Radar Induced Warnings??

[STexan]

Forgive my ignorance, but watching with RadarScope and AllisonHouse data at the storms in the Mobile area today. I see some tornado warning boxes. I look at base velocity, composite reflectivity, and others but nothing looks terribly "tornadic" or even having cyclonic rotation signatures. Granted, I may not have seen the image prior to or at the warning issuance.

I assume many of these warnings are radar induced. Does a human verification precede all tornado warnings where there is no spotter report to trigger the warning or is there certain digital signatures that automatically trigger a "radar tornado warning"?

Does a "wall cloud" and/or "funnel cloud" report automatically get tagged with a warning and posted? Thanks

Robert Keck

 

[Greg Higgins]

There are many considerations regarding issuing a warning. What you typically see on your screen is usually only one of many "elevations". You would need to change heights and check whats also going on at each level. Again, this is only one of many considerations that the NWS goes through before issiung a warning.

 

[Jeff Duda]

As far as I'm aware, there is no software used by the NWS to automatically issue a tornado warning based off the detection of a certain type of radar signature (for example, a threshold value of azimuthal shear). However, I have seen some workstations at NWS offices that will sound an audible tone when a certain radar signature is detected. For example, I witnessed this once when volunteering and there were storms with > 50 dBZ reflectivity at rather high heights. So a sharp ping kept sounding an alarm saying something to the likes of, "HEY...LOOK AT ME...THERE'S PROBABLY HAIL IN THIS STORM...MAYBE YOU SHOULD WARN THIS". (It wasn't literally saying this, but that was the message it was conveying.) But that's as much as I've ever seen. Such alerts are there for the NWS meteorologists to get a heads up and issue a warning in case they aren't actively watching radar.

 

[Bob Hartig]

I'm not a met, but from what I understand about the issuance of tornado warnings, I would be very surprised to learn that any warning is automatically "triggered" by some kind of radar signature without any human involvement. I just can't imagine it. As Greg has indicated, there's more to issuing a tornado warning than just identifying couplets. For instance, how deep is a couplet? How sustained? How strong? What history does the storm have? What kind of environment is it in and what kind of environment is it moving into? You can bet that on days when big weather is happening, warning meteorologists are watching the radar closely and asking questions like these.

By "radar-induced," do you mean "Doppler-warned"? Such warnings get issued a lot, to the point where in many places I think they have a cry-wolf effect. Based on the presence of a couplet plus an environment favorable for tornado formation, WFOs issue such warnings despite the fact that no tornado has actually been sighted. A lot of times nothing happens, but it's better to have the public complain about false alarms than about having been warned too late or not at all of a tornado that did in fact occur.

With QLCSs, watch boxes seem a lot more vague and often not very meaningful. Most of the time, I read them as meaning that rotation got detected somewhere along a section of the line, and it could occur somewhere else, and then somewhere else--impossible to say where or give much lead time, but conditions make tornadoes possible possible even if not necessarily likely. So a big swath gets warned. A couple years ago, I saw a long, skinny warning box that stretched all the way from Kalamazoo to Grand Rapids. Not helpful--all it meant was that a spinup could conceivably occur at any time anywhere along that 50-mile stretch, and the mets had to cover their butts and the public's.

 

[rdale]

To be clear, no warning is ever issued automatically.

 

[STexan]

That's what I suspected but with the QLCS modes that were ongoing a real tornado threat seemed very very minimal at best, however wind gusts, perhaps exceeding 65 MPH would have been highly likely. Yet, there were several small tornado warn boxes issued within 30 about miles up and down a segment of the storm line.

Anyway, I was just curious what the protocol was insofar as doppler based TorWarn thresholds but I suspect it's a rather convoluted system and difficult to explain. For my part, I'd like to input that it would be nice if visual warn boxes could be identified (by clicking and seeing some sort of code) indicating these are spotter initiated (tornado, wall cloud, rotation), or that this box is a radar initiated "cautionary warning".

 

[MClarkson]

"I look at base velocity, composite reflectivity, and others but nothing looks terribly "tornadic" or even having cyclonic rotation signatures."

Are you looking at level 2 data, which is higher resolution?

And some people think (me included) that a few too many warnings are issued on the lower end, for some rather unimpressive echos. Basically the principal that you shouldnt cry wolf, save the tornado warnings for major threats. Does a slight chance of a weak tornado really need the highest level of warning?

 

[Jeff Duda]

I'm pretty sure the NWS operates on the principle that it is absolutely forbidden for a tornado to occur unwarned, even if that means issuing false alarms. POD = 1 is more important than FAR = 0. Granted, that's just the sense I get. If there are any NWS meteorologists on here who would like to opine, please do.

 

[STexan]

I'm wondering if the doppler tor warn standards and thresholds are the same from region to region? To me, it appears there are many more false warnings with the Southeastern systems than with the typical systems in the lower/central plains but I also understand many of the weather dynamics are often different, too. Maybe it's my imagination at work with that statement. I assume these warnings are issued by the regional NWS office?

 

[Bob Hartig]

Within a set of guidelines, every WFO makes its own determinations. I believe that even applies to what triggers a TVS marker in L3. That's the human element, and while it may make a strict, seamless standardization impractical, it's probably a good thing. Theoretically, each WFO understands best how the weather plays out in its area and needs some latitude to call its own shots. And during severe weather, I'd much rather have expert humans committed to protecting the communities in their watch area interacting with the rapidly changing conditions than trust the lives and property of my loved ones and myself to a few algorithms and micromanaged protocol.

It would be nice if visual warn boxes could be identified (by clicking and seeing some sort of code) indicating these are spotter initiated (tornado, wall cloud, rotation), or that this box is a radar initiated "cautionary warning".

Remember that these warnings are intended for the public, not for storm chasers.

 

[STexan]

it would be nice if visual warn boxes could be identified (by clicking and seeing some sort of code) indicating these are spotter initiated (tornado, wall cloud, rotation), or that this box is a radar initiated "cautionary warning".

Remember that these warnings are intended for the public, not for storm chasers.

Understood, but the radio warnings that go out specify if it's a visual spotter, or a doppler caused warn. But other personnel (LE, EMP, and spotters) would seem to be well served to have the information passed to them via computer visuals as they are often out of range of the appropriate noaa radio tower. (What led to the warn) A sheriff deputy may see a warn box that just went up and assume there is supposed to be a visible something or other in a given track if they are not privy to what led to the warn, and be looking diligently for something that was never there.

 

[rdale]

The Impact Based Warnings experiment (which I understand will cover all of the Central Region this year) does have an indication line regarding whether this is a radar-only or spotter-verified warning.

 

[STexan]

That's good and makes sense. The type of warn must be encoded in the message block somewhere, anyway. It's just a matter of the radar app developers retrieving the code and put in the popup message ... it would seem.

 

[rdale]

GrEarth and GR3v2 already do this.

 

[STexan]

Oh, Are those the ones who don't support mac? For mobile use, I've become spoiled on macbook. I would like to run those apps, but I can't understand why they won't support Mac OS and/or I don't know why Base Velocity can get their app up to par with the others.

Will GRLevel run on Windows 8 pads? Their website is kind of weak