Tropical Storm Isaac, Part 1: Moving Through The Eastern Caribbean Sea

[dtk]

Thanks! Is that only for the GFS or are other models available?

That's not the operational GFS nor GEFS, FYI (see my other post).

[LakeEffectKing]

Just some other background on this....this is a higher resolution deterministic run (T878, ~20 km or so?....I need to check the numbers) than our operational GFS (T574, ~27km), but utilizing similar phyics (for better or worse, untuned) and a comparable DA scheme (they are using an all out EnKF, whereas our operational DA is a hybrid EnKF/Var system). The ensemble is run at T382L64 (centered about the T878 EnKF analysis) versus the operational GEFS which is T254L42 (which uses a much more simplistic ensemble perturbation initialization technique). It's all experimental, but is sort of a proving ground for stuff we are thinking about for future operations (though we are actually working on even higher spatial resolution for the GFS with different dynamics [semi-Lagrangian instead of Eulerian] and tuned phyics ).

Thank you. Do we have any verification #'s/scores vs. other output, irregardless of initialization methods?

[dtk]

Thank you. Do we have any verification #'s/scores vs. other output, irregardless of initialization methods?

Not readily available since we are not routinely generating the standardized tracker stuff that we use for verification. Anectodally, it seems to be doing reasonably well, like the operational GFS, but I honestly have nothing quantative to back up that claim. It's also difficult to judge since it is always 12+ hours behind (as it is being run on the ESRL computer in Boulder, and not our operational or dev. supercomputer here out east).

[LakeEffectKing]

Not readily available since we are not routinely generating the standardized tracker stuff that we use for verification. Anectodally, it seems to be doing reasonably well, like the operational GFS, but I honestly have nothing quantative to back up that claim. It's also difficult to judge since it is always 12+ hours behind (as it is being run on the ESRL computer in Boulder, and not our operational or dev. supercomputer here out east).

Wonderful, thank you for your time.

I have been following it since you introduced it on the board here sometime last season....and like you, the "feel" I get is that it is reasonable guidance to glace at.

Looking forward to addition research from your department.

[shaggy]

PRELIMINARY EXTENDED FORECAST DISCUSSION

NWS HYDROMETEOROLOGICAL PREDICTION CENTER COLLEGE PARK MD

1000 AM EDT WED AUG 22 2012

VALID 12Z SUN AUG 26 2012 - 12Z WED AUG 29 2012

...HURCN ISAAC MAY IMPACT FL DURING THE D5-D7 TIME PERIOD...

THE MAJOR THREAT DURING THE MEDIUM RANGE TIME FRAME WILL BE WITH

HURCN ISAAC WHICH IS CURRENTLY CHURNING THE WATERS NEARBY THE

LESSER ANTILLES. THE LATEST NHC TRACK SUGGESTS THIS TROPICAL

SYSTEM WILL BE LOCATED JUST EAST OF CUBA BY 25/1200Z WITH A

GENERAL SHIFT TOWARD THE NORTHWEST CARRYING IT NEAR THE FL KEYS BY

27/1200Z. FCST MODELS SUGGEST A WEAKNESS IN THE MID-LEVEL RIDGE

AXIS SHOULD EVENTUALLY LIFT ISAAC TOWARD THE SERN U.S. BY MID-NEXT

WEEK. UNFORTUNATELY...THE CURRENT 06Z/00Z MODEL SUITE HAS A

VARIETY OF SOLNS PRESENTED THEREBY COMPLICATING MATTERS FOR

INTERESTED PARTIES ACRS THE SERN U.S. THE CURRENT TRACK INDICATED

BY THE NHC GENERALLY FOLLOWS THE 00Z GFS/GEFS MEAN THRU D5 WITH

HPC CONTINUING ITS GENERALLY MOTION TOWARD THE NNW INTO CNTRL GA

BY THE END OF THE PERIOD. OF COURSE...PLEASE VISIT THE LATEST

TRACK OF HURCN ISAAC ON THE NHC WEBSITE AT WWW.NHC.NOAA.GOV.

ACRS THE REMAINDER OF THE CONUS...THE MID/UPPER PATTERN WILL

GRADUALLY BECOME MORE HIGH AMPLITUDE IN NATURE BY EARLY NEXT WEEK.

A MEAN NEG ANOMALY SHOULD BE FEATURED JUST TO THE WEST OF BRITISH

COLUMBIA WITH THIS AXIS OF HGT FALLS GRADUALLY SPREADING SERD

TOWARD VANCOUVER ISLAND/COASTAL PAC NW BY 27/1200Z. CONSIDERING

TELECONNECTIONS UTILIZING THIS NEG ANOMALY CENTER...A BROAD AXIS

OF MID-LEVEL RIDGING IS SUGGESTED OVER THE MIDDLE OF THE CONUS

WITH A GENERAL MAX ACRS THE NRN/CNTRL PLAINS. IN THIS OMEGA BLOCK

SIGNATURE...AN ELONGATED TROF AXIS WILL BE IN PLACE ALONG THE ERN

THIRD OF THE U.S. THIS FEATURE IN PARTICULAR WILL PLAY A CRUCIAL

ROLE IN HOW MUCH LATITUDE ISAAC WILL GAIN THROUGHOUT THE MEDIUM

RANGE PERIOD.

LOOKING AT THE DETERMINISTIC/ENS GUIDANCE AS A WHOLE...THE IDEA OF

THIS HIGH-AMPLITUDE PATTERN IS CONVEYED ACRS THE BOARD.

HOWEVER...THE TIMING/STRENGTH OF THE NRN STREAM TROF MOVING THRU

THE TN/OH VALLEYS AND ERN U.S. DOES VARY WHICH HAS IMPLICATIONS OF

THE EVENTUAL PATH OF ISAAC. CURRENTLY...THE 00Z ECMWF/ECMWF ENS

MEAN MAINTAIN A STRONG ENOUGH 500-MB RIDGE TO KEEP THE SYSTEM

SOUTH OF 30N. HOWEVER...THE REMAINING GUIDANCE...ESPECIALLY THE

GFS WHICH HAS MAINTAINED ITS SOLN THE PAST 4 RUNS...INSIST ON

ISAAC BEING DRAWN NWRD. OVERALL...HPC FAVORED A SOLN COMPRISING

THE 00Z GFS/GEFS MEAN WITH SOME COMPLIMENTS FROM THE 00Z UKMET

EARLY ON GIVEN ITS FAVORABLE HANDLING OF THE NRN STREAM.

[Drz1111]

Not readily available since we are not routinely generating the standardized tracker stuff that we use for verification. Anectodally, it seems to be doing reasonably well, like the operational GFS, but I honestly have nothing quantative to back up that claim. It's also difficult to judge since it is always 12+ hours behind (as it is being run on the ESRL computer in Boulder, and not our operational or dev. supercomputer here out east).

In a prior life, I used to think a lot about graphic representation of data, and that plot is simply a gorgeous way to present the information it contains. It makes it very intuitive and easy to interpret. Whoever designed that should be in charge of making posters for your entire group - they've got a pimp hand when it comes to graphic design.

[dtk]

In a prior life, I used to think a lot about graphic representation of data, and that plot is simply a gorgeous way to present the information it contains. It makes it very intuitive and easy to interpret. Whoever designed that should be in charge of making posters for your entire group - they've got a pimp hand when it comes to graphic design.

Yeah, all credit to our colleagues at NOAA/ESRL on this one (Tom Hamill and Jeff Whitaker) who have been advocating for these kinds of ensemble track graphics for a while now.

[Srain]

For the record:

TROPICAL STORM ISAAC DISCUSSION NUMBER 6

NWS NATIONAL HURRICANE CENTER MIAMI FL AL092012

1100 AM AST WED AUG 22 2012

THE AIR FORCE RESERVE HURRICANE HUNTER AIRCRAFT MONITORING ISAAC

DURING THE NIGHT REPORTED THAT THE INNER CORE WIND FIELD BECAME

DISORGANIZED AFTER 0600 UTC. A COUPLE OF MICROWAVE OVERPASSES

SUGGEST THAT THE CIRCULATION CENTER BECAME ELONGATED DURING THIS

TIME....AND THE CENTER IS NOT READILY APPARENT IN FRENCH RADAR DATA

FROM MARTINIQUE AND GUADELOUPE. THE INITIAL INTENSITY IS HELD AT

40 KT BASED ON AIRCRAFT-REPORTED WINDS NORTH OF THE CENTER. WHILE

THE CENTRAL CORE IS DISORGANIZED...THE OVERALL STRUCTURE OF THE

STORM IS IMPROVING...AS SATELLITE IMAGERY SHOWS A LARGE OUTER BAND

FORMING AROUND THE CENTRAL CONVECTIVE FEATURE.

THE INITIAL MOTION IS A SOMEWHAT UNCERTAIN 280/15...AS IT IS UNCLEAR

HOW MUCH OF THE OBSERVED NORTHWARD NUDGE WAS DUE TO RE-FORMATION OF

THE CENTER. OTHER THAN THAT...THERE IS NO SIGNIFICANT CHANGE TO THE

PREVIOUS FORECAST TRACK OR REASONING DURING THE FIRST 48-72 HR OF

THE TRACK FORECAST. THE EAST-WEST ORIENTED SUBTROPICAL RIDGE TO THE

NORTH OF ISAAC ALONG 30N LATITUDE IS FORECAST BY ALL OF THE GLOBAL

MODELS TO REMAIN INTACT ACROSS FLORIDA AND THE SOUTHEASTERN U.S.

DURING THAT TIME. THIS EARLY PART OF THE FORECAST TRACK IS AN

UPDATE OF THE PREVIOUS FORECAST. AFTER THAT TIME...THE RIDGE IS

EXPECTED TO WEAKEN SOMEWHAT AS A MID/UPPER-LEVEL TROUGH MOVES ONTO

THE EASTERN UNITED STATES. THE MODELS DIVERGE ON THEIR FORECASTS

OF THE RIDGE. THE ECMWF DOES NOT BREAK THE RIDGE...AND THUS

FORECASTS A MORE WESTWARD MOTION NEAR OR SOUTH OF CUBA. THE

CANADIAN AND NOGAPS HAVE A LARGE BREAK IN THE RIDGE AND TURN ISAAC

NORTHWARD EAST OF FLORIDA. THE UKMET HAS A BREAK IN THE RIDGE OVER

THE NORTHERN GULF COAST AND MOVES THE CYCLONE OVER THE FLORIDA

PENINSULA. THE GFS SHOWS A BREAK IN THE RIDGE EAST OF FLORIDA...

BUT FORECASTS ISAAC TO BYPASS THE BREAK AND MOVE NEAR THE SOUTHWEST

COAST OF FLORIDA. THIS DECREASED-CONFIDENCE PORTION OF THE

FORECAST TRACK IS SHIFTED A LITTLE WEST OF THE PREVIOUS TRACK TO

LIE NEAR THE CONSENSUS MODELS.

WATER VAPOR IMAGERY AND MICROWAVE TOTAL PRECIPITABLE WATER DATA SHOW

DRY AIR TO THE NORTHEAST OF ISAAC...AND A LACK OF CONVECTION IN

THIS AREA SUGGESTS SOME OF THIS AIR IS ENTRAINING INTO THE CYCLONE.

OTHERWISE...ENVIRONMENTAL CONDITIONS APPEAR FAVORABLE FOR

INTENSIFICATION UNTIL THE CENTER REACHES HISPANIOLA IN ABOUT 48 HR.

AFTER THAT...THE INTENSITY WILL BE DETERMINED BY HOW MUCH TIME ISAAC

SPENDS OVER LAND. THE NEW FORECAST IS SIMILAR TO THE PREVIOUS

FORECAST THROUGH 48 HR...THEN IS REDUCED A LITTLE BASED ON

INCREASED LAND INTERACTION IN THE NEW FORECAST TRACK.

A HURRICANE WARNING WILL LIKELY BE REQUIRED FOR PORTIONS OF THE

DOMINICAN REPUBLIC THIS AFTERNOON. IT SHOULD BE NOTED THAT WHILE

NONE OF THE FORECAST POINTS ARE OVER LAND...THE CYCLONE IS FORECAST

TO MOVE OVER HISPANIOLA AND CUBA FROM 48-96 HR.

FORECAST POSITIONS AND MAX WINDS

INIT 22/1500Z 15.9N 59.3W 40 KT 45 MPH

12H 23/0000Z 16.3N 61.8W 45 KT 50 MPH

24H 23/1200Z 16.8N 64.8W 50 KT 60 MPH

36H 24/0000Z 17.3N 67.8W 60 KT 70 MPH

48H 24/1200Z 17.8N 70.3W 75 KT 85 MPH

72H 25/1200Z 19.5N 74.5W 60 KT 70 MPH

96H 26/1200Z 22.5N 78.5W 60 KT 70 MPH

120H 27/1200Z 25.5N 81.5W 70 KT 80 MPH

$$

FORECASTER BEVEN

[Isotherm]

Given the current forecasted NHC track, this will be an interesting test case for tropical cyclone track deflection near mountainous islands. The fcsted storm track takes Isaac immediately south of the island of Hispaniola, but I wouldn't be surprised to see a northward deflection into the island (or maybe even around the northeastern/northern periphery of the island).

As tropical cyclones' wind fields interact with the island's topography, the northern quadrants become disturbed due to the mountainous terrain, while the southerlies remain strong, which can have a cumulative effect in turning the cyclone temporarily northward.

The following study examines the phenomenon in-depth. Will be interesting to see if Isaac does a "dance" around Hispaniola, or remains far enough south to be virtually unaffected by the topography.

http://www.gfdl.noaa.gov/bibliography/related_files/mb8702.pdf

Even if it veers nwd into the island, Isaac can easily reintensify thereafter, depending upon its structure and intensity while crossing the Hispaniola.

[deltadog03]

I still think the EURO is a bit too far WEST and the GFS is a bit too far EAST. I think this RIDES along the WEST coast of FL or just into the EGOM and makes landfall near the big bend of FL. *moving North*

[PSUBlizzicane2007]

I still think the EURO is a bit too far WEST and the GFS is a bit too far EAST. I think this RIDES along the WEST coast of FL or just into the EGOM and makes landfall near the big bend of FL. *moving North*

That's not a bad way to go at this point, I think, especially with what the experimental GFS has been showing. If I'm right about diurnal trends, and it will be interesting to see if this is so, the 12Z GFS will probably end up coming in west of the 06Z run, then shift east with the 18Z and/or 00Z runs.

[SDF_Wx]

That's not a bad way to go at this point, I think, especially with what the experimental GFS has been showing. If I'm right about diurnal trends, and it will be interesting to see if this is so, the 12Z GFS will probably end up coming in west of the 06Z run, then shift east with the 18Z and/or 00Z runs.

I tend to agree with you on that. Honestly, it will come down to how the models handle that trough crossing into the central US in the coming days. In addition, NHC has requested 6 hour balloon launches from sites in MS/AL/FL and San Juan so we should get some better data into the model initializations. I also believe we have a G-IV flight coming up as well which may help in getting some forecast convergence.

SDF_Wx

[wxmx]

Keys are Hispaniola interaction/disruption (I'd weight the Euro a bit higher due to it's better resolution) and how much of a weakness do the trough creates between the Central Plains and Bermuda ridge. The trough is rather weak and progressive, but it could still weaken the bridge between the two ridges. After the trough lifts, the Central Plains ridge is forecasted to build east some, so if by that time Isaac is near the Yucatan channel, it should feel the ridge and head north, and maybe a tad NW, while slowing its translation. My guess is that a 06z GFS and 00Z Euro blend is a good approach, with the areas near the FL panhandle being at the higher risk for now.

[wxsouth]

Just an observation regarding the current structure and impact of dry air to the northeast. The 12Z initialization appears to have a good handle on the relatively dry mid-level airmass over the northeast portion of the circulation (700/500 rh), The GFS appears to get much of the dry air worked out in the first 12 hours, and by 24 hours the entire central portion of the circulation is embedded in a higher RH airmass. With that said, my guess is for Isaac to continue struggling with organization for the next 12 hours or so, and then begin to intensify much more rapidly late tonight and Thursday.

In terms of track, I don't think we'll know a whole lot more until it becomes clear what type of interaction the system has with Hispaniola. It appears to me that, in addition to differences in the evolution of the upper ridge, the GFS and Euro are handling the interaction with the island very differently.

[Bryan Wood]

I still think the EURO is a bit too far WEST and the GFS is a bit too far EAST. I think this RIDES along the WEST coast of FL or just into the EGOM and makes landfall near the big bend of FL. *moving North*

This is the forecast I gave this morning at work. I move it right through Key West and parallel the Gulf Coast about 30-40 miles out to sea until it hits the Big Bend.

[CandymanColumbusGA]

I still think the EURO is a bit too far WEST and the GFS is a bit too far EAST. I think this RIDES along the WEST coast of FL or just into the EGOM and makes landfall near the big bend of FL. *moving North*

Great minds think alike Here's what I posted in the SE forum earlier this morning:

I've been thinking the Euro is too far west the past few runs while the GFS was a bit too far east. The 6z run of the GFS this morning is close to what I thought when I first started tracking Isaac after vacation. The weakness in the ridge has traditionally been over the top of AL/GA for much of the summer. I don't see a reason why this won't be the case this go around either... Also, the 0Z run of the Euro has a front to Isaac's north and doesn't pick it up...? That seemed incredibly weird to me when I saw it this morning.

I truly believe Isaac will impact Florida in some way, shape or form. To what extent? Depends how it handles the Greater Antilles and how long it's over Cuba... The picture will become MUCH clearer once it gets closer to Hispaniola and how it handles the island. That's what I've been telling my viewers all along anyway.

[PSUBlizzicane2007]

I tend to agree with you on that. Honestly, it will come down to how the models handle that trough crossing into the central US in the coming days. In addition, NHC has requested 6 hour balloon launches from sites in MS/AL/FL and San Juan so we should get some better data into the model initializations. I also believe we have a G-IV flight coming up as well which may help in getting some forecast convergence.

SDF_Wx

Absolutely... I'm definitely looking forward to seeing the G-IV data ingested into the models. That will be tomorrow at 23/1730Z with a subsequent one at 25/0000Z.

Also, good to see another Louisvillian meteorologist!

[CandymanColumbusGA]

This is the forecast I gave this morning at work. I move it right through Key West and parallel the Gulf Coast about 30-40 miles out to sea until it hits the Big Bend.

The only thing that makes me pause with this (and my forecast too) is the rarity of it. Yes, there have been hurricanes and tropical storms to hit the Big Bend but they USUALLY come from the southwest... not due south. Not saying it's impossible, just something that doesn't happen that often.

[am19psu]

A lot of good red taggers here are coming up with the same solution (me included). That's better than looking at models, imo.

[HurricaneJosh]

The only thing that makes me pause with this (and my forecast too) is the rarity of it. Yes, there have been hurricanes and tropical storms to hit the Big Bend but they USUALLY come from the southwest... not due south. Not saying it's impossible, just something that doesn't happen that often.

Well, the sample set isn't that large, really. There just aren't that many examples of hurricane landfalls in this region. I don't see any physical reason why a cyclone would be less likely to move due N just offshore of FL's W Coast (compared with a cyclone farther W in the Gulf).

[deltadog03]

you know, it seems like clockwork, EURO is biased west with the ridge and GFS is biased East.....

[Bryan Wood]

The only thing that makes me pause with this (and my forecast too) is the rarity of it. Yes, there have been hurricanes and tropical storms to hit the Big Bend but they USUALLY come from the southwest... not due south. Not saying it's impossible, just something that doesn't happen that often.

Certainly, and given it's this far out, it's likely to deviate. I mentioned this as well, saying it could move inland near Naples or even the Panhandle. I work in the insurance industry, so I may be wishcasting Big Bend since it's far less populated than most of coastal Florida.

[ag3]

12z UKMET crosses Hispaniola on day 2 with a continued NNW movement and by day 3 its just north of the eastern point of Cuba.

Waiting for the rest of the run to come out, but it appears it's sticking with its previous forecasts.

Comparing it to the new 12z GFS, the UKMET is a bit further north of the GFS on day 3.

[Analog96]

you know, it seems like clockwork, EURO is biased west with the ridge and GFS is biased East.....

Happens every time.

[Dsnowx53]

wxrisk.com made this point about the 0z GFS run in comparison to Euro which i think is a good argument to make.

So many times, when it comes down to how deep the northern stream trough is, and how far south a shortwave digs, it's the GFS that is over zealous, and it's the GFS that is wrong. I expect the Euro to win that synoptic battle.

I could see the Hispaniola "pulling" in a storm argument, as the wxmann has said, but I ultimately lean much more towards the Euro at this time. I think the Euro will correct east, but not by much, IMO.

[deltadog03]

Looks like the 12z GFS is going to go through the keys.....and on that current heading its heading for the EGOM The ridge looks stronger this run as well.

[Ellinwood]

12z GFS Op. brings Isaac back east a bit compared to the 6z (and just 1mb weaker on the plot), running it into the Keys:

12z:

And the 06z for comparison:

GFS could be trying to center on a solution that some mets have been forecasting recently.

[Analog96]

12z GFS Op. brings Isaac back east a bit compared to the 6z (and just 1mb weaker on the plot), running it into the Keys:

12z:

And the 06z for comparison:

GFS could be trying to center on a solution that some mets have been forecasting recently.

Joyce looks further West on that run I believe, as well.

[Dunkman]

I think Joyce is just faster on the 12z.

[ag3]

12z UKMET crosses Hispaniola on day 2 with a continued NNW movement and by day 3 its just north of the eastern point of Cuba.

Waiting for the rest of the run to come out, but it appears it's sticking with its previous forecasts.

Comparing it to the new 12z GFS, the UKMET is a bit further north of the GFS on day 3.

UKMET ends up pretty similar to the GFS. West coast of FL then into the Panhandle area.

Tough to tell on French site but it might be a little west of the GFS, but the same general idea.