Details on severe weather to be more exact

[Eric Friedebach]

The Wichita Eagle, Stan Finger, 3-14-05

Severe weather warnings will be more precise this spring, thanks to new technology used by the National Weather Service. The software will allow the agency to issue warnings specifically tailored for precise areas and to avoid issuing warnings for areas that will not be threatened.

For example, it will now be possible to issue a severe thunderstorm warning for the northern half of a major thunderstorm and a tornado warning for the southern half, said Chance Hayes, warning coordination meteorologist for the agency's Wichita office.

The change should lead to a stronger sense of urgency for those within a warning area and fewer false alarms for residents who wouldn't be affected by the severe weather, officials say.

"It will place the real emphasis on storms that really deserve the attention," said Dave Freeman, chief meteorologist for KSN-TV in Wichita. He said he plans to use the new technology to improve how his station responds to severe weather events.

"When we use the word 'warning,' we should be talking about something that requires people to do something," Freeman said. "There should be a call to action implicit in that."

While tornado sirens will still be sounded countywide, warning information issued on televisions and weather radios will typically cover much smaller areas.

Mike Smith, the founder and chief executive officer of WeatherData Inc., a private forecasting service based in Wichita, calls the new technology "revolutionary," and predicted that the credibility of warnings "is going to go way up."

http://www.kansas.com/mld/kansas/11130471.htm

 

[Mike Umscheid]

Along with the 1" hail initiative lower-end threshold for SVR criteria, this is another warning operations test this spring: the polygon warning verification. The DDC office, as well as others, will be part of this test this season. The NWS has already been doing polygon warnings, and many software packages are already displaying the lat/lon data, graphically, that is appended at the end of each SVR/TOR/FFW.

The main difference users will notice will be the possibility of multiple warnings in effect for one county... maybe a SVR polygon for northern X county, and a TOR for southwestern X county valid the same time or overlapping times. The way verification works now, one report of SVR in any part of the county would verify a warning even if the polygon only included, for instance, the extreme northwestern portion of a county.

The main motivation in this is to reduce false alarm area. Right now, the way the NWS does warning and verification, even if the polygon includes just a small portion of the county, the entire county is essentially still warned for, but the threat area for a storm may be just 1/6 of the county. Thus, you have 5/6 of the county with a warned, but unverified event: your false alarm area statistic.

Doing warnings and verification this way is far better, especially in those areas where political boundaries are extremely peculiar.

I'm not exactly sure how many offices are testing the polygon warning verification, though.

Mike U

 

[Joe Nield]

I'm not exactly sure how many offices are testing the polygon warning verification, though.

I believe the total is 16. We're on board here in Indianapolis, as are a couple of our surrounding offices. I agree that it's really not a huge change from a warning standpoint, but hopefully we'll see some good results on the verification side.

On an off-topic note, when is your avatar from?! Those are INSANE three body scatter spikes.

 

[rdale]

The problem with the polygons is that all offices use them but not all are ready to use them... So the polygons don't always match up with the text of the warning (sometimes horribly off, I plotted a warning down south that was supposed to be for the eastern half of a county, and the polygon was for the western half as well as the next county to the west - all behind the storm!) So for users wanting to display the polygon data, use at your own risk unless you know the office is "operationally" issuing them.

VTEC seems to have had a BAD start too, as it appears some offices have not yet trained their staff on SVS's and it shows...

- Rob

 

[Mike Umscheid]

On an off-topic note, when is your avatar from?! Those are INSANE three body scatter spikes.

Joe, just a typical svr wx day in western kansas ;-)

Actually, this is the first time I had ever seen the TBSS signature on three separate storms in one volume scan... especially lined up side-by-side-by-side like that. I was working the radar that day, it was sometime in June last year. Run of the mill supercell thunderstorms that remained non-tornadic, but biggie hail, obviously. I think the largest hail reported was baseball size from the western most storm.

Mike U

 

[Mike Umscheid]

VTEC seems to have had a BAD start too, as it appears some offices have not yet trained their staff on SVS's and it shows...

- Rob

The only training that operational forecasters will receive with VTEC is familiarization, and really nothing else. Nothing has changed in the way NWS offices produce SVSs with WarnGen. In other words, the VTEC switch was turned on in February and is transparent to all us radar operators. Radar operators are not allowed to alter VTEC code, i.e. change begin/expiration time manually, etc. We'll get a slap on the wrist for that.

So, what you are saying then, is nothing really related to VTEC at all, you are complaining about how a particular office issues SVSs I think?

Mike U

 

[rdale]

No, Chicago had a two-county SVR. They didn't issue a segmented SVS when the cancelled one county and continued the second, so VTEC coding indicated both warnings were still in effect. Then after the warning expired, they issued a SVS indicating the warning expired, but modified the VTEC code to say the warning was cancelled. And the VTEC code was for only one of the counties. The other county "VTEChnically" still was in effect.

If forecasters do not know how to issue segmented products, and override the VTEC code on their own, it's not going to work well in operation.

- Rob