Localization of the source of ionospheric disturbance...

4. Experimental Results

[50] Using formulae (8)-(12) in the approximation of the spherical disturbance front for the events on 25 September 2003 and 4 June 2000 the values of the coordinates F_s and L_s, time moment t_s, and the delay d_t of the ID source switching relative to the moment of the main shock t₀, and also of the radial velocity V_r of the ID motion and spatial discrepancy s for various values of the height h_s of the assumed ID source were calculated (Tables 1 and 2).

[51] Besides these values, the mean value of the vertical velocity V_v of propagation of acoustic disturbance from the epicenter to the secondary source in the ionosphere was calculated using the simple relation

(16)

One can easily see that at the radial velocity V_r of the disturbance propagation of the order of 700 m s^-1 the calculated moment of switching on of the ionospheric disturbance source for the 25 September 2003 earthquake is late relative to the main shock by 10 min (Table 3). This fact agrees to the concept that the secondary source of the ionospheric disturbance is located in the ionosphere at a height of the order of 300-400 km.

Figure 6

[52] Similar situation takes place for the calculated characteristics of ID during the 4 June 2000 earthquake. Table 4 shows that the secondary disturbance source at a height of 300-400 km switches on 7-8 min after the main shock, the radial velocity of disturbance propagation being 1000-1050 m s^-1. The mean value of the vertical velocity V_v of propagation of the acoustic disturbance from the epicenter to the secondary source was found close to the mean value of the speed of sound in the height range from the ground to h_s (see the vertical profile of the speed of sound in Figure 6 from Li et al. [1994]).

[53] The coordinates of the ID source calculated with an accuracy of 50-60 km agree well with the position of the earthquake epicenter. The difference between the obtained and true (determined from seismic data) values may be explained by the fact that an approximation of the plain Earth was used in the method. This approximation gives corresponding deviation at a distance of about 500 km.

[54] Moreover, the presence of the meridional and zonal winds at ionospheric heights leads to a shift and deformation of the wave front and, respectively, to appearance of the dependence of the acoustic wave intensity on the propagation direction. Here the deciding role is played by the wind velocity gradient [Ahmadov and Kunitsyn, 2003, 2004; Li et al., 1994]. This can explain the shift of the horizontal position of the secondary disturbance source in the ionosphere relative to the earthquake epicenter.

[55] The phase velocity of the ID motion at the height h_d may be more precisely determined in the approximation of the plain front for groups of subionospheric points if the distance between these points satisfies the approximation of the plain wave (Figure 5). In our experiment such approximation is fulfilled for ID generated during the Hokkaido earthquake. For the nearest to the epicenter set of subionospheric points (MIZU, USUD, TSKB; PRN 13) located to the southeast from the epicenter (marked by letter A in Figure 1), the direction a of the propagation wave vector and phase velocity were found 140^o and V_h=1093 m s^-1, respectively. For the more distant GPS grating (USUD, KGNI, KSMV; PRN 24) located to the southwest from the epicenter (marked by a letter B in Figure 1), the value of the phase velocity V_h=1050 m s^-1 almost did not change; however, the direction angle was found equal to a=210 ^o. In Figure 1 the segments A and B mark fragments of the plain wave front on the circles corresponding to the position of the spherical front of the disturbance at the moments 2000:067 UT and 2003 UT. Adjacent to them the values of the direction and phase velocity of the wave are written.

[56] The value of V_h determined in such a way makes it possible to chose between various values of the height h_s of the assumed source of the ID spherical wave presented in Table 3. The value V_r=700 m s^-1 for the source at a height of h_s=400 km is the closest value of V_h for the 25 September 2003 event. The difference between these velocities may be explained by the fact that the Earth spherical form and different geometry of the rays to the PRN 13 and PRN 24 satellites [Afraimovich et al., 1998] are not taken into account in the used procession method [Afraimovich et al., 2002a].