Watching Rita, some models

September 22, 2005 6:15 PM

Noted in the live stream from this TV station This is the "Local2 News" live tv stream (which has been pointed to in three previous MeFi threads about other news stories. Currently they've from time to time been showing storm track predictive models (which they say are their own development). I'd rather have pointers to more models than the TV station's occasional glimpses, but, this is the most varied set of storm track predictions I've seen. Anyone know where they're getting them?
posted by hank (24 comments total)

A bit more about this one:

There have been five or six different paths depending on ... whatever. Not much explanation, no idea where the graphics are coming from, maybe someone knows.

Currently their highest probability is for Rita to head more to the East than NOAA is saying -- and they're mentioning one possible track that goes inland, halts and moves back southwest into the Gulf again as an earlier hurricane did.

I don't know if any of this is archived anywhere, just watching the live stream and seeing them come up with this from time to time.

They do go through the weather systems (two highs to the north, with a band of low pressure further north splitting them, leaving a 'weak spot' Rita seems apt to take) and describe how if the high over the East Coast fades out, the high now over the Rocky Mountain states could move east and block the hurricane.
posted by hank at 6:16 PM on September 22, 2005

It's called guessing.
posted by Balisong at 6:25 PM on September 22, 2005

It's not clear where they get them, but this blog at Weather Underground has these maps. For instance.
posted by smackfu at 6:43 PM on September 22, 2005

You can find the 'spaghetti plot' (so called because the models often go every-whichway) here:
boatus.com

This shows the output of different computer models and their predictions on where the storm is headed. The more they agree, the more likely the path. But hurricanes have minds of their own...
posted by bitmage at 6:44 PM on September 22, 2005

Good ones all. Thank you.

From the wunderground blog suggested above:

"All the global models are forecasting Rita to slow down and become quasi stationary by late Saturday night
somewhere along the TX/LA border about 100-150 miles inland. This occurs in response to the rebuilding of a ridge to the north of the storm over the central U.S.. The ridge then continues to expand towards the east and southeast, and is expected to turn Rita towards the southwest later Sunday and into Monday.....

.... This greatly raises the chances of severe flash flooding over inland areas on Sunday and Monday -- with rainfall totals easily exceeding 25" in some locations - with 30" or more quite possible in some isolated locations.?
posted by hank at 6:50 PM on September 22, 2005

The Houston Chronicle's SciGuy states in this entry that he has two "proprietary" forecasts that suggest that Rita will hit land even further to the east than the NOAA is currently stating. Got my curiosity up, so this post is quite timely. Thanks.
posted by WolfDaddy at 6:54 PM on September 22, 2005

I'd like to know more about this, too, but mainly because I want another reason to hate the local assclown. In their adverts they claim he "does his own forecasting" -- my understanding, maybe wrong, is that very few orgs have the kind of big iron necessary to pump out meterological predictions. True?
posted by docgonzo at 7:04 PM on September 22, 2005

NOAA has a summary of the Guidance Models, that explains--probably in more detail than you want--how these models work.
posted by monju_bosatsu at 7:14 PM on September 22, 2005

The only model missing from the NHC link above is the FSU Superensemble, which has been very accurate.

To quote an FSU research-pimping mag :

FSU's novel approach is to combine--with the aid of three in-lab supercomputers--every model the team can get its hands on and run them through a clever "de-biasing" set of algorithms that eliminate or suppress known quirks in the data. The result is an ensemble of models, or in the parlance, a "superensemble" that coalesces the best aspects of the sum of its parts and spits out as close to a bias-free forecast as exists.

Many news organizations have been claiming that they have their own hurricane models. I have no idea how accurate they are, how much they rely on other models for data, and how much they rely on human intervention to work. Until they release their methods for scrutiny, they should be doubted.

At one time, I had a few public FTP links for model data - provided by FSU, Penn State, and a few other top-shelp meteorological programs - if you felt like plotting each lat/long point by hand (or if you bought/wrote a program to plot them for you). I can't find it now but these sites still exist.
posted by suckerpunch at 8:19 PM on September 22, 2005

I guess I should take the time to say : good luck to those in the storm's path.
posted by suckerpunch at 8:21 PM on September 22, 2005

thestormtrack.com
posted by amberglow at 8:22 PM on September 22, 2005

anyone can do their own forcasting, but they're all going off basically one of two data points, the main one being put out by the Government.
posted by cell divide at 8:37 PM on September 22, 2005

I think it should be clear how FSU does its hurricane predictions. Did you not see the picture in the upper left on their FSU Superensemble explanation page? That's a freaking TARDIS control panel!
posted by jmccorm at 9:08 PM on September 22, 2005

Here is a quick and dirty plot I made last year comparing the actual path of hurricane Ivan with the daily predictions.
posted by euphorb at 9:16 PM on September 22, 2005

If I had to put money on one of the predicted paths, I'd go for the NGPI, CONU, GFDI trifecta. Stalls, then rampages to the west. But then, these paths are dated at this point.
posted by jmccorm at 9:30 PM on September 22, 2005

Please don't advertise "thestormtrack.com" -- they have pretty much hijacked the Stormtrack name (www.stormtrack.org); the real Stormtrack has been around since 1977. Also they are ripping maps off of Colorado State's tropical guidance area and posting them on personal Yale server space.
posted by zek at 9:36 PM on September 22, 2005

More live streams:

KTRK ABC 13
KPRC Channel 2 (direct link to above stream)
KHOU Channel 11 (not really working every so often)
TXCN (apparently not working)

I've been watching ABC 13 most of the day, despite the jitters. They are run by the Big Mouse itself and so can draw on all the ABC resources.

Also, WWL and WDSU, as well as WVUE, in New Orleans also have live streams.
posted by calwatch at 9:49 PM on September 22, 2005

You can find footage from traffic cameras at houstontranstar.org. Those will come in handy when and if the hurricane arrives. In the meantime, here's the furthest traffic camera on I-45, I-10, US 59 Southwest, US 290 Northwest, US 59 North.
posted by calwatch at 10:43 PM on September 22, 2005

The freeways look nice and empty this morning! Looks like everyone decided to leave town at the same time, with obvious results.
posted by rolypolyman at 5:47 AM on September 23, 2005

Please don't advertise "thestormtrack.com" -- they have pretty much hijacked the Stormtrack name (www.stormtrack.org); the real Stormtrack has been around since 1977.

Oh, i didn't know. I got it from Drudge, who jumps on every weather thing.
posted by amberglow at 5:55 AM on September 23, 2005

tut, amberglow! You're practically the anti-Drudge!
posted by mwhybark at 6:11 AM on September 23, 2005

Most of my family still live in Houston; my mom tried to get out with one of my brothers on Tuesday...they drove for eight hours, traveling a total of 30 miles. They turned back at that point; as they were traveling (in two cars) with two cats, a dog, my elderly mom and five kids. They plan to ride the storm out in my brother's office building, which has no windows (on his floor), and its own generator. My dad, meanwhile, plans to hole up at a friend's 2nd story apartment. Another brother who lives on Crystal Beach (near Galveston) evacuated to Beaumont...sort of like evacuating to the breakfast nook when your kitchen is on fire. I have no clue where he is now, or my godmother, who lives in Beaumont.

Cell phone service down there is nonexistent, and land-line service is almost as bad. My dad told me that Houston may turn off all power at some point this evening.

Yikes. I hope* Rita goes further east.

I've read several places that the Texas DOT had no full emergency/disaster evacuation plan for the Houston area. Odd, considering its value as a terror target, and the large number of chemical plants, and oh, the NUCLEAR REACTOR. My understanding is that the roads out of the are were not fully contra-flowed until it was too late for that to do any good. I'm glad that I moved to Chicago; but I'm scared for my family.

* were I not a very happy atheist, I might have used the word "pray"...
posted by weirdoactor at 10:45 AM on September 23, 2005

Um, isn't this more of an AskMe post than a MeFi post?
posted by ChasFile at 10:56 AM on September 23, 2005

docgonzo writes "my understanding, maybe wrong, is that very few orgs have the kind of big iron necessary to pump out meterological predictions. True?"

There's is also a bit of black art in forecasting that manifests itself in a general feel for the kind of weather your predicting and a gut level knowledge of the forecast areas. I've seen meteoroligists all using the same data give wildly different predicitions and I tend to trust the guy who has been living in the area for the longest time.
posted by Mitheral at 11:25 AM on September 23, 2005