LightningCast – Page 4

LightningCast V1 vs V2

Posted in LightningCast, SPG2025 on May 19, 2025 by Kevin Thiel.

When initially getting into the Lightningcast product I noticed just some subtle differences between Lightningcast V1 (left window) vs V2 (right window) around the 1851Z timeframe to 1921Z. Near the Choctow and Pushmataha Counties had jumped up to 30% with even a small area of 50%, while the version 2 had a lower probability of lightning closer to 10-30% during that same timeframe. Will note on both, given the anvil cirrus from a strong storm off to the west may have been obscuring the area which looks like it lead to abnormally higher lightning probabilities than what otherwise would have occurred without the cirrus overhead. That being said, felt the Lightningcast V2 captured the threat better overall and kept the probabilities lower which would have lead me to not mentioning the area for thunderstorm probabilities just yet,

-Sting Jet

LightningCast Comparison

Posted in LightningCast, SPG2025 on May 19, 2025 by Kevin Thiel.

While the training touted cases where Version 2 (V2) of LightningCast had a clear lead time advantage over Version 1 (V1), many of the new cells I tracked today showed the opposite, with V1 showing higher probabilities sooner than V2.

It seemed like many of the cells were producing lightning very quickly after achieving even modest returns at the -10C level. In this case, the significant instability in the region today led to very rapid vertical cloud growth, which likely outpaced significant radar returns aloft. And while it is purely an educated guess on my part, that might have something to do with the weight given to the -10C reflectivity – if it needs a certain reflectivity threshold to really boost the probability, then the V2 product would be artificially slowing the increase in probabilities. Then in addition to waiting for the -10C reflectivities, you have to add in processing and dissemination lag time.

The V1 product, being based entirely on satellite data, was able to key in on just the rapid vertical growth and boost probabilities based on that alone, and not have to wait for the -10C returns to show up in MRMS.

Fig 1: Loop comparing LightningCast V1 (left) to LightningCast V2 (right)

Fig 2: Comparison of LightningCast V1 (left) and V2 (right) at 2041Z on 19 May 2025, showing V1 being first to have a 30% contour over the cell of interest in NW Arkansas.

Fig 3: Comparison of LightningCast V1 (left) and V2 (right) at 2046Z on 19 May 2025, showing both having a 50% at the same time, though V1 is larger in area.

Fig 4: Comparison of LightningCast V1 (left) and V2 (right) at 2051Z on 19 May 2025, showing V1 being first to have a 70% contour, while V2 still only has a small 50% area.

Fig 5: Comparison of LightningCast V1 (left) and V2 (right) at 2056Z on 19 May 2025, with the first GLM Flash Extent Density return (blue square) noted at 2059Z. Note that V1 had a 70% contour nearly coincident with the GLM Flash square, while V2’s highest return was still only 50%, and well displaced from where the lightning actually happened.

While I don’t doubt that the -10C reflectivity can help in many scenarios, on days like today it didn’t seem to help much, if at all, and in many cases the satellite-only V1 seemed to do a bit better. As I said, it’s my hypothesis that this is due to the rapid vertical development outstripping the production of -10C returns. Additionally, I would be curious if testing out MRMS Vertically Integrated Ice (VII) instead of -10C reflectivity would produce better results (if it hasn’t been tried already)

– Marko Ramius

Odd LightningCast Dashboard Reading

Posted in LightningCast, SPG2025 on May 8, 2025 by Kevin Thiel.

An odd or false reading from LightningCast occurred on the Dashboard readout on May 8, 2025. The IDSS event was the Southern Skies Music Festival in Knoxville, TN. The screenshot below shows the Max Lightning Potential (10-mile radius) increasing dramatically after 14:40 UTC while the other LightningCast options don’t suggest any potential for lightning (in next 60 minutes) until after 15:45 UTC and even then, the others didn’t have any probability over 30%. On the plainview map comparing LightningCast 1 vs LightningCast 2 (Figures 1, 2, 3), the contours finally overlap the 10-mile event ring by 20:31 UTC and match up with the Dashboard probability at the same time (See Figure 2). Discussions within the HWT noted the issue could likely be a domain or pixel issue for the Max P.

Screenshot of the LightningCast Dashboard for Southern Skies Music Festival on May 8, 2025 from 19:40 UTC to 21:25 UTC. Interesting to note the blue arrows pointing to high probability of lightning depicted by the Max while the others remained less than 10% until 15:45 UTC.

Figure 1: At 19:46 UTC, the lightning potential approaches the 10-mile radius when the Max probability suggests nearly a 50% probability of lightning in the next 60 minutes.

Figure 2: At 20:31 UTC, the 10 and 25 probability of lightning enter the 10-mile radius.

Figure 3: At 21:26 UTC is when the probability of lightning really tapers off, even the Max P version.

Loop of the LightningCast version 1 and 2 surrounding the Southern Skies Music Festival.

– Podium

Severe thunderstorm warnings ONLY using satellite data

Posted in GREMLIN, LightningCast, OCTANE Speed Sandwich, SPG2025 on May 8, 2025 by Kevin Thiel.

May 8, 2025 was an interesting day for the experiment because “WFO GSP” was not allowed to utilize radar to issue severe thunderstorm warnings. I relied heavily on OCTANE Speed and CTD/cooling products and GREMLIN for making these decisions. At first, I struggled a bit with mentally shifting everything to the southeast due to parallax. After I began to utilize the ENTLN data as a proxy for storm core, it was better to estimate storm location and movement (although still not perfect). By the end of the day, I ended up issuing seven severe thunderstorm warnings. I found myself relying more on the CTD and cloud top cooling panel than the speed sandwich. Since it was my second day using the OCTANE products, I was quicker at picking up on signals in the CTD panel that suggest the presence of stronger thunderstorms. Having the CTD in one number compared to having to mentally calculate it while using the speed sandwich was helpful when working in simulated operations when a couple seconds does make a difference. I also found GREMLIN useful as a situational awareness tool to help distinguish which cells should potentially be interrogated more.

Here is a loop of OCTANE Speed Sandwich and the CTD and CTC products from two severe thunderstorm warnings I issued. The severe thunderstorm warnings were issued for the same storm, with the second one being issued as the first one was expiring. Looking at the OCTANE data combined with ENTLN (Image 1), it was clear that the storm was taking a turn to the right. I don’t know the specific values that CTD was showing, but I do believe this storm had values of 4+.

Image 1: Two severe thunderstorm warnings in OCTANE SS and CTD/CTC with Lightning Cast v1 and v2 overlaid. ~21:20Z to 21:56Z

GREMLIN (Image 2) also subtly shows this change in direction. What is interesting is that GREMLIN using ECONUS actually maxed out with a value of 60.4 dBZ.

Image 2: GREMLIN loop for the same two severe thunderstorm warnings in Image 1.

Image 3: Severe thunderstorm warning with OCTANE SS and CTD/CTC. ~21:56Z-22:22Z.

Image 4: Severe thunderstorm warning for the same time but with GREMLIN.

Another severe thunderstorm warning I issued, was showing CTD values of 4+ that prompted a warning. In GREMLIN, the strength of this cell was not as obvious in the ECONUS version, but slightly more prominent in EMESO-2. GREMLIN shows these cells basically merging, but I don’t know if that was reality. Looking at these loops compared to the movement of the lightning data, I don’t think my strom track was very good. In cases where radar is unavailable, I could see other novice warning forecasters also struggling with identifying storm track and motion when using strictly satellite data and also trying to mentally correct for parallax. This may not be an issue for more experienced warning operators though.

– Golden Retriever Lover

Web based LightningCast for a DSS event

Posted in DSS, LightningCast, SPG2025 on May 7, 2025 by Kevin Thiel.

My office’s internet went down hard for the entire day, so I had to utilize a mobile hot spot. AWIPS in the cloud was very sluggish and unusable on the mobile hotspot so I opted to only utilize the web based tools for today.

With a focus on the Brisket Appreciation Society Annual Bash in Beaumont, TX, storms started to emerge to the west and southwest of the event around 2 PM. I found the DSS dashboard useful when monitoring the specific location, but I wanted to use the LightningCast map as well for overall situational awareness. It was a little difficult for me to find exactly where the location was, so it would be useful for the map to have a dynamic layer for the DSS events.

I created a public graphic around the time lightning started near the to move closer to the DSS event. I generally utilized radar and LightningCast dashboard for the DSS event for the messaging. I did not include an image of the LightningCast for the graphic because it was already above 80% when I created the image. Because of this, there was very high confidence that lightning would occur, with the goal of the graphic to inform people of the approaching storms and the associated hazards.

LightningCast (both v1 and v2) stayed very high during the entirety of the event. When looking at the DSS dashboard, the probabilities within for the event increased to >80% 30 minutes before lightning was within a 10 mile radius of the event. There was an issue with the 1 minute data that caused that dip to 0. However, the 1 minute data is very noisy, and in this case, is not an improvement when compared to the 5 minute data.

– Golden Retriever Lover

GREMLIN and The Orphan Anvil

Posted in GREMLIN, LightningCast, OCTANE Speed Sandwich, SPG2025 on May 7, 2025 by Kevin Thiel.

I experimented with GREMLIN for convective initiation on May 7, 2025. One feature that stood out and was briefly discussed amongst the developers and participants was an orphan anvil. The first animation below is the GREMLIN CONUS (bottom right), MRMS Composite Reflectivity (top right), and GOES-19 Channel 7 (bottom left) and the second animation is Day Cloud Convection overlaid with LightningCast. As you follow the anvil passing south of the Home (or the Hot Pepper Festival), GREMLIN continues to project it as a storm or at least a shower further east. LightningCast did well and held steady and kept the lightning potential with a set of updrafts to the west. Unfortunately, a meso-sector was not available at this time and thus could not compare the 1-minute data to the 5-minute data. On a side note, the MesoAnwhere did fairly well depicting the orphan anvil progressing eastward.

Images above show the orphan anvil being projected eastward and suggested by GREMLIN as a storm/shower but MRMS has no reflectivity. The lack of lightning probably hurt GREMLIN in this scenario and it was suggested by a developer that anvil temperatures can resemble cloud tops and most likely confused GREMLIN in this scenario.

MesoAnywhere showing the orphan anvil.

– Podium

LightningCast Comparison for Cabbage, Potato, and Bacon Festival near Hastings, FL (JAX CWA) on Wednesday May 7th

Posted in DSS, LightningCast, SPG2025 on May 7, 2025 by Kevin Thiel.

LC V2 did better with lightning cessation showing lightning probabilities dropping quicker as convection moved east of the DSS event which it title “HOME”. This allowed for more timely DSS updates to partners mentioning the lightning threat was diminishing, at least over the next hour or so. LC V2 also did better with storms to the south which we will talk about more below.

LC V2 showed higher probabilities for lightning quicker for developing storms near southern portions of the county warning area. V2 also lowered probabilities quicker once the storms temporarily pulsed down. I do believe one or both of these storms eventually pulsed up again after this though so it is possible V1 performed slightly better in this scenario by keeping slightly higher lightning probabilities.

LC V2 had tighter areal extent of lightning probabilities overall which was more helpful and accurate. May just need to be a little more weary of sensitivity to decreasing lightning probs too quickly for pulse thunderstorms in the summer.

– Ricky Bobby

Tulsa Severe Weather Event

Posted in LightningCast, OCTANE Speed Sandwich, SPG2025 on May 7, 2025 by Kevin Thiel.

An upper low across central Oklahoma allowing for an unstable environment across the Tulsa CWA. At 19:38Z, clearing noted across the western half of the CWA allowing for cu development across this region. At this time, limited to only CONUS level satellite imagery as mesoscale sectors are elsewhere.

21:00, towers with anvils noted and are beginning to shear off of the updraft.

I felt that lightningcast overperformed in pinging on lightning that never transpired in real time.

-Jolly Rogers

Day 3 observations

Posted in DSS, GREMLIN, LightningCast, OCTANE Speed Sandwich, SPG2025 on May 7, 2025 by Kevin Thiel.

LightningCast/MesoAnywhere for DSS Event

We were operating as NWS JAX on 5/7/25. There was not a huge severe threat, but there were cells that produced a lot of lightning. Our DSS event was the Cabbage, Potato, and Bacon Festival in Hastings, FL (labeled as “home” in AWIPS). As soon as the day started, there were cells within the range of the festival, with our primary focus on lightning cessation. One of the first things I noticed was that for the cells to the north of the cells, LightningCast v1 had a wider spatial extent of the probabilities than v2 for areas of decaying convection. I assumed this was because of the MRMS data as there were lower MRMS -10C reflectivities in the northern edges of the convection where the lower v2 probabilities were. I was curious how that would impact the probabilities for the event as the convection began to decay there and how it related to notifying the event of an “all clear”.

Image 1: LightningCast data v1 vs v2 from 19:50Z to 20:11Z.

With respect to the event itself, using the lightning dashboard (image 2) was helpful. Probabilities both in v1 and v2 were high for the beginning of the day in the middle of all the convection. Probabilities started to decline around 20:15Z. Initially, v1 was declining faster, but by 20:30Z, v2 was declining at a faster rate than v1. There was still lightning within the 10 mile radius during these declines, but v2 it seems that v2 caught on to the decay a bit faster. This is also shown in the LightningCast loop from AWIPS (image 3) where v1 had a larger spatial extent to the higher probabilities for longer. I used LightningCast probabilities in my DSS messaging. When the probabilities were going down, I used a number in between each version for the probability I used in the message. I also utilized MesoAnywhere 1 minute Channel 13 data once it became available for us to monitor the potential for new convection behind the main cluster of cells. There was not much being shown, which made me more comfortable noting the event of an all clear. Probabilities eventually did start to go back up after 21:15Z because there was a lot of convection to the north and south of the event.

Image 2: LightningCast Dashboard for the Cabbage, Potato, and Bacon Festival in Hastings, FL.

Image 3: Lightning Cast comparison loop from 20:43Z to 21:32Z.

Octane and Gremlin

Today was the first day that I used Octane. After getting some training on it from Jason, my first impression of it is that it would be a valuable resource for situational awareness for developing convection and alerting you to where stronger thunderstorms are. This could alert a forecaster to a cell that they may need to interrogate a little more. During warning operations, if radar data is available, I do not think that a warning operator would have enough time to use this in their data analysis, but the warning coordinator or mesoanalysis person would probably find this tool helpful. We went through a thought exercise about how we would use this tool when radar data is unavailable. During times like that, the radar operator would probably find this tool helpful because the storm top divergence and the speed sandwich offers information that radar operators look in radar when assessing thunderstorm strength. With respect to the individual options, I found myself using the speed sandwich more initially because I was able to infer the storm top divergence in a unit that I am used to. However, the more I get comfortable with the actual CTD units and colors, having it paired with the storm top cooling would provide me more information overall than the speed sandwich during warning operations. If I only had satellite data, I would probably be pretty liberal with my warning issuance. There was a cell near Juniper Springs around 22Z, where the speed sandwich and CTD tools were showing relatively “stronger” signals (stronger is relative since there were not many strong signals today). Looking at radar data (MRMS MESH, VII, dual-pol data) and knowing the environment, I did not think the cell was producing severe weather. I am not very comfortable just yet with the thresholds in these Octane products, but given the environment was supportive of marginal severe weather, I may have issued a warning in this instance. Both the SpeedSandwich and the CTD displays were indicating the storm was weakening though, so I also might not have either. As others have noted the last couple of days, the MedSmooth CTD was the one I preferred.

Image 4: Octane Speed Sandwich and Cloud Top Divergence and Cooling from 21:54-22:21Z=

Looking at GREMLIN for the same storm, it seemed to have picked up on the stronger signals of the specific cell and would be useful for pinpointing the storm with the highest “potential” in this instance since it was the one with the higher reflectivity values. I don’t know if I would make a warning decision specifically off of this product, but if I had it available in addition to other satellite data, I may be more inclined to issue a warning or an SPS if the environment is supportive of it.

Image 4: GREMLIN 20:31Z to 22:21Z

Unrelated to the previous storm, this was another snapshot I took of GREMLIN. There was an outflow boundary (not visible in this image) that was traveling southward and sparking new convection as it interacted with the sea breeze. A small cell that was relatively strong quickly developed to the southeast of the Jacksonville Airport. While reflectivities were picking up on MRMS, satellite didn’t pick up on it as quickly. GREMLIN did not catch this initiation.

Image 5: GREMLIN at 20:01Z.

There was a cell that developed along a outflow boundary very close to the radar. Satellite data was not really picking up on it with little cooling cloud tops so GERMLIN completely missed it.

-goldenretreiverlover

Lake Charles May 6th event

Posted in GREMLIN, LightningCast, OCTANE Speed Sandwich, SPG2025 on May 6, 2025 by Kevin Thiel.

GREMLIN was helpful early in the event to quickly identify most intense thunderstorms developing across SE Texas, but as more convection developed it became less useful. A severe thunderstorm warning was issued for the cell highlighted in blue on the top right panel of Figure 2 which shows MRMS. GREMLIN data hardly shows much lower reflectivities for the severe cell as more frequent lightning producers to the northwest are likely causing it to struggle.

88D radar data showed a tall and strong Z core. Strong STD and lower ZDR/CC values were also noted aloft though were not shown here.

This was right at the end of the day so didn’t have time to issue more warnings, but storm to the NW is likely severe by this point just given supercell characteristics.

OCTANE STD showed the severe cell with cooling cloud tops and then showing a STD signal as the storm quickly grew taller than an anvil from an upstream thunderstorm. The STD signal was rather weak though and was even weaker than some sub-severe storms which developed across East Texas. This could have been due to disrupted flow from upstream convection. The product did still help me quickly realize there was new convection developing over SW LA and led to me eventually issuing a warning after doing further analysis on 88D radar data.

Didn’t see too many significant differences between LC V1 and V2 overall. I did find this image interesting as convection developed across SW Louisiana LC V1 had higher probs over a larger area before V2. Further north across central Louisiana, there was additional convection developing under thicker cirrus where V2 did a better job showing higher probabilities of lightning. LC V1 only had low or moderate probs at most while there were a few lightning strikes already occurring. I forgot to save an image from that example.