Although the radar reflectivity display of the May supercell was about as impressive as the April storm, and it was obviously a dangerous storm, other radar features showed this storm was not nearly as intense. The May storm's observed VIL and computed Vrot (figure 6) were not as great as in the April storm, and it does not show a pattern as dramatic as the April storm. However, it does show a similar correlation between decreasing VIL, increasing Vrot, and the occurrence of a weak tornado. Around 0001 GMT, there was brief F0 tornado damage near Carrollton, Alabama. This was preceded by a slight drop in VIL and peak in Vrot. This pattern repeated near Gordo, in eastern Pickens county, around 0021 GMT, where there was a short F0 tornado track that crossed into Tuscaloosa county. There was a more pronounced pattern beginning around 0052 GMT, but this was over a sparsely populated area of eastern Tuscaloosa county and western Jefferson county, and it is unknown if any damage occurred. Because of the lack of ground access, it would take an aerial survey to detect damage over much of this area, and resource limitations prevented one from being made. After 0100 GMT, both Vrot and VIL decreased as the storm began to weaken. There were sightings of a funnel cloud over Hueytown and Bessemer in Jefferson county, numerous reports of a roaring sound, and some relatively minor wind damage, but no further evidence of a tornado was found.