Show simple item record

contributor authorYarbrough, James W.
contributor authorMeentemeyer, Vernon
date accessioned2017-06-09T17:39:47Z
date available2017-06-09T17:39:47Z
date copyright1978/11/01
date issued1978
identifier issn0021-8952
identifier otherams-9602.pdf
identifier urihttp://onlinelibrary.yabesh.ir/handle/yetl/4233109
description abstractThe goal of this study was to determine the degree of correlation between monthly values of thunderstorm days and tornado occurrences in the eastern United States and to determine the geographic and seasonal patterns of significantly correlated areas. In an initial 8-year analysis (January 1968?December 1975) for each of 124 2° squares of longitude and latitude the number of days with thunderstorms for one city within each square was correlated by month with the number of tornado occurrences. This analysis revealed that 110 coefficients of 1488 (124 squares ? 12 months) were significant at the 95% confidence level. Only five coefficients were negative. The pattern of significant squares suggests that strong correlation is most likely at a place during the month(s) of average passage of the polar front. In a second analysis the 124 squares used in the initial analysis were grouped into 33 regions and into four seasons. Maps of the r values in these 33 regions and four seasons were prepared which show the distribution of the thunderstorm day-tornado relationship. Throughout the entire eastern United States the spring season shows the strongest correlations and summer the least. By using grouped data, areas of inverse relationships are apparent in the Southeast in summer and apparently also fall. In a third analysis squares significant at the 99% confidence level in the initial analysis were researched further for a 26-year period (January 1950?December 1975) to refine tendencies and patterns. Sixteen such squares were examined, but only eight exhibited significant correlations at the 99% level with the highest coefficient equal to +0.712. Thus, the statements in the literature which suggest widespread inverse relationships need to be questioned. Strong positive relationships are somewhat rare; however, for squares for which strong correlations exist, the thunderstorm day per tornado ratio for a particular month and square tends to be numerically smaller when fewer thunderstorm days occur.
publisherAmerican Meteorological Society
titleSeasonal and Regional Variation in the Correlation of Thunderstorm Days with Tornado Frequency
typeJournal Paper
journal volume17
journal issue11
journal titleJournal of Applied Meteorology
identifier doi10.1175/1520-0450(1978)017<1741:SARVIT>2.0.CO;2
journal fristpage1741
journal lastpage1746
treeJournal of Applied Meteorology:;1978:;volume( 017 ):;issue: 011
contenttypeFulltext


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record