Week 1 debrief

We have reached the end of a very successful first week at HWT. Today was a shorter day dedicated to the GOES-R focused 'Tales from the Testbed' as well as a weekly debrief. This debrief covered our activities from yesterday as well as a wrap-up discussion from the entire including topics like which products were used, how they were used, what we can take away from the experience, and what is the next step for the experiment products within in an operational setting. Check it out...

PGLM
- 'We were able to successfully issue a warning in Lubbock based only the PGLM and golf ball sized hail resulted from the storm.'
- 'The one minute update for the total lightning data is a huge benefit over the traditional radar updates.'
- 'There was a storm that barely reached 40 flashes and ended up dropping penny sized hail, so [the PGLM] was fairly successful in this case.'

Cloud Top Cooling
- 'We had upwards of an hour and a half lead time given by the CTC in the Lubbock/Amarillo area.'
- 'The product wasn't useful in areas were there was cirrus contamination.'

GOES-R CI
- 'There was a 97% CI hit on the storms in Lubbock and it was successful.'
- 'The CI is too noisy. THe 10, 20, and 30% values are just a little too much. Maybe it would be better to use values of only 50% and greater.'
- 'The strength of signal concept was good.'
- 'The instantaneous concept of the CI made the product less intuitive, it jumped around. Perhaps it would be better if it focused more on the trend (i.e. a consistently increasing strength of signal on a particular area of cu).'
- 'You could see some of the trend in the CI, but then it faded again. If this happens, as a forecaster you think maybe [storm development] isn't going to happen.'
- 'Some additional continuity would be nice... for example, giving a probability of convective initiation over a specific area at a specific time.'

Forecaster thoughts... i.e. how would do we integrate these into an actual operational setting?
- 'A lot of the experimental products a very useful, but for a warning forecaster their first instinct would be to look back to the base data. It seems that this data would be more useful to a mesoscale analyst, who has time to look at each product in detail.'
- 'It may be hard for a warning forecaster to be looking at the experimental data because there's just so much of it, especially when things are rapidly developing.'
- 'As a warning forecaster I do use some of the experimental data when I can, but don't always have tie to look at all that stuff. That would definitely be useful for a mesosscale analyst and share with the warning forecaster.'
- 'Our mentality is, warning are the most important thing, even in an experimental setting... so it's good to view these new products with that mindset.'
- 'The training had me comfortable with the products when we came in... learning how the interpretation various and also finding out how to work with the products together was a challenge.'
- 'I could see myself using all the products we used this week in an operational setting.'
- 'You have to incorporate the products into an operational mindset. Once you get more used to them it becomes easier, but it's definitely not an immediate process. It take time to become comfortable with them and figure out how exactly it can be incorporated into your own procedures.'

The discussion today, particularly regarding integrating these new products into current forecast operations, was very enlightening. Forecasters have their own mindset when issuing warnings, etc., and  trying to integrate new products into this process is an interesting challenge faced by all in research to operations. However, testbeds like HWT are, I believe the stepping towards towards a successful R to O, and I look forward to continuing this effort next week.

See you then!

Convective Initiation in West Texas

Operations started earlier today, in part to give the chance to use some of the GOES-R products, such as convective initiation, before cirrus contamination.  The two initial locations were Ft. Worth and San Angelo.  The blog post “CI/CTC Identifying Developing Convection in a Favorable Environment” (http://goesrhwt.blogspot.com/2013/05/cictc-identifying-developing-convection.html) summarizes the use of CI earlier in the day in the San Angelo county warning area.

By mid-afternoon, one group shifted to the Lubbock county warning area to investigate the PGLM in that region.  The transfer to Lubbock occurred as storms were already firing, but we took a look back to see what had occurred previously.  

Flipping back we observed the GOES-R CI provide a probability of convection around 20 at 1930 UTC in northwestern Motley County (yellow arrow).

The GOES-R CI (upper left), IR (upper right) and visible satellite (lower left), and Lubbock, Texas radar reflectivity (lower right) at 1930 UTC.

The GOES-R CI remained at this level until 1945 UTC when an upper 70 value was observed at the junction of Motley, Floyd, and Briscoe Counties (yellow arrow).  The visible satellite imagery shows that the features are nearly stationary.

The GOES-R CI (upper left), IR (upper right) and visible satellite (lower left), and Lubbock, Texas radar reflectivity (lower right) at 1945 UTC.

By the next scan, the GOES-R CI had topped 90 for the probability of convection at 2002 UTC (yellow arrow).  Up to this point, the radar was showing now activity in this area.  Also take note of the GOES-R CI in Briscoe County (red arrow) showing moderate strength of signal values around the county.

The GOES-R CI (upper left), IR (upper right) and visible satellite (lower left), and Lubbock, Texas radar reflectivity (lower right) at 2002 UTC.

We then move ahead to 2022 UTC, where radar first observes reflectivities exceeding 35 dBZ and a developing cell can be seen in both the IR and visible imagery in Motley County (yellow arrow).  The GOES-R CI continues to trend upward in Briscoe County (red arrow, and there is no current radar observations of storms.

The GOES-R CI (upper left), IR (upper right) and visible satellite (lower left), and Lubbock, Texas radar reflectivity (lower right) at 2022 UTC.

The situation continues to be interesting by 2045 UTC.  The Motley County storm (yellow arrow) continues to slowly grow.  Also, in Briscoe County (red arrow), where the GOES-R CI was suggesting possible initiation, we are observing several cells in the county on radar, although none have exceeded 35 dBZ just yet.

The GOES-R CI (upper left), IR (upper right) and visible satellite (lower left), and Lubbock, Texas radar reflectivity (lower right) at 2045 UTC.

Finally, at 2115 UTC, the radar signature for the storms in Motley and Briscoe Counties is well defined.  Note that the GOES-R CI stops tracking mature cells.

The GOES-R CI (upper left), IR (upper right) and visible satellite (lower left), and Lubbock, Texas radar reflectivity (lower right) at 2115 UTC.

The GOES-R CI provided the first indication of a potential cell in Motley County at 1930 UTC, giving it a 50 minute lead time.  Even with waiting for a stronger probability of convection, the GOES-R CI provided 35 minutes of lead time (signal of 70+), and 20 minutes (signal of 90+), respectively.  The GOES-R CI followed that with roughly 40 minutes of lead time for the Briscoe County storms.  Once the storms got rolling in this region, they became the focus of the PGLM total lightning blog post “Warning Issued with only PGLM Lightning Jump” (http://goesrhwt.blogspot.com/2013/05/warning-issued-with-only-pglm-lightning.html).

UWCTC detected well before storm turns severe in AMA CWA

UWCTC hit a developing convective element within AMA WFOs CWA at 2045 UTC with a moderate value of -17K/15 min (Figure 1).

Figure 1.  UWCTC valid at 2045 UTC showing cooling rate values of -17K/15 min.

Continued cooling was detected at the following scan at 2115 UTC (of course during the period of the half hour full disk scan!) with a weak cooling rate observed at ~-8K/15 min (Figure 2).

Figure 2.  UWCTC valid at 2115 UTC showing a cooling rate of ~-8K/15 min.

Radar at this time (2119 UTC; Figure 3) showed moderate reflectivity values of ~40 dBZ associated with the detected convective element.

Figure 3.  AMA radar reflectivity valid at 2119 UTC showing moderate (~40 dBZ) reflectivity.

These convective elements did not attain severe levels for quite some time, but the element of interest eventually turned strong to marginally severe with an observed hail report at 2251 UTC of 0.88" (Figure 4).  If one is counting, the CTC signal had a lead time for this convective element of ~110 minutes.  Figure 5 is the radar imagery out of AMA valid at 2249 UTC.

Figure 4.  LSR for hail in AMA CWA

Figure 5.  Reflectivity from AMA radar valid at 2249 UTC during time of above hail report.

This is just one of several other good cases of CTC for today.  Not too bad of an event.

Warning Issued with only PGLM Lightning Jump

At 2226z a storm in the LUB CWA had 1 flash on the pGLM and 10 minutes later the number was up to near 40 flashes.

An experimental warning was issued based only on this and without any other data. About 5 minutes post warning, the HSDA algorithm hit for large hail.

Purple color is large hail

Several minutes later the MESH hit at 1.02″

The pGLM was the first hit of possible severe weather and gave 5-10 minutes more lead time to when the MRMS products hit for severe hail.

Hampshire

PGLM Flash Extent Density in Linear Convection

The PGLM flash extent density product was useful at picking out areas of more severe convection in a linear storm segment. While the lightning “jumps” were not all that impressive (~20 to 30 flashes/min), they were still significant enough in conjunction with MESH near an inch to issue a warning. This is evident in the four panel image below which shows FED, Comp Reflectivity and MESH.

Austin/Frank

PGLM Flash Extent Density Helps with Tracking Storm Intensification

We finally had storms develop within the West Texas LMA. The strongest updraft this evening showed marked ramp up of flashes to 40 just as the storms 50 dBZ core reached 26.25 kft briefly before coming back down about 10 minutes later with the flash density also subsiding back to 15 to 25 flashes. Although the storm didn’t reach severe criteria, it was beneficial that the Flash Extent Density updates in 1 minute intervals, which lets the warning forecaster monitor for rapid updraft development between radar volume scans. This storm resulted in a report of dime size hail.

0.5 reflectivity and Flash Extent Density

Kurtz

Simulated Satellite Imagery, CI, and CTC Forecast Convection Across West Texas

NSSL WRF simulated satellite imagery accurately depicted convection erupting across West Texas as an outflow boundary from a previous MCS across north-central Texas interacted with the dryline. The three following images show a westward propagating outflow boundary interacting with isolated cells along the dryline across portions of West Texas.

NSSL WRF simulated satellite imagery from 21Z. 

Simulated satellite imagery at 22Z.

Simulated satellite imagery at 23Z

The UAH CI and UW-CTC products were also very persistent in developing convection across the extreme southern Texas panhandle and adjacent portions of West Texas.

UAH CI product prior to convective development this afternoon. 

UW-CTC product showing persistent cloud top cooling of 15 to 25 C/15 min.

Satellite imagery verified convective development across West Texas. While impossible to forecast the exact location of initiation, the general area was well represented with lead times of up to around half an hour. Convection can be seen beginning near 2010Z and lasting well into the afternoon hours.

Visible satellite imagery overlaid with analyzed mean sea level pressure and surface observations at 2010Z. 

Same but for 2125Z

Guseman

CTC Starting to Light Up in the TX Panhandle

UWCTC is starting to capture significant convective growth within the Texas Panhandle within the LBB domain.  Current HWT operations are not focused on this area; however, some forecasters have been noticing the development within the region.  This is also an area that was highlighted by some of the experimental models of the EFP and OUN WRF for forecasted development within the current time frame.  The first significant CTC detection was at 2010 UTC with a high moderate value of ~-17K/15 min (Figure 1).  The very next scan (2015 UTC) exhibited a strong CTC rate on the same convective element of ~-20K/15 min (Figure 2).

Figure 1.  UWCTC over visible satellite imagery valid at 2010 UTC.  UWCTC rates of ~-17K/15 min are displayed.

Figure 2.  Same as in Figure 1 but valid at 2015 UTC.  UWCTC rates of ~-20K/15 min are shown.

The radar reflectivity at the time of first detection (2011 UTC; Figure 3)was minimal at < 30 dBZ.

Figure 3.  Radar imagery out of Lubbock, TX showing minimal reflectivity for the storm of interest.

The UWCTC algorithm continued to flag the same convective element with significant cooling rates at the 2025 and 2032 UTC scans (Figures 4 and 5, respectively).

Figure 4.  Same as in Figure 1 but valid at 2025 UTC.

Figure 5.  Same as in Figure 1 but valid at 2032 UTC.

The radar imagery at this time (2031 UTC) was showing a convective element with reflectivity > 40 dBZ (Figure 6).  The VIL values for the respective radar images (not shown) were nothing significant as of yet.  As of the latest satellite scans following the 2032 UTC imagery, the CTC continues to flag newly developing convective elements (not shown).

Figure 6.  Radar reflectivity out of Lubbock, TX valid at 2031 UTC showing moderate reflectivity for the storm of interest.

Monitoring CTC for Possible Initiation Further West than Expected

The image below shows several intense (less than -20 C per 15 min) signals over West Texas in Lubbock’s area. This is a bit of a surprise as most activity was expected to develop farther east. With somewhat limited moisture, we will monitor this area for possible high based severe thunderstorms. If these storms do develop and head east into better low level moisture, they may pose a severe risk across southwest Oklahoma later today.

CTC Performance in Partially Clouded Area

Our forecasters have documented examples of how CTC has performed in areas that are generally clear (aside from cumulus/cumulonimbus development), and we are getting a good feel for how the product behaves. We also understand that areas completely shrouded by high clouds will likely show little to no CI/CTC indications at all.

An “in between” example was found in the SE part of the San Angleo TX CWA near 1930Z this afternoon. This area was west of the enormous cirrus shield in central Texas, but still had some scattered to broken high cloud cover in place. As thunderstorm updrafts developed, CTC signals were indicated on two occasions (pictured below).

CTC (top right) at 1932Z

CTC (top right) at 1940Z

The CT values generated here are extremely (and unrealistically) high. Unlike the “off the charts” error documented in an earlier blog entry, these values appear to be co-located with the actual development of the updraft, and not due to spatial or temporal displacement

At least from this example, we can see how differing cloud environments (clear vs. sct/bkn vs. overcast) will lead to different ways in which the CTC product behaves. It will definitely be useful to learn these differences, and if is determined to be more than just a localized problem here, find ways to separate the developing updraft from the incipient cloud cover in a way that will generate more reasonable and useful CTC values.

–Hatzos

CI/CTC Identifying Developing Convection in a Favorable Environment

The western sections of the San Angelo TX CWA have become quite favorable for thunderstorm development. First, here is a look at the sounding from nearby Midland TX at 12Z.

MAF 12Z Sounding

The values seen here are confirmed by RAP13 (pictured) and OUNWRF soundings just ahead of the developing convection near San Antonio.

RAP13 Sounding nearing San Angelo at 19Z

Lapse rates are extremely steep, and the moisture quality is quite good as well. Deep layer shear is favorable for supercells.

The LAPS 1km (OUN domain) data showed that the dryline in this area would activate explosively, in an area with 3000 J/KG CAPE (first image, top right corner) and LI values near -10 C (first image, bottom left corner). The second image is after the storms are mature, so these values diminish with time due to the convection (perhaps too aggressively ahead of the convection).

1KM LAPS Convective Forecast at 1830Z.

1KM LAPS Convective Forecast at 1930Z.

Because this is an area of clearing behind the large cirrus shield associated with the central Texas MCS, the GOES products are providing information about development in this region.

GOES CI/CTC at 1830Z. Note the areas of enhanced values near the dryline.

GOES CI/CTC at 1845Z. Note the areas of enhanced values near the dryline.

A warning was issued based on this data. The polygon was massaged to incorporate both aggressive CI signal to the south and CTC signal to the north. The UAH CI product was persistent in depicting 90 plus percent probabilities of CI while the CTC product indicated 10 to 15 degrees of cooling over a 15 minute interval. No reports were received to verify the warning, but MESH did end up indicating values of 0.87 inches, which has correlated to severe hail in past events and similar environments this week.

–Hatzos/Guseman

CTC Error Off the Charts

The Cloud-top-Cooling algorithm was off the charts with values of -60 C/15min. This is most likely due to some error in the computing as the anvil moved east. The previous IR frame had no thunderstorm cloud over the location and the cloud top temperature went from +12 to -41 C. This could be a one time error, as other frames do not show this, but something a forecaster needs to be aware of.

Hampshire

Daily Debriefing 5/9/13

Yesterday evening (2-10pm) turned out to be a busy shift, our focus on the OUN and LUB CWAs. Multiple warnings were issued and a lot of good feedback was collected. We even got a bonus and were able to watch a storm roll right overhead from the NWC's observation deck!

Really neat lightning (and hail) show aside, here's some of the forecaster comments from our shift:

Cloud Top Cooling
- 'There was a signal early that rapidly cooled to -26C and I did issue a warning off of that.'
- 'Because of the jet streak moving through central Texas and then the cirrus blowoff from the development of various convective cells, the product struggled due to the cloud contamination.'
- 'Radar can't look through mountains... in the same manner, satellite can't look through clouds.' ...from a product PI

GOES-R CI
- The GOES-R CI had the same struggle at the the CTC. Because of the cirrus contamination forecasters weren't able to utilize it aside from a few cells early on.

PGLM and associated lighting applications
- 'There were some pretty impressive lightning jumps in storms that eventually produced strong winds.'
- 'The lightning jumps did seem to be associated with storm top divergence, storm top growth... so there was a jump on the strengthening of the updraft and the storm itself.'
- 'I liked the idea of the tracking tool, but it was difficult to use with the storms that were barely moving.   It became a very labor intensive task.'
- Forecasters also noted that the number of strikes in the OKLMA network were fewer than anticipated for the strength of the storms. However, after some exploration it was discovered that not all of the LMA sensors were functional and this limited the amount of flashes detected.

Today's focus area will be slightly further south, covering San Angelo and Fort Worth, TX with possible coverage of OUN later in the afternoon. Cloud cover is already widespread over a broad area in this region, but we are hopeful for some clearing and development on the back end where the GOES-R CI and the CTC can be utilized again.

Evaluating the Total Lightning Tracking Tool

Posted by Geoffrey Stano – NASA SPoRT

The 2013 Experimental Warning Program (EWP) has offered a great opportunity to test the NASA SPoRT / MDL total lightning tracking tool. Part of EWP’s efforts are to demonstrate future GOES-R capabilities, such as the Geostationary Lightning Mapper. This is being done with the pseudo-geostationary lightning mapper products being produced by NASA SPoRT in collaboration with seven ground-based total lightning networks. With total lightning observations (intra-cloud and cloud-to-ground) forecasters are looking for lightning jumps, or very rapid increases in total lightning activity ahead of severe weather. In the past, this has been done by visually inspecting the PGLM data. From previous evaluations by both SPoRT and the EWP, the number one request was for a way to visualize the time series trend of total lightning in real-time. With AWIPS II and the ability to create custom plug-ins, the first effort for this has been developed and is now here in operations at the EWP. One design feature was to make this a manual tool. In other words, it is the forecaster who selects the cell track and not an automated tracker, which traditionally has difficulty with merging and splitting cells or not activating at the proper time. Once the forecaster places two points, the tracker activates and will create a track based on these points.  Each point can be moved individually to adjust the track and the tracker will attempt to maintain the track as new data are ingested.  We wanted to evaluate the utility of the total lightning tracking tool as well as its impact on operations. In only a couple days, the feedback has been great. First, here is an example of the PGLM with the total lightning tracking tool this evening.

The PGLM flash extent density observing a maximum of 49 flashes with the total lightning tracking tool time series (pop-up to left).  The lightning jump can be seen early in the time series as it goes from 20 to 49 flashes in two minutes.

Overall, feedback has been very positive as forecasters have  appreciated being able to visualize the time series instead of creating a mental picture.  Also, the time series plot is pinned to always be on top, which prevents it from being lost behind the D2D display.  Also, the feedback has been very constructive to help improve the look and provide ways to minimize the time impact.  Some of the commentary has focused on modest improvements to the layout and look.

• Add a minimize button to the time series display pop-up.
• Have the time series trend color match the active cell point color for the track.
• Have a way to change the default size of the cell point radius of influence (which determine how much lightning data to query).

The feedback also has included very interesting options to save time in implementation.

• The tracker initiates after two points are plotted.  However, it attempts to extrapolate the cell track after placing one point.  This results in the cell points being very spread out.  Suggest waiting to extrapolate the track until after two points placed.
• Also, when the storms are more discrete, instead of using individual cell points created a small “domain” around the storm cell of interest.  The forecaster selects a “polygon” of where the storm will be and the tracker selects the PGLM value within this domain at each time step. This may be able to be done by using the tools similar to how forecasters plot a warning polygon as well as how that can be edited.  Furthermore, the storm motion could be added to further aid the plot.  The big advantage here is that this polygon may be much quicker to place than several individual cell points.

All of these are very good suggestions and this feedback will be taken back to see what can be implemented. The polygon suggesting is very interesting to consider, while the current cell point method would be most viable in a multi-cellular environment.  Additionally, this feedback has been very useful in that it helps refine the total lightning tracking tool ahead of its next evaluation as part of the Operations Proving Ground later this summer.
Tomorrow should be another good day to gather additional feedback.

UWCTC Captures Developing Severe Convection in SE Kansas

CTC captured developing convection that eventually turned severe in the ICT WFO.  The initial and only UWCTC cooling rate was diagnosed to be strong at approximately ~-22K/15 min (see Figure 1) during the 2332 UTC scan. 

Figure 1.  UWCTC over visible imagery valid at 2332 UTC.  The cooling rate was ~22K/15 min.

The base scan radar imagery out of ICT at 2329 UTC had a precipitation core on radar with approximately 50 dBZ radar reflectivity (Figure 2).  No severe warning was issued at that time (The warning shown on the screen near the echo was issued at a later time).

Figure 2.  Radar imagery from ICT WFO.  Storm of interest is NE of radar site.

A severe warning was issued by the ICT WFO at 0016 UTC for this storm with the threat of quarter size hail and strong winds (Figure 3).  No hail reports were received as of this blog entry for verification of this severe warning.

Figure 3.  SVR from ICT concerning storm of interest.

Figure 4 shows radar imagery valid at 0032 UTC for the same storm.  Notice the stronger reflectivity in the core of the storm.  The strong UWCTC rate had a lead time of ~45 minutes over the severe warning that was issued.  Again, no hail reports were received so far.

Figure 4.  ICT base reflectivity valid at 0032 UTC.

Flash Extent Density Jump

One cluster of storms over southwest Oklahoma produced the most impressive FED signal of the day. While we’ve had a warning out for this storm for some time, no reports have been received, and MESH has kept values generally at or below one inch.

**Update**

One rapidly intensifying cell has shown a lightning jump with up to ~50 flashes/min and is likely producing severe hail per MESH.

PGLM Total Lightning Data Aids in Warning Issuance

A rapid ramp up in total lightning data led to increased lead time for a Severe Thunderstorm Warning this afternoon. The Flash Extent Density product increased to 19 flashes per minute prior to warning issuance.

Flash Extent Density product showing an increase to 19 flashes per minute prior to warning issuance.

Flash Extent Density further increased following the warning as severe hail was reported.

Flash Extent Density product increased to 32 flashes per minute as severe hail was reported.

Frank/Guseman

High and Mid Cloud Contamination Limiting Use of GOES-R CI and CTC Products

Early high and mid cloud contamination from an approaching upper-level jet limited use of both the UAH CI and UW CTC products across northwestern Oklahoma this afternoon. Storms initiated around 3 pm with the first Severe Thunderstorm Warning issued around 345 pm. The UAH CI product never showed any signs of CI due to this cloud contamination, but the UW CTC product did briefly show some signs of severe storm development as it provided an approximate 20 minute lead time on the first warning issuance. Previous shifts had found longer lead times utilizing the CTC product, but the extent of upper and middle cloud contamination was not as pronounced during those cases.

4 panel imagery showing nothing from the CI product. The CTC product did briefly pick up on cooling rates near 15 C/15 min at 2025Z.

Same as above at 2032Z, with CTC showing cloud top cooling rates greater than 20 C/15 min.

CTC product continued increasing to greater than 25 C/15 min with storm motion toward the northeast at 2040Z.

Both the UAH CI and UW CTC products show nothing as mid and high clouds continue streaming in with mature storms at 2045Z.

Frank/Guseman