Comparison of NUCAPS and Observed Sounding at Corpus Christi

NUCAPS 19z soundings show a similar temperature and moisture profile as the special 18z observed sounding at Corpus Christi but there is some lack in detail in NUCAPS. For example, an elevated mixed layer just above the boundary layer is evident in the Corpus Cristi sounding but is not shown in the NUCAPS. Higher in the atmosphere there is a pronounced dry layer in the observed sounding that the NUCAPS captures fairly well. This could be important for synoptic patterns where dry air aloft advects into the region and the temporal and spatial resolutions of observed soundings are insufficient to show this. NUCAPS can supplement these observed soundings and provide a clearer picture. We noticed the mid level lapse rates are steeper in the observed soundings as compared to NUCAPS though with about 1C difference between the two. This has implications for analysis of how much instability is present and how the CAPE is distributed.

EDIT: Added images below. The first two are the nearest high quality NUCAPS soundings near Corpus. As mentioned above, the temperature discontinuity between the observed and NUCAPS sounding was operationally significant (~1.5C) at 500mb. However, the second NUCAPS sounding that we observed showed the 500mb temperature to be nearly the same as the observed sounding. We learned that these differences across small distances are typical near the coast due to land/sea surface emissivity and temperature differences.

-dryadiabat

2110z CRP Update:

2110z Update:

Interesting to see the difference between the observed 18z KCRP sounding and a 21z NUCAPS sounding south of the stronger convection.  The NUCAPS sounding implies a weakening of the earlier cap around 850 mb.  Is this weakening real, or is cloud contamination playing a role?  This detail would be key in determining plausibility of convection to become more surface-based.

                                                        18z CRP sounding

                                                        21z NUCAPS

Lightning Jump Algorithem

Is there a way to display the lightning jump algorithem with more transparancy?  I like to look at it however I do not like how it masks the relfectivity underneath.  Just a thought...

torman1

Moving to Brownsville

A quick update, one of our forecast groups is moving to the Brownsville WFO in the anticipation of convective development and potential severe weather.

-Sieglaff

As of 2015z

As of 2015z:

Also, special 18z KCRP sounding quite instructive.  Strong cap in place near 850 mb, with most of the steep lapse rates and deeper shear in the mid-levels above this feature.  This substantiates our earlier notion that tornadic/wind producing storms are less likely, with hail the main threat.  Stronger/potential severe storms in the NW portion of the forecast area at this time are seemingly not surface-based.

Are the weaker lapse-rates/shear in the surface-850 mb layer biasing CI-severe downward? ;) 

Updated Mesoanalysis

18z special sounding from CRP has arrived with well-pronounced capping noted near the 850-hPa level.  Radar and satellite trends continue to show severe convection north of the now stalled outflow induced frontal boundary.  Considering the 18z sounding, along with observed radar and satellite trends, all severe convection so far this afternoon appears to be non-surface based convection.  As a result, svr hail remains the main concern.

torman1

Prob Severe in Complex Convection

This may be a rarely small percentage of cases, and probably more common in high CAPE / low shear environments where multicell storm structures dominate. But, in the case of the cells in western Webb County we have noticed two distinct cores on radar. Prob Severe tracks these distinctly to start and gives useful values for the southernmost storm, but then merges the two later before splitting again. This leads to inconsistency in the lightning flash rate values. When the two cells are merged, lightning flash rates are combined and much higher. When they split, the lightning flash rates are separated and lower. Obviously there are limitations in how the algorithm can identify separate cells in these complex environments, but it is something the user should be aware of.

-dryadiabat

20z update

As of 20z:

Additional severe thunderstorm warning issued for flanking line developing SW of initial storm in Northern Webb county, TX.  Large hail is again quite probable, with deep, tall core through the mid-levels.

Interestingly, severe CI has seemingly remained fairly low with this storm, while ProbSevere has elevated values.

Suggestion for CI probability:  Have the algorithm mark a particular cloud element as "already convected" (or some other wording to that effect), in order to let the user know this particular cell/cloud element is no longer being sampled by the algorithm.  - J-Money 

20Z Radar Update...

Initial cell as weakened below severe limits with a secondary storm strengthening directly to its south.  Overall thoughts on PROBSvr remain  unchanged as it seems to lag behind observed radar trends.  Traditional radar interrogation methods (50 dBZ core above the -20C line) continue to provide greater lead-time for warning decisions.

Convective initiation algorithem actually did hint at general region of southern TX for new storm development over 1-hr in advance of new storms forming.  Once storms form, is there something that could be displayed to show that the CI algorithem is no longer tracking the area?

torman1 

Storm moving into KCRP WFO

Both forecaster groups are setup in the Corpus Christie, TX WFO.  Thus far the only storm of interest was over Webb County near the Rio Grande river.  The forecaster groups found ProbSevere to confirm what they had observed in their all-tilts radar analysis--a strengthening storm in the 1850 to 1900 UTC time frame (Figure 1).  The ProbSevere information helped increase confidence, but did not much provide lead-time since the forecasters said, "...we were already on this storm, examining all-tilts".  They speculated ProbSevere output would be even more useful in a region with multiple storms that require interrogation and a warn/no-warn decision.

Figure 1.  NOAA/CIMSS ProbSevere and MRMS composite reflectivity valid 1846 - 1900 UTC 18 April 2016.

As of this writing no reports have been received from this storm.  We've had some conversations related to the difficulty in validating storms in very rural regions.

-Sieglaff