ProbSevere Percentages Associated With Storm Clustering?

[21:00 UTC] As strong to severe convection continues to progress just south of the immediate DC area, it was interesting to note that ProbSevere percentages increased as the algorithm(s) clumped the separate storms together as if they were "one" storm. When this happened, ProbSevere showed an unrealistic jump in severe percentages that then decreased again once the algorithm identified the cluster as "multiple" storms. Interestingly though, in the following couple of radar scans following the "separate clustering", the 0.5 degree V from KLWX showed an enhanced area of outbound wind of approximately 40 to 45 kts at several hundred feet AGL. For ProbSevere, but more specifically ProbWind, values did not respond and remained fairly low (on the order of 40-60%). Therefore, the ProbSevere "increase" seemed more tied to the false clustering of storms rather than the specific increase/enhancement in outbound/outflow from the storms themselves.

 Fig. 1: Note the sudden increase in ProbSevere as it falsely "clusters" the storms together, but then returns to "regular" values as it identifies the storms as separate entities.

Rosie Red