LightningCast for DSS at HWT

Several forecasters at the HWT were using the ProbSevere LightningCast meteograms to help inform communication and decision-making in simulated DSS events.

In Albuquerque, NM, we simulated their famed International Balloon Fiesta. Several areas of terrain-based convection were in the vicinity, when agitated cumulus clouds started heading toward the event from the southwest (Figure 1; red dot is the event location). Despite partial obscuration of the boundary layer clouds from upper-level outflow of another cell, LightningCast could still "see" the growing cumulus clouds and gradually increased the probability of lightning in the next 60 minutes, with a rapid rise in probability from 20:55 to 21:00 UTC (see Figure 2). The first flashes within 10 miles were at 21:23 UTC.

Figure 1: LightningCast probabilities of lightning in the next 60 minutes (contours), GOES-18 ABI daytime cloud phase distinction RGB (background), and GOES-18 GLM flash-extent density (foreground). Red dot is location of the Albuquerque International Balloon Fiesta. 

Figure 2: Top panel: LightningCast probability time series for GOES-18 5-minute (purple) and 1-minute (green) sectors; Bottom panel: Time series for ENI lightning (orange) and GOES-18 GLM flashes (blue).

Near Boise, ID, the DSS location was the American Aquarium Concert. A weakening upper low still supported sufficient deep-layer shear to produce strong storms heading for the concert from the east-southeast (Figure 3). As the storms edged closer, the probability of lightning continued to increase gradually, until a rather rapid increase from 21:45 to 22:15 UTC, from 20% to 80% (see Figure 4), as the anvil of the storm and some adjacent cumulus clouds moved right towards the event location (red dot in Figure 3).

Both GOES-East and GOES-West captured the increasing lightning potential at the simulated DSS event about 20-25 minutes prior to the first flash within 5 miles of the event. HWT forecasters at one point noted that the probabilities in AWIPS were not lining up exactly with the probabilities in the time series tool. We explained that in AWIPS, the contours are parallax-corrected, while the time series LightningCast data are not corrected. Based on their feedback (and that of other forecasters), we aim to provide parallax-corrected data in the time series capability very soon. 

We hope these cases illustrate the potential benefit of using an on-demand time series capability where lightning potential can be quickly ascertained, for both lightning initiation and advection scenarios.

Figure 3: LightningCast probabilities of lightning in the next 60 minutes (contours), GOES-18 ABI daytime cloud phase distinction RGB (background), and GOES-18 GLM flash-extent density (foreground). Red dot is location of the American Aquarium Concert in Boise, ID.

Figure 4: Meteograms of LightningCast probabilities (top panels) and lightning observations (bottom panels) for GOES-East (left) and GOES-West (right), centered at the American Aquarium Concert simulated DSS event.