4. Discussion

[20]  Earlier the dependency of the appearance time of the precursors on the earthquake magnitude and epicenter distance was found on the basis of the measurements of the surface geophysical fields for a wide energetic class of earthquakes ( K = 10-17 ): lg(TR) = 0.48K -3 [Sidorin, 1979]. Later, on the basis of the observations of the sporadic E layer in Ashkhabad, Korsunova et al. [1999] obtained the following formula for the earthquakes with K = 11-13: lg(TR)= 0.49K - 3.29. It should be noted that the ground-based measurements and ionospheric observations were performed at different times and different places. The common is the fact that the epicenters of the considered earthquakes were located on the land though in different regions. The similarity of the obtained empirical dependencies shows that the ionospheric precursors are a manifestation in the atmosphere of the processes of earthquake preparation with conservation of their main features, and also that the performed identification of ionospheric precursors of earthquakes on the basis of simultaneous measurements of the F2 and E_s parameters is correct.

[21]  In this paper the precursor effects in the ionosphere were studied for stronger earthquakes than in [Korsunova et al., 1999] and also for epicenters in the sea. The behavior of the appearance time on the basis of 33 earthquakes in the vicinity of Japan shown in Figure 2 (top) differs from the behavior presented by Sidorin [1979] for the fields of surface geophysical parameters: lg(TR) = 0.72(M- 1). The comparison of the appearance time of ionospheric precursors at stations Kokubunji and Ashkhabad for similar epicenter distances and magnitudes shows that over the land the precursors appear earlier than over the sea surface. Since the method of revealing ionospheric precursors in this paper is the same as used by Korsunova et al. [1999], the difference in the coefficients in this dependence for Kokubunji and Ashkhabad stations located in the same latitudinal zone manifests some real differences related to the features of earthquake preparation processes in the Central Asian and Pacific regions.

[22]  We explain the empirical formula for the appearance time of precursors obtained by Sidorin [1979] on the basis of the model presentations of the epicenter of an earthquake in preparation as a zone with an increased fracturing. As far as the fracturing develops, changes in the deformation processes are manifested at longer distances. The time of the earthquake precursor appearance on the Earth's surface at the given epicenter distance is determined by the time of the deformation development up to some threshold value depending on the earthquake magnitude. The time of the appearance of the precursor in the ionosphere is a manifestation of this process in the upper atmosphere, though the mechanism of seismoionospheric interaction explaining quantitatively all the features of the observed precursors is not yet developed.

[23]  Currently, two main mechanisms of the seismoionospheric relations are considered: one is related to acoustic gravity waves [Karimov et al., 1990; Khegai et al., 1997; Pertsev and Shalimov, 1996], the other is an electromagnetic one [Gokhberg et al., 1988; Golovkov, 1983; Pulinets et al., 1998b]. Pulinets et al. [1998b] developed a model making it possible to explain the observed 30% variations in the electron concentration in the F2 layer caused by anomalous vertical electric fields ( E_z ) generated in the period of strong earthquake preparation near the Earth's surface. Under the field value of E_z 1 kV m ^-1 at the surface level, the electron concentration in the sporadic E layer increases by an order of magnitude 4 hours after the switching on the field. This fact agrees with the disturbances in E_s observed during preparation of earthquakes with M > 7. However, such anomalously high values of the electric field are observed rather seldom [Smirnov, 2005], the same as earthquakes with M > 7 in concrete region. Moreover, according to model simulations [Kim et al., 1993], because of the influence of seismogenic electrostatic fields, the height of formation of the sporadic E layer grows at ~15 km in comparison with usual conditions. As it follows from the table presented in the text, the growth of the E_s height for the singled out groups of spikes, considered as precursors, is indeed close or exceeds the theoretical value for all earthquakes.

[24]  On the other hand, such changes in the electron concentration (especially in the sporadic E layer) may be caused by atmospheric gravity waves with periods of 2-3 hours. Karimov et al. [1990] detected an increase in the power of AGW in seismically active periods. In the absence of quantitative calculations of the IGW influence on the electron concentration distribution in the F2 and E_s layers, it is difficult to prefer this or that mechanism of seismoionospheric disturbance generation in the ionosphere. One should not exclude a possibility that there is a joint action of both mechanisms if in the process of earthquake preparation both AGW and anomalous electric fields are generated in the epicenter zones. Nevertheless, estimates for the velocities with which seismogenic perturbations propagate in the ionosphere (based on the logarithmic dependence for time of appearance of medium-term precursors) give the value ~3 km h ^-1. This value deviates substantially from the known velocities of the AGW at the ionospheric heights, but it is close to the values of the velocities with which perturbations propagate in the geophysical fields on the surface of the ground [Sidorin, 1979]. The problem of seismoionospheric interaction will be considered later.

Powered by TeXWeb (Win32, v.2.0).