SPG2026 – EWP Blog

Cirrus over I-10 Cold Front, Cirrus, Anvils: Lake Charles, LA

Posted in Lightning Stoplight, LightningCast, OCTANE Speed Sandwich, SPG2026, Synthetic GXI, SZA Imagery on April 30, 2026 by Kevin Thiel.

Today was a very marginal day as a majority of the CWA was covered in cirrus and then later anvils from nearby thunderstorms. This inhibited much of the solar heating potential with the exception of a small corridor from Lake Charles to Lafayette along I-10. The thermodynamics was not impressive across the area with only 1000 to 2000 j/kg of CAPE. Much of the convective initiation relied on a weak cold front and a sea breeze. There were some thunderstorm complexes that had moved through the area which brought lightning concerns to our two DSS points. Towards the end of the afternoon we were finally seeing convective initiation along the cold front.

With clouds contaminating many of the satellite products, today was mainly focused on lightning cast, lightning stoplight, and seeing how they interact with each other. We found some issues with the stoplight product today. The stoplight product was putting an active grid box that was not co-located with the lightning strike that reported from the lightning detection system. In the image below there is a small purple cross outside of the lone red box.

Final thoughts:

Every product we worked with this week has potential to make an impact. From extending daylight rgb products using a solar zenith adjustment, to using lightning cast and lightning stoplight to help make DSS decisions. OCTANE is a massive improvement for helping to forecast the state of convection. While there are limiting factors like clouds, overall the product could quickly make an impact diagnosing convection and being able to see how correct the models are handling the shear profile. The 0.91 and 5.95 products when operational could be used as ground truth in order to identify the boundaries, and being able to compare that with the models will help with getting those first few storms correct.

-Blizzard

Lightningcast V2 Over Terrain

Posted in LightningCast, SPG2026 on April 30, 2026 by Kevin Thiel.

With a lull in activity over our assigned CWA. I decided to go into CONUS view just to see what was going on across the country. An area that drew my attention was in the Grand Junction CWA. In this area, there were multiple examples of regions with observed lightning from FED and ground based networks, where lightningcast version 2 had little to no probabilities.

Figure 1: Example of FED and ground based lightning detections with contoured lightning cast. Shown are the Grand Junction and Denver CWAs

Figure 2: Example of FED and ground based lightning detections with contoured lightning cast. The Denver CWA is the centered CWA.

-wxboi

Final Ops Day – LCH

Posted in Lightning Stoplight, LightningCast, OCTANE Speed Sandwich, SPG2026, Synthetic GXI, SZA Imagery on April 30, 2026 by Kevin Thiel.

Started off the day with a cold front moving north towards the Louisiana coast. As discussed previously, GeoXO products don’t do as well with an already cloudy area, however we were able to clearly pick out where the cold front was over the gulf using the 5.15 imagery. Beyond this, I did not use this product again for ops.

Figure 1: GeoXI Imagery on top, GOES imagery on bottom

When looking at the Day Cloud Phase Distinction around 1830Z, I noticed the SZA imagery was much easier to view in terms of picking out cloud top textures vs the MesoAnywhere imagery. On the same topic of SZA, we got to see SZA really shine by looking at an area of bubbling convection that occurred around sunrise. The difference SZA made was just remarkable. In real time operations, I would’ve viewed the SZA imagery for at least 45 minutes, maybe longer, compared to what we typically use.

Figure 2: MesoAnywhere (left), SZA added (right)

Applied the Stoplight product in AWIPS today, really enjoyed using that with the Earth Networks lightning and GLM. I did observe some spatial difference between the desktop version and the version in AWIPS. I realize this could be because I approximated the range ring in the desktop viewer since there is no lat/lon input option.

Figure 3: Stoplight Desktop (left), Stoplight AWIPS (right) – 1 minute apart timestamped 18:49 and 18:50 Z, respectively

I also observed that the Stoplight expanded in coverage once storms got stronger, which increased my confidence in using this product to assist decision markers. There was a substantial jump in lightning activity highlighted by all of the lightning products, and Stoplight’s footprint expanded in response.

Figure 4: 18:41Z (left), 19:11Z (right)

LightningCast also did a great job picking up lightning potential before seeing evidence of cloud top cooling.

Figure 5: LightningCast probabilities

Finally, there were numerous cirrus clouds/anvil debris over our CWA, so we did have a more difficult timing utilizing some of the OCTANE products, but I did want to highlight that the CTC product detected areas of cooling along a boundary that was pushing away from the complex of storms into an area not obscured with anvil debris. CTC also showed the core areas of the storm where more rapid cooling was occurring.

Figure 6: OCTANE CONUS Cloud Top Cooling

-simoom

Lightning Stoplight in AWIPS

Posted in Lightning Stoplight, LightningCast, SPG2026 on April 30, 2026 by Kevin Thiel.

While utilizing stoplight for IDSS activities. I noticed a few mapping errors in the AWIPs version. The first example comes from a lightning strike detected by ground-based networks underneath the anvil ahead of the storm. Stoplight correctly highlighted this strike, however on the AWIPS display it highlighted the adjacent grid (Figure 1). The strike occurred just northwest of McDade.

Figure 1: ENTLN lightning plot with lightning stoplight and lightningcast (contoured).

While comparing the CAVE lightning stoplight with the webviewer, it appeared as if they did not match one another. Below is an example of what was seen while viewing the CAVE version (figure 2) versus what was seen on the web viewer (figure 3). The webviewer indicated that the radius of concern had been lightning free for 30 minutes, whereas the cave version held on to some for longer.

Figure 2) AWIPS lightning stoplight valid at 2018z.

Figure 3) Lightning stoplight webviewer at 2018z.

-WxBoi

Thursday HWT …the last operational day…

Posted in Lightning Stoplight, LightningCast, SPG2026, SZA Imagery on April 30, 2026 by Kevin Thiel.

FYI: The image above shows my LightningCast forecast setup on my AWIPS workstation at NWS Pueblo. All Ltg data plots in the upper left, and the lightningcast is in the upper right. I prefer lightningcast as an image (vs contours), as the image shows ALL of the probabilities.

Today we are tasked with both IDSS and Warnings for the Austin TX area. I have been doing warnings while my colleague Nick has been doing DSS (2 sites). Once again Lena picked two excellent sites as the strongest of the day went right over the 2 DSS sites. As the storm approached the first DSS site, it intensified, and I decided to issue a warning based on subtle cloud top cooling, the gradient with the storm and the MRMS trends in hail size. At the time I issued the warning a 1 inch hail report came in.

Late yesterday, Kevin was able to put Stoplight in the AWIPS instances. During luchbreak today, I was able to create a procedure which married the ltg stoplight product with LightngCast. We used this today for our DSS. As mentioned earlier, Nick was taking care of the IDSS so I did not look at the ltg data as much as I wanted to. Below is a loop of the data and the storm we were warning for….

Image above…the married “ltg cast” and “ltg stoplight” (upper left panel).The marginal severe storm (vcnty Austin) is in the lower right image. I really like the fact that we can marry the two lightning products into one (unfortunately, this can only be done cleanly with AWIPS, although it may be doable within the GR environment (but would be really cluttered))

Just some thoughts…A lot of the ltg data above is based primarily off of the GLM. The GLM has an exaggerated footprint of 8 x 8 km. Hopefully, with the new ltg mapper which will eventually come on line in several years will have a smaller footprint. Based on the way things are now, We are likely “overwarning” for the lightning threat when we do DSS based off of ltg stoplight and LightningCast.

Some more thoughts, with the HWT ending tomorrow, the most obvious product that needs to be put into operations in the SZA sat pix products. THE SZA has the biggest bang for the buck and should be easily do-able (but I am no AWIPS software expert). The lightning products (Ltgcast and Stoplight) are also very valuable tools, but need to be simplified as these tools are not ready for public consumption (even though they are public facing). The OCTANE products are impressive, but need alot of work before they are placed into AWIPS (The OCTANE site on the CIRA website is the place to go to see this data the best.

Here are some suggestions on making things a bit easier for the forecasters for the next HWT:

Have IDSS parameters defined ahead of time, i.e., ‘For IDSS point A, wind criteria is 40 mph, any lightning, any hail. Let us know if any of these parameters get within 10 miles of my IDSS Site. Once the weather threat is 10 miles outside my circle, let me know”. In a nutshell, pretend you are the emergency manager for that site.

Set up AWIPS so it defaults with the maps of the CWAs, cities, county names already plotted. Make sure the density and magnifications are appropriate.

A 4 panel radar display of R/SRM would help. Set this up for all radars that will be used on this day. See below for a good display setup

I think it would be better if the teams were either doing IDSS or Warnings, but not both at once. That way we can work together and learn (bounce ideas off each other in realtime). If one is doing IDSS and the other is doing warnings we really can’t share knowledge amongst ourselves (we are too focused on our separate responsibilities).

Some last second things….we noticed that ltg cast was having issues of not detecting lightning in Colorado….

-Mesovortex

Overview: Week 1, Day 4

Posted in Daily Summaries, Lightning Stoplight, LightningCast, OCTANE Speed Sandwich, SPG2026, Synthetic GXI, SZA Imagery on April 30, 2026 by Kevin Thiel.

10 AM

Day 4 kicked off with another debrief session, with considerable focus on LightningCast and Lightning Stoplight for decision support. Near the end of the operations period yesterday we got the Lightning Stoplight tool into the HWT’s AWIPS-II instances (thanks to Justin, Jonny, Roger, Kris, and the rest of the Stoplight team for the hard work to get that done!). We talked about what products in AWIPS-II you could display with it, and some forecasters preferred to have the lightning information (GLM, ENTLN) while others didn’t. I’m hoping another day with DSS will better inform forecasters of these products, and using OCTANE and Synthetic GXI imagery in marginal/messy environments.

Picking our WFO locations was a bit easier today, but it was unfortunately from how limited the areas for convection were. A cold front had washed out much of the moisture and instability down to the Gulf Coast, leaving us with marginal convection from the 13 Z SPC outlook. You know it’s a meager setup when we had to request an ABI MESO domain for our area, and we were glad the SAB approved. We settled for Lake Charles, LA (LCH) and returned to San Antonio (EWX) in hopes of getting some semi-discrete convection in each. Both offices were again given two DSS events and encouraged to use the LightningCast Super Dashboards.

1PM

Operations kicked off and forecasters spent a good bit of time playing with LightningCast and the Lightning Stoplight tool together, and providing feedback to both sets of developers. Working with forecasters and developers, we made a few ‘flavors’ of the Stoplight tool in AWIPS by modifying the colorbar. Here’s the descriptions from the animation below:

Top left is default
Top right colors the flashes in the last minute as dark red
Bottom left changes the green to blue
Bottom right colors the flashes in the last minute as pink followed by a red, orange, yellow scheme

Forecasters had a positive response to the top right panel, and said that coloring the 0-1 minute flashes from GLM and ENTLN dark red in some ways removed the need to overlay those products in AWIPS, simplifying their display. I saved this display off as a procedure and hope to show future weeks and get their feedback too.

3 PM

The Synthetic GXI folks decided to show the forecasters WVT data from the METEOSAT-12 FCI, since it has the 0.91 µm band that GXI plans to add. Features such as fronts and low level moisture boundaries were identified and discussed.

4PM

We found an interesting case of CI near the Lake Charles, LA CWA, with anvil debris from upstream convection obscuring the convection from visible and infrared satellite imagery. The OCTANE cloud top cooling product had a hard time showing cooling until the updraft produced its first lightning. LightningCast v2 and v1 were also compared for their performance in this challenging scenario.

5PM
To end the day I went back and made another animation of the SZA product. This time I got sunrise over the CONUS (thanks Justin!) and showed some early morning convection across central TX, with a different cloud types and layers.

Even in the MRGL, we still had a full day!

-Kevin

LightningCast V1 vs V2

Posted in GLM, LightningCast, SPG2026 on April 29, 2026 by Kevin Thiel.

With an ongoing storm, LightningCast version one featured a “lobe” of higher probabilities extending to the south/southeast. Version 2 didn’t have as pronounced of a “lobe” at the same time (Figure 1). Later on, GLM did pick up on flashes extending down to the south/southeast (Figure 2). To me, the signal for the lightning potential was clearer in the version 1 of lightning cast.

Figure 1. Conus Day Cloud distinction, GLM Flash Extent Density, NLDN CG Flashes, ENTLN Cloud Flashes, Lightning cast V1 (left contours), and lightning cast V2 (right/contours) at 1901z.

Figure 2. Conus Day Cloud distinction, GLM Flash Extent Density, NLDN CG Flashes, ENTLN Cloud Flashes, Lightning cast V1 (left contours), and lightning cast V2 (right/contours) at 1911z.

-wxboi

HWT day 3

Posted in GLM, LightningCast, SPG2026 on April 29, 2026 by Kevin Thiel.

Realtime blogging today….(taking notes on the fly vs waiting until the end of the day)

IDSS begins….

Ltg occurring in PA but not being detected by Ltgcast without MRMS over over central PA at 1852 Apr 29

MRMS ltgcast with MRMS is clearly superior to ltgcast without MRMS as the MRMS version clearly showed the ltg occurring over central PA at 18:52

Use the planview ltg data on awips to verify ltgcast as it shows up quicker then on the dashboards.

Something is not correct…lots of ltg data showing up over our circle(KCRW) point, but it is NOT showing up on our ltgcast display board!?!? (See AWIPS plot below and match that up with the LTG cast board)

Lena gets another star as she is very good at picking active weather points for IDSS locations for this HWT.

2053 UTC – ltgcasat seems to be once again “cheating on itself” as AWIPS GLM is showing lots of ltg occurring over our IDSS site, but th eltg data is NOT plotting in the ltgcast board. Screenshots below: (note the GLM is ramping up again)

White Circle above is our IDSS point (KCRW airport)

In summary,

The MRMS version of lightningcast is superior to the non MRMS version. MRMS picked up the ltg in PA while the non MRMS version failed to pick up lightning (i.e., zero probs indicated and GLM flashes were occurring.

HOWEVER….GLM data was showing over our IDSS site as seen in the AWIPS imagery, but th eltg data was NOT plotting in the lght cast ltg boards. Something is not right and this needs to be looked at (Another way to say this is ltg cast is cheating itself….the ltg is occurring but it is not plotting the data on the ltbg boards.

I need to add that I used multiple boards at once

I wish it wasn’t so time consuming to set up things when we go into ops

Lena is very good at picking IDSS points.

Thanks to Dr Apke for going over the OCTANE in a bit more detail with me (and my group, I learned more today).

-Mesovortex

Lots of Lightning

Posted in Lightning Stoplight, LightningCast, OCTANE Speed Sandwich, SPG2026, Synthetic GXI, SZA Imagery on April 29, 2026 by Kevin Thiel.

Got a little more time in with both LightningCast and Stoplight today being over RLX’s area. Learned today that you can add range rings in the Stoplight product. Kelley walked us through how to make adjustments to our display, such as changing the basemap and adding overlays if preferred. Feedback: One thing I noticed while using Stoplight today was I changed tabs for a little bit, came back and noticed that most of the pixels had disappeared. However when I zoomed out some, they re-appeared. I gave an early “all clear” because of this. Granted lightning was definitely on the downward trend, but once I zoomed out and the pixel repopulated, it repopulated within the 10-mile radius. Not sure what happened there if it was user error or what. Also not sure if this is something that can be added or not, but the ability to input a lat/lon would be awesome!

LightningCast: LightningCast did a really great job today, especially V2 with the ABI+MRMS. We observed that this version detected lightning much earlier over eastern Pennsylvania compared to V1. Our storms were more low-topped with much less instability, and we saw LightningCast pick up on the probabilities very quickly even when there was hardly a radar echo. The biggest feedback from today’s session was a lack of GLM data appearing in the Super Dashboard even though GLM data was directly over our DSS event in AWIPS. Lena explained that there are different GLM sources, and since my group was using density, that’s why it wasn’t showing up in the dashboard (at least I think that was the reason?) Regardless, we shared the feedback that it would be nice if those things matched to prevent confusion for the forecaster. On the plus side – I learned even more about GLM today!

OCTANE– Since we were in an area with lower instability, Jason recommended viewing CONUS OCTANE Speed as shear was good in the area. I was amazed at how well this product highlighted the convection. At a quiet point, we jumped down to Texas to observe the firing supercells in the SW, and noticed OCTANE had cooling cloud tops around 5 minutes before lightning cast picked up on it, so that was also an interesting observation. I continue to really, really enjoy using the CTC product with LightningCast overlaid. I would integrate this today into my normal every day DSS procedures in AWIPS if I could.

Figure 1: OCTANE CONUS Speed over WFO RLX

SZA played a huge role today, too, in helping me see the clouds at a higher brightness for a longer time than I normally would! The images below are just down right impressive. I don’t know any forecaster who wouldn’t want more time to have better visualization on satellite.

Figure 2: Pictured left: OCTANE Meso (no SZA), pictured right: OCTANE CONUS (with SZA) at the same time of 22:37Z

I had a harder time using GXI imagery today given all of the cloud cover already in place at the start of ops. It was difficult to make out areas of moisture on the 0.91. However I was still able to pick up on better areas of moisture with the 5.15

Figure 3: GXI imagery at 18Z focusing over WFO RLX

Just to add: Graphic created today for DSSm posted in Slack.

-simoom

Impressive Lead Time A weak cold front: San Antonio, Texas

Posted in LightningCast, SPG2026, Synthetic GXI on April 29, 2026 by Kevin Thiel.

This afternoon had the potential to be fairly active in the San Antonio area with a cold front moving through the area and 3000 to 4000 J/Kg of CAPE. There was weak convection along the frontal boundary in the northern part of the CWA at the start of the “forecast period”. These thunderstorms remained north of the San Antonio Airport which was the site of an airshow. As the afternoon progressed it remained generally quiet for the CWA with the exception of a thunderstorm moving into the western portion of the CWA.

This aforementioned thunderstorm was handled extremely well by both OCTANE CTC and lightning cast. When CTC had begun to pick up, lightning cast quickly elevated lightning probabilities to above 50%. When both products began to highlight the area of concern, it was roughly 25 to 30 minutes before the first lightning strike was reported. This is highlighted in the GIF inserted below.

After the forecast period ended this thunderstorm finally moved southeast into an area where we were providing DSS. There was a similar relationship with the lightning cast probabilities as it moved into the area of concern. When it crossed the 50% threshold it was 35 minutes until the first strike. There was a weird artifact with V2 lightningcast where it was on an upward trend but briefly dropped before immediately increasing again. If we can continue to get these 25 to 35 minute lead times on lightning, especially for outdoor events, it will greatly increase the service we could provide to our partners and the public.

While cloud cover and model performance made it hard to use Synthetic GEOXO imagery. It was still correctly showing the cold front in roughly the correct location. There wasn’t a large change of moisture across the front, so it was very hard to identify in the 0.91 channel. This imagery still provides us another tool to look at, in order to identify areas of concerns for the afternoon.