4. Discussion

[15]  One should note that many zonal and morphological features of the Pg pulsations [Brekke et al., 1987; Chisham and Orr, 1991; Chisham et al., 1997; Green, 1979] and "pearls" [Guglielmi and Troitskaya, 1973] have mutual similarity. For example, (1) both pulsations type have a characteristic modulation of the amplitudes (view of wave packets) and belong to relatively rare events; (2) the greatest probability of observation of Pg and "pearls" falls on the dawn sector of the auroral zone; (3) the probable region of Pg and "pearls" generation is the vicinity of the plasmapause; (4) the excitation of these pulsation is typical for weak and moderate magnetic disturbances ( Kp 1-3 ) at the recovery phase of geomagnetic storms; and (5) the seasonal and solar cycle behavior of the pulsation observations coincide. The performed study showed that the giant pulsations not only are observed against a background of excitation of "pearls" series, but there are correlations between some characteristics of these pulsations. This fact assumes apparently some genetic relation (of the generation mechanisms) between the Pg and "pearls" pulsations.

[16]  According to Guglielmi and Troitskaya [1973] the oscillation frequency of "pearls" is by the magnitude close to the gyrofrequency of the ring current protons. Usually as a mechanism of "pearls" generation the cyclotron instability of the protons of the radiation belt is considered [Guglielmi, 1979; Trakhtengerts et al., 2000] and that assumes a relation of Pc1 to energetic proton precipitation. In the current publications there are different points of view on estimation of the parameters of the precipitating protons related to Pc1. It follows from the theoretical evaluations and analysis of the solar wind plasma parameters [Guglielmi, 1979; Matveeva et al., 1972] that the protons with energies of about 10-30 keV, injected from the magnetosphere tail to the daytime side, are responsible for generation of "pearls". At the same time the comparison of the direct satellite measurements of the proton precipitation characteristics to the probability of the Pc1 pulsations observations (on the ground) shows that Pc1 are closely correlated to the proton fluxes with energies > 30 keV [Yahnina et al., 2002].

[17]  Currently the most popular point of view on the Pg pulsation origin is that they are excited on the plasmapause because of the bounce instability of the ring current protons [Chisham, 1996; Chisham and Orr, 1991; Poulter et al., 1983]. If one accepts that the "pearls" excitation is related to the energetic protons and the Pg pulsations are observed against the background of "pearls", one may assume that the giant pulsation generation is caused by protons of the same energies as excitation of the structured Pc1. Apparently, the linear relation between tan a and Dst (Figure 3) manifests the relation of these pulsations to the process of the ring current decay at the recovery phase of a storm. It is known that the intensity of Dst variations at the recovery phase of a storm is determined by the energy of the protons injected from the tail into the daytime sector of the magnetosphere.

[18]  On the basis of the data available it is hard to provide a particular evaluation of the energy of the injected protons responsible for simultaneous excitation of Pg and Pc1. Taking into account Yahnina et al.'s [2002] results, the observation of the giant pulsations on the background of “pearl” series, and the found relation between the characteristics of Pg and "pearls", one can assume that both types of pulsations are related to the protons with the energy in the 10-80 keV range. Thus the results obtained in this paper do not contradict to the conclusions of Poulter et al. [1983] and Chisham [1996] and may serve as an indirect experimental confirmation of the Pg pulsation relation to the protons injected during the recovery phase of a storm from the tail into the daytime side of the magnetosphere.