On the storm south of Fort Stockton, the prob severe values increased to >70% before significant depth of the 50 dBZ core above -20C was seen on radar. In the loop below the prob severe circles the MRMS 50 dBZ echo height above -20C. Notice the higher values in the pink color show up before significant 50 dBZ depth above -20C does. This suggests that the prob severe product may be able to provide lead time over legacy severe detection techniques such as the Donavon criteria for severe hail. Prob severe values are quite high on the storm to the north which had much less 50 dBZ core depth above -20C, although may still have met minimum severe thresholds.

UPDATE: At 5:50, golf ball size hail was reported with the northern storm. This encouraging. Perhaps the false alarms from prob severe are relatively minimal. This is something that I am particularly interested in assessing throughout the rest of this experiment. Prob severe reached >70 approximately 16 minutes before my warning threshold was reached using MRMS 50 dBZ depth above -20C. If assessing with All-Tilts on a single radar I may have caught it and decided to warn a few minutes earlier though. Still, two cases suggest prob severe offers better lead time. Another cell was clearly non-severe based on radar assessment in eastern Ward County moving into Crane County. Prob severe values remained low with this storm, further supporting the suggestion that prob severe may offer better lead time with relatively low false alarms. Prob severe seems to be performing much better in this environment characterized by stronger deep layer shear as compared to yesterdays low-shear / high-CAPE environment in South Texas.

-dryadiabat