GOES-W & GOES-E GLM comparison

Posted in Live Blogs, News on by Michael Bowlan.

 

Looking at GLM for GOES-W in Full Disk (L)  and GOES-E in meso sector (R) the flash extent density and average flash area compare well to one another. Max FED values were 62 and 64 fl/5-min respectively. They also showed similar overall footprint with even the smallest flash values. Parallax is obvious with the peak  values separated by ~15 mi at this longitude.

GLM FED from GOES-W full disk (UL), GLM FED from GOES-E meso sector (UR), 10.3 um from GOES-W full disk (LL), 10.3 um from GOES-E meso sector (LR)

— S Coulomb

CPTI in Ranges for Analysis

Posted in HWT, Live Blogs, News on by Tiffany.

CPTI product shows very little change in the probabilities from one level to the next, therefore i recommend doing ranges with this analysis.  This should highlight the threshold ranges of interest in a better way.

For example:  Cat 1 (80-110mph), Cat 2 (125-155mph), Cat 3 (170+ mph)

-Jake Johnson

 

Cell merger – AzShear, CPTI and DMD comparison

Posted in Live Blogs, News on by Brandon Smith.

Gonna take a look at SC merger  noting two small SC storms WNW of KINX radar @ 2058UTC. Prob TOR with these storms ~ 25% at this time with generally weak AzShear with stronger northern storm.

2106 UTC – just before storm merger see uptick in Azshear (0-2 km) with slight increase in CPTI products (50-52 %).  DMD output not providing any useful information on these storms at this time probably.  ProbTor ~ 10%

After merger @ 2114 UTC although not much change in AzShear, CPTI probabilities do increase some to 55 to 60%. Note that only the XDMD mesocylone algorithm displays a mesocyclone icon at this time while legacy DMD products do not.

2126 UTC – ProbTor rapidly increased to 77% with marked increase in AzShear. CPTI products also near 70% (110 mph).  Now all three DMD products display icon by this time.  BUT the XDMD seemed to identify mesocyclone sooner than legacy products.  This was encouraging.

 

2142 UTC – looks like storm looses organization as inflow may have been disrupted. AZshear becomes elongated with CPTI values falling notably.   Encouraging to note the character of AzShear (became elongated ) was consistent with the less organized storm (velocity structure).  ProbTor falls to ~ 65%.  Note the xDMD algorithm did still correctly identify shear couplet (SRM not shown) where legacy algorithm fails.   XDMD performed much better and and had greater temporal continuity than legacy product in this example. In this more complex storm interaction example appeared XDMD was able to better identify and retain areas of shear/rotation compared to legacy.

-Quik TWIP

 

 

 

 

 

 

NMDA

Posted in News on by Brandon Smith.

It can be a bit hard to see from the color scale, but take a look at the bottom left screen. The NMDA is shown there. One thing that I have noticed about it is it was able to better track the Mesocylone compared with the normal one. However, I have not gotten to test this extensively yet.

Notice in the time-step below that the NMDA is the able to hold onto the mesocyclone when the normal one cannot.

South Beach

CPTI and Azshear

Posted in AzShear, Live Blogs on by Brandon Smith.

Tracking this northern most SC @ 1926 UTC prior to tornadogenesis.  Azshear 0-2 km product did a nice job highlighting shear with developing storm very early on in storm evolution and was very persistent in highlighting area of shear throughout.  Temporal continuity was very good.  Lower two panels are 80 and 110 mph intensity probabilities.  Used this 4-panel to monitor evolution.  Also displayed is legacy mesocyclone (upper-right), DMD (lower right) and XDMD l(lower left).  This SC also exhibited well-organized ZDR arc and KDP foot (not shown) that whose centroids were normal to storm motion suggesting MC intensification was likely.

1938 utc Azshear still showing strong signal with CPTI gradually increasing to near 60%.  ProbTor increased to 78% as well.  DMD, XDMD and MD not showing much and did not seem to be very helpful.

 

CPTI was insightful with increasing probabilities of tornado intensity probabilities at higher speeds (110 mph lower right).   Note that XDMD was also indicating MXRV of 26.

 

1448 UTC – would occasionally see dual maxes in AZShear and CPTI products that was a bit of an issue at times.

CPTI 110 probablities up to ~ 75 % at this time (lower right)

Overall…0-2 km Azshear was quite helpful in quickly grabbing attention to areas of concern.  0-2 km product was temporally very consistent.  Its location relative to developing Mesocyclone was typically off…but still very useful.    XDMD, did seem to perform better than legacy DMD products but not much at least with this storm.   Early in storm evolution, examining ZDR/KDP centroids relative to storm motion was also helpful as a reality check on MC intensification and increasing SRH with this particular storm.  Bottom line was generally pleased wiith  AzShear and CPTI performance for this storm. – Quik TWIP

ProbSVR

Posted in ProbSevere on by Tiffany.

ProbSVR time series graphs with the online version continues to be one of my favorite for tracking trends.  Uncertain how to display this cleanly in AWIPS II?   – Jake Johnson

Low-biased MESH

Posted in Live Blogs on by Tiffany.

At 336pm CDT, a supercell storm near Electra, TX produced golf ball sized hail despite MRMS MESH showing sub-severe hail (~0.85″). Despite the low MESH, ProbSevere hail was progged at 75%. KFDR data had 50+ dBZ above 34K feet MSL and the storm was highly tilted downshear (a known limitation of MESH – underestimates in this case). This storm highlights the need to couple MESH/ProbSevere products with a thorough analysis of the base radar data and a fundamental understanding of strengths and limitations of supplementary datasets.

–Stanley Cupp