< political rant>
Before we get to Ike, here's my Sarah Palin diary (doesn't everyone have one?): Can we stop the obsession? 90% of Americans don't vote based on the VP. Maybe a few do, and maybe a handful of Republican are now fired up. Big deal. If we spend half as much time fighting for Obama as we've done handwringing over and railing against the irrelelvant Sarah Palin, we'd be making a real difference and walking away with this thing.
I agree with nyceve's diary that this election isn't a game. For some (people lacking health care or reasonable wages, in harm's way due to poor environmental regulation, in potential need of FEMA assistance in the event of a storm like that discussed herein, or any other number of issues), this election is critical... possibly life or death. Never mind the very real fight for "the soul of America". We need to stop with the trite, silly, ridiculousness and focus on the real issues and get our fellow Americans to understand what's at stake, so they'll make the correct choice and pull that lever for Obama.
</political rant>
...okay, now on to Ike...
Speaking of rants, yesterday's was a rant-filled diary. The primary rant was about how the models could be unanimously wrong, as it appears they were three days ago (or, if they weren't, then they're going to end up unanimously wrong now ...either way, at some point, 100% of them were wrong - and all in the same direction). Then, in an update, I had to rant about NHC's premature downgrade of Ike. I understand why they did it - satellite presentation of Ike was not very good - but had they waited a mere 5 to 15 minutes they would have had aircraft reconnaissance data to definitively say how strong Ike was (and as I mentioned yesterday, they knew that data was coming... the recon "fixes" are scheduled). Oh, well, as it was done by one of their best forecasters, I'll just chalk it up to an honest mistake. Lord knows, I make my share of errors - like buying in too strongly to the models three days ago (and worse yet, when presented with that same unanimity in the models, I'm not sure I'd do anything differently now - except emphasize uncertainty a bit more - reminding people of what happened earlier with Ike). So, we move on...
First, an interesting little lesson on how these forecasts can be so confusing... It would seem reasonable, even to forecasters, to expect that a further south track for Ike early in the period may allow him to miss an incoming trough and remain further south thereafter. (A "trough", by the way, is exactly what it sounds like - a southward dip in the jet stream; these are what come along and pick up tropical cyclones, if they can reach far enough south to "grab" the storm... I've often called them "fronts"; that's technicaly inaccurate, but works for simplicity as these "troughs" are accompanied by cold fronts.) Well, it doesn't always work that way. Check out this image of model forecasts from yesterday afternoon:
...notice that, except for the outlier blue line labeled "EGRR", that most of the tracks running more northward early on, end up being some of the furthest west tracks at the end, and vice versa. For example, the GFDL, GFDT, GFDN, NGX and NOGAPS are all among the most southerly tracks early on, but then end up the most northeastward; while the HWRF, CMC and AVNO are just the opposite (note that the COAL and NGM do not go through the entire period and, so, should not be examined on par with the others). How can this be? Simple, the air is fluid and these more northward tracks of the GFDL, GFDT, GFDN, NGX and NOGAPS require stronger troughing to "grab" Ike. That stronger trough "dipping" down forces the air in advance to push upwards more, expanding the "ridge" (precisely the opposite of a trough) to the east of the trough. That ridge is what's guiding Ike now. If it is forced to expand more, as in those aforementioned models, it will expand in all directions (most likely). The expansion on the south side will force Ike further south, so long as he is under the influence of that ridge. The result? More southward motion now, followed by a sharper northward turn in the eastern or central Gulf. To be clear... I'm honestly unsure which solution is correct (we still have considerable spread this morning), but my point here is to explain how what seems like such an inconsistency can exist. That's how. A more "amplified" pattern forces Ike to go more south now, but more north later. A "flatter" pattern allows him to stay on a more due west track - which is more north now and more south later.
And now, with Ike, there remains much uncertainty, but the situation is slowly beginning to come into better focus. Yesterday's afternoon models began the process of eliminating some of the more extreme solutions. That has continued this morning, with even the stubborn UK model (UKMET, labeled EGRR on the plot below) beginning to come in line. You'll see on that plot that that UK model (the southern blue line) is still south of most of the other forecasts. However, that plot only goes out to 120hrs (5 days); the UK model goes out to six days. So, I can tell you that from day five to day six the UK model turns Ike more northward - heading for landfall just beyond day six along the upper Texas coast. Anyway, here's that plot:
What this plot doesn't show is models beyond day 5, the new early morning runs of the GFDL, GFS, HWRF and NOGAPS, and the Canadian and Euro models. Going beyond day 5, the GFS (AVNO on that plot) shows landfall just south of Houston on Saturday. However, the newer, early morning run is into central Louisiana on Saturday. The Euro model landfalls near Houston on Saturday. The Canadian is a little faster and east, with landfall near Mobile or Pensacola on Friday. Beyond day 5, the Navy model (NOGAPS ...labeled NGPS on that graphic) has landfall near New Orleans on Saturday. Its new, early morning run is about the same - just a shade west in eastern Louisiana. The new GFDL run has Ike heading for landfall in central Louisiana late Friday or early Saturday. Finally, the early morning HWRF aims towards Pensacola, roughly, late Friday or early Saturday.
Bottom line on the track, there are a handful of points to be made:
- No discussion of Cuba here because, sadly, it's a done deal. Cuba will get hammered by Ike. As will the Turks and Caicos and southern Bahamas. Haiti and the Dominican Republic will receive a more glancing blow (well, they already are), but a "glancing blow" is disastrous for them... it means heavy rain, especially through northwestern Haiti. As mentioned in previous diaries, they are struggling to recover from Hanna; Ike is going to seriously compound the nightmare there.
- Landfall remains 5-6 days away, and I can't emphasize enough how large the typical error is in that time range (on the order of hundreds of miles). So, great uncertainty remains and all interests throughout the entire Gulf Coast need to stay on alert.
- That said, there is the slightest increase in confidence this morning. The outlier solutions are fading away (UKMET still a bit of an outlier, but not so much this morning), with all models showing landfall between just south of Galveston (as implied by the UKMET and the overnight GFS run) and Appalachicola (as shown on the GFS Ensemble mean ...and "ensemble" model is where the model is tweaked in a any number of ways and run over and over with these various tweaks). And the real "focus" when you look at the model majority, is really between Houston and Mobile. Moreover, this represents not only agreement, but some level of consistency since yesterday afternoon's model runs.
As for Ike's intensity... well, let's take a look at the storm right now:
...a strange occurence here... Ike looks much stronger than yesterday in terms of the eye. However, because convection (heavy thunderstorms... the deep red in that image) is less than yesterday, Ike is actually no stronger. The aircraft reconnaissance data backs this up. After a brief drop in pressure (indicating a stronger storm) Ike's pressure rose back up to 949mb... identical to what it was last night. Meanwhile, flight level winds have actually decreased slightly. To be clear, Ike is still a very powerful hurricane and the National Hurricane Center's most recent discussion setting him at 115kts (135mph) - a Category 4 - looks about right. Ike shouldn't change much before striking Cuba, so he will likely landfall on the eastern part of the island late today as a Category 4.
How much Ike weakens over Cuba is virtually impossible to say. But this is what the models show:
...and, frankly, it's very difficult to argue with either the HWRF or GFDL on that plot. They are the two "best" intensity models (though the HWRF has overshot the intensities several times this year), they are in excellent agreement, and they have been quite consistent since yesterday. They both show Ike dropping to a Category 1 hurricane while over Cuba. This seems reasonable... favoring severe weakening is that Ike will be over Cuba for quite a while - traveling up the island, and the fact that Ike is a small, easily disruptable storm; favoring less weakening is that Cuba is narrow and has only a small area of significant mountains - keeping much of the circulation over water and preventing huge mountains from disrupting the inner core of the storm. Given Ike's small size, I'd favor the more significant weakening, but wouldn't go overboard with it... so, dropping him to a Category 1, as in the GFDL and HWRF, seems reasonable. Thereafter, both the GFDL and HWRF get Ike back to a Category 3 in the Gulf of Mexico, but cap him off as a strong Cat 2 or weak Cat 3 prior to landfall. This also seems reasonable given the lesser heat content and increased shear over the northern Gulf. The official NHC forecast concurs with slowly bringing Ike back to a minimal Category 3 by the end of the forecast.
Speaking of which, here's the NHC's official track forecast:
...quite frankly, it's as good as anything I could come up with. It's pretty much right down the middle of the guidance. They note in the discussion linked above that they are on the north side of guidance. But this forecast came out before the early morning model runs. In those runs, the GFS and HWRF shifted significantly to the right. The GFDL also nudged right, while the NOGAPS shifted only very slightly left. Given those changes, the NHC is pretty much right down the center of the guidance. Again, though, keep in mind, forecast errors are large that far out.
That's about it for this morning. I'll be away for a good part of the late morning and early afternoon. Still, I plan at least one update later this afternoon, after the midday model runs come in. So, in case we roll off of (or don't make) the rec list, feel free to link or subscribe to this, as I'll do the afternoon update (between 3 and 5PM EDT) whether we're on the list or not.
In closing, check out this image from the Holguin, Cuba radar site, as Ike approaches:
Not sure why, but the Flickr interface appears to reformat that to a jpg file. It's supposed to be an animated gif. Well, if you want to see the updated radar loop from Holguin
go here.
UPDATED 4:50PM EDT Okay, I promised an update between 3PM and 5PM. Running late, very nearly 5PM already, and I want to make sure I get this out to those who are interested in a timely manner. So, I'm going to try to keep this short. And, sorry for not replying to many comments/questions yet. I haven't read them all. I'll try to read and get to them once I post.
The basic story this afternoon is that the model guidance remains in good general agreement and with decent consistency, but definitely still with some shifting around. Most of the shift in the afternoon models is westward. Here's how most of them look:
Please note that these images come from Allan Huffman's nice weather site. You can see that there are two groups of model solutions here, with most in the western group. Despite the decided "split" in the two groups, the range still covers the previous range of solutions - from around Pensacola to the upper Texas Coast. But, generally, you can see that the bulk of solutions has shifted west a bit, say, roughly, between Houston and central Louisiana. The models not on that image have also shifted west. The Euro model is now down towards the mid-Texas coast... so it is beyond the aforementioned range, but only slightly. And the Canadian model now shows landfall near the LA/TX border.
Be forewarned... I've already said it many times, that the forecast this far in advance typically has a fairly large error, and given the model jumping around we've seen, they could easily just jump right back east again. But I do think the consistent focus between Galveston and Pensacola should put that range of coastline on a slightly higher alert. But, really, the entire Gulf Coast needs to continue monitoring.
I should also point out that I've ignored the Florida Keys herein. That's largely because Ike has moved just far enough south so that he is unlikely to provide a major impact on most of the Key. But if Ike makes a bit of a tighter turn than anticipated, the lower Keys, including Key West could get into his "core"... probably not in the eyewall, but into some pretty serious conditions.
As you can see, lots of "what ifs". I'm not trying to waffle here. Just still too far out to know for sure where Ike is headed. The best advice is to really stay on top of this.
As for intensity, here's what Ike looks like now:
That looks more impressive than earlier in the day, with a more well defined eye, and more deep convection (though not a tremendous amount) surrounding the eye. The latest recon data, however, is a mixed bag. The pressure has dropped down to 945mb (and that's down quite a bit from a few hours ago when 950mb was reported), but the winds have been, comparative, extremely low. The flight level winds were "only" 107kts, supporting surface winds down around 90kts. But that seem excessively low, given the pressures. So, the latest National Hurricane Center advisory correctly, in my view, holds him at 105kts.
...and no change in thinking for how strong Ike gets once in the Gulf of Mexico. Models largely unchanged, and the linked NHC advisory above still brings Ike back up to 100kts over the Gulf (after significant weakening over Cuba).
Bottom line: Not any significant enough developments this afternoon to really narrow anything down. There has been a westward shift in the models. But the typical errors, and recent performance (bouncing back and forth) makes this rather small shift pretty uninformative. And no change in the intensity forecast - except that the lower current intensity may, thankfully, spare Cuba more severe conditions... though I emphasize may ...with the recent pressure drop those winds may come back up (winds are caused by pressure difference, so the winds respond to pressure and the response isn't instantaneous... so, the pressure drop could be a harbinger of higher winds in the net 6-12hrs, prior to Ike's Cuba landfall).
Well, I'll wrap it up there... already typed enough, since there's not a whole lot new this afternoon.