GOES 16 - Full Disclosure

Ok, last post....honesty time.
I didn't make this a secret in my application that I had been out of meteorology for about 4 years between 2014-2018. During that time, GOES-16 launched. As I returned to broadcasting, I didn't adequately spin-up on the new capabilities of this new generation of satellites. I was aware of the spatial and temporal resolution improvements of course - what meteorologist - currently working in or out of the field, didn't get excited about 1 minute imagery?



However, I wasn't aware of some of the RGB combinations that I've been exposed to here at the HWT. The features that can be picked out by applying coordinating colortables to multiple channel views is simply astounding. I've used the simple water vapor, day cloud phase distinction, day convection, differential water vapor RGBs this week....I'll be making a phone call to my broadcast weather vendor. I know I'll need to get some other broadcasters on board with me to lead a charge, but count me in. Until then, I'll be livin on the CIRA RMMB Slider site!

Color (RGB) Me A Fan!

-icafunnel

Merged TPW Composite Values running low

A quick comparison at 21Z reveals that Merged TPW Composite is running about .3 inches lower than Blended Total PW.

NUCAPS Comparison with Special 17Z Sounding

Today we were tasked with the LIX CWA (New Orleans/Slidell, LA), and the NUCAPS sounding went overhead at 1824Z. Luckily there was a special sounding out of the LCH (Lake Charles, LA) office at 1700Z so there was at least a loose chance to compare the data. So, let's do just that! First the LCH sounding...



Shows a relatively unstable atmosphere with decent lapse rates in the mid-levels.  And now the nearest good-value NUCAPS sounding (taken just north of Vermilion Bay)...



The two soundings aren't an exact comparison, they aren't in the same location or taken at the same time, but synoptically it should be a pretty similar environment ahead of the broken line of storms moving into the area. The biggest problem comes in the boundary layer with values that are way off. The profiles were showing a 2m value of 76/70 when in the observed sounding it was 82/75. Given that the sounding was taken earlier in the day, it's troublesome that the temps were much higher than the NUCAPS sounding. Unfortunately, modified soundings were not available on this day and given the unmodified soundings inability to correctly analyze the boundary layer, it makes trusting any surface derived data near-impossible. The only application I can see for these soundings are to analyze the upper-level synoptic environment (i.e., upper-level lapse rates, relative humidities, etc.) For this, you can compare some values... 850-500 Lapse Rate was 6.1 C/km in the NUCAPS sounding it was 6.7 C/km, some variability is expected and it's not too bad but the difference is worthy of caution. In the gridded product somehow the Lapse Rate was 6.48 C/km, much better given the expected variability but I'm not sure these values are different. Are the gridded data not derived from the soundings or vice versa? Mid-level RH in our LCH sounding is 90%, yet in our NUCAPS sounding it's 79%. I'd expect these values to be a bit closer given the strengths of NUCAPS, but it's close enough to perhaps chalk up to differences in space and time.

Afterwards, I saw that LIX also had a Special Sounding that could be comparable, I'll leave the comparisons below...



#ProtectAndDissipate

This was interesting

Since things were kinda boring over HGX today I had to look for things to blog about. Here is something interesting that I noticed with the Vis/IR Sandwich RGB. At the beginning of the loop the RGB didn't show any IR brightness temperatures. As the TCU continued to develop we saw pixels of IR brightness temperatures starting to show up indicating cold cloud tops and the potential for glaciation to start. Low and behold, if we continue to watch the loop, more pixels start showing up...and a few minutes later both GLM and ground based networks picked up on a couple of CGs. As the thunderstorm developed, cloud tops cooled and more pixels started getting displayed we began to see more lightning.  This was something that I didn't expect the RGB to pick up on and potentially helpful for situational awareness and IDSS in the west where most lightning tends to come from single cell thunderstorms. If you are able to pick up on these ahead of time you may be able to some lead time before lightning occurs.



Update...it happened again, pixels starting being displayed and then lightning occurred within a couple of minutes. If this generally hold true for single cell thunderstorms this would be awesome for outdoor IDSS.

Life and Death in One Loop

Well, the convection is trying. And dying.  Case in point, the full life cycle of a pair of storms across the  northern suburubs of Houston.  Thoughts below:


You are looking at the 1-minute Mesoscale Sector scan with GLM Flash Extent Density, the Vaisala GLD dataset, and ProbSevere overlays. Some things to look at:

ProbSevere struggled when the storms were most intense by merging two individual storm objects, splitting them apart, then re-merging back into one object.  Trends were a mess because of this.

GLM lightning activity ceased a full 10 minutes before ProbSevere dissipated the tracking associated with the cells.  However, it continued to show flashes taking place for 7-minutes after the last CG was detected in the GLD data.  Good use for IDSS to let people know that cloud-to-ground strikes are still possible.

The storms initially were moving northwest along a boundary then deviated to the northeast as they strengthened.  As soon as the moved off the boundary...bye bye!

-Dusty

NUCAPS FCAST

Knowing that the NUCAPS FCAST information was in today for the first time this week, I decided to do some analysis of the fields. Initial thoughts are that the missing blocks of data (not sure if that is typical - more visible in the CAPE field in top image - similar data holes observed in LCL, LFC and EL) make it hard to have confidence in it at first glance.

CAPE (Top Image)
NUCAPS FCAST CAPE data isn't good in the pre-storm environment along the LA coast. RAP13 initialization is overlayed for comparison. RAP13 values are 2500-3500+ and much more representating thatn the NUCAPS values that in some cases are less than  500 ahead of the line.

CINH (Lower Image)
While these values appear to be more closely representative compared to the RAP13, the inability of the product to adequately represent the boundary layer at times during the week don't provide me much confidence for these surface based stability parameters.

-icafunnel

ProbSevere Helped to Hold Off On TOR Warning

Here is an example of a long lived supercell where the trends are very important to monitor. Notice the uptick in Prob TOR but then backed off.  AZShear has been impressive for several volume scans.  Prob TOR trends kept me from issuing a TOR warning, and provides some security in not issuing a warning along with other products.  As a side note,  it would be nice to sample some sort of maximum wind gust product, much like we get from MESH with the estimated max hail size.




ZDR_Arcophile

Warning Decision Influenced by ProbSevere

Since our shields have been engaged all afternoon I decided to look at a storm to our east. Below are a series of reflectivity images from the three lowest scans (0.5, 0.9, and 1.3)...







I won't lie to you, if I were back in my home CWA I would have probably issued a SVR based on these scans, but I decided to hold off on issuing one today to see how ProbSevere did.  So what did ProbSevere show? Here is a sample of various MRMS products overlaid with ProbSevere contours...


And here is a time series for this storm looked like...

These trends looked pretty impressive, with ProbHail peaking around 86%. The increase in probabilities make sense, with MESH peaking around 1.4 in (below in purple).


We will have to wait and see if my decision to not issue a SVR based on a ProbHail is justified, but based on the lack of any reports, I think it may have been a good call.

NUCAPS_Ahead of Line

NUCAPS data in advance of convective line today. Iberia County METAR station coincides with NUCAPS operational sounding (modified soundings not available today).



As you can see from the plotted NUCAPS sounding below, the surface values are too low. From the METAR in western Iberia County, you can see in the first image the the T/Td values are 81/76. Sounding surface values are 76/70. Other sounding points in the pre-storm environment are also about 5 degrees too low on temp and dew. While I can modify these, it doesn't give me confidence in an accurate representation of the boundary layer and on any day expected to be busy I'd likely use other products.

-icafunnel

Four Views of TPW

After making some adjustments to the color scaling, we're able to mix model and analysis products of TPW.   The GOM was mostly clear, while clouds covered most of the land.  A very moist airmass was onshore, the Hammond LA GPS site measured 2.25" of TPW.
The GFS and HRRR forecasts both indicated this maximum.  The All-Sky LAP also did well, although it's field is a reflection of the good GFS forecast.  The Merged TPW did not generate these very large purple values (> 2" of TPW) except in a very small area.  The polar orbiter TPW retrievals used in Merged TPW must not have sampled these large values to advect in.  The operational blended TPW (which relies heavily on GPS sites for the analysis over land) also showed a maximum.  This points to the possibility that including the surface GPS network into Merged TPW and All-Sky LAP might be a way to increase the precision of these analyses.

Cloud Mask from CIRA Merged TPW (blue = clear; yellow = cloudy)

[caption id="attachment_16647" align="alignnone" width="1422"]
Four Views of TPW along the Gulf Coast at 19 UTC 09 May 2019: Upper Left: CIRA Merged TPW: Upper Right: All-Sky LAP. Lower Left: GFS 6 hour forecast @ 18 Z, Lower Right: HRRR 1 hour forecast @ 19 Z.

Analyses and forecasts all agreed well on the drier air over the central GOM.

JohnF

Prob Severe New Cell Detection

Here's an example where Prob Severe was slow to detect a new cell on the southwest flank, which is a fairly typical severe scenario.  And when it finally does detect the cell, it is combined with the already large conglomeration of convection.



ZDR_Arcophile

Wherefore art thou Convection?

Two hours in and we are STILLLLLL waiting for convection to fire in the KHGX CWA.  Front is stalled across the northern quarter of the CWA and the sea breeze has pushed inland from the southeast.  Come on storms...really?



O.K., why no convection?  KLCH launched a sounding at 17Z which looked like this from GOES-E (why this is important in a bit):


Congrats to Lake Charles getting a special sounding off before those storms arrived!  This is what the lowest half of the sounding looked like with boundary layer conditions guessed from the near the center of Houston's CWA at KTME (Houston Executive Airport):


No Cap, ~2500 J/KG MLCAPE, surface moisture isn't mixing out too much....so, where are the storms?

Let's take a look at a NUCAPS sounding taken from this point almost an hour after the KLCH sounding:


I know you are going to ask...why not the next closest dot further west?  Well, it is on the edge of the instrument swath and that has a whole host of problems.  Second thing is that the soundings along the edge looked...wonky...compared to the next row further east.  The sounding at that point, again modifying the surface conditions for KTME (88/75 T/Td):


Things that make you go; hmmmm.  Using the MLCAPE profile, the NUCAPS sounding shows that there probably is a residual EML behind the departing storms that the KLCH didn't capture ahead of the storms.  That may be just enough to keep the atmosphere capped and storms at bay.  There isn't much of a cap, so hope continues we will get convection over the KHGX CWA.

NUCAPS sounding profiles - thumbs up for helping figure out what is going on.  And get more polar orbiters up and going so we can have more than two soundings per day...preferably on different orbital inclinations for better spatial coverage. Please?

-Dusty

Prob Severe Combines Discrete Cells

Here is a prime example of two warned storms with distinct updrafts that were combined due to their close proximity to each other. As opposed to being slow to recognize splitting cells, ProbSevere actually took two established discrete cells with separate updrafts and combined them.



ZDR_Arcophile

MRMS and ProbSevere - SAD Match made in the EWP

Wanting a good way to keep up on the environment while also monitoring for storm development?  Give this combo a tryout (and note, I totally stole this from another EWP participant this week.  Why?  Because it works!)



Top left: MRMS Reflectivity at Lowest Altitude (RALA), Reflectivity at -20C, ProbSevere All

Top right: MRMS Maximum EstimatedSize of Hail (MESH), MRMS 60 Minute Hail Tracks (50%             Alpha), ProbHail

Bottom left: MRMS Vertically Integrated Liquid (VIL), MRMS Vertically Integrated Ice (VII,                    100%), ProbWind

Bottom Right: MRMS 0-2km Merged Azimuth Shear, MRMS 0-2km Rotation Tracks 60-minute          accumulation (35% Alpha), ProbTor

The nice thing with this display is that you can quickly sample any of the objects to get details on what is going on with any given storm to keep situational awareness.



Yeah, that's a lot of text on the 4-panel but each one is specific to the threat in the pane and the ProbSevere (top left) is a bit of everything mixed in.  So far, it has worked with all convective modes we've seen so far in the Testbed.

-Dusty

GLM Minimum Flash Area Indicates Convective Weakening

GLM Minimum Flash Area indicating potential weakening in convection. Larger area/extent is likely due to lightning along the anvil, and less localized/shorter distance also indicates potentially weaker updraft. Reflectivity in this example also confirms weakening of the convection.

All Sky CAPE dropout between time steps

I would like to be able to load ALL-Sky LAP Data Type and CAPE on top of radar data. An unfortunate annoyance is that if the time stamp is too far from the radar product time stamp, the All-Sky products completely disappear,  and in fact, if too far away, it won't load at all. I would like the ability for the latest All-Sky products to display, despite the time lags between the the two. In other words, I would rather have the latest "old" data from All-Sky rather than display nothing.



ZDR_Arcophile

Downward Trend in ProbSevere Time Series & GLM Results in Warning Cancellation

A noted downward trend in the ProbSevere Time series along with a downward trend in the GLM products (not pictured) resulted in this warning forecaster to cancel a SVR for Kerr County early. This early cancellation may not have occurred had these products not been available, potentially leading to areas being unnecessarily warned for a longer amount of time.

Comparison of ATPW and Blended TPW

I was finally able to take a look the ATPW (top left) and compare it with the operational blended TPW (top right). My first impression was that both products show similar large scale patterns with a large plume of moisture extending from the Gulf of Mexico into the Ohio River Valley. The ATPW also tends to show a little more detail in the moisture field compared to the blended TPW.

At 15Z ATPW shows an area of lower moisture over the Texas Gulf coast which is sort of fills in by 17Z, likely due to upstream moisture being advected into the area. In contrast, this moisture is continuously present in the blended TPW. ATPW generally had PWATs around 1.5 in, while the blended TPW had PWATs around 1.7-1.8 in. In contrast, 12Z RAOBs along the coast had PWATs that ranged from 1.6 inches at KBRO to around 1.8 at KLCH. These values seems to match better with the blended TPW product. I suspect that the upstream convection over Louisiana and lingering clouds over the coast impacted the ATPW that was corrected in the blended TPW.

HWT Issues SVR 3-Min Before Field Office Does Due to GLM & ProbSevere Data

A strengthening thunderstorm in Val Verde County was noted by Prob Severe and GLM data. The Average Flash Area and Minimum Flash Area values quickly increased in the area of a strengthening updraft across Val Verde County. The combination of these factors, along with a noted uptick in ProbSevere  data, prompted the HWT warning forecaster to issue a SVR for Val Verde County at 1:25pm CDT with WFO EWX issuing a SVR for Val Verde at 1:28pm CDT.

ProbSevere Time Series Used to Hold off on Warning Issuance

While the storm looked strong to possibly severe on radar, the trends in the ProbSevere Time Series were indicating a downward trend/weakening. Thus, I chose to held off on issuing a SVR. The storm did ultimately weaken (middle image compared to bottom image), as the ProbSevere Time Series was indicating.

Tornadoes in northern Louisiana

With a QLCS moving through yesterday, many local storm reports (LSRs) were delayed. The time series below show how ProbTor (Figure 1) and the other ProbSevere products (Figure 2) performed with respect to official NWS warnings and reports. Several forecasters at HWT this week have noted that they look for jumps in ProbSevere and ProbTor to help identify threats and provide confidence in their warnings. For the storm segment below, there was a jump in ProbWind about 8 minutes before the first tornado report ("widespread damage" reported) and a jump in ProbTor 4 minutes prior to the tornado. Increases in lightning activity and 0-2km AzShear helped jump the probabilities.

Figure 1: ProbTor time series for a tornadic storm. The jump in ProbTor was due to a sharp increase in 0-2km AzShear and and increase in ENI lightning density.

Figure 2: ProbSevere products time series compared to NWS warnings and LSRs.