Relation between daytime and nighttime values of the critical frequency foF2

1. Introduction

[2] The morphology of the F2-layer critical frequency f_oF2 has been studied in detail in numerous publications. However, the author knows of no publications dedicated to a detailed analysis of the relation between the daytime and nighttime values of f_oF2 for the same day.

[3] Vanina and Danilov [2003] were the first to draw attention to the fact that there is a negative significant correlation between the f_oF2 values in the daytime and at night. They obtained this conclusion as a byproduct of the study of the relation between the stratosphere (the parameter h(100) which presents the height of the 100 hPa level (about 17 km)) and ionosphere (the critical frequency f_oF2 ). The moments 0200 LT and 1400 LT were taken as the daytime and nighttime moments.

[4] To analyze the relation between the daytime and nighttime values of f_oF2, a correlation coefficient r(f_oF2 ) which shows the correlation between the values of f_oF2 at 0200 LT and 1400 LT for the same calendar day over the entire data set chosen was calculated. Initially all the data for each year (but only for quiet days with Ap < 8 ) were analyzed.

Figure 1

[5] Figure 1 shows the values of r(f_oF2 ) for the period 1979-1989 for two ionospheric stations: Moscow and Kaliningrad. The first fact attracting attention is that the time behaviors of r(f_oF2 ) for both stations are similar (they almost coincide). The correlation coefficient between the r(f_oF2 ) values for both stations is 0.94 for 11 points (11 years considered). Since quite independent data sets are used (ionosonde observations at different ionospheric stations separated by more that a thousand km) this coincidence may be taken as an indirect proof of the reality of the effect considered.

[6] One can see in Figure 1 that the statistically significant values of r(f_oF2 ) at the confidence level above 99% according to the Fisher criterion (the boundary value of r(f_oF2 ) for 100 points (that is approximately the number of days with Ap < 8 per year) is shown by the horizontal dashed line) are obtained for the years before 1984 and after 1987.

Figure 2

[7] In order to study the behavior of r(f_oF2) during the year, Vanina-Dart and Danilov [2006] performed the following procedure. For the given year they took a running three-month window (January-March, February-April and so on). The r(f_oF2) value was calculated within each window and was plot as a point for the middle month (for example, the value shown for April corresponds to the value of r(f_oF2) calculated for the March-May period). Seasonal variations of r(f_oF2) in 1980 for three stations are shown in Figure 2. The boundary values of r(f_oF2) for the statistical significance of 99% according to the Fisher criterion for the number of quiet days per month equal to 15 are also shown in Figure 2.

[8] Two conclusions are evident from Figure 2. First, the r(f_oF2) value strongly varies during the year and reaches negative values (0.5-0.6) at the significance level of 99% at the end of the spring and in the beginning of the summer. In other months the statistical significance of the r(f_oF2) values is below 99% by the Fisher criterion. The latter means that (at least in the 1980 data considered) there is only one time interval (approximately March-June) when the negative correlation between the daytime and nighttime values of f_oF2 does exist actually. Vanina and Danilov [2003] noted that during this very interval the highest correlation between the stratospheric and ionospheric parameters was detected. However, the fall minimum in r(f_oF2), though at the 90-95% level of significance, should be noted. It may appear important while looking for possible mechanisms of the daytime/nighttime relation of f_oF2 (see below).

[9] Second, there is again a very good agreement in Figure 2 between the seasonal behaviors of r(f_oF2) for all three stations considered (for Gorky station the picture is absolutely similar, but the points are not shown to avoid overloading the figure). Again, it may be interpreted as a proof of the reality of effects detected.

[10] A similar picture of r(f_oF2) behavior was obtained by Vanina and Danilov [2003] for all 6 stations (Kaliningrad, Moscow, Gorky, Tomsk, Khabarovsk, and Alma-Ata) considered. The choice of the stations and the 1979-1989 period was determined by the availability of the proper data of stratospheric sounding used for the search of stratosphere-ionosphere relation. Other stations and years will be considered below.

[11] It should be noted that statistically significant values of r(f_oF2) in spring were obtained by Vanina and Danilov [2003] not for all years. The most pronounced effects were detected for 1980, which was a year of high solar activity (the mean value of the F(10.7) solar index for March-June was 201). Also pronounced effects were seen at the other end of the solar cycle in 1989. In the intermediate years the effects were less pronounced, the r(f_oF2) values were often below the 95% significance level, and the seasonal behavior (shown in Figure 2) was distorted. That gave the ground for the statement that the effect in consideration depends on solar activity. We will come to this problem below.