I emphasize the word "threat" in my diary title as the forecast for any landfall in the U.S. by Irene remains four to five days away (the storm will dawdle its way through the Bahamas over the next couple of days). As such, the forecast isn't even close to being etched in stone. I would say that it's higher confidence than usual for a five day forecast, with the computer model guidance converging dramatically overnight and this morning. But there still remains a general eastward trend to the tracks on some models. And with the models homing in on North Carolina now any more significant eastward trending and Irene could slip out to sea, giving the U.S. a brush-by with little notable impact. But some of the better models have stopped trending much, so it appears to me as though an eventual landfall is likely. But best just called a "threat" at this point.
UPDATE 11AM EDT: Just a quick update to let you know that, as expected, the 11AM NHC track has nudged a bit more to the east. I've not changed any of the discussion below. My thinking is a bit to the east of their's as they never get Irene quite off the coast after hitting NC. But they do have her essentially right on the coast. Since that's the Day 5 position, the difference between their forecast and mine is splitting hairs - though that minor detail could have a huge bearing on Irene's intensity once (IF!) she makes it to NY/New England. That major caveat aside, since our differences are so small, I think NHC's forecast is excellent.
Before getting into all of the model data and how it's converged, let's take a step back and look at Irene's current status. Unfortunately, there's no new recon data within the past couple of hours. But that's ok since previous recon reports clearly indicated that Irene hasn't changed much in the last 12 hours. That's likely due primarily to the Hispaniola mountains disrupting some inflow, plus some slight shear was helping keep Irene in check as well. Unfortunately, she maintains a strong core. Check out this satellite image (the dark reds are very strong convection):
With little sign of shear anymore (there were only some minor hints of it before - it was not strong shear), and that solid core, I cannot imagine any reason why Irene wouldn't begin intensifying within the next 12-24 hours. Earlier recon data suggested the 85kt setting may be overdone, but pressures hadn't risen. So, The National Hurricane Center, correctly, in my opinion, opted to leave Irene at 85kts, acknowledging in their discussion that it might be generous. The scientific stickler in me says they should've downed Irene, but realistically, with little change in central pressure (and, in fact, it actually dropped very slightly - and in case you don't know how a hurricane's dynamics work, it is the pressure that drives the wind) the only good explanation for the wind drop-off is the circulation being inhibitted by Hispaniola. So, the winds should bounce back up almost immediately once Irene gets a bit further away from the Hispaniola coast.
Now, as for Irene's future, let me start with the NHC track. Here's where they have Irene going:
...but keep in mind that I'm writing this around 9AM EDT. So, this is before their 11AM update. And since their 5AM update there was an important eastward shift in both the GFDL and HWRF models (their early morning runs) which dramatically reduced the model spread even further. Their 5AM forecast (above) doesn't look bad at all, really, but don't be surprised if they tweak it even further east again (and you can see why I mention in the intro that there's a real chance that Irene could actually go out to sea with any further significant eastward shifting of the models).
So, what are these models showing? Well, here's a plot from Clark Evans' page:
...with one important caveat, the far west HWFI and GFDI (a little hard to see) plots are the old runs of these models. The newer runs of both of these models have shifted up to a southeastern North Carolina landfall. Here's the HWRF:
...showing a landfall just up the coast from Wilmington, NC. The upper left box showing a max wind of 119kts is well off the ground. Surface sustained winds would be more like 95-100kts... a very strong Category 2, borderline Category 3 hurricane. And here's the GFDL:
...its landfall point is around Morehead City or the Outer Banks. Again, that upper left box is misleading (and hard to read the numbers anyway - but it says max wind is 132kts); the surface maximum sustained winds would be roughly 105kts, a solid Category 3 hurricane. (Note: Even though hurricanes are most intense in the lower atmosphere, the very lowest levels are inhibitted by friction... so, surface winds are generally 20% lower than the 900mb winds shown on those two plots.)
Now scroll back to that track plot and check out all of the lines hitting North Carolina between Wilmington and the Outer Banks. What's striking is what this also means upstream. Only one of those track lines (dark yellow, CGUN) keeps Irene inland all the way up to New England, and only one (GFNI, the purple line) turns Irene hard enough to miss NY/New England after a North Carolina hit. And I can tell you for having already seen it (though the old plots are on that track map), the angle at which the new HWRF and GFDL are striking NC would strongly lean against a track post-NC that's similar to the GFNI. In other words, they would strongly favor Irene coming back offshore near the NC/VA border and then going up to strike NY/New England. Furthermore, also referencing that larger track map, the outlier models are some of the most routinely poor performing (the BAMS, for example, is built for very weak storms; the NGXI has not been a stellar performer; the LBAR is generally not too good either). So, among the reliable performers, accounting for the new HWRF and GFDL the agreement is pretty astonishing for a five-day forecast. That does not mean it's right! But it certainly gives one pause to take this current forecast seriously.
Basically, in total from these models we're looking at a landfall in eastern North Carolina. Then Irene slips offshore on the VA/NC border. From there she continues NNE to a landfall in southern New England - possibly hitting Long Island first, if she tracks west far enough.
And how about Irene's intensity upon landfall. We already saw the the new HWRF was at about 95-100kts while the GFDL was at about 105kts. Here's a plot of various models from Brian Tang's site:
...The strong downturn after 108hrs on several of these models is due to landfall. So, looking at the 108hr time slot all models except one have Irene between 95kts and 120kts. That's a pretty clean range, since it's 5kts either side of a Category 3. Therefore, it seems the wisest bet - based on the models at least - is that Irene landfalls as a Category 3. This also makes good dynamic/historical sense. Conditions look pretty ideal for intensification, and Irene is already a Category 2. On the other hand, the far SW Atlantic isn't quite the bathtub of the Caribbean or Gulf, so it's difficult to sustain a Category 4. In fact, out of 18 U.S. Category 4 landfalls only three have occurred on the East Coast north of Florida (and one of those was borderline - an 1898 hurricane on the GA/FL border). So, a Category 4 on the East Coast is extremely rare. I should note, though, that one of the global models (the European) actually suggests that a Category 4 is not out of the question:
...that shows a landfall near Wilmington, NC (note - importantly for the Northeast - the Euro is a second model that keeps Irene inland after her NC landfall, but it's not on the tracks map above). It does identify 850mb winds as "only" 117kts - which translates to "only" a low end Category 3, having surface winds of 100kts. But its central pressure of 926mb is supportive of a 130-135kt hurricane... a high end Category 4.
Now, if the model consensus if correct - still a HUGE if, keep in mind, we also need to consider the NY/New England landfall. We don't have a lot of intensity data from the models on that yet. But here's a GFS projection:
Plain and simple... that's nothing to laugh at. Those max winds are 106kts... converted to the surface it comes out to 85kts - a Category 2. Now, granted, most of those winds are out over the ocean on the east side of the storm. However, there's plenty of wind on the back side. Plus, of the solutions in the NC to NY/New England track the GFS is one of the more eastern ones (meaning a track more to the west is plausible - putting more of SE New England into those hurricane force winds and extending those lesser - but still strong - winds back further into the NYC area). Furthermore, cross referencing to the Dvorak technique, that 965mb pressure at landfall equates to a 95kt hurricane - slightly stronger than the actual winds the GFS is producing - and approaching minimal Category 3 level. Category 3 is unlikely in New England as Irene will not be accelerating enough to avoid some weakening or partial extratropical transition as she reaches the coast there.
...of course, that all assumes this model consensus is correct. There is way too much room for error this far in advance. The landfall in NC is still about 4+ days away, and in NY/New England it's 5+ days away. At this point, though, you notice I haven't given my own thinking... that's because I can't come up with any logical reason to vary from this consensus. I could argue that we're still seeing an eastward trend... and, so, I should lean more to the east. But the eastward shift in the consensus is largely the result of the far west UKMET, GFDL and, to a lesser extent, HWRF models coming in line with the others. The GFS has pretty well settled in, just bouncing back and forth a bit, perhaps nudging east on balance over the past 24 hours - but only slightly. The Euro seems settled in too - on the western edge of the "best" models. Plus, we saw this jump with the UK/GFDL/HWRF while the GFS/Euro/Canadian remained locked in at a time when a lot of extra data got ingested into the models - because NOAA sent a Gulfstream jet out to sample the environment around Irene with a bunch of dropsondes... which should further enhance confidence in this consensus forecast.
So, my own thinking is just to go with the model consensus as I see no compelling evidence to go in any other direction. That means a southeastern NC landfall as a Category 3 hurricane on Saturday morning, followed by a second NY/New England landfall mid to late Sunday as a Category 1 or 2 hurricane. Though the superb model agreement should make this a high confidence forecast, I'll emphasize yet again that it's not. Is it higher confidence than usual for a Day 4/5 forecast? Yes, most certainly - let's be honest. But even if this verifies with only half the typical error out at Day 4 or Day 5, that's be enough to push landfall back into South Carolina (with no second landfall of course) or, in the other direction, miss making any landfall altogether. So, this is still a low confidence forecast... stay tuned... it's just higher than normal confidence.
I'll try to update this diary or at least post updates in the comments as the day goes on. Of course, I'm working my "real" job, so I can only post as time permits.