The NOAA/CIMSS ProbSevere model observed strong vertical cloud growth and glaciation rates in the northern arc of storms at 0100Z in a highly unstable and moderately sheared environment. The southernmost storm only observed weak normalized growth (in the midst of some cirrus cover), but still had > 60% probability of severe before the first 1" hail report (0137Z) and the first severe thunderstorm warning (0143Z). This storm also recorded a 75 mph wind gust and brief tornado reports. The lead-times from ProbSevere ≥ 60% to the first NWS warnings for these four storms were 7, 7, 8, and 18 minutes, and to first reports were 1, 17, 48, and 48 minutes -- so there may be considerable variability depending on the nature of the storm and it's local environment.
Another storm in Wabaunsee Co., KS proved to be a high probability (> 90%) false alarm, owing to moderate to strong satellite growth rates, > 1" MESH, and strong MUCAPE. No warning or report occurred with this storm, which occurred in the middle of the night in a low population density county (3.40 people/km^2). Nevertheless, ProbSevere sometimes has an overforecasting bias, and work is ongoing to incorporate other satellite and radar predictors to reduce any bias.
At around 5AM local time a second arc of storms formed in eastern Kansas. The southern two clusters of storms exhibited moderate to strong normalized vertical growth rates. Each of these storms had at least silver dollar hail reported. The storm northwest of Kansas City observed weak satellite growth, though the MRMS MESH was very high (> 2"). In this case, the MESH heavily weighed the model to produce a high probability of severe (> 80%) over 30 minutes before the first hail report, which was 2" (at 10:45Z), and the first severe thunderstorm warning (10:50Z).
John Cintineo
UW-CIMSS