Аскаплоты советских станций за 1982 год : (оперативно-информационный материал) / Акад. наук СССР, Кол. фил. им. С. М. Кирова, Поляр. геофиз. ин-т. – Апатиты : Кольский филиал АН СССР, 1989.

12 Auroral PA index. The index of auroral activity calculated with the help of ascaplots was suggested in /9/. It equals to PA = N'KCCJD Л '), where N is the number of blackened squares within the time interval, and К is the coefficient depending on the station's latitude. The connection between N and the auroral magnetic index AL for midnight hours was determined in /9/ from ascaplots of 35 stations during IGY /1,2/. Using these data the dependence of the К-coefficient on the latitude was obtained for two longitud inal intervals. This allows to adjust the data of the stations located at different latitudes to a common value. Since the К-index was calculated for midnight aurora, the PA index is estimated only for a 4-hour interval in the region of the local geomagnetic midnight. AL scales are chosen so that the corresponding values of the PA index be linearly proportional to the energy of auroral electrons of the auroral oval /9/. This permits to make any arithmetic operations with PA index, such as averaging, without losing of the physical content. Table 2 includes PA values for each day in 1982. For a separate station PA can be determined only for a 4-hour interval which may not coincide with the maximum of auroral activity. That is why Table 2 presents PA values obtained by averaging the indices of all the stations operated during that day. The numeral in brackets following every index gives the number of operating stations the data of which were averaged in calculating the index. The trough in the table denotes absence of observations on that day at local midnight, Auroras registered at cloudy conditions were taken into account in calculations of the PA index. This change increases the representativity and statistical veracity of the data, but on the other hand, this leads to the rise of the PA average value. The correlation coefficient of the PA value obtained according to different procedures is 0.7. According to the new procedure, on the average, the value of the PA index is higher by a factor of 1.3. Auroral activity in 1982. Figure 3 presents variations of the PA and Kp daily values. For PA the dashes link the points with gaps of observ ations. This plot characterizes auroral activity during the whole year that provides operative preliminary data selection. In general,variations of auroral and magnetic activity are similar in general but the degree of similarity is slightly worse than during the previous years. Figure 4 shows the same data as Figure 3 but after averaging for three neighbouring days. The data obtained were smoothed according to the trapezium formula. An analogous procedure was applied to the data for 1975 (see "Ascaplots for 1975", /5/) but unlike for 1975 data, averaging and smoothing did not lead to a significant improvement of the similarity of the curves. The correlation coefficient is only 0.27, The best similarity is observed for the first half of the year. Averaging for a longer time intervals leads to a better coincidence of variations of auroral and magnetic activity. Figure 5 shows the changes of the value of the correlation coefficient for various smoothed time intervals. Point 1 denotes the comparison between the curves from Fig.4, point 2 shows the correlation coefficient for monthly mean values for 1982; whereas point 3 3tands for annual averaged values for 1972-1982. The correlation