ATL: IRMA - Models

[galaxy401]

Lol on the Weather Channel, the forecast models for Irma has one of them going northeast. Which one is that?

[RL3AO]

12 GFS Loop for the record

[KWT]

HMON dives Irma so far WSW that it is actually south of its current location by 114hrs, quite the WSW dive.

Meanwhile HWRF does its own thing and ignores all other models and carries it on W/WNW through 72hrs. Upto 21N.

I know what model I will back given the weight of evidence so far...

I think HMON is out to lunch on intensity though...has a hurricane in 21 hours.

I actually think that is looking quite possible at the moment. Convection is a little weaker at the moment but the system has a cracking shape and has a classic look to it. All it will need is one big burst and it will be on its way.

Anyway ECM will be rolling soon enough, my gut says it will be a little further north on this run but who knows!

We need to watch the Gulf on the 12z ECM and how it handles any lower of pressure in the W.Gulf and the interaction it has with the upper trough.

[gatorcane]

Pretty solid recurve from the GFS ensembles between US and Bermuda though still in the long-range. There are two ensembles that take a hurricane into Florida though and two ensembles into the Carolinas (not shown below)

[chaser1]

I will say this much, i'd give a bit more consideration to the GFS at least come this evening's 0Z run. Part of the issue with the GFS might be a poor initialization of Irma given that it seemingly did not have a very good grasp of the system altogether. This goes as well for the HWRF given that it's run is "GFS data" based. Lets see how/if today's EURO run maintains a more Southwest and west bias or begins to show Irma gaining increased latitude. For that matter, i'll be very curious to see if this evening's 0Z GFS runs suddenly back off on a quick pole ward departure. One obvious thing that might be telling, is simple seeing what the storms present motion appears .... right here, right not. At least from this standpoint, i'd think the more dependable models "should" at least be better capable to forecast the immediate 72-96 hours with reasonable accuracy

[chris_fit]

I believe the 12z Euro should be starting to come in in the next 10-15 min

[RL3AO]

I'll probably be proven wrong, but I think the GFS is on the eastern edge of where I expect Irma to be. I still think it passes west of Bermuda, but a lot can change in 10 days.

[KWT]

Thanks for the ensembles gatorcane.

Having had a closer look, some of the ensemble members that do hit land are among the furthest south of the runs, the ones that broadly take Irma below 17N. Alot of the early recurvers are already north of 20N by 120hrs, which seemingly looking at the other model consensus isn't going to happen. So it perhaps gives more weight to the landfall idea than the pure raw image would suggest.

[tolakram]

12Z Euro Init

[chris_fit]

12Z Euro Init

[forecasterjack]

12Z ECMWF coming shortly. My typical reminder that y'all can get full res ECMWF data for free: https://weather.us/model-charts/euro use the menus to adjust domains/parameters/time steps etc. You can create custom "floater" style domains by selecting custom zoom under map selection menu.

It takes slightly longer for the full res to come in compared to WMO essential stuff on TT, but it's full res worth the wait.

In GFS news, ensembles look fairly similar. Still plenty of spread. Someone tweeted about ridge orientation. ECMWF is E-W whereas GFS is NE-SW. Interesting to think about what impact, if any, that would have. Still so far out though that little if anything is changing at the moment. The storm continues its slow development.

[chaser1]

Looking at the present satellite, then at the 12Z EURO initialization..... even here, its just too hard to see and tell if the storm is being property initialization in terms of true COC and motion. I think' we're all gonna need a few model runs and of course a more clearly defined COC to properly see what its' doing in the present, so to better and more accurately anticipate how close to the Islands it will/might get. As for beyond that, that's a real crap shoot at the moment.

[tolakram]

[KWT]

12Z ECMWF coming shortly. My typical reminder that y'all can get full res ECMWF data for free: https://weather.us/model-charts/euro use the menus to adjust domains/parameters/time steps etc. You can create custom "floater" style domains by selecting custom zoom under map selection menu.

It takes slightly longer for the full res to come in compared to WMO essential stuff on TT, but it's full res worth the wait.

In GFS news, ensembles look fairly similar. Still plenty of spread. Someone tweeted about ridge orientation. ECMWF is E-W whereas GFS is NE-SW. Interesting to think about what impact, if any, that would have. Still so far out though that little if anything is changing at the moment. The storm continues its slow development.

Outstanding website, I'm going to be using when doing deeper analysis that is for sure, especially in the winter time as well, so many useful properties!!

[banksmanforever]

Should Barbados be worried

[gatorcane]

12Z ECMWF already looks south of GFS by a decent amount by hour by 48.

[tolakram]

[AutoPenalti]

12Z ECMWF already looks south of GFS by a decent amount by hour by 48.

Looks like a stronger ridge.

[TheStormExpert]

With the way Irma looks now on satellite the Euro sure is not in any hurry to intensify her.

[tolakram]