![](https://csdnimg.cn/release/download_crawler_static/10943042/bgb.jpg)
FEJER AND KELLEY: IONOSPHERIC IRREGULARITIES 411
bJ
180
160
140
I00
80
! I !
DAY
i i i i i
ß
%
ß
ß
ß
!
ß
ß
ß
ß
ß
ß
ß
ß
e e ß
ß ß
, ,
I I I I I I , I I I • I
NIGHT
, I I I I I i [ I ' , • • • I I I I I I I ' , I I
10 4 I0 5 0 1.0 I0 2
ELECTRON I-I,5 rn ELECTRON
DENSITY (rn -s} IRREGULARITY DENSITY (m
AMPLITUDE %
,
i i i
_
:!.'.;..
eeee
--
--
--
I I I
104 0 I0 • 30
30-300rn
180
160
140
120
ioo
80
IRREGULARITY
AMPLITUDE %
Fig. 5. Height variations of the electron density and irregularity amplitudes measured at Thumba, India, around noon
and midnight (adapted from Prakash et al. [1972]).
cant 30- to 300-m irregularities above about 92 km during the
daytime, a result also reported by Klaus [1978]. At night, how-
ever, the Indian data show that the spectrum peaks in the 30-
to 300-m range. This also holds true for the nighttime rocket
electric field data reported by Kelley et al. [1979a].
A second difference occurs in the spectral index n, deter-
mined by fitting a power law relation (fin/n) 2 oc k-" to the ob-
servations in the range where fluctuations were observed dur-
ing both the daytime and the nighttime flights (1-15 m).
Prakash et al. reported an index n near zero in the daytime
data, while Klaus [1978] reported a daytime index of 0.5.
However, Klaus was considering the quantity (&n/n) per unit
frequency instead of the more standard (fin/n) 2 per unit fre-
quency, as had been done by the Indian group. Thus these
two results are in agreement with a nearly fiat spectrum. The
Peruvian daytime data actually extend to X = 0.5 m, and the
signal level there falls off with respect to the spectral density
in the range (1-15 m) [Klaus, 1978].
In the nighttime data, Prakash et al. reported wave number
spectral indices for 1 m < X < 15 m ranging from 1.5 to 4,
most of the values falling between 2 and 3.5. As was men-
tioned above, the spectral density is even higher in the range
from 30 to 300 m. The electric field results reported by Kelley
et al. [ 1979a] are in general agreement with these results. The
power spectrum of one component of their electric field data
is reproduced in Figure 9. Peak intensity was found at the
lowest wave numbers (240 m _< X _< 640 m). The spectrum
fiattens appreciably in the middle range and then obeys a
power law with n = 2.5 _+ 0.2 for h _< 40 m. The instrument
quantization level was reached at about 6 m in the data chan-
nel plotted. The power law determined from a higher gain
channel extended to wavelengths near 1 m. The spectrum of
&E2(k) is equivalent in form to the electron velocity spectrum,
8V2(k), since 8V = 8E/B. In turn, from the continuity equa-
tion it can be shown that (Sn/n)2(k) also has the same spectral
form. Thus the electric field data are in good agreement with
the Indian results and the theory discussed below.
VHF radar studies have shown that the type 1 phase veloc-
ity increases slightly with decreasing wavelength but that the
mean type 2 phase velocity is wavelength independent [Bal-
sley and Farley, 1971]. At HF frequencies the type 1 phase ve-
1ocity decreases more significantly with increasing wave-
length, while the type 2 average phase velocity is independent
of frequency [Crochet, 1977; Hanuise and Crochet, 1977]. The
type 2 echoes dominate the spectra at about 9 m (even at large
zenith angles), but only type 1 echoes are ever detected at 1 m
[Balsley and Farley, 1971 ]..
5. Electrojet turbulence: High-resolution radar studies
have been used to study in detail the temporal and spatial
characteristics of the electrojet irregularities [Balsley and Far-
ley, 1973; Farley and Balsley, 1973; Fejer et al., 1975b, 1976a;
Farley et al., 1978]. The large- and middle-scale patches are at
least 2 orders of magnitude larger than the irregularities re-
sponsible for the VHF scattering and have quite variable ve-
locities. Figure 10 shows examples of large-scale structures in
the electrojet observed at 50 MHz using the RTI diagrams.
The local conditions within the patches can be significantly
different from the average electrojet conditions [Farley and
Balsley, 1973]. Figure 11 shows a typical set of the daytime
spectral data. Type 1 echoes are only observed in a narrow
range between about 103 and 107 km, as was also the case
with the oblique measurements reported previously. The fast
spectral changes are possibly associated with the convection of
large- and middle-scale turbulent structures through the radar
beam. During daytime, upward moving type 1 irregularities
are more frequently observed and occur over a larger range of
altitudes than the downward traveling irregularities. In the
lower part of the scattering region, below about 103 km, the
spectra are considerably narrower, have smaller Doppler
shifts, and are much less variable in time.
The nighttime electrojet echoes are appreciably weaker
than those during daytime due to the greatly reduced electron
densities. The instabilities, however, are more strongly driven,
and the spectra indicate a higher level of turbulence [Fejer et
al., 1975b; Farley et al., 1978]. Figure 12 shows some typical
nighttime spectra during a period of frequent type 1 echoes.
Large and fast changes in phase velocity can occur in time
scales of a few seconds and in height intervals of 1 km or less.
The downward type 1 waves are more common at night and
are observed over a larger altitude range than the upgoing
type 1 waves. Therefore the up-down type 1 asymmetry re-
verses from day to night. The asymmetry is also more pro-