Dual Polarization Radar - Coming Soon!

[Kevin]

Things are beginning to move very fast in the dual-pol upgrade. Morristown now complete....well ahead of schedule...making TN completely covered in the upgrade process (first state outside the Pacific NW).

Many other sites are now moving up their upgrades because how smoothly things are going. Locally...Birmingham (BMX) is coming up fast now with the upgrade now coming next week.

[Nashville_wx]

Kevin to view the dual pol radar, do you have to use Grlevel2? Then a add on? What is the best way to do so, I normally use GRlevel3.

[Eric]

Kevin to view the dual pol radar, do you have to use Grlevel2? Then a add on? What is the best way to do so, I normally use GRlevel3.

Dual-Pol isn't yet available for GR3...only GR2A and Radarscope.  Last I heard the update for GR3 will be available in March for $40.

[Thundersnow]

Danel Lamb has a nice writeup on TW on dual pol and use cases of it during Friday's outbreak, specifically in north Alabama:
http://www.talkweather.com/forums/index.php?/topic/57908-dual-pol-is-awesome/

I still need to assimilate all this information to gain a better understanding.  On Friday, I saw some CC anomalies that I thought might be TDSes.  In some cases they were.  But, in other cases, storm surveys haven't confirmed tornadoes in those areas.  It's apparently key to be able colocate the CC anomalies with high reflectivity and velocity couplets before you can say it's a TDS, otherwise, some false readings are possible.

An example... I snapped the following CC1 image from RadarScope on my IPhone of a CC anomaly over southern Dickson County, which would spawn a confirmed EF1 tornado just a few miles downstream in southern Cheatham County.

http://tennesseewx.com/index.php?action=dlattach;topic=3176.0;attach=3190;image

At this time, there were reports of a tornado on the ground in Dickson County.

When I heard that report, I was then pretty sure that this was indeed a TDS I was looking at.

Well, NWS storm surveys have currently assessed out of all the damage reports in the path of that cell that tracked from Waverly to Nashville, only one 1-mile tornado track has been confirmed in Cheatham County near Kingston Springs.  The rest of the wind damage in Dickson and Humphreys Counties has been assessed as straight line wind damage, even with the above mentioned CC anomaly.  And, that wind damage includes such things as substantial roof damage to buildings in the Dickson area.

My conclusion is either that the CC anomaly wasn't a TDS as I thought it was at the time, or the storm surveys didn't correctly conclude a long tornado path well before the confirmed path in Cheatham County (though, we're obviously not talking about anything over an EF1 rating on any of it anyway).  I'm going with the assumption that the storm surveys resulted appropriate conclusions, and the blue/green "spot" within the purple CC activity wasn't really debris picked up by a tornado, unless someone is convinced otherwise.  Could it have been large hail that was tripping those colors?  There certainly was large hail involved in that storm.  I just saw my dad's dented up car yesterday after it was subjected to the golfball to baseball sized hail in the storm in Pegram.  The siding on many, many homes in that area is shattered, along with some windows, including at my sister's house, also in Pegram.

Of course, if the anomaly was a TDS, then there's some explaining to do about the storm survey... as it potentially raises some questions about assessed straight line wind damage possibly deserving a designation as tornado damage.

Just some thoughts... I would appreciate any input on this from Kevin or anyone else who can weigh in on analysis of the new dual pol capabilities of radar as well as the accuracy of these storm surveys.

[Kevin]

That's a great post by Daniel. Very well explains how to differentiate between TDS and non-TDS when you may be expecting one.

He's right...there should be some reflectivity co-located with a velocity couplet AND low CC. It doesn't have to be the 60dbz debris ball needed in the past but something should be there. If you have low CC in an inflow region/hook within the "dry" part of it...assume its NOT a TDS as most likely it's not. Those "low signal" areas are always unreliable and will probably have low CC anyway...especially if there's things being picked up as a result of inflow into the storm which is an interesting research topic for folks to look into.

I've seen several examples already where its easy to think its a TDS but its not. Low CC is going to happen in a lot of situations in severe storms...those "low signal" areas...hail and if you're above the melting layer (freezing level) TDS is probably not going to be noticeable at all unless its a very strong signature with debris being lofted up many thousands of feet...which will probably be only in your situations when you have the infamous reflectivity "debris ball" already.

Thundersnow...your image above...that's NOT a TDS. Even without reflectivity/velocity...I can tell that because of the "texture" of the CC. That's due to low signal areas and non-uniform beam filling which is an always present radar limitation. Don't worry...that will fool people...it even appeared to fool a few WFOs Friday. That's why...again...make sure you have some decent reflectivity signal co-located with the CC/Velocity. IF you have that...confidence is very high that a TDS is present and thus a tornado in progress. If not...that not a 100% guarantee there's no tornado...but you should assume not.

Look at Daniel's images...most TDSs are relatively small and rounded (or square if the data is not smoothed) because that's the extent of the lofted debris. If the low CC is spread out across many square miles...that's not consistent with typical debris in an active tornado. One thing I did notice in the MO storms early Wednesday...after the TORs dissipated often then the debris spread out as it was gradually returning to ground-level...but also CCs were rising...so you weren't getting those deep blues anymore seen in the image above.

ETA...Here's two TDS images I've captured this past week...just to add some confidence perhaps...

[Eric]

That's a great post by Daniel. Very well explains how to differentiate between TDS and non-TDS when you may be expecting one.

He's right...there should be some reflectivity co-located with a velocity couplet AND low CC. It doesn't have to be the 60dbz debris ball needed in the past but something should be there. If you have low CC in an inflow region/hook within the "dry" part of it...assume its NOT a TDS as most likely it's not. Those "low signal" areas are always unreliable and will probably have low CC anyway...especially if there's things being picked up as a result of inflow into the storm which is an interesting research topic for folks to look into.

I've seen several examples already where its easy to think its a TDS but its not. Low CC is going to happen in a lot of situations in severe storms...those "low signal" areas...hail and if you're above the melting layer (freezing level) TDS is probably not going to be noticeable at all unless its a very strong signature with debris being lofted up many thousands of feet...which will probably be only in your situations when you have the infamous reflectivity "debris ball" already.

Thundersnow...your image above...that's NOT a TDS. Even without reflectivity/velocity...I can tell that because of the "texture" of the CC. That's due to low signal areas and non-uniform beam filling which is an always present radar limitation. Don't worry...that will fool people...it even appeared to fool a few WFOs Friday. That's why...again...make sure you have some decent reflectivity signal co-located with the CC/Velocity. IF you have that...confidence is very high that a TDS is present and thus a tornado in progress. If not...that not a 100% guarantee there's no tornado...but you should assume not.

Look at Daniel's images...most TDSs are relatively small and rounded (or square if the data is not smoothed) because that's the extent of the lofted debris. If the low CC is spread out across many square miles...that's not consistent with typical debris in an active tornado. One thing I did notice in the MO storms early Wednesday...after the TORs dissipated often then the debris spread out as it was gradually returning to ground-level...but also CCs were rising...so you weren't getting those deep blues anymore seen in the image above.

You say a TDS could be misidentified due to inflow...could the same be said with the RFD on the backside?

[dwagner88]

Is Chattanooga too far from htx for cc to be useful? We had a pretty strong tornado producing a lot of damage here, and I never saw anything on cc. I was only using radarscope, so that might have been part of the problem.

[Thundersnow]

Thanks for the input!  Now, that I look at the image captured, I do see sort of a junky "cone" projecting out in the radar beam.  Unfortunately, it goes right through the most dangerous part of the cell in that case.

This will obviously take a some experience and training to accurately interpret the subtleties of these radar modes.

[Kevin]

You say a TDS could be misidentified due to inflow...could the same be said with the RFD on the backside?

Probably not...because RFD usually is associated with very heavy precipitation either as it crosses or shortly thereafter...so the reflectivity signal should be strong and that makes CC more reliable. If the outflow starts racing ahead of the parent cell...and you're in a real gust front/OFB situation...CCs will lower dramatically then most likely but of course that's not a scenario where a tornado would develop.

[Eric]

Is Chattanooga too far from htx for cc to be useful? We had a pretty strong tornado producing a lot of damage here, and I never saw anything on cc. I was only using radarscope, so that might have been part of the problem.

IIRC, TDSs can only be picked up within 40nm of the radar site.  Anything over that, the return gets funky.

[Kevin]

Is Chattanooga too far from htx for cc to be useful? We had a pretty strong tornado producing a lot of damage here, and I never saw anything on cc. I was only using radarscope, so that might have been part of the problem.

I saw a definite TDS in that first morning tornado once it moved a bit NE of CHA. I think the issue there was it needed to be a little stronger/persistent to be picked up because of the distance at first because it took a while to become apparent. I'll see if I can get the image of that one and post it.

[Kevin]

Thanks for the input!  Now, that I look at the image captured, I do see sort of a junky "cone" projecting out in the radar beam.  Unfortunately, it goes right through the most dangerous part of the cell in that case.

This will obviously take a some experience and training to accurately interpret the subtleties of these radar modes.

My personal guidelines as of this moment...though this could change as we see new case events...

If you have ALL three co-located:
Reflectivity above 35dbz
Velocity couplet exceeding 60kts
CC below 0.85

Confidence is 75% or greater that a TDS has been identified.

Then...If you have ALL three co-located:
Reflectivity above 50dbz
Velocity couplet exceeding 85kts
CC below 0.80

Confidence is 100% that a TDS has been identified.

[Thundersnow]

My personal guidelines as of this moment...though this could change as we see new case events...

If you have ALL three co-located:
Reflectivity above 35dbz
Velocity couplet exceeding 60kts
CC below 0.85

Confidence is 75% or greater that a TDS has been identified.

Then...If you have ALL three co-located:
Reflectivity above 50dbz
Velocity couplet exceeding 85kts
CC below 0.80

Confidence is 100% that a TDS has been identified.

Great tips.  Thanks.

[Kevin]

Here's the TDS in the CHA area...I was watching for it but didn't appear until the 12:48p EST scan. This was at about 4,000ft and a distance of 50nm from HTX and continued for a few more scans up to at least 6,000ft and a distance of 60nm as the tornado moved NE.

The TDS in this one is in category two...so confidence was 100% a tornado was in progress...


Looking at the survey from MRX...looks like the tornado first touched down near Harrison...so it appears HTX picked it up pretty quickly...even at that distance. Of course...this was a strong tornado...I doubt it would be visible if it were just an EF-1.

[Bigm33]

Looking at the survey from MRX...looks like the tornado first touched down near Harrison...so it appears HTX picked it up pretty quickly...even at that distance. Of course...this was a strong tornado...I doubt it would be visible if it were just an EF-1.

The fact that it wouldn't be visible if it was an EF-1 or weaker, is that due to the distance from the radar site itself?