Newbie model question

[RickyR]

I have been watching the NHC forecast through several storms lately, and they have usually been pretty accurate.. The NHC projects the track, the other models are somewhat close, then about 12-18 hours later, the models start to agree with the NHC track.
Does the NHC use a model that they don't show. I am not talking conspiracy here, just wondering if they have something we may not know about????
With Ivan, they have been kinda off by moving the track to the left several times, but I attribute this to forward speed playing a factor in the tracking models..
It would be great to know that they may have something, but keep it under wraps, so they don't get blasted if they are wrong...
Keep up the diligence NHC..
Ricky...seeya...

[Wthrman13]

Does the NHC use a model that they don't show. I am not talking conspiracy here, just wondering if they have something we may not know about????

No, there is no super-secret mega-model out there that the NHC uses that they don't talk about. However, there are many other NWP models out there that aren't as much discussed or well-known, primarily because they have yet to completely prove their usefullness in tropical cyclone track prediction, are non-operational, and/or are "experimental" in the sense that they are under development, and have many different versions run by different agencies. Examples include the MM5, ARPS, and WRF models. The WRF model especially is an "experimental" model that is being developed as a joint effort by many different government agencies and universities. It is eventually slated to replace the Eta model in a few years. I daresay that you will see this model being talked about more and more in the coming years, including as it pertains to tropical cyclone track prediction.

Even if the NHC was taking a look at runs of these models (which they probably aren't paying much attention to, they probably stick with the tried and tested operational models most of the time), I very much doubt it is what is influencing the behavior that you are seeing. In fact, what the NHC usually relies most heavily on before making their final, official track, is a consensus of some of these models, called the GUNA (stands for GFDL,UKMET,NOGAPS, and AVN (now the GFS)). The average of these model-predicted tracks is usually better than any of the individual model tracks themselves, and it thus appears sometimes that the other models are "converging" around it's, and thus the NHC's, track. That is likely all that you are seeing.

The reason Ivan's projected track keeps shifting west is not as easy to figure out, but is certainly not surprising. 3-5 day predictions are very difficult for hurricane tracks, and only show a great deal of skill when the steering flow is well-established, obvious and/or relatively unchanging, which is not the case for Ivan (and wasn't for Frances either). In this case we have a diving trough that is eroding the ridge to the north. Small changes in the models to the position, size, and speed of this trough can have major repercussions on how the steering flow around the ridge is affected, and thus the track of Ivan. Also, the representation of Ivan in the models is lacking, because none of the models (with the exception of the GFDL (which has other issues, but that's another story) and other high-resolution regional models) have the resolution to properly capture the intensity and inner-core behavior of the storm. Since Ivan is so powerful, these are likely having important effects on the track that the models simply are not capable of capturing because of resolution issues.

With Ivan, they have been kinda off by moving the track to the left several times, but I attribute this to forward speed playing a factor in the tracking models..

The forward speed of the cyclone is always taken into account, either implicitly as in the case of the dynamical models, or explicitly in the case of the statistical models, so that is really not an issue.[/quote]

[wjs3]

The other "super-secret" model you'll see mentioned on this board--and in the news--is the FSU super-ensemble. It is not released to the public, but is (apparently) a tool the NHC relies on a great deal. Its most famous call (thus far) was with Floyd a few years back. In plain english, it takes a whole bunch of models, adjusts for their known biases and releases a forecast of its own.