Initiation times of warm season precipitating storms in the southern Great Plains

In previous work we showed (Tucker and Li 2008) that individual storms could be extracted from gridded multisensor precipitation data. We divided these storms into those made up of small single cells (a.k.a. Air Mass thunderstorms), small multiple cell storms and Mesoscale Convective Systems (MCS). We have been examining these data for the Arkansas-Red River Basin. In this work we will show that the single cell storms and small multiple cell storms have a preferred initiation time of 22 UTC, and the MCS at 19 UTC. The single cell storms were more likely to initiate after 22 UTC than before whereas the reverse was true for the small multiple cell storms. Note that although 19 UTC is not the most favorable time to initiate storms, the storms that form at this time are more likely to become MCS than storms that form earlier or later. This pattern is most pronounced in August and less so in April. The pattern has some variation from year to year with years with the most variation being those with less than average precipitation

Single cell and small multiple cell storms were much more likely to form near mountainous regions than on the plains, even late in the day. The single cell and small multiple cell thunderstorms that did form on the plains were more likely to form 12-15 UTC. We found 31.5% of MCS originated west of 103 degrees west longitude. Thus the Rocky Mountains region accounts for more than its share of these storms. The MCS generated in and around the Rocky Mountains were concentrated in a few specific hours whereas those generated in other parts of the domain had less preference for the time of day of initiation. The spatial density of the origin locations of the MCS storms is much greater than the spatial density of the origin locations of either the single cell storms or the small multiple cell storms. Thus, MCS are more likely to form from storms originating at specific locations.