1 Introduction

The study of periodicities in different solar data is important for the understanding of solar magnetic activity. For over a decade many authors have reported various periods other than those at 11 yr and 27 days, of which the one near 154 days is the best known (Rieger et al. 1984; Dennis 1985; Bogart & Bai 1985; Lean & Brueckner 1989; Bai & Cliver 1990; Dröge et al. 1990; Pap et al. 1990; Carbonell & Ballester 1990, 1992; Bai & Sturrock 1991, 1993; Kile & Cliver 1991; Bouver 1992; Oliver et al. 1998; Ballester et al. 1999). This investigation was designed to show that accurate daily radio observations can be very useful for studying the solar periodicities generated in different layers of the solar atmosphere, and in different phases of the solar cycle. We used a new approach to the daily measured radio fluxes which allow us to separate to some extent the radio emission generated in the strong magnetic fields of active regions from that emitted by large but weaker magnetic structures. All the solar data from the minimum and the rising phase of solar cycle 23 were processed separately using the Scargle periodogram technique (Scargle 1982). The statistical significance levels of all the periods found were estimated through the FAP (false alarm probability - the probability that the given periodogram value z is generated by noise), as well as by a Monte Carlo approach to the periodograms obtained (Horne & Baliunas 1986; Bai 1992b; Özgüç & Ataç 1994; Oliver & Ballester 1995).

Papers published by Das & Chatterjee (1996); Das & Nag (1998, 1999) presented some periods in radio data without any discussion about their significance level. Our analysis of radio observations however, gives the majority of the known periods, reveals a clear difference between periodicities observed in two phases of the solar cycle and shows that daily measured radio fluxes at various frequencies are very useful for the systematic study of solar periodicities observed in the different layers of the solar atmosphere.

 

 

Table 1: Parameters of the two different models of the solar radio emission used to eliminate the basic component B from the daily observed flux F_i.

Linear model Fi = B + h(ISN)i
frequency in MHz	405	810	1215	1620	2800	4995	8800
B- basic component [su]	$22.9 \pm .2$	$39.8 \pm.3$	$48.2 \pm.4$	$53.5 \pm.4$	$67.3 \pm.5$	$117.0 \pm.5$	$199.4 \pm.1$
h- radio flux production	$0.138 \pm.003$	$0.337 \pm.005$	$0.508 \pm.007$	$0.611 \pm.007$	$0.785 \pm.008$	$0.633 \pm.008$	$0.752 \pm.017$
% of the data variance	71.0	78.6	81.9	85.7	87.5	83.5	61.6
explained by the model
Boltzman formula model   $F_i = B_i + h_\mathrm B (ISN)_i = A_2 +(A_1 - A_2 ) / (1 + \exp ((t_i - t_\circ) / \Delta t ))+h_\mathrm B (ISN)_i$
frequency in MHz	405	810	1215	1620	2800	4995	8800
parameters of   A1 [su]	$22.1 \pm.5$	$38.9 \pm.8$	$48.0 \pm1$	$53.1 \pm1.2$	$67.4 \pm 2$	$117.8 \pm 1.4$	$200.5 \pm 7.6$
the Boltzman   A2 [su]	$31.4 \pm.5$	$62.5 \pm.7$	$75.0 \pm.9$	$78.9 \pm 1.1$	$100.7 \pm 1.8$	$137.1 \pm 1.5$	$292.5 \pm 7.1$
sigmoidal      $t_{\circ}$ [days]	$807 \pm15$	$855 \pm 7$	$879 \pm 8$	$880 \pm 10$	$926 \pm 12$	$962 \pm 11$	$1064 \pm 20$
formula       $\Delta t$ [days]	$132 \pm 13$	$111 \pm 7$	$80\pm 7$	$91 \pm 9$	$106 \pm 12$	$45 \pm 12$	$143 \pm 16$
$h_{\rm B}$- radio flux production	$0.067 \pm.004$	$0.154 \pm.006$	$0.291 \pm.008$	$0.408 \pm.010$	$0.542 \pm.012$	$0.487 \pm.012$	$0.261 \pm.025$
% of the data variance	81.0	91.8	91.4	91.5	92.5	87.5	79.3
explained by the model