ATL: ISAAC - Models

[CYCLONE MIKE]

going to have to go due north from there or its pucker up time for those of us further west.

[Jevo]

0z GFS +57

[MiamiensisWx]

GFS pattern at H5 would mean a landfall SE of NOLA due to the rebuilding ridge...

[HurrMark]

Now at 51 hrs, it is slightly west of 18Z 57 hrs...

Assuming this continues for the rest of the run, this definitely is not good for NO and the central Gulf Coast...we can junk the 18Z all we want, but the 00Z is a major run, and continues the trend.

[bayoubebe]

[quote="Hurricane CheeseThis could get very interesting for the N.O. areas and south. Supposed to have 72hr evac plan. They cant sit on this too much longer before begin having to make decisions.

Agreed. The NHC HAS to at least be giving them their best guesses privately tonight to the emergency planners. I'm sitting here watching the Saints pre-season game, wondering how many people in the Superdome known the growing danger that could be coming toward them...

HUH?? When did this all change? Sorry, I did not read through this whole looong thread. I'm one of those people who watched the Saints preseason game without a worry, last I heard Isaac was headed to Destin?[/quote]

The models showed it 6 hours ago, and the NHC is now taking them into effect, and they are starting to let the general public know that things can change in the near future.[/quote]

Thank you for the reply. It's the NHC that I have been hearing or seeing reported. When should we expect to see an update in track from NHC?

[Jag95]

Looking a little further west to me. This might be interesting.

[ROCK]

Yep close enough on the GFS initialization.,,might ad GFS is notorious for underestimating ridges....again SW LA needs to be on guard...

[Jevo]

0z GFS +60

[HurryKane]

The 0z GFS should be running soon, correct?

Technically, the GFS had been running for about 20 minutes when you made that post. The images we see posted are only the last portion of the running of the model. First, it has to remove old data files and prep the system for the new run (takes about 10 minutes). Then it needs to make its analysis; that's taking in all the observed station data, radiosondes, satellite data, etc. and creating its best guess at what reality is like within its grids. This takes about another 25 minutes. Then it gets started on running through the forecast. As it gets to the determined times, it will dump out and upload output files, which are then used to generate the maps.

Just thought I'd mention it, since it seems like there's this general sentiment (may be real, I may just be imagining it) that somebody manually puts in some numbers representing "the data", then hits a button and out pops some maps, just like that. Really, its a bunch of scripts and programs that pull in huge batches of data automatically, run through some intense algorithms to try and reproduce the world, then numerically approximate solutions to a ton of equations, which some more scripts then use to make the maps we see. It's a process that, despite using large clusters of extremely powerful computers, still takes quite a while (edit - for the GFS, about two hours from start to finish). Hopefully this was interesting to someone out there. Otherwise, I guess it was a decent way to kill some time.

By the by, if you ever want to see how the NCEP models are progressing, you can visit http://www.nco.ncep.noaa.gov/pmb/nwprod ... index.html. Tonight, we're running a little bit ahead of schedule!

As someone who assists a group of oceanographic modelers with their runs on our supercomputers, I very much appreciate your succinct write-up and may crib from it in the future when describing to the general public what we do. Thank you!

Now back to chewing off my fake nails while waiting for the rest of the GFS.

edit:well hell. Based on the image above this post I'll be digging the plywood out of the garage tomorrow.

[ATCcane]

Mouth of the Mississippi at 60hrs Yikes ! Also looks to be slowing

[Jevo]

0z GFS +66

[BigB0882]

Not looking good for LA! Let's see what the last few give us. Will we see that major slow down again?

[BigB0882]

Looks to be more WNW at the end there, might this ride the coast West???

[Countrygirl911]

Looking more liks a SE LA/SW MS land fall do u think they will shift the cone

[CYCLONE MIKE]

Not looking good at all for those of us in SE LA. I really don't want to stay up for euro but I think i might have to

[Ixolib]

....But I can't help but find myself reading your posts in the voice of the Character that Phillip Seymour Hoffman played in Twister.. I smile everytime I catch myself doing it.

Now you made me smile too. But, agreeably, his info is great stuff and much appreciated!!

[StormClouds63]

I thought Georges might be a comparable track, but holy cow ... farther west than that.

[Jevo]

0z GFS +72 (If it takes 72 hours to Evac if needed.. it looks like they may miss their window if this verifies)

[HurrMark]

Looking more liks a SE LA/SW MS land fall do u think they will shift the cone

Don't even see MS in this...looks like LA all the way with this run.

[BigB0882]

Not looking good at all for those of us in SE LA. I really don't want to stay up for euro but I think i might have to

I know. I am thinking of just setting my alarm and waking up to see it. Don't think I can stay up the entire time.