Question about GR Level 3 and gate to gate shear

[George Tincher]

This may seem like a dumb question (and probably is), but which of the choices on GR Level 3 gives you the gate to gate shear numbers? For example, if you hold the mouse pointer over a specific portion of the storm, it will show shear at a variety of levels and types. Here's an example:



Ok, you can see the following features listed:

AVGDV: 76 kts
LLDV: 106 kts
MAXDV: 110 kts
MAXSHR: 58 /ks

Which of those numbers should one zero in on for low level rotation? And which, if either represents the gate-to-gate shear?

Granted, I could ignore all of that stuff, simply zoom in on the area using 0.5 tilt and look at the color chart and figure out the gate to gate that way. So it's really no problem. I'm just more interested in knowing what all of that stuff means. If you have all those tools, you should at least understand how to use them.

BTW, I am not yet a registered user of GR Level 3 but soon will be. So I've not had a chance to read any of the user instructions or anything like that. These features may be explained in detail there. I just had a friend share some of those images with me and I'm just curious about that feature of GR.

Thanks,

-George

 

[Greg Stumpf]

Originally posted by George Tincher
AVGDV: 76 kts
LLDV: 106 kts
MAXDV: 110 kts
MAXSHR: 58 /ks

Which of those numbers should one zero in on for low level rotation? And which, if either represents the gate-to-gate shear?

LLDV = Low-Level Delta-Velocity. This is the gate-to-gate velocity difference (GTG) at the base of the TVS detection (usually 0.5 degree tilt, but not necessarily).

MAXDV = Maximum Delta-Velocity. This is the max GTG for all elevation scans in the TVS.

AVGDV = this is new to me - was not in the NSSL version of the TVS Algorithm (where the WSR-88D version originated). I'm guessing that this is the average GTG across all elevation scans.

MAXSHR = Maximum Shear, the maximum at all elevation scans of the quantity GTG / diameter of the TVS. This value is pretty much useless, since it is very dependent on radar sampling limitations. TVS diameters, since they are defined as gate-to-gate, are always the distance across a 1 degree radial, and is smaller closer to the radar. Thus, shears are for TVSs with similar GTG always higher close to the radar.

Editorial - The TDA (Tornado Detection Algorithm) is a misnomer. It is not designed to "detect tornadoes". At NSSL, the algorithm was called the TVS Detection Algorithm, and bear in mind that not all TVSs are assocaited with tornadoes (only about 35-40% in the data set we had, assuming a LLDV > 25 or MXDV > 36, and depth > 1 km). Folks using inverted red traingles to indicate the locations are doing a disservice to the NSSL folks who originally developed this algorithm (yours truly was one of the scientists on the team) since it looks like a tornado when it isn't necessarily a tornado.

As for Tornado Warning Guidance in using the TDA to diagnose tornadoes - bear in mind that NWS warning forecasters don't use algorithms to make automatic decisions. They are only used as safety nets, and the mets intergrate a lot more information about the storm based on radar data, NSE data, etc. See:

http://www.wdtb.noaa.gov/resources/PAPERS/...wg02/index.html

 

[George Tincher]

Thanks Greg. That tells me which set of numbers to pay the most attention to. A very nice, detailed reply.

And yeah, I am aware that these TVS signatures aren't the most reliable sometimes. But didn't they use to be far less sensitive back around 1997 or 1998, then for some reason the algorithm was changed? I hear in the old days, the TVS was a pretty reliable indicator that something likely was taking place.

 

[Walker Ashley]

George:

Sorry for the thread jack, but how did you create the 4-panel display in grlevel 3? I see no option for this. PM me if need be.

Thanks ...
Walker

 

[Greg Stumpf]

Originally posted by George Tincher
And yeah, I am aware that these TVS signatures aren't the most reliable sometimes. But didn't they use to be far less sensitive back around 1997 or 1998, then for some reason the algorithm was changed? I hear in the old days, the TVS was a pretty reliable indicator that something likely was taking place.

The former (pre-1998) TVS algorithm wasn't really a separate algorithm. It was part of the form WSR-88D Meso algorithm, and basically "declared a TVS" when a mesocyclone reached a certain intense strength. I can't recall what values and thresholds turned the meso into a TVS.

The reason it seemed "more reliable" was because the TVS detections were very rare, and only triggered when the mesocyclones were extremely strong, which usually meant the meso was tornadic. Using the presence of the TVS to diagnose a tornado (not *detect*, but *diagnose*), the false alarm rate (FAR) was very low. Almost every TVS detection was associated with a tornado. But correspondingly, the probability of detection (POD) was also very low (<5%) - LOTS of missed tornadoes. Low FAR combined with low POD gives low overall success rate (measured by Critical Success Index, or CSI).

The replacement TDA was designed to address two things. 1) better CSI for tornado diagnosis. Since CSI is a combination of POD and FAR, better CSI does not neccessarily mean both better POD and better FAR. In fact, the POD went way up, but correspondingly, so did the FAR, and 2) separate TVS detection from Meso detection, since you can have TVSs without mesos, as in the case of non-supercell tornadoes (e.g., "landspouts"). The new TDA is a more realistic algorithm if the users are aware of that one important note - only about 35%-40% of all TVSs in Doppler radar data are associated with tornadoes.

 

[Tyler Allison]

Originally posted by Walker Ashley
George:

Sorry for the thread jack, but how did you create the 4-panel display in grlevel 3? I see no option for this. PM me if need be.

Thanks ...
Walker

It's one of Mike's (guy who writes GRLevel2/3) betas. You have to be an owner of GRLevel2 or GRLevel3 to get on his customer board...he posts beta versions with various features to that board.

 

[Patrick Ware]

Where can I find a detailed explanation for every feature on GR Level 3 software. I have looked at the website and use the software (not yet purchased but going to) and did not find any type of FAQ or manual.
Thanks,

 

[Tyler Allison]

Dont know if there's a feature list.

You can click on 'help' and then click on various parts of the image that are given.

Also you may want to post these questions to the GRLevel3 forum:

http://grlevelx.com/forums/index.php