computer models

[msbee]

I know this is being explained in the other thread, but I just need a simple explanation. which models are more reliable?
in other words. which one should you pay the most attention to?
and which ones should you dismiss?
thanks
Barbara

[GeneratorPower]

There is no quick answer to that question. All the models provide a piece of the puzzle. Knowing which models are probably closer to reality is a decision which is made by very experienced forecasters with lots of technical knowledge about the models and how they arrive at their conclusions.

Best thing to do is to listen to others discuss the model output and rely heavily on the National Hurricane Center forecast cone.

[x-y-no]

The most simplified answer I can manage is this:

CLIPER and NHC98 are of very little worth - their main purpose is in evaluating the performance of others in post-analysis.

LBAR is also of very little worth.

the BAM models - BAMS (shallow), BAMM (medium), BAMD (deep) are useful for their corresponding strength systems in the deep tropics, but not much good as systems move further north.

The dynamical models - GFDL, HWRF, GFS, UKMET, ECMWF, NOGAPS (not so much ETA and CMC) are all useful. Best overall is probably the GFDL (don't have a baseline yet on HWRF which is new) but no one model outperforms others in all cases.

Hope that helps a little.

[Lowpressure]

There is no clear answer to your question. A lot depends on how well models initialize. Some work better with developing systems while others work better with stronger systems. You will see spreads on many models runs based on several environmental factors. My best advice would be to watch and learn each models character, that is what are their tendancies. NHC paths usually are blends and tend to be rather accurate. Hope this helps some.

[msbee]

OK
Thanks everyone
Barbara

[WindRunner]

The most simplified answer I can manage is this:

CLIPER and NHC98 are of very little worth - their main purpose is in evaluating the performance of others in post-analysis.

LBAR is also of very little worth.

the BAM models - BAMS (shallow), BAMM (medium), BAMD (deep) are useful for their corresponding strength systems in the deep tropics, but not much good as systems move further north.

The dynamical models - GFDL, HWRF, GFS, UKMET, ECMWF, NOGAPS (not so much ETA and CMC) are all useful. Best overall is probably the GFDL (don't have a baseline yet on HWRF which is new) but no one model outperforms others in all cases.

Hope that helps a little.

In particular, the UKMET and NOGAPS are usually the poorest of the higher-end dynamicals listed above. The GFDL and HWRF models are only run once a cyclone forms, and therefore offer no insight as to the possibilities of tropical cyclogenesis. The GFS, in its old version, had a bias that in which it was excessive in forming tropical systems, though the extent of this bias with the new GFS is not yet known. The ECMWF is the top-of-the-line model in the mid-latitudes, so I would assume it would perform similarly in the tropics as well. Its limited accessibility, however, makes its use a little uncommon, especially for the public.

[x-y-no]

The GFS, in its old version, had a bias that in which it was excessive in forming tropical systems, though the extent of this bias with the new GFS is not yet known.

Another well-known bias with the old GFS which I'm curious to see if it has been cured is a tendency to break down mid-level ridging north of storms way too fast.

[Regit]

LBAR is also of very little worth.

Oh come on now. If not for the LBAR, we'd have much less off-season entertainment on this board.

[Berwick Bay]

XY wrote
Another well-known bias with the old GFS which I'm curious to see if it has been cured is a tendency to break down mid-level ridging north of storms way too fast.


You got that right. Ivan and Katrina tracked quite a bit further west then they were supposed to. Even Jeanne was originally thought to stay east of Fl, but instead crossed the state before making its turn further west than anticipated. 06 had no Gulf storms of real note. Interesting to see the performance in 07.