Let's hope this goes to the fish. My gas station just rose its price from 2.31 to 2.69
OT: Official Hurricane Irma Thread
- Thread starter RU4Real
- Start date
The 12Z GFS is pretty consistent with the previous run - a tiny bit south and west on the track, with a landfall somewhere around Myrtle Beach. The major difference between this run and the last is that intensity is up.
Bearing in mind the fact that we're a week out, so there's still considerable change that can happen, this looks to be a whopper of a storm. There's plenty of model support at this time for a Cat 5.
Bearing in mind the fact that we're a week out, so there's still considerable change that can happen, this looks to be a whopper of a storm. There's plenty of model support at this time for a Cat 5.
Folks should keep in mind that the GFS has a known bias for overdoing intensity and that there's never been a Cat 5 landfalling north of Florida in recorded history; Hugo was a strong Cat 4 though, when it hit Charleston. Euro is running. We're starting to get to the point where a US landfall between north Florida and the Outer Banks is becoming at least a 50-50 probability and an out to sea solution is looking less likely. It would be catastrophic if some of these model scenarios play out, especially with FEMA so focused on SE Texas and SW LA.
True enough WRT the GFS pressure bias.
Euro coming in now, so far no real substantial changes from last. At 120 hours, north of Cuba just moving into the Bahamas.
Annnnnnnnnnddddddd........
At 240 hours the 12Z Euro is out to sea.
Just goes to show ya, folks - operational models in the 5+ day range are for entertainment purposes only.
At 240 hours the 12Z Euro is out to sea.
Just goes to show ya, folks - operational models in the 5+ day range are for entertainment purposes only.
And another suite of runs that mostly show a hurricane making landfall on the US East Coast, with almost all of them battering the Bahamas as a likely cat 4/5 before doing so. GFS crushes the Myrtle Beach area and the CMC hammers Savannah (very unlikely - they haven't sustained a major hurricane hit in a long time, although hurricanes hit the GA coast more frequently in the 1800s for some reason), with the storm then wreaking havoc inland for days.
The Euro makes a close pass to the entire Florida coast (reminiscent of last year's Matthew), before moving back NE and making a close pass (~100 miles or so) to the SC/NC coasts, but not making landfall, before starting to head ENE, out to sea. Unfortunately, the operational model only goes out to Day 10, where it's several hundred miles off the NC coast. However, despite the "miss" it's so close to FL and then the SE coasts that it would only take small changes for the Euro to still show a hit. .
Still 8-10 days out from a landfall, depending how far north it is, but the consensus (except for the Euro, now, although it's so close to still "count" as a potential landfall this far out) is certainly growing for a possible landfall somewhere between northern Florida and the Outer Banks, with south Florida and up to New England still possible. Out to sea is becoming less likely, but is not out of the picture, as the evolution of an approaching trough across the US and the Western Atlantic Ridge (high pressure), among other players, is still pretty uncertain this far out.
But it's sooooo close to FL and then the SE coast that, to me, it's still in the "almost landfall" camp. Completely agree on the Day 5+ comment, which I've tried to stress many times in my posts. Huge uncertainties beyond that, especially out to Days 9+.
#'s / Real - I believe the Euro Model tends to be the most reliable with snowstorms etc - however is it or has it been as reliable for hurricanes etc - given it appears per your posts above that the Euro is currently the only model taking Irma out to sea. Thanks, Phil
My personal opinion is that the Euro and the GFS are interchangeably accurate depending upon the synoptic conditions prevalent during the period.
The Euro got high praise for "seeing" Sandy with pretty good accuracy, from a pretty good distance. This past winter, however, it got roasted on our big snowstorm, while the shorter-range NAM - a model ridiculed by many as "useless" - had that storm pegged for every minute of its duration.
In a very general sense the Euro is better for tropical systems the last several years but, in Irma's case, the GFS has been more consistent and has also verified more accurately, thus far.
This is anybody's ballgame, and will be until about Wednesday or so.
Yes, the Euro is generally the best model for cyclones, including hurricanes, but it's not like it's hugely better, just a bit better, and it's sometimes wrong. But it did nail Sandy from 7-8 days out, while others struggled with Sandy - that's the most famous "win" for the Euro with hurricanes. Also, we're 9 days out - these model runs are mostly for entertainment value as the uncertainty that far out is 500+ miles on track. It is fascinating to watch the evolution of each model in time, though, and to compare them for each suite. But people really only ought to be looking at the NHC 5-day froecast, which shows a huge impact of a Cat 4 on the SE-most Bahamas at that time.
Thanks as always. Hope all is well. Regards,PhilYes, the Euro is generally the best model for cyclones, including hurricanes, but it's not like it's hugely better, just a bit better, and it's sometimes wrong. But it did nail Sandy from 7-8 days out, while others struggled with Sandy - that's the most famous "win" for the Euro with hurricanes. Also, we're 9 days out - these model runs are mostly for entertainment value as the uncertainty that far out is 500+ miles on track. It is fascinating to watch the evolution of each model in time, though, and to compare them for each suite. But people really only ought to be looking at the NHC 5-day froecast, which shows a huge impact of a Cat 4 on the SE-most Bahamas at that time.
Unless, of course, you happen to reside in the Bahamas.
Actually, the Euro has been more accurate than the GFS (AVNO in the graphic below) and every other model, by far, with track verification, to date, on this storm, especially beyond 48 hours.
Actually, the Euro has been more accurate than the GFS (AVNO in the graphic below) and every other model, by far, with track verification, to date, on this storm, especially beyond 48 hours.
That's very true. However... We're just coming up on 120 hours with this storm, so the verification sample size is incrementally small. Plus, there's simply not as much variation to this point in a storm's track.
What people are criticizing is the wild variations in endpoint prediction from run to run. People are focused on the landfall projections. With respect to that, the GFS has been more consistent. We won't know which will have been more right for another week.
Now, having said that, I will also say that I believe the 12Z Euro is, in my opinion, the most likely solution at this point. I don't like the way the GFS is handling the interaction with the ridge over the Ohio Valley in the 168ish range.
I have people in Orlando until Saturday morning flying back here and friends about an hour west of Myrtle.. what are the odds/concerns at this point for those spots?
I totally agree with respect to the synoptic conditions prevalent during the period. Good point.
So, the 0Z model suite is out and all of the major models have a US landfall, although we're still 7+ days out from that and uncertainty is still pretty high on the track, such that a US landfall is not a given (out to sea is still possible). . The GFS and CMC both take Irma straight up the center of the Florida peninsula from the bottom (in the Keys around Day 7/8) to nearly the top (by Day 8/9), i.e., about 50-75 miles inland from either coast. Highly unlikely, but potentially devastating for a cat 4 hurricane, which won't lose much punch from interaction with the very flat landscape of Florida. The Euro shows Irma again just missing South Florida at Day 7/8 and riding up, just off the east coast of the peninsula, followed by a landfall near Myrtle Beach SC on Day 9 (Tuesday, 9/12). For all 3 scenarios, a landfalling major hurricane will obviously have devastating effects near the coast, as well as hundreds of miles inland from wherever that landfall takes place.
However, what I find very interesting is that both the Euro and GFS show Irma coming very close to Cuba in the 120 hour range, whereas the NHC predicted location at 120 hours is the Central Bahamas, about 75-100 miles NE of where the Euro/GFS have Irma at that time; see the graphics below (if Irma does get close to or hit Cuba, the land interactions could have a significant weakening impact on the storm). And the NHC has seen these model outputs, so clearly they don't quite agree with the Euro/GFS tracks at 120 hours or so. May not sound like much, but 75 miles is the difference between a storm coming up the center of the Florida peninsula vs. coming up just off the coast of Florida, a la Matthew last year. The CMC looks to be almost in the exact same location in the Central Bahamas at 114 hours, 12 hours earlier than when the NHC has Irma in that location (120 hours from the NHC is 126 hours for the 0Z model suite, as the NHC has a 2 am EDT initial condition, vs. the models' initial condition of 8 pm EDT, I believe). And yet the CMC also comes up the center of the FL peninsula.
The point of all of this isn't so much to debate where Irma might make a US landfall, but that a US landfall is now probably more than a 50% probability, somwhere between Florida and the Carolinas. An out to sea solution is still quite possible, though, especially given that the "players" that will ultimately drive the track of the storm aren't even in North America yet, such as the eventual trough that will be approaching the east coast in a week and an upper level low feature in SE Canada; in addition, the eventual position of the Western Atlantic ridge, which is in place, is in question). Actually, TWC guys have done a great job explaining how these features can greatly influence a landfall scenario vs. an out to sea scenario. One near certainty at this time: much of the Bahamas is going to be raked by a Cat 3/4 hurricane (within the 5-6 day period), with the potential to reach Cat 5.
FYI, the odds on a NJ/NYC landfall are now very low, as the storm will come too far west before the turn to the NW/N - we need a track where the turn happens sooner (further east) so that the storm can miss hitting FL/GA/SC/NC and then be "pulled" west into our area, a la Sandy. However, we could certainly be in for a lot of rain from the remnants of any storm that makes landfall, depending on where it strikes (if it does) and how far inland it goes. Some of our worst storms in history, especially from a flooding perspective, were remnants, like Agnes in 1972 or tropical storms after hurricanes weakened after landfall, like Floyd in 1999.
Last edited:
So, the 0Z model suite is out and all of the major models have a US landfall, although we're still 7+ days out from that and uncertainty is still pretty high on the track, such that a US landfall is not a given (out to sea is still possible). . The GFS and CMC both take Irma straight up the center of the Florida peninsula from the bottom (in the Keys around Day 7/8) to nearly the top (by Day 8/9), i.e., about 50-75 miles inland from either coast. Highly unlikely, but potentially devastating for a cat 4 hurricane, which won't lose much punch from interaction with the very flat landscape of Florida. The Euro shows Irma again just missing South Florida at Day 7/8 and riding up, just off the east coast of the peninsula, followed by a landfall near Myrtle Beach SC on Day 9 (Tuesday, 9/12). For all 3 scenarios, a landfalling major hurricane will obviously have devastating effects near the coast, as well as hundreds of miles inland from wherever that landfall takes place.
However, what I find very interesting is that both the Euro and GFS show Irma coming very close to Cuba in the 120 hour range, whereas the NHC predicted location at 120 hours is the Central Bahamas, about 75-100 miles NE of where the Euro/GFS have Irma at that time; see the graphics below (if Irma does get close to or hit Cuba, the land interactions could have a significant weakening impact on the storm). And the NHC has seen these model outputs, so clearly they don't quite agree with the Euro/GFS tracks at 120 hours or so. May not sound like much, but 75 miles is the difference between a storm coming up the center of the Florida peninsula vs. coming up just off the coast of Florida, a la Matthew last year. The CMC looks to be almost in the exact same location in the Central Bahamas at 114 hours, 12 hours earlier than when the NHC has Irma in that location (120 hours from the NHC is 126 hours for the 0Z model suite, as the NHC has a 2 am EDT initial condition, vs. the models' initial condition of 8 pm EDT, I believe). And yet the CMC also comes up the center of the FL peninsula.
The point of all of this isn't so much to debate where Irma might make a US landfall, but that a US landfall is now probably more than a 50% probability, somwhere between Florida and the Carolinas. An out to sea solution is still quite possible, though, especially given that the "players" that will ultimately drive the track of the storm aren't even in North America yet, such as the eventual trough that will be approaching the east coast in a week and an upper level low feature in SE Canada; in addition, the eventual position of the Western Atlantic ridge, which is in place, is in question). Actually, TWC guys have done a great job explaining how these features can greatly influence a landfall scenario vs. an out to sea scenario. One near certainty at this time: much of the Bahamas is going to be raked by a Cat 3/4 hurricane (within the 5-6 day period), with the potential to reach Cat 5.
FYI, the odds on a NJ/NYC landfall are now very low, as the storm will come too far west before the turn to the NW/N - we need a track where the turn happens sooner (further east) so that the storm can miss hitting FL/GA/SC/NC and then be "pulled" west into our area, a la Sandy. However, we could certainly be in for a lot of rain from the remnants of any storm that makes landfall, depending on where it strikes (if it does) and how far inland it goes. Some of our worst storms in history, especially from a flooding perspective, were remnants, like Agnes in 1972 or tropical storms after hurricanes weakened after landfall, like Floyd in 1999.
So, the NHC did adjust their track significantly SW towards Cuba in the Day 4/5 forecast, as per below; not quite clear what this means for the long term, although it has to make a US landfall a bit more likely. Here's what they said in their most recent discussion on Irma's track:
A strong, stationary mid-tropospheric high centered over the central
Atlantic is forcing Irma on a west-southwestward course, and the
initial motion estimate remains 255/12 kt. Irma will begin
rounding the southwestern edge of the high soon, which will allow
the hurricane to turn westward later today and then west-
northwestward in 36-48 hours. Down the road, a large mid-latitude
trough is expected to dig southward over the eastern United States
during the next 72 hours, but the global models have been trending
toward quickly lifting the trough out over New England and eastern
Canada on days 4 and 5, with the subtropical ridge building westward
toward Florida. As a result, it's becoming increasingly likely that
Irma would maintain a west-northwestward heading on days 3 through
5, and the track guidance shifted significantly westward on this
cycle during that period. Remarkably, the track models are very
tightly clustered through day 5, which increases the confidence in
the westward shift of the latest NHC forecast.
The 06Z GFS has Irma's WNW motion continuing past the northern coast of Cuba and over the Keys before turning north for a landfall on Florida's southwest corner, basically right up into Everglades City / Marco Island.
At this point all of the major models agree on the general track - what they don't agree on is the precise placement of that track. Yesterday the GFS was a tad north and east of the Euro, today they've swapped positions.
There's one thing they DO agree on, though - land interactions with Cuba are disrupting the storm and once it does make its U.S. landfall it decays pretty quickly.
At this point all of the major models agree on the general track - what they don't agree on is the precise placement of that track. Yesterday the GFS was a tad north and east of the Euro, today they've swapped positions.
There's one thing they DO agree on, though - land interactions with Cuba are disrupting the storm and once it does make its U.S. landfall it decays pretty quickly.
It means that one run of one model is looking like a Florida landfall. And, as we've been saying all along, that's not the sort of thing you want to get wrapped up about at this point. Give it a day or so - another 3 runs of the Euro, 5 or so more GFS - and see if there's any consistency.
Certainly possible it goes into the Gulf of Mexico, as that's beyond Day 6/7, where track acuracy usually is poor, although with none of the global or hurricane models currently showing a track into the GOM, it's highly unlikely.can this track push the storm into the gulf with two potential landfalls in the US?
One of the things we can be pretty certain of, at this point, is that any effects of Irma in this region will be minimal. Any model support for a Sandy-like solution has evaporated. Not that it was likely to begin with.
I fear for all those Virgin Islands and Bahamas weddings and cruises.. maybe event the Keys.. certainly looks like Puerto Rico will be hit. Thanks for all the updates guys.
I've always wanted to do a bareboat charter in the Bahamas in September. It's pretty near the top of my to-do list. The charter rates are really low, for obvious reasons, and there's the slight chance you get to see some serious weather from the perspective of a 40' sailboat. Which is awesome.
ahhh.. the high-variance lifestyle... wouldn't the boat just stay in port?.. or is it more like a cruise for storm-chasers?
Well, since I'm in command of the boat it would go where I took it.
Realistically, there's an almost impossible window of opportunity. If, for example, you were to board and depart on Saturday and a storm was tracking right at you in the 48 to 72 hour range, they would just cancel your trip and that would be that.
In order for it to align properly you'd need a forecasted miss in the 96 hour range with a track change after you'd already left port. That way you'd have to find a hole to hunker down in. There's hurricane holes all over the Bahamas, it's how they keep all the boats from being destroyed every time one rolls through.
#'s or 4Real....
If as a couple models show, Irma passes over Cuba for 12 hours would it have a chance to significantly weaken before it curves toward Fla or is it just too powerful to have much effect?
Edit: Sorry didn't see the post on this.
If as a couple models show, Irma passes over Cuba for 12 hours would it have a chance to significantly weaken before it curves toward Fla or is it just too powerful to have much effect?
Edit: Sorry didn't see the post on this.
Last edited:
Numbers, can you post the updated Global Model Runs. I'd like to see the overall shift in trend from up the US coast to GOM. As of 12z, as you posted yesterday, only the CMC was predicting this path.While the model spread from today's 12Z runs might look large, this spread at 8-9 days is less than usual. Quite a few landfalls in the Florida to OBX area, but not all, of course. So, we watch.
