ARTICLE IN PRESS
Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
www.elsevier.com/locate/jastp
Planetary wave signatures in the equatorial
atmosphere–ionosphere system, and mesosphereE- and F-region coupling
M.A. Abdua,, T.K. Ramkumarb, I.S. Batistaa, C.G.M. Bruma, H. Takahashia,
B.W. Reinischc, J.H.A. Sobrala
a
National Institute for Space Research—INPE, Sao Jose dos Campos, 12245-970 SP, Brazil
b
National MST Radar Facility, Gadanki, India
c
University of Massachusetts, Lowell, USA
Available online 8 November 2005
Abstract
Upward transport of wave energy and momentum due to gravity, tidal and planetary waves from below and extratropics controls the phenomenology of the equatorial atmosphere–ionosphere system. An important aspect of this
phenomenology is the development of small- and large-scale structures including thin layers in the mesosphere and Eregion, and the formation of wide spectrum plasma structures of the equatorial F-region, widely known as equatorial
spread F/plasma bubble irregularities (that are known to have significant impact on space application systems based on
trans-ionospheric radio waves propagation). It seems that the effects of tidal and gravity waves at mesospheric and
thermospheric heights and their control of ionospheric densities, electric fields and currents are relatively better known
than are the effects originating from vertical coupling due to planetary waves. Results from airglow, radar and ionospheric
sounding observations demonstrate the existence of significant planetary wave influence on plasma parameters at E- and Fregion heights over equatorial latitudes. We present and discuss here some results showing planetary wave oscillations in
concurrent mesospheric winds and equatorial electrojet intensity variations in the Indian sector as well as in the
mesospheric airglow and F-layer height variation in Brazil. Also presented are evidences of planetary wave-scale
oscillations in equatorial evening pre-reversal electric field (F-region vertical drift) and their effects on equatorial spread F /
plasma bubble irregularity development and day-to-day variability.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Planetary wave effects; Equatorial ionosphere; Pre-reversal vertical drift; Equatorial electrojet; MF radar winds; Equatorial
spread F
1. Introduction
Corresponding author. Fax: +55(12)3945-6990.
E-mail addresses: [email protected] (M.A. Abdu),
[email protected] (T.K. Ramkumar),
[email protected] (B.W. Reinisch).
1364-6826/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jastp.2005.03.019
Vertical coupling in the atmosphere–ionosphere
system through atmospheric waves (tidal gravity
and planetary waves) that propagate to ionospheric
heights from the lower regions (tropo–stratosphere)
of their generation, and the associated interactive
ARTICLE IN PRESS
510
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
processes, controls the dynamics and phenomenology of the ‘quiet time’ ionosphere. Over the
equatorial latitudes the unique feature of the
horizontal orientation of the Earth’s magnetic field
lines leads to a condition in which the dynamo
electric fields generated by these waves play the
leading role in associated ionospheric response
sequences extending to the upper layers of the
ionosphere. While the vertical coupling processes
driven by tidal forcing and the consequent dynamo
electric fields are basically responsible for shaping
the major phenomenology of the equatorial region,
the interactive processes involving planetary waves
(PW) and gravity waves are believed to play a
significant role in the day-to-day and short-term
variabilities widely observed in the important
aspects of this phenomenology: equatorial electrojet
(EEJ) current, F-layer plasma drifts and evening
pre-reversal electric field (PRE) that controls the
equatorial spread F (ESF) and related processes.
However, evidences on direct association between
the lower atmosphere wave forcing and the ionospheric response signatures continue to be sparse in
this respect. Major sources of perturbations that can
obscure a clear identification of the cause–effect
relationship in the vertical coupling processes, in
this context, are known to be solar ionizing
radiation flux variations and enhanced solar wind–
magnetosphere–ionosphere
interaction
during
events of space weather disturbances when the
penetrating/disturbance dynamo electric fields and
the disturbance wind system dominate the equatorial region. Thus, only quiet time observational data
sets will be useful in the investigations of the
variability arising from the vertical coupling processes. The entire vertical coupling process can be
seen as a two-stage process: (a) dynamic processes
by which the wave energy from the lower heights/
regions (troposphere and stratosphere) propagates
to higher levels in middle atmosphere and
lower thermosphere (MLT) regions (including the
lower E-region heights); (b) the subsequent sequences in which the electrodynamic processes take
over at height levels near and above the E-layer
extending well into the F-region. In this paper, we
intend to present some new results, and briefly
evaluate some relevant results available recently on
these two aspects and discuss their possible consequences to our investigations towards a better
understanding of the quiet time variability of the
equatorial F-region plasma dynamics and ESF
irregularity processes.
2. PW oscillations in the MLT E- and F-region
Evidence of PW in the mesosphere has come from
extended wind observations by medium frequency
(MF) radars operated at different geographic
locations (Harris, 1994; Haris and Vincent, 1993;
Gurubaran et al., 2001). These waves are very large
horizontal scale atmospheric oscillations in neutral
wind, density and pressure that propagate zonally
and vertically from the troposphere–stratosphere
regions of their generation to the MLT region and
perhaps beyond. Their oscillation periods have been
found to lie in the 2- to 20-day range and beyond
(Forbes, 1996, Forbes et al., 1995), the main periods
being near the normal Rossby modes, i.e., 2-, 5-, 10and 16-days, as well as at other (secondary) periods
resulting from the superposition of these periods. 2day waves in the MLT region is now well
established (see, e.g., Pancheva et al., 2004), as are
the 5-, 10- and 16-day periods (Forbes et al., 1995).
An important feature of the PW is that their
occurrence is of episodic nature so that attention
needs to be focused on specific intervals. For
example, a study of the 2-day waves in mesospheric
zonal and meridional winds over Adelaide by Harris
(1994) showed their occurrences for 2–3 week
duration generally following mid-summer, and
Clark et al. (2002) presented mesospheric wind
measurements showing occurrence of 7-day wave of
zonal wave number S ¼ 1 that lasted more than
20–25 days during August–September 1993. PW of
quasi-2-day and 3- to 5-day periodicities in equatorial mesospheric winds have been reported by
Gurubaran et al. (2001) and Vincent (1993),
respectively, and in mesospheric airglow intensity
by Takahashi et al. (2002). Influence of PW on
ionospheric parameters had been indicated earlier
when Brown and Williams (1971) showed correlation between stratospheric pressure variations and
electron density. More recently, direct evidence of
PW role on mid-latitude sporadic E-layer generation, based on f0Es data from an extended longitudinal chain of stations was provided by
Haldoupis and Pancheva (2002).
Manifestation of the PW effects in the equatorial
ionosphere came to be widely recognized after Chen
(1992) demonstrated 2-day oscillations in the
intensity of the equatorial anomaly and Forbes
and Leveroni (1992) demonstrated 16-day oscillations in the EEJ intensity and the F2-layer peak
density. In view of the theoretical difficulty to
predict their upward propagation to ionospheric
ARTICLE IN PRESS
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
heights (Hagan et al., 1993; Forbes et al., 1995)
these results were attributed to the electrodynamic
signature of the PW interaction with the dynamo
region (lower E-region) of the ionosphere. The wave
oscillations in the EEJ over Huancayo, Peru,
analyzed by Parish et al. (1994) for an interval in
1985 confirmed the similar periodicities, namely, 2-,
5-, 10- and 16-day, to those found earlier by Forbes
and Leveroni (1992) for an interval in 1979. Thus
the oscillation at some prominent PW periods in the
EEJ appears to be reasonably well established. We
have attempted in this paper, as a first step, to verify
the possible association of such oscillations in the
EEJ strength with the oscillations in the MF radar
winds using their concurrent long duration measurements realized in the EEJ region over India.
Fig. 1 shows, as an example, the power spectrum
obtained for the EEJ strength (lower panel),
represented by DH, the difference of magnetic field
H-component variation over an EEJ station (Trivandrum: 8.481N, 76.951E; dip lat: 0.281) and that
over an off-EEJ station (Alibag:18.681N, 72.861N;
Power spectrum of winds & EEJ strength
[15 March to 26 April 1994]
Power (ms-1)2
35
zonal wind at 88 km
35
30
30
25
25
20
20
15
15
10
10
5
5
0
0
0
5
2
4
6
8 10 12 14 16 18 20 22 24
EEJ strength (ABG-TRD)
Power (nT2)
4
5
4
3
3
95 % Confidence level
2
2
90 %
1
1
0
0
0
2
4
6 8 10 12 14 16 18 20 22 24
Period of oscillation (days)
Fig. 1. Power spectrum of the zonal winds at 88 km (top) and
EEJ strength (bottom) during the period of 15 March to 26 April
1994.
511
dip lat: 12.961). In the upper panel is shown the
power spectrum of the zonal wind at 88 km,
concurrently measured by an MF radar at Tirunnelveli (located close to Trivandrum). We may note
in these spectra (which correspond to an observational interval from 15 March to 26 April 1994)
prominent spectral peaks near the periodicities at
6.5-day and 14-day in both the EEJ strength and
zonal wind. The prominent periodicities vary with
the observational epoch as other such plots have
demonstrated. These results might seem to indicate
that PW found at mesospheric heights could
propagate up to the dynamo region where they
modulate the ionospheric current system associated
with EEJ at the PW time scales (more discussion on
this aspect will follow later).
Additional results based on concurrent data on
EEJ and MF radar winds for extended intervals are
presented in Fig. 2. Here, the contour plot shows the
time evolution of normalized power spectrum
determined by ‘‘Morlet wavelet’’ analysis for EEJ
strength (upper panel) and zonal and meridional
winds at 88 km (lower panel) for the first half of
1999 (left panels) and second half of 1993 (right
panels). The data availability dictated the selection
of these intervals. We may note in this figure that
during the late northern summer month of August
1993 the EEJ strength has significant spectral power
for periodicities near 2-, 3-, 5- and 6.5-days.
Correspondingly, the zonal wind also presents
significant power at these periodicities, suggesting,
thereby, that the mesospheric PW observed at 88 km
might have propagated to dynamo region and/or
influenced the wind fields there, so that the EEJ
strength is modulated at the PW time scales similar
to the example presented in Fig. 1. Also to note is
the prominent 12–18 days oscillation in the EEJ
strength from the last week of September till the end
of October that is accompanied by a prominent
12–20-day oscillation in the zonal wind. We note
further that the oscillation in zonal wind precedes
by 12 days that in the EEJ strength, which might
suggest that PW of 12–20 days period observed at
88 km could take some time (of the order of 12 days)
to propagate to the dynamo region heights
(105 km) in the ionosphere. There are also
frequent cases of oscillations present only in one
of the two parameters. Examples are the zonal wind
periodicities of 5-day in July, 8–12-day in August
and 14–20-day in November–December, and
meridional wind periodicities 14-day in August
and 6–11-day in September. Conversely, the
ARTICLE IN PRESS
512
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
Normalaized Morlet wavelet power spectrum (1993 &1999)
EEJS (nT) (top), Zon. (mid) & Mer. (bot.) wind (m/s) at 88 km
22
20
18
16
14
12
10
8
6
4
2
22
20
18
16
14
12
10
8
6
4
2
Fourier period of oscillation (days)
1
31
181
211
241
271
301
331
181
211
241
271
301
331
181 211 241 271 301
61
91 121 151
(1 Jan. to 20 Jun.1999) Day number (21 Jun. to 8 Dec. 1993)
331
61
91
121
151
22
20
18
16
14
12
10
8
6
4
2
22
20
18
16
14
12
10
8
6
4
2
1
31
61
91
121
151
22
20
18
16
14
12
10
8
6
4
2
22
20
18
16
14
12
10
8
6
4
2
1
31
Fig. 2. Time evolution of the power spectra determined by Morlet wavelet analysis method for the period of 1 January to 20 June 1999
(left panels) and 21 June to 8 December 1993 (right panels). The top, middle and bottom panels show, respectively, the results for EEJ
strength, zonal and meridional winds at 88 km. The contours inside the thick dashed contour lines are above the 95% confidence level. The
vertical axis denotes the Fourier periods of oscillation corresponding to the ‘‘Morlet wavelet’’ scales ranging from a fraction of a day to 22
days. The horizontal scale shows the observational period in terms of the day of the year.
enhanced 8–11-day oscillation in EEJ is not
associated with corresponding mesospheric PW,
which might point to the possibility of in situ
generation of these oscillations in dynamo region
winds. As regards the results for the year 1999, we
may note that the EEJ strength presents pronounced oscillations with periodicities near 2-day
in the first week of February and near 5–6-day in
April which are accompanied, respectively, by
significant and less significant oscillations in zonal
wind. Some times, dominant oscillations in mesospheric winds are accompanied by less significant or
no oscillations in EEJ strength. For example, the
dominant 12-day oscillation in zonal wind in May
ARTICLE IN PRESS
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
a clear sky nights and there are 10–14 nights of
observation surrounding each new moon period (see
also, Takahashi et al., 2005), whereas the h’F data
has (usually) 15-min interval on a continuous basis.
For the present analysis we have used h’F values
averaged over 21–24 LT of each night. The result of
Morlet wavelet power spectral analysis performed
on this averaged h’F values are shown in the upper
panel of Fig. 3, wherein shaded zones indicate the
missing data gaps. Due to the relatively discrete
nature of the airglow data its time series was
subjected to Lomb–Scargle (LS) spectral analysis
for each monthly group, and the results are shown
in the lower panels of Fig. 3 (diverging arrows
joining the upper and lower panels indicate the
intervals of the h’F data series for which the airglow
was analyzed). A comparison between the two sets
of results brings out some cases of reasonably good
coincidences in the oscillation periodicities. A 3–5day oscillation in the h’F seen during the interval
9–26 March can be easily identified with the airglow
oscillation showing an 4-day period in both the
OI5577 and OH rotational temperature. During the
interval of 6–19 June a relatively weaker 2–3-day
is accompanied by less significant 14-day oscillation in EEJ, and the enhanced 14–20-day oscillations in zonal winds during February and March are
not accompanied by any signatures in EEJ strength,
which might indicate that the waves might be
getting damped before penetrating to the dynamo
region. It is interesting to note in the results of Fig. 2
that there are more cases of correspondence of EEJ
oscillations with the oscillation periods of zonal
winds, than with those of meridional winds, which
might suggest that the EEJ processes are controlled
more by zonal wind than by meridional winds.
Next we consider some evidences on associated
PW oscillations in the mesospheric airglow emission
and F-layer height in the equatorial region over
Brazil. The upper mesosphere and lower thermosphere night airglow intensities in the OI5577 and
OH(6,2) bands monitored over Cariri (7.51S,
36.51W) and the virtual height of the F-layer
monitored by a Digisonde (Reinisch et al., 1989)
over Sao Luis (2.61S, 44.21W) during the period
from March to November 1999 are used in this
analysis. It should be pointed out that the airglow
data time series has 10-min interval during 6–10 h of
21-LT 24-UT
100
4
50
8
0
OI55
July
40
OH-T
20
0
2 3 4 5 6 7
Period (days)
40
OI55
Nov
OI55
20
0
2 3 4 5 6 7
PSD
PSD
2 3 4 5 6 7
60
40
20
0
Sept
PSD
May
2 3 4 5 6 7
40
OH-T
PSD
PSD
16
Mar-99
60
40
20
0
151
2
PSD
Period (days)
1
513
20
0
2 3 4 5 6 7
Period (days)
OH-T
20
0
2 3 4 5 6 7
Period (days)
Fig. 3. Spectral distribution of the F-layer height oscillations obtained from Morlet wavelet analysis of h’F (F-layer base height) data over
Sao Luis, Brazil, during the interval March–November 1999 (Top panel). The grey areas indicate data gaps. The h’F data represent
averages of 21–24 LT values. Power spectral distribution of the airglow emission intensity at OI5577 (OI55) and OH rotational
temperature (OH-T) are presented in the lower panels.
ARTICLE IN PRESS
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
2.1. PW scale modulations in the evening zonal
electric field and spread F
It is well known that equatorial F-layer height
undergoes a large increase at sunset due to the zonal
electric field enhancement, PRE (Woodman, 1970)
that develops under the action of the F-layer wind
dynamo that intensifies at this time. The height of a
specific plasma frequency can reach values
4300 km and, therefore, its rate of change with
time is a good/reliable indicator of the vertical
plasma drift of the F-layer as discussed in detail by
Bittencourt and Abdu (1981) and Abdu et al.
(1981). Thus the nighttime F-layer height (at a
given plasma frequency) is an approximate representation of the time-integrated vertical drift of the
post-sunset F-region (subject to the recombination
effects that set in for heights o300 km). We have
calculated the vertical drift velocity (eastward
electric field) over Jicamarca using the F-layer
height information available from digisonde data.
The result for the entire month of March 2000
(using the almost uninterrupted half-hourly data) is
presented in Fig. 4 (the day numbers on the x-axis
mark the end of the day in UT, ‘0’ being the
beginning of the day ‘1’). The large increase in the
vertical drift velocity that falls nearly on the day
markings, represents the evening vertical drift
velocity enhancement (PRE) that maximizes near
19 LT (note that the local time (LT) over Jicamarca
is 5 h behind the UT). We note that the vertical drift
velocity peak, Vzp, has relatively large values
typical of the seasonal/equinoctial maxima of a
solar maximum epoch for this location (Fejer et al.,
1991). Large variation in the Vzp amplitude is
observed during this period, the lowest and highest
values attained being 20 and 70 m/s, respectively.
Oscillations of a few days periodicity can be clearly
identified in them. The magnetic activity level, in
terms of the hourly Dst values plotted in the bottom
panel, indicates generally quiet conditions for this
period, as confirmed also by the K p indices that
80
JICAMARCA - March 2000
60
40
Vz (m.s-1)
oscillation in h’F is associated with strong oscillation of the same period in the airglow emission. In
the interval 3–14 September, when the airglow
emission presents strong oscillation of 4–5-day
period the h’F shows a tendency for 4–8-day
oscillation which gets intensified, staying strong
during the later half of September. These results do
suggest associated oscillations in the equatorial
mesosphere and F-region on PW scales.
20
0
-20
-40
-60
Dst (nT)
514
15 - min Vz values
-80
40
0
-40
-80
0
5
10
15
20
Day of the month
25
30
Fig. 4. F-region vertical drift calculated as dhF/dt (hF being the
true height of a specific plasma frequency of the F-layer) using the
true heights at 4 and 5 MHz plasma frequencies obtained from
digisonde data over Jicamarca. The average of velocities obtained
from the two frequencies is plotted in the figure. The lower panel
shows the hourly Dst index variation (downloaded from the
Kyoto WDC for geomagnetic data).
attained slightly higher values only on 1 and 31
March, the corresponding SK p (diurnal sum of the
3-h values) being 28 and 29+, respectively. Thus it
appears evident that the large amplitude oscillations
in the Vzp are unlikely to have been caused by any
disturbance electric field. Consideration of Vzp
variations during some specific intervals such as
2–6, 17–21 and 22–29 March, when the Dst
remained stable and very low, testifies to the above
affirmation. Thus we may attribute these oscillations in the Vzp amplitude to PW-induced effects in
the ionosphere similar to the oscillations in the h’F
presented in Fig. 3.
The Vzp values of March 2000 (from Fig. 4) are
plotted in the top panel of Fig. 5. These were
subjected to Morlet wavelet analysis and the result
is shown in the lower panel of this figure. The cone
of influence due to edge effect in the data series
analyzed is indicated by the thick (parabolic) curves.
Significant amplitude of 3- and 7-day waves is
present during the second week of March and the
wave periods evolve to 2–3-day, 4–5-day and 7–8day towards the third week of the month. These
periodicities are in good agreement with the widely
observed PW periods discussed before (see, e.g.
Forbes, 1996; Forbes and Leveroni 1992; Gurubaran et al., 2001; Haldoupis et al., 2004; Pancheva
et al., 2003, Takahashi et al., 2002). Thus we note
that the PW disturbances are capable of modulating
ARTICLE IN PRESS
515
JICAMARCA
0.075
2
0.075
4
6
8
10
80
1.05 1.351.5 1.20.9
0.75
0.075
0.75 0.9
0.45 0.3
1.6
0.3
0.45
0.75 0.9
1.2
0.8
0.3 0.15
0.075
1.05 1.2
0.6 0.75 0.9
0.4
POWER
Period (days)
Vzp (m.s-1) Morlet mother function
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
1.35
0
60
40
20
1
4
7
10
13
16
19
MARCH 2000
22
25
28
31
Fig. 5. Vzp values (the peak vertical drift velocity) time series for March 2000 over Jicamarca taken from Fig. 4 (upper panel), and the Vzp
oscillation periods obtained by Morlet wavelet spectral analysis (lower panel).
JICAMARCA ESF MAR 2000
Vzp (m.s-1)
UT
UT
33
5.5
30
27
4.5
24
3.5
33
30
27
24
80
2.5
1.5
0.5
60
40
1
20
61
64
2
67
3
70
4
73
76
79
DAY OF THE YEAR
82
5
85
88
91
Fig. 6. Equatorial spread F intensity from ionograms quantified (as a rough indicator) in terms of the range of the echo spreading (one
unit representing 100 km of range spreading) is shown in the top and middle panels (white areas mark missing data). The top panel shows
the echo spreading intensity at the higher frequency range of the F-layer trace and in the middle panel is plotted similar parameter for the
lower frequency range. The bottom panel shows the corresponding Vzp variations. The data set is for March 2000 over Jicamarca.
the intensity of the evening equatorial F-region
vertical drift velocity (zonal electric field).
The evening F-layer uplift due to the PRE, of
which Vzp is the best indicator, is known to be the
main cause for the generation of the post-sunset
ESF/plasma bubble irregularities (see e.g., Abdu
et al., 1983, Fejer et al., 1999). We have plotted in
the top two panels of Fig. 6 the ESF intensity as a
function of UT and day of the year for the same
interval as in Fig. 5 (March 2000). The corresponding Vzp values are plotted in the bottom panel. It
should be mentioned here that we have used as a
rough indicator of the ESF intensity the spread
range of the ESF echoes in the ionograms, one unit
of intensity being represented by 100 km of range
spreading around the F-layer trace. The middle
panel shows the spread F intensity at the lower
frequency limit of the ionogram which indicates the
first onset of the bottomside spread F. The onset of
SF at the higher frequency range that follows in
sequence (with some time delay, usually of
15–30 min) marks the evolution towards topside
ARTICLE IN PRESS
516
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
bubble formations when the intensity (spread range)
generally increases, as can be verified from the plots
in the top panel. We note that there is significant
control of the SF intensity by the Vzp amplitude,
higher intensity of the SF, in general, corresponding
to larger amplitude of the Vzp. For example, during
two of the lowest values of Vzp identified by the
vertical lines 2 and 5, corresponding to VzpE20 m/
s, the SF was nearly totally absent, and during three
of the higher Vzp values identified by the vertical
lines 1, 3 and 4, which corresponded to
VzpE65–70 m/s, the SF was clearly more intense.
Of course a closer examination of this figure would
show some important exceptions to a direct
relationship between the two parameters, of which
we will be discussing later. Independent of the
detailed nature of the relationship between Vzp and
ESF intensity we may affirm that PW scale
oscillations are clearly present in the evening PRE
(vertical drift) which in turn modifies/controls the
formation and intensity of the ESF.
3. Discussion
We have presented observational results in
evidence of vertical coupling processes involving
PW scale oscillations in the mesosphere E- and Fregions of the equatorial atmosphere–ionosphere
system. The results call our attention to two
important aspects of the coupling processes underlying these results: (1) Mesosphere E-region dynamic coupling (MEDC) through upward
propagation of PW and (2) E- and F-region
electrodynamic coupling (EFEC) that produces
manifestations at PW oscillation periods.
Both these aspects pose an important question as
to the accessibility of the PW to the ionospheric
dynamo region and to higher levels. In the case of
the MEDC, the penetration of the PW into the Eregion could result in dynamo action by the
associated winds, leading to the generation of an
electric field responsible for the effects observed in
the E-regions, such as the PW scale oscillations in
DH observed in Figs. 1 and 2. However, the existing
model calculations do not seem to support direct
penetration of PW into the E-region capable of
inducing wind amplitudes required for a viable
dynamo mechanism (Forbes, 1996; Hagan et al.,
1993; Forbes et al., 1995). Further, recent analysis
results by Pancheva et al. (2003) on MF radar winds
seem to show a certain amplitude attenuation in
meridional wind with height increase in the
91–103 km region. Several mechanisms have been
proposed to explain the PW period oscillations
observed in the dynamo region (100–170 km) and
at higher levels of the ionosphere. They include, as
explained by Forbes (1996), (a) stratosphere/mesosphere PW modulation of the accessibility of gravity
waves to the upper levels, involving mechanisms
that may lead to secondary source of PW excitation
at higher levels and (b) PW modulation of the
upward propagating tides that participate in the
dynamo action to generate electric fields at higher
levels. The latter mechanism seems to be especially
promising and more easily verifiable, in view of the
fact that the forcing by upward propagating diurnal
and semidiurnal tides is basically responsible for the
phenomenology of the quiet time ionosphere,
especially over equatorial latitudes. In fact, evidence
on the influence of PW modulated atmospheric tides
of diurnal and semi-diurnal periodicities on midlatitude sporadic E-layer formation has been
provided recently by Pancheva et al. (2003). Our
results in Figs. 1 and 2 showing cases of associated
planetary scale oscillations in DH and mesospheric
winds (especially zonal winds) can be examined in
the light of this discussion. The DH represents the
diurnal range of the magnetic field H-component
variation due to the EEJ, which is a direct measure
of the product of dynamo electric field intensity and
electron density/conductivity at E-layer heights.
Assuming that the PW oscillations in the E-layer
electron density are small (there is no independent
evidence of such oscillations so far) we attribute the
periodicities observed in DH to the corresponding
oscillations in the dynamo electric field, which is
produced dominantly by a diurnal tidal wind
dynamo. In view of the observation by Pancheva
et al. (2003) that mid-latitude Es-layer formation is
affected indirectly by PW through their modulation
of the diurnal and semidiurnal tides, we may expect
that the equatorial DH oscillations at PW periods
may arise in the dynamo region also through the
action of PW modulated tidal winds. This point
needs to be checked, however, from further analysis
of the data. In any case, a more direct role of PW in
the dynamo region cannot be ruled out either,
without examining additional concurrent data of
winds in the equatorial dynamo region.
As regards the aspect of the EFEC mentioned
above, the PW scale oscillations in the F-region
parameters that result from the EFEC processes
have important implications to our understanding
of the causes of the widely observed day-to-day
ARTICLE IN PRESS
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
where Uy is thermospheric zonal wind
0 is the
P and BP
Earth magnetic field intensity, and F and E are
the integrated conductivities, respectively, of the Eand F-regions (Abdu et al., 2003). P
Due to the faster
decay
of
the
E-layer
conductivity,
E , as compared
P
to F , in the post-sunset hours, E z tends to increase
towards the nightside. The application of a curl-free
condition to such an electric field could lead to the
enhanced zonal (eastward followed by westward)
electric field as proposed originally by Rishbeth
(1971) and modeled by Eccles (1998) (for other
different approaches used in modeling this problem
see e.g., Farley et al., 1986; Batista et al., 1986;
Haerendel et al., 1992; Crain et al., 1993, Fesen et
al., 2000). Thus it appears evident that the
magnitudes of the zonal wind as well as that of
the longitudinal conductivity gradient across the
terminator could control the intensity of the PRE.
Therefore, the PW oscillations observed in the PRE
(Figs. 4 and 5) could be caused by such oscillations
either in the intensity of the thermospheric zonal
wind or in the degree of the conductivity longitudinal gradient. From the discussion on the
vertical penetration of the PW presented above it
looks less likely that significant oscillations in the
neutral winds at thermospheric height could be
directly induced by the PW. There is a possibility of
PW modulation of the solar diurnal thermal tide
that is mainly responsible for the winds at higher
levels of the thermosphere, which is difficult to
verify however. A more likely source of these
oscillations may be sought in the possible dependence of the E-layer conductivity longitude/LT
gradient on the lower thermosphic zonal wind as
explained below. Fig. 7 taken from Abdu et al.
(2003) shows the E-layer integrated Pedersen conductivity as a function of time calculated for diurnal
tidal zonal winds of different amplitudes ranging
from 0 to 100 m/s. The time t ¼ 0 is adjusted to
coincide with 18 LT, close to local sunset (for details
of the calculation see Abdu et al., 2003). We note
that the LT gradient (which is equivalent to
longitudinal gradient), positive towards dayside,
increases significantly with increase of the zonal
wind amplitude in the E-region (90–140 km). On the
basis of this dependence we may expect that the
PRE amplitude which depends upon the longitudinal gradient in the E-layer integrated conductivity (as a dominant controlling factor) could
undergo fluctuations due to corresponding oscillations in the amplitude of the E-layer zonal wind of
diurnal or semi-diurnal tidal mode. The amplitude
of the tidal wind itself could be modulated by PW
through nonlinear interaction, as mentioned before.
Here again we cannot rule out the possibility of a
direct role of the PW in causing the zonal wind
fluctuation of the required magnitude at E-layer
heights.
The aspect (b) concerns the day-to-day variability
in the ESF occurrence and intensity due to
oscillations in the PRE amplitudes as one of its
causes. It has been known from previous studies
integrated Pedersen conductivity
1.0
Integrated conductivity, S
variability in the equatorial F-region evening
vertical drift (PRE) and the related spread F
development processes. This problem is complex
enough that a detailed discussion is not attempted
here. For a brief discussion we may simplify this
problem into the two following aspects: (a) PW
scale oscillations in the evening PRE/vertical drift
and (b) ESF variability arising from the oscillations
in the PRE as one of the causes.
Regarding the aspect (a), it is known that the
equatorial evening electric field enhancement arises
from the F-layer wind dynamo, which involves
electrical and electrodynamical coupling of the Eand F-regions of the equatorial ionosphere (Rishbeth, 1971; Heelis et al., 1974). Briefly, the thermospheric zonal wind (eastward in the evening hours)
in the presence of the strong longitudinal/LT
gradient in the E-layer conductivity (caused by the
conductivity decay towards night side of the
terminator) produces by dynamo action vertical
(downward) electric field increasing towards the
night side, as can be verified with the help of the
following equation:
hX .X
X i
E z ¼ U y B0
þ
,
(1)
F
F
E
517
0.8
0 m/s
0.6
100
0.4
50
0.2
20 m/s
0.0
0
2
4
6
time, hours
Fig. 7. One-dimensional simulation results of height integrated
E-layer Pedersen conductivity in unit of Siemens (S) as a function
of time for zonal (westward) wind amplitudes 0, 20, 50 and
100 ms1.
ARTICLE IN PRESS
518
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
that ESF can develop and increase in intensity for
Vzp amplitude of 20 m/s and higher (see for
example, Abdu et al., 1983; Fejer et al., 1999). This
relationship is evident also in the plots of these two
parameters shown in Fig. 6. The ESF development
and intensity depend also on other conditions
besides a direct dependence on the PRE amplitude.
To understand its variability we need to examine the
linear growth rate, g, of the ESF/plasma bubble
instability by the generalized Rayliegh–Taylor
mechanism that uses the background ionospheric
parameters expressed in terms of their values
integrated along the magnetic flux tube in which
the instability begins to develop at the bottom-side
gradient region of a rapidly rising evening F-layer.
Based on flux tube integrated quantities, as proposed by Haerendel (1973, unpublished report)
(also, Haerendel et al., 1992), a generalized form
of the linear growth rate derived by Sultan et al.
(1996) is given by
gFT ¼
SFP
E
g
1
P
U
þ
bFT .
FT
neff LFT
SEP þ SFP B
(2)
Here, SE;F
is the field-line integrated conductivities
P
for the E- and F-region segments of a field line; U PFT
is the conductivity weighted flux tube integrated
vertical wind; b is the recombination loss rate; neff is
the effective ion-neutral collision frequency. The
subscript FT in Eq. (2) stands for flux tube
integrated quantity. The influence of the different
terms in the ESF variability has been discussed
briefly in Abdu (2001). We note from Eq. (2),
especially, that larger flux tube integrated conductivity (the denominator of the conductivity term)
reduces the instability growth rate. Even if gFT is
sufficiently positive for the initiation of the instability the ensuing nonlinear growth of the
instability into developed topside bubble structures
can be retarded or even totally inhibited depending
upon the magnitude of the flux tube integrated
conductivity. It has been shown by Maruyama
(1988) based on model calculations that transequatorial winds (that blow from summer to winter
hemisphere) can cause significant increase of the
flux tube integrated conductivity. Its effect will be to
reduce instability growth rate as well as to inhibit
the polarization electric field responsible for the
nonlinear development of the plasma bubbles. Thus
possible day-to-day variability in the intensity of
trans-equatorial winds could be a significant source
of ESF day-to-day variability although this effect
has not been adequately established/investigated so
far (see e.g. Mendillo et al., 2001; Abdu, 2001).
Another source of variability for the field line
integrated conductivity seems to reside in sporadic
E-layers when present at the extremities of the flux
tubes that participate in the ESF development
processes which include also the development of
the PRE. Model calculations have shown that
increase of field line integrated conductivities of
sufficient magnitude to affect the instability growth
rate can occur due to the presence of Es-layers of
significant intensity (Stephan et al., 2002). However,
observational evidence on the effect of Es-layers on
ESF development so far available lead to diverging
conclusions. For example, while an isolated case of
ESF inhibition in the presence of sporadic E-layer
over the equatorial station Jicamarca has been
reported by Stephan et al. (2002) an exactly
opposite effect from the statistical data set of ESF
enhancement in the presence of sporadic E-layers
over an equatorial anomaly (low latitude) station,
Chung-Li, has been shown by Bowman and
Mortimer (2003). Such apparently ambiguous/contradictory manifestations of the Es-layer–ESF
relationship can be understood in the light of a
direct relationship between the PRE development
and the evening Es-layer formation conditions that
has been recently demonstrated by Abdu et al.
(2003). It was shown based on analysis of extensive
data sets that Es-layer formation in the evening
hours over a sub-equatorial station (such as
Fortaleza, Brazil) can be disrupted during the
development of the PRE of significant amplitude.
An example of this result (taken from Abdu et al.,
2003) is presented in Fig. 8 in which the right panel
shows the interruption of an ongoing Es-layer at the
time of the maximum F-layer vertical drift (zonal
electric field) near sunset, the Es-layer formation
resuming after 2–3 h. When the vertical drift is
relatively less intense such disruption of the Eslayers does not occur, as can be seen in the results
plotted in the left panel. This was explained (by
Abdu et al., 2003) in terms of the effect of a vertical
electric field in causing vertical plasma transport at
E-layer heights that can act, depending upon the Efield direction, to retard or enhance the vertical ion
convergence driven by a wind/wind shear mechanism that is basically responsible for the Es-layer
formation. An upward-directed vertical electric field
can disrupt the vertical ion convergence while a
downward electric field can enhance it (see Abdu
et al., 2003, for further details). The vertical
ARTICLE IN PRESS
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
519
FORTALEZA, BRAZIL (2001)
h'F (5MHz) (km)
2001-Nov.2,4,6,8,
9,10,12,14,18,20,
23,25,27,29,
Dez.3,4,7,8,12
2001-Mar.6,10,
12,18,20,21,22,
23,24,25,28,29,
31,Apr.3,4
600
500
400
300
Vz (m.s-1)
200
Vzp=37m.s-1
40
Vzp=62m.s-1
0
h'Es (km)
ftEs (MHz)
-40
12
8
4
0
200
160
120
80
9
12
15
18
21
0
3
6
9 12
Local Time
15
18
21
0
3
6
9
Fig. 8. F-layer height versus Es-layer relationship using digisonde data over Fortaleza for March–April and November–December 2001.
Left panels: the case of post-sunset Es-layer non-disruption; and right panels: the case of Es-layer disruption. (Note the large difference
between the values of Vzp for the two cases).
structure of the evening vertical electric field over
the equator, which is associated with the PRE, as
modeled by Haerendel et al. (1992) consists of an
upward directed field in the height region
90–300 km which reverses to a downward directed
field above 300 km (see Haerendel et al., 1992, for
further details). When this electric field is field-line
mapped to off-equatorial latitudes we have an
upward-directed electric field over Fortaleza
(3.91S, 38.451W, dip angle: 91) and a downwarddirected electric field at a location farther away in
latitude, such as Cachoeira Paulista (22.61S, 3151E;
dip angle:281). As explained by Abdu et al. (2003)
this situation leads to disruption of Es-layer
formation over Fortaleza while uninterrupted, or
even enhanced, Es-layer formation over Cachoeira
Paulista during the PRE development. Results in
support of this explanation are presented in Fig. 9,
which shows, as an example, a few days of
observations during December 1988 of simultaneous Es-layer variations over Fortaleza and
Cachoeira Paulista, together with the F-layer height
and the associated vertical drift variations over the
former station. It is clear that Es-layer disruption
over Fortaleza is accompanied by its formation over
Cachoeira Paulista when the PRE development was
at its peak over the former station, which is in
complete agreement with the vertical electric field
control of the Es-layer just explained above. We
have verified that ESF was present over Fortaleza
on all these nights as expected for Vzp amplitudes
420 m/s, in agreement with the results of Fig. 6 as
well. Thus we note that the nature of the association
between the ESF and Es-layer occurrences to be
expected in observational data would depend upon
the latitude region for which the data analysis is
performed. Further, it appears that a possible effect
on ESF development and intensity due to an
ARTICLE IN PRESS
520
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
FORTALEZA-December 1988
495
h´F
h´F (km)
420
345
270
Vz (m.s-1)
195
120
40
Vz
0
-40
20
ftEs (MHz)
ftEs
15
10
5
0
h´Es (km)
180
h´Es
160
140
120
100
80
ftEs (MHz)
10
ftEs
8
6
4
2
h´Es (km)
180
h´Es
160
140
120
100
80
9
12
15
18
21
LOCAL TIME
0
3
6
9
Fig. 9. F-layer height h’F, vertical drift velocity Vz, and Es-layer parameters, f t Es and h0 Es, over Fortaleza, plotted in the top four panels
and the Es-layer parameters over Cachoeira Paulista in the bottom two panels.
enhanced field line-integrated conductivity arising
from the mere presence of Es-layers is a challenging
question warranting experimental/observational
verification. In any attempt to verify this point it
is necessary, as a minimum requirement, that the
Es-layer, to be associated with ESF occurrence/
intensity, be located in common areas in the
conjugate E-layers at the feet of a potentially
unstable flux tube. Further, the direct association
(in an interactive manner) that exists between the
ARTICLE IN PRESS
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
PRE and the Es-layer, as shown by Abdu et al.
(2003), needs to be isolated in order to establish any
possible effect on the ESF arising from the mere
presence of the Es-layers. These considerations
show that while the PW scale oscillations in the
PRE amplitude is an important factor contributing
to the day-to-day variability in the ESF, other
factors such as the meridional/ trans-equatorial
winds, field line integrated conductivities, etc.,
mentioned before, as well as the Es-layer connections, also contribute to its variability in ways
difficult to evaluate from limited data bases.
4. Conclusions
The following conclusions may be noted from the
present study: PW scale oscillations of different
periodicities with episodic nature occur simultaneously in the mesospheric winds and EEJ intensity.
It is not clear from the present analysis if the EEJ
effects are due to direct penetration of the PW up to
the dynamo region or indirectly by a mechanism of
tidal wave modulation by PW interaction at lower
height as found to be operative in the case of midlatitude Es-layer responses. Nighttime F-layer
height oscillations at PW periodicities are found to
be associated with such oscillations in the mesospheric airglow intensity and OH rotational temperature. In this case the F-region oscillations
represent the time-integrated effect of the evening
vertical drift (zonal electric field) due to F-region
dynamo. For the first time we have provided
evidence for oscillations at PW periods in the
equatorial evening zonal electric field believed to
result from the combined action of thermospheric
zonal wind and the evening longitudinal gradient in
the E-layer Pedersen conductivity. Considerations
on the relevant electrodynamic processes suggest
that PW oscillations in zonal winds at E-layer
height, rather than such oscillations in the zonal
winds at F-region height, could be a more likely
cause of the observed PW scale oscillations in the
evening zonal electric field /F-region vertical plasma
drift. A direct consequence of the PW scale
oscillations in the evening electric field is its role in
the quiet time day-to-day variability of the ESF/
plasma bubble occurrence and intensity. However,
other factors, such as the relationship between the
PRE field and sporadic E-layer formation in the
evening hours, to mention as an example, should be
considered in any investigations aiming for a better
understanding and evaluation of the causes of the
521
widely observed day-to-day variability in the ESF
phenomenon.
Acknowledgements
The authors wish to acknowledge the supports
from the Fundac- ão de Amparo a Pesquisa do
Estado de São Paulo- FAPESP through the projects
1999/00437-0, 04/07695-5 and 05/52654-8, Conselho
Nacional de Desenvolvimento Cientifico e Tecnologico- CNPq through Grants no 502804/2004-1.
We thank the Jicamarca Radio Observatory for the
use of their Digisonde data that we have retrieved
from the UML DIDBase. BWR was supported by
AF Grant F19628-02-C-0092. We thank Maria
Goreti dos Santos Aquino for assistance in processing the digisonde data and in preparing some of the
figures.
References
Abdu, M.A., 2001. Outstanding problems in the equatorial
ionosphere-thermosphere electrodynamics relevant to spread
F. Journal of Atmospheric and Terrestrial Physics 63,
869–884.
Abdu, M.A., Bittencourt, J.A., Batista, I.S., 1981. Magnetic
declination control of the equatorial F region dynamo field
development and spread-F. Journal of Geophysical Research
86, 11443–11446.
Abdu, M.A., Medeiros, R.T., Bittencourt, J.A., Batista, I.S.,
1983. Vertical ionization drift velocities and range type spread
F in the evening equatorial ionosphere. Journal of Geophysical Research 88, 399–402.
Abdu, M.A., MacDougall, J.W., Batista, I.S., Sobral, J.H.A.,
Jayachandran, P.T., 2003. Equatorial evening pre reversal
electric field enhancement and sporadic E layer disruption: a
manifestation of E and F region coupling. Journal of
Geophysical Research 108 (A6), 1254.
Batista, IS., Abdu, M.A., Bittencourt, J.A., 1986. Equatorial F
region vertical plasma drift: seasonal and longitudinal
asymmetries in the American sector. Journal of Geophysical
Research 91, 12055–12064.
Bittencourt, J.A., Abdu, M.A., 1981. A theoretical comparison
between apparent and real vertical ionization drift velocities
in the equatorial F region. Journal of Geophysical Research
86, 2451–2455.
Bowman, G.G., Mortimer, I.K., 2003. Spread–F/sporadic E
coupling at Chung-Li, especially for postsunset periods of
sunspot maximum years. Journal of Geophysical Research
108 (A4), 1148.
Brown, G.M., Williams, D.C., 1971. Journal of Atmospheric and
Terrestrial Physics 33, 1321.
Chen, P.R., 1992. Two-day oscillations of the equatorial
ionization anomaly. Journal of Geophysical Research 979
(A5), 6343–6357.
Clark, R.R., Burrage, M.D., Franke, S.J., Manson, A.H., Meek,
C.E., Mitchell, N.J., Muller, H.G., 2002. Observations of
ARTICLE IN PRESS
522
M.A. Abdu et al. / Journal of Atmospheric and Solar-Terrestrial Physics 68 (2006) 509–522
7-day planetary waves with MLT radars and UAR/HRDI
instrument. Journal of Atmospheric and Terrestrial Physics
64, 1217–1227.
Crain, D.J., Heelis, R.A., Bailey, G.J., Richmond, A.D., 1993.
Low latitude plasma drifts from a simulation of the global
atmospheric dynamo. Journal of Geophysical Research 98,
6039–6046.
Farley, D.T., Bonelli, E., Fejer, B.G., Larsen, M.F., 1986. The
prereversal enhancement of the zonal electric field in the
equatorial ionosphere. Journal of Geophysical Research 91,
13723–13728.
Eccles, J.V., 1998. Modeling investigation of the evening
prereversal enhancement of the zonal electric field in the
equatorial ionosphere. Journal of Geophysical Research 103,
26,709–26,719.
Fejer, B.G., de Paula, E.R., Gonzales, S.A., Woodman, R.F.,
1991. Average vertical and zonal plasma drifts over Jicamarca. Journal of Geophysical Research 96, 13901–13906.
Fejer, B.G., Scherliess, L., de Paula, E.R., 1999. Effects of the
vertical plasma drift velocity on the generation and evolution
of equatorial spread F. Journal of Geophysical Research 104,
19854–19869.
Fesen, C.G., Crowley, R.G.R., Richmond, A.D., Fejer, B.G.,
2000. Simulation of the pre-reversal enhancement in the low
latitude vertical ion drifts. Geophysical Research Letters 27,
1851–1854.
Forbes, J.M., 1996. Planetary waves in the thermosphere–ionosphere system. Journal of Geomagnetism and Geoelectricity
48, 91–98.
Forbes, J.M., Leveroni, S., 1992. Quasi 16-day oscillations in the
ionosphere. Geophysical Research Letters 19, 981–984.
Forbes, J.M., Hagan, M.E., Miyahara, S., Vial, F., Manson,
A.H., Meek, C.E., Portnyagin, Y., 1995. Quasi-16-day
oscillation in the mesosphere and lower thermosphere.
Journal of Geophysical Research 100, 9149–9161.
Gurubaran, S., Sridharan, S., Ramkumar, T.K., Rajaram, R.,
2001. The mesospheric quasi 2-day wave over Tirunelveli.
Journal of Atmospheric and Terrestrial Physics 63, 975–985.
Haerendel, G., Eccles, J.V., C
- akir, S., 1992. Theory of modeling
the equatorial evening ionosphere and the origin of the shear
in the horizontal plasma flow. Journal of Geophysical
Research 97, 1209–1223.
Haris, T.J., Vincent, R.A., 1993. The quasi-two-day-wave
observed in the equatorial middle atmosphere. Journal of
Geophysical Research 98, 10481–10490.
Hagan, M.E., Forbes, J.M., Vial, F., 1993. Numerical investigation of the propagation of the quasi-two-day wave into the
lower thermosphere. Journal of Geophysical Research 98
(D12), 23193–23205.
Harris, T.J., 1994. A long-term study of the quasi-two-day wave
in the middle atmosphere. Journal of Atmospheric and
Terrestrial Physics 56, 569–579.
Heelis, R.A., Kendall, P.C., Moffet, R.J., Windle, D.W.,
Rishbeth, H., 1974. Electrical coupling of the E and F
regions and its effect on the F region drifts and winds.
Planetary and Space Science 22, 743–756.
Haldoupis, C., Pancheva, D., 2002. Planetary waves and
midlatitude sporadic E layers: strong experimental evidence
for a close relationship. Journal of Geophysical Research 107
(A6), 1078.
Haldoupis, C., Pancheva, D., Mitchell, N.J., 2004. A study of
tidal and planetary wave periodicities present in midlatitude
sporadic E layers. Journal of Geophysical Research 109,
A02302.
Maruyama, T., 1988. A diagnostic model for equatorial spread F
1. Model description and application to electric field and
neutral wind effects. Journal of Geophysical Research 93,
14611–14622.
Mendillo, M., Meriwether, J., Biondi, M., 2001. Testing the
thermospheric neutral wind suppression for day-to-day
variability of equatorial spread F. Journal of Geophysical
Research 106 (A3), 3655–3663.
Pancheva, D., Haldoupis, C., Meek, C.E., Manson, A.H.,
Mitchell, N.J., 2003. Evidence of a role for modulated
atmospheric tides in the dependence of sporadic E layers on
planetary waves. Journal of Geophysical Research 108 (A5),
1176.
Pancheva, D., Mitchell, N., Manson, A.H., et al., 2004.
Variability of the quasi-2-day wave observed in the MLT
region during the PSMOS campaign of June–August 1999.
Journal of Atmospheric and Terrestrial Physics 66 (6–9),
539–566.
Parish, H.F., Forbes, J.M., Kamalabadi, F., 1994. Planetary
wave and solar emission signatures in the equatorial electrojet. Journal of Geophysical Research 99 (A1), 355–368.
Reinisch, B.W., Bibl, K., Kitrosser, D.F., Sales, G.S., Tong, J.S.,
Zhang, Z.–M., Bullett, T.W., Ralls, J.A., 1989. The digissonde 256 ionospheric sounder. In: Liu, C.H. (Ed.), World
Ionosphere/Thermosphere Study, WITTS Handbook, vol. 2,
SCOSTEP, IL, pp. 350–366.
Rishbeth, H., 1971. Polarization fields produced by winds
in the equatorial F region. Planetary and Space Science 19,
357–369.
Stephan, A.W., Colerico, M., Mendillo, M., Reinisch, B.W.,
Anderson, D., 2002. Suppression of equatorial spread F by
sporadic E. Journal of Geophysical Research 107 (A2),
101029–101034.
Sultan, P.J., 1996. Linear theory and modeling of the Rayleigh–Taylor instability leading to the occurrence of equatorial spread F. Journal of Geophysical Research 101,
26875–26891.
Takahashi, H., Lima, L.M., Wrasse, C.M., Abdu, M.A., Batista,
I.S., Gobbi, D., Buriti, R.A., Batista, P.P., 2005. Evidence on
2–4 day oscillations of the equatorial ionosphere h’F and
mesospheric airglow emission. Geophysical Research Letters
32, L12102, doi:10.1029/2004GL022318.
Takahashi, H., Buriti, R.A., Gobbi, D., Batista, P.P., 2002.
Equatorial planetary wave signatures observed in mesospheric
airglow emissions. Journal of Atmospheric and Terrestrial
Physics 64, 1263–1272.
Vincent, R.A., 1993. Long-period motions in the equatorial
mesosphere. Journal of Atmospheric and Terrestrial Physics
55, 1067–1080.
Woodman, R.F., 1970. Vertical drift velocities and East–West
electric filed at the magnetic equator. Journal of Geophysical
Research 75, 6249–6259.