Methods of developing a sequence climatology for a three-category division of temperature data

Abstract

There are a variety of studies dealing with the persistence of many meteorological variables, including temperature. The resulting sequences of days with similar conditions have also been studied along with these persistence features. Most of the studies have dealt with conditional probabilities which are determined from a two-category division of the data. A method was developed from which sequence characteristics of a three-category division of temperature dab could be determined and utilized. This method allowed for the analysis of not only the two extreme category sequences (cold and warm), but also for the near-normal sequences which have been, to this point, overlooked. It was determined that although the occurrence of each temperature category is equally probable in this study, these occurrences tend to cluster together differently for the different categories. Near-normal sequences significantly differ from the extreme category sequences in both slumber and duration. Slimmer certainly seems to have many differences from the rest of the annual cycle. Lower numbers of sequences and longer sequence lengths, compared to the annual median, are characteristic of the summer season - especially for maximum temperatures. The sequence characteristics that are unique to the summertime are not surprising due to the persistent air masses which reside over Texas during these months. While Waco and Victoria data produce results which are quite similar, Amarillo differs in characteristics for maximum temperatures. A significantly higher clamber of sequences occurs at Amarillo, while the warm sequences are shorter, and cold sequences are about the same as the other locations for maximum temperatures. Also, Amarillo has greater conditional probabilities for minimum temperatures than for maximum temperatures, while the opposite is true for both Waco and Victoria. In addition, the maximum cold and warm sequences observed in a month are highly dependent upon the monthly temperature category, while maximum near-normal sequences are not. There is very little promise in attempting to improve upon climatology for near-normal sequences independence. For extreme categories however, such a venture would improve upon the by utilizing monthly outlooks due to this available information quite significantly.