3. Results
[15] Using the above described method,
Korsunova and Khegai [2005]
earlier processed the ionospheric data for 17 earthquakes under
more hard conditions for their choice: the absence of strong
geomagnetic disturbances in the earthquake preparation periods.
The results obtained in this paper confirm the regularities in the
appearance of medium-term ionospheric precursors of earthquakes
found for magnetically quiet conditions: prevailing of anomalous
positive spikes 2-3 hours long in all ionospheric parameters
exceeding the chosen value for the deviations determined in the
previous section. Most often the ionospheric precursors are
detected around the noon and midnight, but the daytime is
prevailing (29 cases out of 40). The disturbances in the regular
F2 layer most often goes ahead of the disturbances in the sporadic
E layer parameters (32 cases). No increase in the spikes amplitude is
observed to the earthquake moment. The most important parameter
characterizing the process of earthquake preparation is the time
interval between the appearance of the anomalous spikes and the
shock moment. We call it the advance time ( T ) of the appearance
of an ionospheric precursor, the value of T being measured in
diurnal intervals (24 hours).
[16] Table 1 shows the data on the term of appearance of ionospheric
precursor and corresponding values of the deviations in the
parameters of the
F2 and
Es layers for the epicenter distances
R 
600 km in the conditions of the absence of strong geomagnetic
disturbances. The performed analysis showed that at large distance
and under disturbed magnetic field, the value of seismogenic
disturbances decreases considerably, especially in the
F2 layer, so
it is difficult to detect them on the background of the general
ionospheric variability. For powerful earthquakes with
M > 7, the
precursor effects are detected even at distances of the order of
1000 km. The value of the effect is different in the
F2 and
Es layers: in the
Es layer it is higher by a factor of 1.5-2.
Taking into
account the comments made, one may state that the analysis of the
changes in the critical frequency of the
F2 layer only cannot in the
real conditions provide complete enough description of the
ionospheric reaction to an earthquake preparation.
[17] On the other hand, the sporadic
E layer parameters are in a
significant degree influenced by the dynamic processes in the
environment (IGW, turbulence)
[Chimonas, 1974;
Chavdarov et al., 1975;
Ovezgel'dyev et al., 1977],
so the use of the
Es layer only
for revealing of seismoionospheric effects is also able to lead to
wrong conclusions. This is why we emphasize the need to consider
simultaneously changes in the
F2 and
Es layer parameters. This
makes it possible to identify the ionospheric effects of earthquake
preparation with more reliability. In our study, changes in the
height of the sporadic
E layer are a determining factor satisfying
the above presented requirements, as the virtual heights of the
sporadic
E layer are close to the real heights because of the small
width of this layer.
|
|
Figure 2
|
[18] The analysis in the ionospheric precursors for all the considered
earthquakes showed that the advance of their appearance and
quantitative characteristics are determined by the earthquake
magnitude and the epicenter distance: the higher the magnitude
and the closer the epicenter, the earlier the precursor is observed
and the larger is its value. Figure 2 shows the obtained empirical
dependencies of the advance in precursor appearance and values of
the deviations in the ionospheric parameters on the earthquake
magnitude at Kokubunji station for 33 earthquakes with the
epicenter distances
R < 1000 km. The earthquakes for which the
preparation period coincided with strong geomagnetic disturbances
(the daily mean index
Ap > 20 nT) were excluded, because in these
conditions (as it has been noted above), detection of the precursors
in
foF2 becomes problematic. The empirical dependence for the
precursor appearance time for 33 earthquakes at
Ap < 20 nT may
be presented by the following expression
lg(TR) = 1.14M - 4.72 obtained using the least square method. This expression agrees
qualitatively with the expression obtained by us earlier on the basis
of 12 earthquakes for magnetically quiet conditions:
lg(TR) =1.33M - 5.96 [Korsunova and Khegai, 2005],
though the
coefficients in the right-hand side of the approximating lines differ
significantly. The inclusion into the analysis of the earthquakes
with magnetically disturbed preparation period ( Ap > 20 nT) leads
to the dependence
lg(TR) = 1.15M- 4.78 which almost coincide
with the previous one. The latter fact means that the differences in
the coefficients of the approximating lines manifest the errors in
determination of the precursor appearance time in the real
conditions. For the sporadic
E layer, which is the determining
parameter while choosing the precursors, this difference leads to
the change in the product
T 
R by a factor of 1.2 in the magnitude
range 6.0-7.8. Therefore the error in determination of the
appearance time of a moderate-term precursor for the known
distance is about 20%.
[19] The empirical dependencies for the assumed seismogenic
disturbances in the
Es and
F2 parameters shown in Figure 2
(second through fourth panels from the top) are obtained for the
first time. It follows from these dependencies that the stronger the
earthquake and the closer the epicenter the more the value of the
deviation in ionospheric parameters, this fact agreeing also with
the results for the time of precursor appearance
T. The slopes of
the approximating lines for all three parameters are fairly close to
each other ( 0.65 
0.1 ) but are almost by a factor of 2 less than for
lg(TR). Moreover, it could be noted that the coefficients in the
right-hand side of the logarithmic dependencies for the frequency
parameters of the
F2 and
Es layers differ in a lesser degree than for
the virtual heights of
Es
and the time of precursor appearances. The
latter means that the same effects in
foF2 and
f bEs may be caused
by close but weak earthquakes or by strong but remote ones (see
lines 2 and 19 in Table 1). So for forecasting purposes, one can use
the empirical dependencies for
T,
Dh'Es, and
Df b/f m Es, whereas the
F2 -layer parameters are needed to control the correctness of the
ionospheric precursors identification. Having determined such
dependencies for magnetically quiet conditions and particular
observational point using known earthquakes, one can evaluate the
magnitude of the earthquake in preparation, the time of the shock,
and the distance from the observational point to the epicenter.
Approximate estimates of the epicenter location may be made on
the basis of a map of the Earth crust fractures where earthquake
hypocenters are located most often.
Powered by TeXWeb (Win32, v.2.0).