Another severe storm in the Midland WFO

In a story of close calls within the HWT today, another severe thunderstorm developed in the Midland WFO, just one county away from in the San Angelo TX WFO team.  The NOAA/CIMSS ProbSevere model was able to diagnose strong vertical cloud growth rates and moderate glaciation rates--which able to provide good lead-time for this storm (Figure 1).  The first ProbSevere value over 50% occurred at 2140 UTC and over 80% at 2142 UTC--providing 5 and 3 minutes of lead-time ahead of the first severe thunderstorm warning, respectively.

Today the GOES imager was operating in routine scan mode--or 15 minute updates, in 2016 GOES-R ABI will be launched and the 5 minute scanning over CONUS will help increase lead-time in rapidly developing storms like these.  Additionally of interest, these storms are producing very little CG lightning strikes, but the total lightning products showed a considerable increase in total lightning (not shown) during the times the ProbSevere probabilities were also rapidly increasing.

Figure 1.  NOAA/CIMSS ProbSevere and MRMS composite reflectivity valid 2134 UTC - 2150 UTC 05 May 2015.

-Sieglaff

Intensifying Storm to the NW of SJT Area

While monitoring the storms just to the northwest of the SJT area, noticed a limitation of the ProbSevere tool, largely due to the processing time. At 2139Z on the KSJT radar, a TBSS was noted on several levels of the radar. With the time lag on the ProbSevere, it was still only outputting a 26% severe probability. But looking at the details below, you can see that the satellite growth around the same time was indicating both moderate and strong and leading to increased confidence in the strengthening storm.

Jumping ahead 2 minutes to 2141Z, the ProbSevere has jumped to 58% with the radar data being processed.

Now moving ahead to the next scan at 2144Z, the TBSS has moved closer to the surface and the ProbSevere has quickly jumped up to 94%.

While this jump has increased confidence, I felt that looking closely at the radar data I would have issued the warning (if it was in our area) before the ProbSevere had made the jump to 94% (would have probably been issued at the same time as the jump).

Finally, looking at the Lightning Cell Time Series data for that cell, a significant jump in the Total Lightning and Cell Area was seen around 2140Z, which matches the significant increase in the radar reflectivity data.

While our group is unable to use the lightning jump algorithm today, I find my self both wishing I was able to use that data but also glad it was able to utilize the time series data to pick up on the significant jump manually. -SRF

Convective Initiation 30+ min lead time

The Convective initiation product showed an area of 51% probability in the south central part of Parmer county in the far northwestern part of the LBB CWA at 2015 UTC after showing some 10% probabilities at 2000 UTC.  By 2030 UTC the convective initiation probability increased to over 60% and also identified another area with a 74% probability.

Image loop of Convective Initiation

Looking back at the radar data in the loop below the first 10+ dbz reflectivity values appeared at 2037 UTC.

Radar image at 2031 UTC

The first 35 dbz value occurred at 2048 UTC in the west central part of Parmer county.

I think the CI product successfully identified this initiation with over 30 minutes of lead time again highlighting it’s situational awareness value.

Jack Bauer

Moved from ABQ to SJT

With scattered thunderstorms and limited severe threat in the ABQ area, we have been moved to SJT to hopefully capture convection moving northeast into the area from the dry line and leftover outflow from last nights convection.

Taking a look back at the visible satellite and convective initiation product, it highlighted the development of convection to the southwest of KBPG very well at 2000Z, with an area of 91% prob of development (and 67% to the south-southwest of that area).

Moving forward 30 minutes, you can see that area developed experienced development in several locations along that line.

-SRF

NUCAPS sounding near strong storm south of LBB

I didn’t get a chance to look at the NUCAPS soundings on Day 1 but as soon as the pass occurred today we looked at a sounding that was about 60 miles east of the storm of interest that we have been following in the northern portion of the MAF CWA.  The NUCAPS availability point was located in a brief area of clearing within the cumulus field but we were a little uncertain whether or not clouds interfered or not.
















NUCAPS availability plot. The point we used is in the center of the image.

However, the data did look reasonable as the temp was close to surrounding obs, though the dewpoint appeared to be low compared to surrounding observations.  The SBCAPE in the NUCAPS sounding was about 245 J/kg which was significantly lower than the GOES-R LAP CAPE observation which was around 1000 J/kg.

















Unedited NUCAPS sounding

We modified the sounding to increase the dewpoint to 59 degrees to be better in line and this resulted in SBCAPE of just under 1000 J/kg and lined up well with the GOES-R CAPE.


















Edited NUCAPS sounding to increase the dewpoint.
















GOES-R CAPE image

Jack Bauer

Prob Severe and Lightning…developing storm in Midland CWA

An area of convection (70% probability in GOES CI) continued to develop around 1900 UTC.  Between 1922 and 1928 UTC, the Prob Severe increased from 18% to 63%, and by 1935 was up to 86%.  The lightning jump increased to 2 sigma during this time.  The 1 minutes lightning density increased over a 2 minute time span from 1 to about 4 flashes/gridbox.   By 1945, the lightning density was up to 9 flashes/gridbox.  All of this evidence points to convection which is increasing in intensity.

Evidence from the NUCAPS sounding as well as the GOES LAP CAPE suggest the storm is in an environment with around 1000 J/kg of CAPE.  By 20:06, the storm is a well-organized supercell with a hook echo and well defined rotational velocity.

The combination of lightning data, Severe Prob, NUCAPS and GOES LAP CAPE were good indicators of strengthening storm worth watching.  The storm produces a tornado.

NUCAPS compared to 18z AMA sounding

Pass came across with NUCAPS available at 20z. Looked at 19z data from a point close to AMA which launched an 18z sounding. First image is NUCAPS, Second image is the 18z AMA soundings. Differences notices were cooler temperatures above 300mb on the NUCAPS sounding. There was a small but more distinct cap near 800 mb on the AMA sounding. The freezing level and -20 level had dropped on both soundings compared to the 12z AMA sounding.

Severe Thunderstorm in Midland WFO

An isolated thunderstorm began to develop during the afternoon hours northeast of Midland, TX, near our Lubbock WFO team.  The NOAA/CIMSS ProbSevere model has proved to be very useful on this storm.  The loop below demonstrates how the ProbSevere values increase as satellite vertical growth rate increases (1922 UTC), MESH increases (the entire loop), and satellite glaciation rates increase (1936 UTC).

The ProbSevere model first exceeded 50% at 1924 UTC and 80% at 1932 UTC.  The Midland WFO issued a tornado warning on this storm at 2013 UTC.  The lead-time offered by ProbSevere for this example is 49 minutes (using 50% threshold) and 41 minutes (using 80% threshold).  This storm produced at tornado at ~2032 UTC.

Figure 1.  NOAA/CIMSS ProbSevere and MRMS composite reflectivity valid 1920 - 1926 UTC 05 May 2015.

-Sieglaff

Conflicting Information

The Probability Severe and ENI Thunderstorm Alerts seem to have differing opinions on a couple cells in the ABQ CWA.

At 1932Z, the cell to the north has a Dangerous Thunderstorm box, yet a prob severe of only 17%. Meanwhile, the southernmost cell has a prob severe of 78% but is only considered a "significant" thunderstorm.

At 1940Z, the prob severe of the south cell went up to 82% with no DTA, and the cell the the north went up to 43%. It's interesting to note these cells are approaching the east/west boundary of the ENI Thunderstorm Alerts. No alerts showed up for the "east" algorithm either.



















The storms are still pulsing down just as quickly as they develop. They really only look impressive for 1-2 volume scans. No hail or wind reports yet, just the funnel cloud near Roswell earlier.

-V. Darkbloom

Convective Initiation temporal resolution concern

Looking just beyond our LBB CWA to the south I noted a CI change from at 1845 UTC which showed 10% probability of convective initiation to a 70% probability at 1900 UTC.

Animated loop

However, when looking back at the radar loop starting at 1845 UTC which matches the first satellite image and continuing past 1900 UTC, we can see the development of an isolated convective cell.

Animated radar loop

At 1900 UTC a 39 dbz core was noted, so in this case the temporal resolution of the CI product (15 minutes) effectively resulted in a miss as the 35+ dbz was already present at the time the 70% probability was given.  This does not diminish the situational awareness value of the product.

Jack Bauer