4. Data Analysis
4.1. Longitudinal Variations
|
|
Figure 2
|
|
|
Figure 3
|
|
|
Figure 4
|
[9] Figures 2,
3, and
4 show the results of measurements at altitudes of 75-85 km
at solar zenith angles
45o 
c 60o and
60o c 90o in the summer
and equinox periods. One can see a small number and irregularity in the
spatial distribution of the measurement points do not make it possible to
reconstruct the fine structure of the longitudinal behavior of
ne.
Nevertheless, even at such unfavorable for the given analysis conditions,
deviations from the zonal mean values are clearly manifested in the
longitudinal behavior of
ne. These deviations show a pronounced
minimum in the longitudinal behavior of
ne falling on the Pacific
longitudinal sector. Approximating the longitudinal behavior of
ne by a
periodic function, we obtain that at
h = 75 km the ratio
l = ne max/ ne min =3.8-4.7,
where the indices max and min refer to the electron
concentrations in the vicinity of longitudes
l = 340o and 160o,
respectively. At
h = 80 km and 85 km
l = 1.5-5.6 and
l = 3-3.3,
respectively. The longitudinal dependence of
ne at the zenith angles
c 
60o,
as a rule, is less pronounced than at
c 60o.
One should note that
the extremes in the longitudinal distribution of
ne demonstrate some
relation to the spatial structure of the NO at
h = 105 km which is
governed by the dipole latitude. The dip equator in Figures 2-4 is shown
by solid curve. According to
Gravens and Stewart [1978]
the maximum
and minimum of [NO] at middle latitudes of the northern hemisphere fall
on the longitudes
l = 280o-300o and
l = 100o-120o,
respectively. One can
see in Figures 2-4 that the maximum in the longitudinal behavior of
ne is
actually located close to
l =280o-300o. As for the minimum in the
longitudinal behavior of
ne it is shifted to
l = 160o-180o. This fact is,
most probably, explained by the absence of enough measurement data in
the Oceania longitudinal sector.
4.2. Dependence on Local Time
[10] The least deviations between the available lower ionosphere
models are in the dependence of the shape of the
ne(h) profile on local
time. Nevertheless they still exist. For example, the value of
n(h) for
c 70o in the IRI model has the maximum value at a height of 65 km,
whereas according to
Belikovich et al. [1983]
the maximum is located at
a height of 90 km.
Knyazev et al. [1993]
performed a comparative
analysis of the expression for
ne(h) presented in the IRI with their own
data. They came to the conclusion that the discrepancies between their
results and the data tabulated in IRI at some altitudes reach 100%.
Danilov and Smirnova [1994]
showed that the
ne(h) dependency on
c postulated in IRI is not confirmed by the experimental data for
c >70o.
An alternate description of the
ne(h) dependency on
c was proposed in
the
Danilov et al. [1991]
model:
where the
A(h) coefficient is tabulated. However, again according to the
conclusions of
Knyazev et al. [1993]
the
Danilov et al. [1991]
model
values of
ne are underestimated as compared to the empirical data up to
an order of magnitude. An attempt to analyze the diurnal behavior of
ne on the basis of the absorption measurements by the A3 method was made
by
Pancheva and Mukhtarov [1997].
This model shows that in winter at
heights below 60 km
ne is subject to no changes with the decrease of
cos c. At altitudes 65 km and 85 km
ne increases linearly toward the noon,
whereas at a height of 70 km the
ne dependence on
cos c have a
pronounced nonlinear character. For summer conditions the picture stays
nearly the same, but at heights of 65 km and 80 km the dependence
becomes of a power character and at 70 km it becomes even more
complicated.
|
|
Figure 5
|
[11] The results of our analysis for 80 km performed separately for the
Eurasian and American longitudinal sectors are shown in Figure 5. One
can see that the
ne dependence on
c in these sectors is significantly
different. Similar difference takes place also at heights of 75 km and 85 km.
Figure 5 shows also the approximating curve of the
log ne(c) dependence from the IRI model for the middle of the American
( l =280o ) and Eurasian ( l = 40o )
longitudinal sectors. One can see that the
log ne(c) dependence from the IRI demonstrates no longitudinal features. On
the other hand, at the solar zenith angles
c 60o it is close to the
log ne(c) dependence for the Eurasian region and at the angles
c 60o the
dependence is close to the
log ne(c) curve for the American region.
4.3. Relation to the Solar Activity Level
[12] A compilation of the dependencies of the shape of
ne(h) profiles
on solar activity obtained by different authors was presented by
Danilov [1998].
The compilation shows that there are considerable differences in
this shape. For example, the ratio
r = neh/ne l,
where the indices "h" and
"l" correspond to high and low solar activity, respectively, at a height of
60 km vary according to different authors from 0.3 to 1.6, that is by a
factor of more than 5. This ratio at 70 km and 80 km varies within
0.9-2.6 and 1.0-3.6, respectively. Only at 90 km the scatter of the ratio in
question decreases down to the range 1.2-2.0. The directly opposite
conclusion was drawn by
Bremer and Singer [1977].
On the basis of the
data of ionospheric radio wave absorption they claimed that the effect of
the solar activity impact (while
F10.7 changed from 75 to 150) on the
electron concentration may be considered negligible small at all altitudes
in the range 60-100 km. However, the vertical structure of the
ne distribution is almost identical in both publications. The ratio
r in the
Bilitza [1990]
model is taken equal to 1.9-2.0 for altitudes of 70-90 km,
whereas
Danilov et al. [1991]
accepted
r = 1 because the available
material did not make it possible to detect any
ne dependence on
F10.7.
|
|
Figure 6
|
[13] The allowance for the influence of the longitudinal effects on the
log ne(F10.7 )
dependence made it possible to reveal an astonishing fact.
Figure 6 shows separately the results of the measurements obtained by
rockets and PR method in the Eurasian region. The corresponding
dependence from the IRI model for similar solar and geomagnetic
conditions is also presented. One can see that the dependence from IRI is
close to the
log ne(F10.7 )
dependence obtained on the basis of rocket
data but is opposite to the dependence obtained on the basis of the PR
data. This inverse relation between
ne and
F10.7 revealed on the basis
of the PR measurements in the Eurasian sector is especially astonishing
|
|
Figure 7
|
because it is direct in the American sector as one can see in Figure 7.
The explanation to this fact should be looked for in peculiarities
of the method used in the PR measurements by the
Belikovich et al. [1983]
group. Their data present the major part of the set of
measurements by PR method in the Eurasian sector. In favor of this
conclusion is the fact that the
ne( F10.7) dependence presented in their
paper was based only on the data obtained by rockets. At the same time
the question on the cause of the appearance of such unusual reaction of
this group of measurements to variations in solar activity needs a special
consideration what is out of the scope of this paper. Here we emphasize
the fact that everywhere according to the results of rocket measurements
and in the American longitudinal sector according to the measurements
by the PR method the relation of
ne to
F10.7 may be presented by the
linear formula of the type:
4.4. Seasonal Variations
[14] The strong variability of the
ne(h) profile in winter period often
leads to a distortion of the
ne diurnal behavior in the major part of the
D region and considerably complicates its description. Therefore currently,
modeling the winter anomaly of the lower ionosphere on the whole, it is
reasonable to consider some general tendencies of the dependence of the
ne(h) profile features on the season, bearing in mind that the profile for
each particular moment may differ considerably from the mean one. This
makes it clear why out of all the models considered above the winter
anomaly is taken into account only in the
Danilov et al. [1991]
and
Pancheva and Mukhtarov [1997]
models. In the former publication the
phenomenon is considered discretely: "no WA", "weak WA", and
"strong WA". The corresponding function entering the expression for
log (ne) takes the values 0, 0.5, and 1, and the value of the coefficient A5 at
the function varies from 0.1 at a height of 65 km to 0.7 at a height 90 km
and 1.0 at heights of 80 km and 85 km. Moreover, the model takes into
account such events as stratospheric warmings (SW) leading to a
decrease of
ne at altitudes of the
D region rather than to an increase. This
factor is also divided at three steps: "no SW", "weak SW", and "strong
SW".
The effect of WA in the
Pancheva and Mukhtarov [1997]
model is
described in a slightly different way. In this model the electron
concentration at the same values of
c is subjected to only weak variations
in different seasons. At the same time in the
ne(h) profiles an additional
intermediate layer is clearly seen at altitudes of 55-65 km. The IRI
model also demonstrates no significant seasonal variability of
ne at fixed
solar zenith angles
c (see Table 2).
|
|
Figure 8
|
|
|
Figure 9
|
|
|
Figure 10
|
|
|
Figure 11
|
|
|
Figure 12
|
[15] Our analysis for the Eurasian and American regions (the results
are shown in Figures 8 and 9) indicates to a strong and weak seasonal
variability of
ne at
60o c 75o and
75o c 90o in the former and latter
longitudinal sectors, respectively. One can see, for example, that in the
Eurasian sector the winter values of
ne at the same zenith angles
c exceed
the summer ones by about a factor of 5. Similar difference between the
winter and summer values of
ne exists at altitudes of 75 km and 85 km.
No such difference is detected in the American sector. To clarify the
problem of possible relation of the seasonal effect in the Eurasian region
to the peculiarities of the PR method, we show in Figure 10 separately
the results of the measurements at a height of 80 km by this method and
by rockets. One can see that the anomalous character of the seasonal
variations of
ne cannot be explained by the peculiarities of the PR
method, because the amplitude of the seasonal variations corresponding
to the rocket measurements is almost by a factor of 2 higher than the
amplitude of the same variations according to the PR data. Therefore it
follows that the location of the anomaly in the Eurasian region has a
natural cause most probably related to the asymmetry of the mechanisms
of horizontal and vertical transport influencing the nitric oxide content at
heights of the
D region. The annual behavior of
ne in the near-noon hours
indirectly also indicate to this fact. The maximum of the electron
concentration values in the summer and equinox periods falls on
~175 day of the year (the summer solstice, see Figure 11),
whereas in the
American sector it falls on ~110 day of the year (see Figure 12), that
means it is in advance of the summer solstice by about 2 months.
4.5. Dependence on Geomagnetic Activity
[16] According to
Danilov et al. [1991]
the increase of the
Kp index
from 0 to 2 units leads at altitudes of 70-85 km to an increase of
log ne by
0.1-0.2 and stays unchanged with further increase of
Kp. However, it is
noted that the problem of the dependence of
ne in the
D region on
geomagnetic conditions needs further consideration. In the other models
it is accepted that this dependence is so weakly pronounced that one can
neglect it. According to our estimates the
ne dependence on geomagnetic
activity has a value of minor corrections what may be neglected at all
altitudes.
Powered by TeXWeb (Win32, v.1.6).
