4. Discussion

[15]  Studies of Pg pulsations performed during recent years showed that still there is no commonly accepted point of view at the mechanism of Pg pulsation excitation. The majority of authors note that, most probably, Pg pulsations are a consequence of plasma instability within the magnetosphere in no way related to fluctuations of the solar wind. In numerous publications dedicated to Pg pulsation analysis several approaches to the explanation of the generation mechanism of giant pulsations are suggested. For example, Green [1979] considers as a source of Pg the fundamental mode of the Alfvén waves generated as a result of the drift instability at the outer boundary of the ring current. The other approach [Rostoker et al., 1979] suggests that Pg is a consequence of the resonance of the asymmetrical poloidal mode in the plasmapause region where the azimuth component of the electric field changes its direction from the westward to eastward. Most often the excitation of drifting waves of Pg pulsations is considered in publications as a result of the bounce instability of the ring current protons [Chisham and Orr, 1991; Poulter et al., 1983; Southwood, 1976; Southwood et al., 1969]. The approximate scenario of Pg pulsations generation was presented by Chisham [1996] in the following manner. The energetic protons (10-20 keV) drift westward from the nightside magnetosphere into the dawn sector. Because of the bounce resonance at the plasmapause, Alfvén waves are formed with characteristics similar to Pg pulsations. In particular, as the experimental studies show, the bounce resonance frequency coincides with the carrier frequency of Pg pulsations [Chisham, 1996].

[16]  One should note that no matter what mechanism of Pg pulsation generation is considered, the appearing Alfvén waves would apparently have a nonlinear structure. Therefore their dynamics can be described by the nonlinear Schrödinger equation with a derivative (NSCED) the solution of which is a soliton of the envelope.

[17]  Our studies showed that the envelope of the Pg pulsation amplitude has a similarity to the shape of the envelope specific for a soliton. Moreover, the data (Figure 2) suggest that there may be a linear dependence of the Pg amplitude on 1/t f^1/2. This makes it possible to state that Pg pulsations are nonlinear wave packets having the properties of shock solitary waves. Thus the experimentally derived relation between the amplitude, filling in frequency, and duration of a wave packet of Pg fits the concepts of the MHD solitons theory [Mio et al., 1976a, 1976b].

[18]  However, according to Takahashi et al. [1992] the wavelength of Pg pulsations is of the order of the length of a magnetosphere field line. This prevents propagation of Pg wave packets in the form of Alfvén solitons. On the other hand, there exist theoretical models what principally solve the problem of the existence of envelope solitons, the carrier wavelength of which is comparable to the length of a magnetosphere field line. For example, Scott et al. [1973] showed a possibility of existence of solitary waves in limited systems. We are not going to consider here any particular model explaining the detected particularities (it is out of the scope of this paper), but we note that the nonlinearity of Pg wave packets should be in the future taken into account while modifying the generation mechanisms of giant pulsations. The latter statement leads to the conclusion that, for example, the Schrödinger nonlinear equation for Alfvén waves may be a successful model and its solutions may describe many the most interesting properties of giant pulsations.

[19]  If one takes into account that Pg demonstrates soliton properties, one can explain some of their morphological features. Actually, solitons are waves propagating without changing of the shape. Their formation is related to fulfillment of some particular conditions which do not always realized in the magnetosphere. Studies of wave packets in the near-Earth plasma and in the solar wind in front of the magnetosphere showed that solitons are observed very seldom [Guglielmi, 1979]. It is known [Green, 1985; Rostoker et al., 1979] that to generate Pg pulsations, very specific conditions in the region of their generation are also needed. Thus, at least rare cases of Pg observations and conservation of their picturesque drop-like shape during their propagation to the Earth surface may be interpreted in the scope of the MHD soliton theory.