper level patterns, along with remotely sensed wind data, cloud top temperatures and so forth. The assimilation systems will smear the information from the observations over an influence radius, along with nudging the previous forecast information towards the observations. Along with that, the forecast error covariances (the errors that usually occur in the forecast compared to a self-analysis) will nudge the whole system again.

The information provided by the soundings is mixed in with a lot of other information, not all of it completely accurate. If it was a day when there was a strong thickness gradient over the region, it would be a concern, and you'd have to keep this in mind when looking at the short-term forecasts (first few hours or so). However if the thickness gradient wasn't too great, then it would be less of a problem.

#263104 - 05/10/2005 13:24 Re: Model Analysis
Daryl Offline

Registered: 29/03/2001
Posts: 3553
Loc: Mt Nelson , Hobart, Tasmania
Thanks Craig. This issue has given me plenty to ponder. As well as what you say about short term forecasting, I guess that long term forecasting could at times, need to take this into account as well, as errors amplify with time step in a model run. I can think already of one possible example where a surge of polar air results in a large blocking low pressure system on it's eastern flank. If the Analysis was compromised significantly it could alter the models' projections considerably on following systems, over a few days.

In light of what you said above, it seems to me that the cause of the error in the LAPS Analysis chart was probably largely due to the error LAPS made 12 hrs out in forecasting the intrusion of cold air north. In this case, around 50m of thickness too high.

In areas like the Southern Ocean with such limited sounding coverage I presume the only way to get atmospheric data for fields like thickness, would be through satellite soundings and it would be interesting to know what error margins are involved in such data.

#263105 - 05/10/2005 16:26 Re: Model Analysis
Seabreeze Offline
Weatherzone Moderator

Registered: 18/09/2005
Posts: 10665
Loc: SWR
Does anyone know which model out of Meso LAPS, LAPS, GFS, and GASP, is usually the most accurate? i.e. the model which usually has the highest success rate of predicting the outcomes of weather events.
Also, Do some certain models specialise in certain weather patterns or tend to be more accurate during certain seasons?


Who's Online
7 registered (Yowie, Steve777, bd bucketingdown, Thunderstruck, Warwick Eye2Sky, snowbooby, 1 invisible), 51 Guests and 3 Spiders online.
Key: Admin, Global Mod, Mod
Today's Birthdays
mikesox, muddy
Forum Stats
29947 Members
32 Forums
24185 Topics
1528891 Posts

Max Online: 2985 @ 26/01/2019 12:05
Satellite Image