Sudden commencements and sudden impulses in geomagnetism: Their hourly frequency at Cheltenham (MD), Tucson, San Juan, Honolulu, Huancayo, and Watheroo

The diurnal variation in the frequencies of sudden commencements of magnetic storms (SCs) and sudden impulses (SIs), defined in Section 1, is examined at six stations, Cheltenham, Tucson, San Juan, Honolulu, Huancayo, and Watheroo. The data for SCs and SIs are analysed separately for the period 1926–1946, which is common to all six stations. The results show that any local time variation in the hourly frequency of SCs seems likely to be small, and it is suggested that SCs may be more frequent in the afternoon hours, with a maximum around 13h local time, in agreement with the observation made by Moos from his analysis of the Bombay SCs. The hourly frequency of SIs for the six stations considered separately do not exhibit any marked local time effect. Indeed, their U.T. hourly frequencies are not greatly different from one another. When the local mean hourly frequency for the combined data at the six stations is considered, this curve shows similarities to the curve obtained by Newton in his analysis of the Greenwich records, with minima around 08h and 20h local time, this second minimum being much more pronounced than the Greenwich one. Much the same variation is obtained when the ratios of the observed frequencies of SIs at any one station to the greatest frequency at each U.T. are arranged according to local time and the local time mean plotted. The curve exhibits a variation which is very nearly semi-diurnal. It is suggested that there is an essential difference between SCs and SIs in spite of the similarity of their appearance on the magnetic traces. The diurnal variations in the frequency of sudden commencements and sudden impulses with a preliminary movement in the opposite direction (SCs* and SIs*) exhibit a local time effect aad they appear to be most frequent in the afternoon. The frequency of occurrence of SCs* and SIs* may also depend on the longitude of the particular station, though this dependence may be more complex than the one suggested in [6]. The curves for the annual mean number of SCs and SIs (combined) and the curve giving the annual mean sunspot numbers show striking similarities.