angiest's WunderBlog

GFS Futures - 12Z 12/29

By: angiest, 5:28 PM GMT on September 29, 2010

I will not spend any time on newly named Tropical Storm Nichole.  She is pretty well on track and is not long for this world as, within 24 hours or so from now, she is forecast to be absorbed into a non-tropical low.
Beyond Nichole, GFS is still showing a very interesting pattern in the Caribbean and surrounding waters through the first half of October. The first storm forecast to form after Nichole appears to be on October 2nd in the SW Caribbean:
I do not think the small low shown in the Bahamas is actually a tropical cyclone, and I am discounting it for the time.  As of this time steering indicates a path similar to the one taken by Nichole, roughly to the NNE or NE, though a little further right in the short-term.  However, the storm does not move very quickly, and it takes until mid-day on the 4th to move off the northern coast of Cuba:
During this time I am not sure that we ever have a tropical storm.  If this does become a named storm (potentially Otto) it doesn't look like it will be very powerful, and the slow crossing of east-central Cuba certainly doesn't bode well for the survival of the cyclone. I am unsure as to why the storm moves so slowly, as there was a pretty decent weakness for it to move into.  However, by this time, that weakness is getting squeezed between a strong A/B high (1027mb, near the Azores) and a 1022mb ridge over the Ohio Valley.
GFS is showing this storm meandering for a time off the north coast of Cuba without much change in strength, following by a drift roughly to the SSW, taking it slowly back across Cuba, where it is shown emering on the 6th.  As a side note, it looks like GFS still thinks there will be one or more CV storms this late in the season.  Timing of such storm has been inconsistent, however.
A slow drift to the SW with little change in strength is shown until very late in the run, at which point the cyclone finally seems to become a tropical storm and moves toward the Gulf of Mexico:
The storm was apparently moving toward a trough depicted over Canada, but there doesn't seem to be much weakness present over the southeastern US, and the storm makes a very sharp right turn for its third crossing of Cuba:
The cyclone then proceeds rapidly through the Bahamas and into the open Atlantic, giving Florida a glancing blow.  
Of late, GFS seems to be backing off both the number of storms in the Caribbean and on the intensity.  However, the Caribbean still needs to be watched as there is a lot of energy out there for a tropical cyclone, if atmospheric conditions improve.

GFS Futures - 12Z 12/27

By: angiest, 6:09 PM GMT on September 27, 2010

Here is an update on what GFS shows for the Caribbean, Gulf, and SE US over the next couple of weeks.

First, GFS continues to forecast the development of Nichole within the next 1-2 days. It is hard to pick up on an exact time for cyclogenesis, since there is already a very broad area of low pressure in the NW Caribbean. However, in today's 12Z run this seems to be the most likely time:


This is essentially 24 hours from now. However, NHC only places this feature in the medium chance development category for the next 48 hours:


Given how broad and, apparently, still disorganized this feature is, that is not unreasonable. I am not sure why GFS currently forecasts this system to develop into a tropical cyclone so quickly. Regardless, GFS is forecasting this system to cross Cuba by mid-week, and likely be a hurricane in that time. It is worth noting that there is still a tremendous amount of warm water in that area, so rapid development once a cyclone forms is not out of the question, assuming favourable atmospheric conditions.

Shortly after crossing Cuba, Nichole is shown hitting far-south Florida, most likely as a minimal hurricane:


In this GFS run, the A/B high is shown very strong, 1030mb, and positioned near the Azores. In response to this, a strong trough develops and is shown in Canada, extending down through the Appalachians. This trough appears to be what is steering Nichole in this direction. Steering currents are apparently quite strong, and Nichole moves rapidly to the NNE. By Thursday, the cyclone, still a hurricane, is showing making landfall in the Carolinas, squeezed between two ridges and moving quickly toward the trough to its north:


By Friday, Nichole, still likely a hurricane, makes final landfall in the New York area, moving quickly into New England. Meanwhile, another tropical cyclone is shown forming in the Caribbean:


By the time this new Caribbean storm forms, steering currents have weakened, and it does not feel the weakness left in Nichole's wake. Indeed, that weakness already seems to be closing. By Sunday, this new cyclone is shown hitting western Cuba as a hurricane:


Unlike Nichole, which was picked up quickly by a strong trough, this new storm gets stuck in an area of weak steering, which a rather large ridge forecast to be positioned over the Carolinas by the beginning of next week. Therefore, this cyclone moves slowly through the extreme eastern Gulf, just off the coast of Florida, as shown on Tuesday of next week:


Steering is so weak, that for the next couple of days, this cyclone only meanders off the coast of Florida, possibly executing a slow loop before finally making landfall in the Thursday-Friday timeframe, again next week. What I find interesting is that this storm is shown weakening, and may not even be a tropical storm by the time it does make landfall. It is unclear to me why this should be. At landfall, the pressure is around 1002mb:


This storm dissipates over Florida, and appears to be drawn into what I think is a developing Nor'easter off the Carolinas. I won't rule out this this is a tropical or subtropical storm, but Nor'easter looks most likely to be at this time. A respite follows in the Caribbean, and all that is shown for awhile is another large area of low pressure that does eventually develop, but it is hard to pinpoint just when. However, it appears that, by October 12th, there is a solid looking tropical cyclone in the NW Caribbean:


Because this storm forms so late in the forecast period, it is yet unknown where it is going, however, it appears to be heading toward the Yucatan, as depicted in the 384 hour forecast frame:


At this point, it is hard to say where else this forecast cyclone may go. It appears there is a very broad area of low pressure over the southern and southeastern US, which would indicate a turn more toward the north, but steering is not very clear.

As I wrote this, a new Tropical Weather Outlook came out. Currently, the system expected to become Nichole is listed as having a 40% chance of becoming a tropical cyclone in the next 48 hours, and the following warning is in the text:







TD 15 and GFS Futures

By: angiest, 6:07 PM GMT on September 23, 2010

It seems that 95L will soon be updated to a tropical cyclone.  Recon has been investigating the system today and it seems they have now found a closed circulation.  A Vortex Data Message was issued for 95L:

URNT12 KNHC 231542
A. 23/15:26:30Z
B. 13 deg 52 min N
  075 deg 37 min W
D. 32 kt
E. 009 deg 17 nm
F. 090 deg 36 kt
G. 025 deg 26 nm
H. EXTRAP 1008 mb
I. 25 C / 193 m
J. 25 C / 196 m
K. 16 C / NA
N. 1345 / 01
O. 0.02 / 2 nm
P. AF309 01FFA INVEST             OB 09

The VDM does not support tropical storm intensity, but it is possible Matthew will be declared.  Satellite presentation is including, and some banding features may even be visible:
95L satellite from WU
The future of 95L/Matthew is not looking too good, for the storm that is.  Models are coming into better agreement that Matthew will make landfall somewhere in Central America or Mexico and die:
95L models from WU
HWRF remains a northern outlier, and is the only model that seems to offer much of a future for Matthew.  SHIPS seems to peak Matthew as a Category 2 hurricane with winds of 95-100mph.   Regardless, this will be a potentially major rainmaker for areas that dealt with Alex and Karl earlier this season, particularly if the remains stall out, as GFS has forecast for a few days.
By September 28th, GFS is showing the remains of Matthew sitting in the Bay of Campeche, with a new storm forming off the coast of Brazil.  I am not sure what the source of this new storm is, as there didn't seem to be a recognizable low associated with it on previous frame.
I am not sure what GFS is doing here.  This looks like a new tropical cyclone forming.  However, it is shown drifting slowly off to the east while another, stronger, cyclone develops to its east.
The cyclone forming near Jamaica is, I believe, supposed to become Nichole. I am not sure the weak cyclone near Belieze ever becomes named in this scenario.  Whatever the name, the Jamaican cyclone appears to eat the one to its west as it moves in that direction before turning north toward Cuba.  The crossing of Cuba doesn't seem to weaken the cyclone, and it appears that by this time presumptive Nichole will be a category 3 hurricane:
Once again, the cyclone looks to be quite large.  This could reduce the intensity, in much the way of Ike and Alex, which had central pressures indicative of category 4 intensity but winds of "only" category 2.  Notice, there is a fairly strong ridge located over the Northern Plains with a very significant trough located well north in Canada.  At this point there is only one route for the cyclone to follow, and the track is very similar to Charley of 2004.
October 3rd has a large, intense hurricane making landfall near Tampa.  A second landfall is forecast to occur two days later near the North Carolina/South Carolina border by a much weaker hurricane.  By the 6th, presumptive Nichole is shown moving rapidly into the northern Atlantic and undergoing extratropical transition as a very powerful cyclone.
Meanwhile, two more tropical cyclones seem to be forming in the Caribbean, one near Jamaica and one well to the south of Puerto Rico.  The previous hurricane (Nichole) has left a weakness between the A/B high and the ridge over the northern US.  However, because these two cyclones are so close to each other and are of relatively similar strength, GFS currently depicts a Fujiwara effect, with the Jamaican storm drifting somewhat to the south and the eastern of the two storms moving northward and then being ejected from the Caribbean into the Central Atlantic:
This is where the run ends.  It looks like the cyclone in the Central Atlantic may possibly best back to ward the west here, as there is a trough located in Canada.  That may only be temporary, however.  Meanwhile, for the storm that has now been sitting very close to Jamaica for a day or two, I do not see anywhere for it to go except to follow the first storm.  Obviously, this far out it is not even certain which, if any, of these storms might form, let alone where they will go.
I have been following GFS close for nearly two weeks now, since I first noticed a storm forming in the Caribbean that was, at that time, forecast to enter the Gulf.  I have stressed repeatedly, first at Weather Underground, as well as here, that the track and intensity forecasts this far out are not as important as the cyclogenesis.  That remains true.  Even for 95L/Matthew, its demise is not forecast to occur for 5 more days.  A lot can change in that time.  And given the rapidity with which is trying to forecast tropical cyclones in the Caribbean, what happens to Matthew is likely to impact the remainder of the storms GFS is showing.  Basically, the situation is still quite fluid and people living along the coast should monitor the development of these systems.
As I was writing this, the Tropical Weather Outlook was released, which indicates 95L has become a tropical depression and a special advisory will be released shortly.  Also, the remains of former tropical storm Julia, located near the Azores, have a 10% chance of redeveloping in the next 48 hours.

Invest 95L update 1

By: angiest, 6:30 PM GMT on September 22, 2010

The original entry is at Link
Well, this looks interesting.  Model guidance on 95L is rather divergent right now.  First, Weather Underground's model summary:
invest 95l models from wu
Right now, all the models except for one take this system into Central America in the short-term.  This is reflected in the SHIPS intensity forecast, which now takes 95L only to a 90mph hurricane before weakening it to a tropical depression.  The notable outlier is one of the hurricane models, HWRF.  Over the course of the next five days, HWRF takes 95L through the gap between the Yucatan Peninsula and Cuba, apparently toward the Gulf of Mexico.  For the next couple of days, the BAMM model seems most likely until a proper tropical cyclone forms.  Interestingly, GFS currently following BAMM for the first three days.  My own opinion is to discard models like BAMM after we have a tropical cyclone, and I am pretty sure WU will stop showing that one after we have a depression or tropical storm at the latest.  Same goes for NAM.  It would appear NOGAPS has not locked on to 95L yet, as the legend indicates it should be there but I do not see its track.  The only other model on here, GFDL, at least seems to have a reasonable track, though by days 4 and 5 I have low confidence, especially given that the GFS ensembles have a very interesting spread:
WU 95L GSF ens
The operational GFS is the white line.  The others are GFS runs with different parameters used.Most of the ensemble members take a much more northerly track in the day 3-5 period than does the operational GFS run.  Most of them are clustered on Florida or just off Florida's east coast, with a few outliers in the Bay of Campeche and some further east into the Bahamas.  As far as the operational GFS, it takes 95L into Central America on the 25th as a weak system:
12Z gfs
The Azores/Bermuda high is fairly strong at this time, and positioned much closer to Bermuda.  95L is now trapped against a ridge to its west, and strong ridging remains to the north.  Over the next day or so, 95L is shown moving along the coast of Central America, and eventually gets stuck just off Belieze.  This continues, with 95L shown looping off Belieze until the 30th, when it finally makes landfall on the Yucatan Peninsula:
Notice that there is now a second tropical cyclone, situated between Cuba, Hispaniola, and Jamaica.  This is where things become truly interesting. Previously, GFS had shown this second system fizzling and being absorbed by 95L which moves toward it. However, this run is much different.
Over the next couple of days, GFS shows 95L and this second system moving west together, and as 95L moves into the Bay of Campeche, it appears that the two storms do a Fujiwara, with 95L moving to the SW and then SE back onto the Yucatan Peninsula from the Bay of Campeche side.  This marks the death of 95L as far as I can tell.  Meanwhile, skirting along side the south coast of Cuba, the second storm has become very large (and possibly consuming the weakened 95L) and intensifying into a category 2 or 3 hurricane.  As these two storms have interacted, a trough has eroded the western side of the A/B high, and GFS shows this new storm taking that path:
On October 6th, GFS shows this storm (which is not 95L) making landfall in North Caolina, and it looks like it is a very large storm.  What is even stranger is that, after moving up through North Carolina, the storm exits into the Atlantic and starts to drift to the south:
And by the end of the run, this storm looks poised to make another landfall, though as a very weak system:
By the looks of things, GFS says we still aren't done, with three active systems in the Atlantic, plus the one off the Southeast.
GFS is getting more consistent with a storm moving out of the Caribbean and hitting somewhere along the East Coast.  What is troubling for the Gulf is that either of the two storms, 95L or the second storm is that either of these could be a risk, and depending on timing, it is not inconceivable the US will receive a one-two punch, with 95L in the Gulf and another storm off the East Coast.  Obviously, trends need to be watched, and the next 3-5 days will be very important to what happens for the US in the beginning of October.

Invest 95L

By: angiest, 6:01 PM GMT on September 21, 2010

The original is at Wotan's TheXtremeWeather Blog.
So I see the NHC has tagged a new Invest in the Caribbean.  This would be the storm GFS has been forecasting for well over a week now.  Let us see if it becomes a tropical cyclone.  This is the initiation information for 95L from ATCF:
Since there is now something "real" to track I will shift focus slightly. Although I will continue to monitor GFS, especially since it is still developing multiple systems in the Caribbean in the coming weeks, I will start including more models for 95L.  Since this is a newly designated investigation there aren't many models we can look at.  Here is what Weather Underground is showing:
invest 95l models
This is a standard suite of models for Weather Underground to show for new or weak systems.  At least for the next two days or so I don't see much of a problem with these tracks as this matches what GFS has been showing consistently.  As the system gets stronger (SHIPS saying it will be a hurricane by Friday), the more sophisticated models will get a better grasp on where this is going.  In particular, models such as the GFDL and HWRF will be useful.  Although these models are now being run for 95L, I won't pay too much attention to them just yet.  The global models, such as GFS and ECWMF will also become more useful for track guidance as this system becomes established.  As it looks right now, if the cyclone doesn't die in Central America (as some of the above models indicate might happen), then it is likely to get pulled toward a strong trough forecast to move into the Central US in the next several days, allowing the cyclone to enter the Gulf.  This is in line with most runs of GFS in the last week.
Speaking of GFS, it does show 95L making landfall in Central America, but as previously the system re-emerges into the Western Caribbean, and stays there for 2-3 days, just off the coast of Belieze.  Meanwhile, another tropical cyclone is forecast to form near and move over Hispaniola.  At this point, 95L finally moved out into the Caribbean, and apparently will absorb the second storm:12Z GFS
I am curious why 95L is not drawn further north here, as there is a pronounced weakness over the NE Gulf Coast.  Instead, 95L is rapidly drawn to the NE, over Cuba and through the Bahamas.  It then parallels the East Coast until landfall in the New York/New Jersey area:
12 GFS
The huge spread in where GFS has taken this system, from Mexico to the NortEast US, should underscore what I have been saying about not paying much attention to the track this far out.  It is still too early to say it will go this way.  It may not make it past its encounter with Central America. GFS shows at least one more Caribbean system in this run.  It starts forming south of Cuba (the low is visible in the image above) shortly after 95L makes landfall.  After drifting between Cuba and Jamaica for a time, it follows 95L, and at the end of the run it winds up here:
Just off the east coast of Florida.  So, it would appear we will be in for an interesting few weeks in the tropics, as it does appear conditions are more conducive for US impacts.  I will make new entries periodically, as time allows and as changes warrant.


Update 4 for future Caribbean/Gulf storm

By: angiest, 6:11 PM GMT on September 20, 2010

Original post located at Wotan's TheXtremeWeather Blog.

My my, the 12Z GFS is quite "interesting."  At this point I am almost ready to ignore it given the extremely strange scenario it presents.  It looks like GFS is seeing no fewer than three systems develop in the Caribbean over the next couple of weeks, and one of those appears to dissipate and then regenerate!
When I first started seeing this storm on GFS well over a week ago, the system was being developed in the far easter Caribbean, near the southern Leeward Islands.  However, for the last several days (and as recently as the 6Z run on September 20th), this storm doesn't develop until it reaches the Southwest Caribbean.  In today's 12Z run, the system develops in the central Caribbean on 9/23:
12z storm 1 cyclogenesis
The subject is the 1008mb low just off the South American coast.  Of special note is that GFS is showing much stronger ridging over the Atlantic and SE United States, although there is a pretty significant trough located over the KS-NE-CO region.
Contrary to most prior runs, which show the storm moving to the NW or N (with some exceptions showing a NNE to NE motion), 12Z today takes the storm more or less to the west at first, until a landfall, or grazing, of Central America:
12Z GFS storm 1 landfall1
Notice the very strong ridging extending all the way into the Central Caribbean!  No route to the north at this time.  Furthermore, the storm stays weak at this time, perhaps no more than a tropical storm.  But this is where things really start to get interesting.  The cyclone continues generally WNW toward the Yucatan Peninsula and Belieze:
12Z GFS storm1 dissipation
And at this point the cyclone has dissipated or degenerated into a remnant low.  Ridging remains strong across the Gulf of Mexico.
Interestingly. GFS indicates this system will regenerate:
12Z storm1 regeneration
By the 27th, the cyclone has reformed into a depression or weak tropical storm.  Ridging has also begun to breakdown ahead of another strong trough located somewhere in Canada.  And the shenanigans are only just beginning, because GFS now develops storm number 2:
12Z storm 2 formation
The first storm is getting organized again, apparently as a strong tropical storm or perhaps a minimal hurricane, and now there is a new depression off to its east, just south of Haiti.  Although overall weaker than it had been, the ridging present over the US is still at this time exerting over the Gulf of Mexico, and ridging over Mexico seems to be pushing the first storm back to the east slightly.
The second storm does not last very long, and appears that the energy from that one will be absorbed by the first cyclone:
12Z storm 2 dissipates
The first tropical cyclone is now strengthening into a pretty solid category 2 or stronger hurricane.  It is difficult to tell exactly how strong the core is on a global model such as GFS, since the center of the storm is a small-scale feature and GFS doesn't resolve quite that fine.  Ridging off the east coast is strengthening at this time.
By October first, it looks like this hurricane is now category three and GFS show it making another landfall on the Yucatan Peninsula:
12Z storm1 landfall Yucatan
The cyclone appears penned in between several ridges, but a new trough looks to be moving in from the Western US.  With very little room to move, the hurricane continues moving across the Yucatan Peninsula, and storm three forms just north of Hispaniola:
12Z storm 3 formation
Storm 3 appears to move off into the Central and Eastern Atlantic, with storm 1 finally making landfall in Florida on October 6th as what appears to be a very large cyclone:
12Z storm 1 landfall FL
As I said, a very interesting run. I doubt the 18Z run will look anything like this.

Update 3 for future Caribbean/Gulf storm

By: angiest, 1:19 PM GMT on September 20, 2010

Original is located at Wotan's TheXtremeWeather Blog.

I was out of pocket over the weekend so I was unable to do any updates.
It appears there have been some rather interesting changes to how GFS develops this storm.  I am looking at the 6Z GFS run on 9/20.  As is usual, the storm first appears around 9/26:
6Z GFS cyclogenesis1
Still in the southwestern Caribbean.  There is a rather large weakness between the Azores/Bermuda high and another ridge located over the Great Lakes.  There is also a center south of the Florida Panhandle.  This has some significant impacts on the track of this storm, as this run of GFS now shows the storm tracking into the Atlantic and making landfall near the Georgia/South Carolina border:6Z GFS landfall
But GFS is not done yet.  At the very end of the run there is a new storm in the Caribbean:
6Z GFS cyclogenesis2
The previous storm, as well as a trough over southeastern Colorado has opened up a large weakness over the entire Gulf of Mexico.  It seems likely that this storm, if it persists, would track towards the WNW or NW.  There is also another central Atlantic storm present.
Obviously, GFS continues the active season to persist into October, and there really seems no reason to doubt this.   It is also worth realizing that, with a week left until the first storm is forecast to form, it is still not possible to make any guesses as to the future track of this storm, should it form.  I'm also not ready to buy into this storm tracking through the Greater Antilles and Bahamas.  While not unheard of, that is not a common track for storms forming in the Central Caribbean to take.  I will continue to monitor this future storm.

Update #1 for future Caribbean/Gulf hurricane.

By: angiest, 7:49 PM GMT on September 17, 2010

 Original is at Wotan's TheXtreme Blog.
Here's the results of the 12Z GFS for 9/17/10.
Cyclogenesis occurs in the deep SW Caribbean (which seems a tad odd, but that is not important at the moment):
12Z clyclogenesis
This is on September 25th.  As the GFS run continues, instead of impacting the Yucatan as the 6Z run did, the future hurricane passes between the Yucatan and Cuba:
12z img1
This is September 30th.  The strengthening hurricane continues through the Gulf of Mexico, seeming to head for a landfall in Louisiana:
12z img2
That is on October 2nd.  At this time there is a fairly strong ridge over the central plains, but there is a small weakness located over Alabama which would seem to offer a route to the NE for this cyclone(probably a major hurricane).  However, the cyclone starts drifting to the SW, and by the end of the run winds up here:
12z img3
That is October 3rd, and marks the end of the GFS.   At this time, the ridge over the plains has either moved east or split in two, as there is a weakness now located over Nebraska, and again offering a route for the cyclone to move north, this time to Texas.  Based on history that would be unlikely, as hurricanes in the western Gulf are rare after mid to late September.  However, there is no reason to believe it can't happen, especially since La Nina is supposed to lead to warmer winters in this part of the country, which may change things around for tropical cyclone landfalls, as well.

Upcoming Caribbean/Gulf Hurricane

By: angiest, 2:07 PM GMT on September 17, 2010

The following is a repost of my full blog at TheXtremeWeather. Images are available there.

For pretty well a week now, GFS has been developing a hurricane in the Caribbean, and in most runs it has moved into the Gulf. As I have been stressing in Dr. Master's blog on Weather Underground, it is far too early to pay attention to a particular track or to the intensity. Indeed, I have seen this storm go anywhere from the Texas/Mexico border area to heading potentially to the SE coast of the US. Intensity is also apparently anywhere from rather weak to a major hurricane, although it is relatively hard to estimate intensity using a global model.

I understand other models beside GFS have started developing this storm, adding further confidence to the development of this system. As of this time, I would strongly suggest anyone in the Caribbean and Gulf of Mexico to follow tropical weather updates. If you haven't already gotten together hurricane supplies (but you have, right?), you might want to get a few essentials to avoid any rush. Of course, if you live along the coast, this should already have been done before the season began.
I am going to start tracking (as much as I can) the development of this system as forecast by GFS. As of the 6Z GFS run on 9/17/10, it appears the system develops in this image:

And in this particular run, landfall occurs just south of the Texas/Mexico border as what appears to be a major hurricane:

Since this storm is a potential threat to me in Houston, I will definitely follow this closely.

The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.

angiest's WunderBlog

About angiest