BAM Model Spread for 91L...Stacking Issues?

[MWatkins]

I was looking through the 12Z model guidance suite, and I wandered across the BAM guidance.

The BAM models use a vertically averaged layer for steering in the shallow (BAMS), medium (BAMM) and deep (BAMD). It is reasonable to assume that if the model has a uniform background pattern in all 3 layers, the result sets should come out somewhat close together.

In 72 hours here is the BAM guidance forecast plots:

BAMS 72, 17.7N, 64.6W
BAMM 72, 14.5N, 58.9W
BAMD 72, 16.4N, 55.3W

The shallow BAM is already way out in front of the medium layer run. But by 120 hours, it's even worse:

BAMS 120 23.7N 75.6W
BAMM 120 17.9N 70.8W
BAMD 120 18.1N 60.4W

BAM shallow is way up near the Bahamas...or about 300+ nautical miles from the BAM medium.

This spread suggests the steering flow is not lining up very well in the GFS model...and if 91L organizes it is probably going to have some trouble with vertical stacking.

MW

['CaneFreak]

That's an interesting point there Mike...the models have really been having a hard time with this season thus far....hopefully that will change when we begin to develop large, and more moisture laden systems later in August and definitely by September. I am really concerned about the models...they have done a poor job thus far...

'CaneFreak
Intern H.I.R.T.

[Jim Cantore]

I'm not suprised with the LBAR though

[clfenwi]

That's an interesting point there Mike...the models have really been having a hard time with this season thus far....hopefully that will change when we begin to develop large, and more moisture laden systems later in August and definitely by September. I am really concerned about the models...they have done a poor job thus far...

'CaneFreak
Intern H.I.R.T.

While I kind of read it that way at first, I don't think Mike was saying that there was a problem in the models... when he said"This spread suggests the steering flow is not lining up very well in the GFS model" he's not saying that it's linging up well due to something wrong in the model, he's saying that the flow in the atmosphere is not stacked, which would imply an unfavorable enviroment for the storm. (Of course, Mike, if I'm misunderstanding you and it is indeed CaneFreak who understands you correctly, please set me straight.)

The shallow BAM actually performed well during Chris; a rare situation in which the nature of the storm combined with the fairly static steering current favored it over the others.

[x-y-no]

That's an interesting point there Mike...the models have really been having a hard time with this season thus far....hopefully that will change when we begin to develop large, and more moisture laden systems later in August and definitely by September. I am really concerned about the models...they have done a poor job thus far...

'CaneFreak
Intern H.I.R.T.

While I kind of read it that way at first, I don't think Mike was saying that there was a problem in the models... when he said"This spread suggests the steering flow is not lining up very well in the GFS model" he's not saying that it's linging up well due to something wrong in the model, he's saying that the flow in the atmosphere is not stacked, which would imply an unfavorable enviroment for the storm. (Of course, Mike, if I'm misunderstanding you and it is indeed CaneFreak who understands you correctly, please set me straight.)

The shallow BAM actually performed well during Chris; a rare situation in which the nature of the storm combined with the fairly static steering current favored it over the others.

Yes, that's it. He's saying that if the low-level steering is really that much faster than the mid-level steering, the system is going to have a very hard time getting properly stacked (i.e. mid-level circulation directly over low-level circulation).


I have to say that is a pretty impresive difference in speed. The BAMs don't usually diverge that dramatically.

[Hybridstorm_November2001]

Has the BAM ever correctly predicted a TC's track? It always seem to be a outlier in the final analysis.

[vacanechaser]

is it due in part to the sytrength of the azores high???? thats kinda what i have been thinking most of the year thus far... azores seems a bit strong



Jesse V. Bass III
http://www.vastormphoto.com
Hurricane Intercept Research Team

['CaneFreak]

I agree jesse....azores high is way too strong....allowing sal (dry air) and high shear values thus far....

[x-y-no]

Has the BAM ever correctly predicted a TC's track? It always seem to be a outlier in the final analysis.

Well, even if none of the BAMs were ever right (and I don't think that's a fair statement) that's not really material to the point Mike was making.

[clfenwi]

Has the BAM ever correctly predicted a TC's track? It always seem to be a outlier in the final analysis.

When I said The shallow BAM actually performed well during Chris" I could have said "perhaps was the best performer"... that's the conclusion I came to a couple of days ago when looking at the forecast plot arfchive.

I did a couple of different spot verifications. One of the 00Z forecasts from the 1st onward using the estimated position of the remnants on the 6th as the verifying position, and another from the 06Z forecasts from the first using the 06Z position on the 5th.

Here's how it turned out for the BAMS, GFDL, and the official forecast

BAMS

120h 152 nautical mile error
96h: 50
72h: 43
48h: 23
24h: 18

GFDL

120h: 126 nm
96h: 238 nm
72h: 205nm
48h: 74 nm
24h: 89nm

NHC

120h: (dissipated the storm)
96h: 64 nm
72h: 62
48h: 18
24h: 17


and the 06Z forecasts using the 96 hr from the 1st, 72 hr from the second...

BAMS:
96h: 124
72h: 114
48h: 62
24h: 45

GFDL
96h: 106
72hr: 321
48h: (dissipated the storm 18hrs early)
24h: 62

NHC:
96h: 86nm
72h: 124
48h: 89
24h: 17


Apologies for making a long post... but I felt I should post some data backing up my thought.

Chris happenedthe kind of situation that was "perfect" for the shallow BAM. (Probably) because of the storm not strengthening, the medium and deep BAM performed as poorly was worse as the GFDL. Overall, in the course of the season the BAMs are near the back of the performance list (especially true of the shallow BAM), but there are situations in the course of the season that allow it to perform well.

[P.K.]

thats kinda what i have been thinking most of the year thus far... azores seems a bit strong

Just to make a quick comment about this. The Azores High has been a bit above the 1961-1990 average for the last two months. June, July.

['CaneFreak]

thanks for the info P.K. Thats very interesting....we'll have to watch that over the next month or so...

[MWatkins]

Quick update: The 18Z BAM suite (run off of the 12Z GFS background) is in even worse shape.

BAMS 120 22.3N 78.1W
BAMM 120 19.8N 69.8W

Err...500 nautical miles apart.

MW

['CaneFreak]

geez...not good....

[SouthFloridawx]

Not trying to be negative or anything but, I think the BAMM models tend to differ most of the time anyway.

Let's take a look at these EPAC model plots.

Some of those BAMM models are even farther off than those show above.


http://euler.atmos.colostate.edu/~vigh/ ... early3.png

http://euler.atmos.colostate.edu/~vigh/ ... early4.png

http://euler.atmos.colostate.edu/~vigh/ ... early5.png

[MWatkins]

Not trying to be negative or anything but, I think the BAMM models tend to differ most of the time anyway.

Let's take a look at these EPAC model plots.

Some of those BAMM models are even farther off than those show above.


=Models

Models

Models

Yep...and in these examples the system in question fell apart.

They are almost never the same...but as a general rule the further apart the two are...in both distance and trajectory...the greater the difference in the forecast flow patterns between the shallow and medium layers.

Just something to look for. I can't get to them now but tonight I will try to dig up some data from 2003 for Isabel or 2004 for Ivan...both relatively well-behaved and vertically stacked systems...

MW

[SouthFloridawx]

Yep...and in these examples the system in question fell apart.

They are almost never the same...but as a general rule the further apart the two are...in both distance and trajectory...the greater the difference in the forecast flow patterns between the shallow and medium layers.

Just something to look for. I can't get to them now but tonight I will try to dig up some data from 2003 for Isabel or 2004 for Ivan...both relatively well-behaved and vertically stacked systems...

MW

i Fixed the links above as I had a couple of them the same.

Just trying to get were your coming from here.

[SouthFloridawx]

Plus I can hardly remember when the BAMM models were ever in fine tune with a newly developing invest into a TC. I know they have on some occasions but, I don't think we can base this verticle stacking issues soley on the fact that the bamms are not matching up during Invest Model runs.

Perhaps you can show me some examples of when they have in the past on a newly forming invest.

Thanks Mike

[Javlin]

I take it then Mike if the models are relatively close in agreement for the most part the better chance the system has in development.this being in part that the flow in the low,med and upper layers flow is uniform thus stacking can take place.The research on the other two storms would be interesting if it verifies.

[MWatkins]

Here's an example for Ivan when it was in almost the exact same place in the basin as is this system (nearest 10 and 40). I could pick any run but this is a good example (initial 9.5 43.4) of the forecast tracks and verify times (120 hour)

AL, 09, 2004090506, 03, BAMM 120, 18.5N 70.9W
AL, 09, 2004090506, 03, BAMS 120, 20.2N 72.1W

Note these are less than 2 degrees of lat/long apart...whereas the 91L run is almost 10.

Here's the previous run:

AL, 09, 2004090500, 03, BAMS, 120 19.4N 69.8W
AL, 09, 2004090500, 03, BAMM, 120 18.4N 66.6W

Remember these run against the GFS background so they will do goofy things if the GFS does as well...but the trend should be closer than in 91L.

Isabel (almost same spot) initial postiion 16.6N 40.8W

AL, 13, 2003090806, 03, BAMM, 120, 18.9N 59.6W
AL, 13, 2003090806, 03, BAMS, 120, 17.7N 60.7W

Next run

AL, 13, 2003090812, 03, BAMM 108 19.9N, 61.2W
AL, 13, 2003090812, 03, BAMS 120 18.8N, 65.2W

Hope this helps

MW