June 15, 2023 – Role playing as DDC, I was tasked with providing DSS for a (fictitious) grass fire near Meade, KS.
DSS: Grass fire near Meade, KS
For this DSS, requested information included lightning within 10 miles of the site, any significant changes in wind speed and direction, as well as other hazardous weather that would pose a risk to emergency personnel containing the fire.
There was ongoing severe convection within the western half of the CWA by the start of my shift, and there was a high likelihood of this convection approaching the DSS site. As shown in Figure 1, my datasets/tools of choice for tracking severe convection and lightning were as follows: GOES-East Mesosector LightningCast, MRMS Composite Reflectivity, ENTLN intracloud and cloud to ground lightning, surface observations, Time of Arrival Tool, Distance Bearing Tool, and Range Rings Tool.
Ground based lightning observations and LightningCast complimented each other nicely when assessing the potential for lightning at the site. Additionally, LightningCast picked up on additional agitated Cu well ahead of the main line of thunderstorms closer to the DSS Site. Using the Time of Arrival tool to track the main cluster of cloud to ground lightning associated with the severe convection was also very useful in providing information on potential to see most lightning via advection, in the absence of additional convective initiation and/or a rapid change in forward speed in ongoing convection.
Some consideration was made to not “overwarn” on lightning potential as the main breadth of lightning would likely come from the severe convection still well off to the west. So with this particular scenario, I set an internal threshold of 80% within LightningCast to send a DSS message. The data readout of the parallax corrected LightningCast offered within AWIPS (not shown) was favored over the non-parallax corrected time series (Figure 2), giving higher confidence in the true probability of occurrence used within the DSS message. This gave around a 35 minute lead time before the first strike was detected within 10 miles of the DSS site. Had we used the non-parallax corrected readout values, lead time would have been much shorter, around 10 minutes using 1-minute imagery and less than 10 minutes using 5-minute imagery. This clearly demonstrates the value of using parallax corrected data compared to non-parallax corrected data when performing DSS.
Here was the DSS message sent at around 21:10 UTC:
Severe thunderstorms have developed around 50 miles to your west, and will likely move over your site between 5:15 pm to 6:30 pm CDT. There is a high chance for storms to remain severe by the time they reach your site, bringing very strong winds over 70 mph out of a direction ranging between northerly to westerly, large hail, heavy rainfall, frequent lightning. We still cannot rule out the potential for a brief tornado, although the chance for a tornado is much lower than previous hazards mentioned. Because of the approaching thunderstorms, the chance for lightning to occur within 10 miles of your site within the next hour (5:15 pm CDT) is over 80%.