Sunteți pe pagina 1din 13

RESEARCH ARTICLE On Deriving Gravity Wave Characteristics in the Daytime

10.1029/2019JA026723
Upper Atmosphere Using Radio Technique
Special Section: Subir Mandal1,2 , Duggirala Pallamraju1 , Deepak K. Karan1 , Kedar A. Phadke1,3 ,
Equatorial Aeronomy: New
results from the 15th Ravindra P. Singh1 , and Pradip Suryawanshi1
International Symposium on 1
Equatorial Aeronomy (ISEA‐ Physical Research Laboratory, Ahmedabad, India, 2Department of Physics, Indian Institute of Technology, Gandhinagar,
15) and beyond India, 3Now at Department of Astronomy, University of Illinois at Urbana‐Champaign, Urbana, IL, USA

Key Points: Abstract The importance of neutral wave dynamics in the understanding of the upper atmospheric
• New approach using digisonde for
determining gravity wave time
processes is well known. Conventionally, optical methods are used to derive information on the neutral
periods, vertical propagation speeds, wave dynamics by obtaining gravity wave (GW) characteristics. Optical measurement techniques use
and scale sizes is presented airglow emissions as tracers to obtain such information that correspond to altitudes from where the
• Gravity wave time periods from
digisonde match with those
emissions emanate. However, in this paper, we describe a method using radio wave measurement technique
obtained using optical (digisonde) to obtain information on the neutral GW behavior. It involves monitoring of variations in the
measurements, indicating the heights of isoelectron densities as a function of time, and their phase shifts, if any, to derive vertical
presence of a common driver
• This approach opens up new
propagation speeds and scale sizes of GWs. The daytime values of GW time periods, vertical phase speeds,
possibilities for obtaining gravity and vertical scale sizes obtained for the duration of 16–21 May 2015 are in the range of 1.47 ± 0.05 to 2.64 ±
wave characteristics in all weather 0.07 hr, 30.06 ± 4.35 to 45.69 ± 11.84 m/s, and 183.21 ± 39.23 to 393.07 ± 66.38 km, respectively. Further, we
conditions
have used the GW dispersion relation to make a first‐order estimation of the horizontal scale sizes. This
Supporting Information:
method of deriving neutral GW characteristics through radio measurement technique is effective for the
• Supporting Information S1 daytime conditions and opens up new possibilities of investigations of the wave dynamical behavior in the
• Data Set S1 upper atmosphere during all weather conditions.
• Data Set S2
• Data Set S3
• Data Set S4
• Data Set S5
• Data Set S6
1. Introduction
• Data Set S7 Upper atmospheric oscillations are present in a range of scale sizes and periodicities. While periodicities of
• Data Set S8
• Data Set S9 planetary scale oscillations are of the order of several days and spread over lengths of several thousands of
• Data Set S10 kilometers, the oscillations of diurnal scales of periodicities of 24 hr and its subharmonics (12, 8, and 6 hr)
• Data Set S11 have scale lengths of several thousands of kilometers. Further, the upper atmosphere also experiences the
forcing due to the lunar rotation around the Earth with orbital time period of 28.4 days (and its
subharmonics). Sometimes this forcing is intertwined with the planetary scale oscillations in the Earth's
Correspondence to:
D. Pallamraju, atmosphere. While all these dynamics describe the large‐scale behavior of the Earth's upper atmospheric
raju@prl.res.in oscillations and their effects that span over large distances, the class of GWs whose effects extend to short dis-
tances from the place of their origin, play important roles in modifying the atmospheric processes around the
Citation: source regions within a couple of thousands of kilometers. In general, GWs are characterized as those oscilla-
Mandal, S., Pallamraju, D., Karan, D. tions in the atmosphere whose scale sizes extend from several tens to a few thousands of kilometers and of
K., Phadke, K. A., Singh, R. P., & periodicities smaller than the quarterdiurnal component of atmospheric tides (6 hr). After the existence of
Suryawanshi, P. (2019). On deriving
gravity wave characteristics in the GWs in the atmosphere was postulated by Hines (1960), there have been several experiments using optical,
daytime upper atmosphere using radio radio, and magnetic measurements from ground, balloon, and spaceborne instruments to characterize them.
technique. Journal of Geophysical GWs are known to be generated due to changes in orography (e.g., Alexander, 1996), convective activity (e.g.,
Research: Space Physics, 124. https://
doi.org/10.1029/2019JA026723 Singh & Pallamraju, 2016), presence of sharp longitudinal gradients, as in the case of solar terminator (e.g.,
Forbes et al., 2008), and wind shears (Pallamraju et al., 2014; Pramitha et al., 2015). Further, auroral processes
Received 20 MAR 2019 in the high latitudes (Hocke & Schlegel, 1996) and equatorial electrojet in the low and equatorial latitudes
Accepted 10 JUN 2019 (Raghavarao et al., 1988) are known to generate GWs. Electron density irregularities moving at supersonic
Accepted article online 26 JUN 2019
speeds in the nighttime F region also generate GWs (e.g., Hunsucker, 1982). Various features and conse-
quences of the GWs generated by any of the processes as mentioned above have been studied through mod-
eling and simulations (e.g., Fritts & Vadas, 2008; Hocke & Schlegel, 1996; Vadas, 2007).
There exist several naturally occurring airglow emissions that emanate at different altitudes of the Earth's
©2019. American Geophysical Union.
upper atmosphere. The emission rates of airglow depend on the densities of the reactants that produce
All Rights Reserved. the excited atomic and molecular states. Also, several of the reaction rates are temperature dependent. As

MANDAL ET AL. 1
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

propagation of any wave disturbs the density and temperature of the reactants, by passively sensing the var-
iations in naturally occurring airglow emissions as a function of time, information on wave movement can
be obtained. Optical and incoherent scatter radar measurements can be used to derive information on GWs.
However, being relatively inexpensive and easily portable, optical measurement techniques have been con-
ventionally used for obtaining GW characteristics (time periods, scale sizes in all the dimensions, propaga-
tion speeds, propagation directions, etc.) in the mesosphere and the thermosphere. In addition to airglow
intensities, some of the optical measurements can yield information on the variation of temperatures, and
winds, which also have been used to obtain information on GW characteristics. While GW periodicities
can be obtained from a single point measurement (e.g., Chakrabarty et al., 2008; Pallam Raju et al., 1996),
large field‐of‐view optical measurements are required for deriving information on scale sizes and propaga-
tion directions of the GWs (e.g., Karan & Pallamraju, 2017, 2018; Lakshmi Narayanan et al., 2010;
Pallamraju et al., 2013, 2014, 2016; Shiokawa et al., 2009).

Large‐scale motions, such as traveling ionospheric disturbances that are usually generated during geomag-
netic storms, have been studied using radio measurements by monitoring the time variations of ionospheric
parameters such as ionospheric electron content (IEC), base height of ionosphere (h′F), and critical F2 layer
frequency (foF2; e.g., Hunsucker, 1982; Rastogi & Klobuchar, 1990; Mendillo, 2006; Rama Rao et al., 2006).
IEC is obtained essentially by integrating the measured electron densities on the bottomside of the iono-
sphere as possible by a digisonde, whereas for the topside, an exponential decaying nature above peak F2
layer height (hmF2) is considered with a peak electron density corresponding to that of foF2. In the past cou-
ple of decades due to the availability of several ground‐based GPS receivers, GPS‐based total electron content
(TEC) measurements are being used copiously for the understanding of different processes of the ionosphere
(e.g., Lin et al., 2007; Tulasi Ram et al., 2012; Yadav & Pallamraju, 2015).

While the radio measurements of ionospheric parameters, such as the base height, peak ionospheric fre-
quency, and IEC, can be used for the investigations of wave‐like phenomena, the information obtained is
generally limited to time periods of the order of tidal scales or their harmonics. This is because the iono-
sphere moves up or down as a bulk during tidal scale oscillations and so it is reflected in all the ionospheric
parameters at all heights simultaneously, and hence, the presence of such oscillation(s) shows up clearly in
the spectral analysis. However, smaller‐scale oscillations (less than 3 hr) at the ionospheric altitudes can be
submerged in the waves due to larger time periods. Even in the spectral analysis, the smaller timescale oscil-
lations may not have enough power to show up beyond the false alarm limit (FAL; described in section 3.2
and can be seen in Figure 2b). Further, these periodicities correspond to the movement of ionosphere as a
whole; therefore, vertical propagation characteristics of GWs cannot be inferred. The situation is even worse
in the case of TEC data that are obtained by GPS as in that measurement both the ionospheric and plasma-
spheric electron contents are integrated. Thus, especially during geomagnetic storms, there exists a greater
ambiguity in the relative contribution of the plasmasphere versus the ionosphere in the GPS‐derived TEC
data (Ravindran et al., 2014). It is therefore important to consider the variation of ionospheric parameters
at different heights independently and analyze them for the existence of smaller‐scale periodicities. One of
the initial attempts in this regard was carried out by Klausner et al. (2009) wherein variations in virtual
heights for six fixed transmission frequencies were obtained which showed quasiperiodic variations.
However, being variations in virtual heights, ambiguity exists in the values of the derived phase speeds, time
periods, and vertical scale sizes of GWs.
In the present work, we report on the results obtained after rigorous analyses of digisonde data for the inves-
tigations of GWs. The data are analyzed to obtain real height profiles of electron densities. Comparison is
made between the GW time periods obtained from the analysis of the data obtained by radio measurements
and those from the well‐established optical daytime airglow emissions to investigate the relation, if any.
Simulations have been carried out using the GW dispersion relation to estimate the horizontal scale sizes
that vary as a function of time period, ambient wind speeds, and their directions. This method, as described
in this work, opens up new possibilities of deriving GW parameters globally at different locations as digi-
sondes are available aplenty as compared to optical measurement techniques. Also, investigations can be
carried out round the year including during monsoon seasons as radio measurement techniques are not hin-
dered by the opacity of clouds, as in the case of optical measurements. Thus, this method of deriving GW
parameters fills in much needed gap in the investigations of neutral wave dynamics.

MANDAL ET AL. 2
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

2. Experimental Techniques Used


2.1. Digisonde
A Digital Portable Sounder (DPS‐4D) digisonde is operational from
Ahmedabad (23°N, 72.5°E; 14.9°N magnetic latitude), India, wherein
ionograms are obtained at a cadence of 7.5 min. In order to match the data
corresponding to different transmission frequencies in shorter timescales
to obtain phase shifts more accurately, an interpolation in 2.5‐min inter-
val has been carried out.

2.2. Multiwavelength Imaging Spectrograph using Echelle


grating (MISE)
MISE is a large field‐of‐view imaging echelle spectrograph (Pallamraju
et al., 2013) capable of obtaining high spectral resolution images simulta-
neously at three prominent wavelengths of neutral oxygen dayglow emis-
sions in the thermosphere (namely OI 557.7, 630.0, and 777.4 nm). These
emissions originate from around 130‐, 230‐, and 300‐km altitudes (Tinsley
et al., 1997; Zhang & Shepherd, 2004, 2005), respectively. In this work,
630.0‐nm dayglow data obtained at a cadence of 5 min are used. MISE is
in operation from Physical Research Laboratory's optical aeronomy obser-
vatory at a low‐latitude location, Hyderabad (17.5°N, 78.5°E; 8.9°N mag-
netic latitude), India.

3. Data analysis and Results


With regard to the analysis of optical data, spectral images obtained by
MISE at varying times of integration are coadded for a period of 5 min.
Figure 1. (a) OI 630.0‐nm dayglow (red colored dotted line) diurnal varia- MISE spectra are compared with that of the solar spectrum in both wave-
tion on 18 May 2015 along with the 3‐hr running average (blue colored length and continuum. The difference obtained between them corre-
solid line) and the residual (black colored dash‐dotted line). (b) Lomb‐
sponds to the contributions of the dayglow emissions and atmospheric
Scargle periodogram of the residual shows the presence of dominant peri-
odicities (power greater than the 90% FAL) of 1.07, 1.3, 1.87, and 2.66 hr scattering. This scattering contribution (also known as the Ring effect) is
which are in the gravity wave range. FAL = false alarm limit; PSD = power taken into account, and the information on dayglow emission intensity
spectral density. is obtained as described in detail in our earlier works (Pallamraju et al.,
2013). Such analyses are carried out for each of the three emission wave-
lengths measured by MISE. To compare the wave dynamics in the iono-
sphere as obtained by the digisonde with those in the thermosphere as derived by optical measurement
techniques, the height region from where optical OI 630.0‐nm dayglow emission emanates (around 230
km) was considered to be most optimal. The peak heights of the other two dayglow emissions measured
by MISE are 130 and 300 km, which are either too low or too high when compared to the measurements pos-
sible by a digisonde. Diurnal distribution of OI 630.0‐nm emission intensity (red colored dotted line)
obtained for a sample day, 18 May 2015, is shown in Figure 1a.

3.1. Estimation of Time Periods (τ) of GWs in Optical Data


Figure 1a shows the typical diurnal variation in OI 630.0‐nm dayglow emission rates for 18 May 2015. This
variation in emission rates is governed by the three major daytime OI 630.0‐nm airglow production mechan-
isms, namely, photoelectron impact on atomic oxygen, photodissociation of molecular oxygen, and dissocia-
tive recombination of molecular oxygen ion (Pallamraju et al., 2004). Thus, a broad solar zenith angle‐
dependent variation is expected as all the three production mechanisms are dependent on solar flux. It
can be noted that in addition to the solar zenith angle variation, there exist other smaller fluctuations as well
in the behavior of the dayglow emissions. These fluctuations are known to be due to the GWs which are con-
volved with those emissions engendered by the variation in the solar zenith angle. A 3‐hr running average
(blue colored solid line in Figure 1a) is subtracted from the original data to obtain the residual (black colored
dash‐dotted line in Figure 1a). The residual thus obtained contains information on the fluctuations of wave
periods that are smaller than 3 hr and so enables comparison with the results reported in the literature.
Figure 1b shows the results of Lomb‐Scargle (Lomb, 1976; Scargle, 1982) periodogram analysis (red

MANDAL ET AL. 3
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

colored solid line) carried out on the residual along with the 90% FAL (black colored dashed line). It is noted
that GW time periods of 1.07, 1.30, 1.87, and 2.66 hr exist in the daytime thermosphere on this day. The
uncertainty in the time period is estimated using the following relation as given by Bretthorst (1988)

1:1σ
δf ¼ pffiffiffiffiffiffiffiffiffiffiffi (1)
A:T total : N total

where A is the amplitude of the signal, Ttotal is the total observation time, Ntotal is the total number of obser-
vations, and σ is the standard deviation of the data obtained after the signal has been subtracted (which cor-
responds to the noise). Thus, the time periods along with the corresponding uncertainties are 1.07 ± 0.03,
1.30 ± 0.02, 1.87 ± 0.04, and 2.66 ± 0.06 hr.
Such analyses of daytime neutral GW time periodicities during geomagnetic quiet and disturbed epochs
have been carried out earlier, which revealed various aspects of neutral wave dynamics in low latitudes
along with their dependence on the effects of processes at equatorial latitudes and of solar activity (e.g.,
Laskar et al., 2015; Pallamraju et al., 2010).

3.2. Method of Analysis of Radio Data


The raw ionograms obtained from a digisonde are traces of the virtual height of the ionosphere as a function of
transmission frequency which is swept in a certain frequency range. These ionograms yield information on the
variation of electron density and ionospheric height as a function of time. The height of the ionosphere depends
on the magnitude and sense of meridional neutral winds, electric field, and neutral temperature (Rishbeth,
1967). Thus, over the geomagnetic equator, the height variation of the F layer in the nighttime was used as a
proxy to the equatorial electric field, as the meridional neutral winds do not alter the height of the F layer over
this latitude (e.g., Chakrabarty et al., 2008) (with the assumption that temperature is constant).
Just as integrated dayglow data were used to derive GWs (Figure 1), we have considered IEC on this day for
deriving the GW time periods. Here, the IEC data have been calculated by integrating the electron densities
from 100 to 800 km as obtained from the true height analysis of the ionograms obtained by the digisonde at
Ahmedabad, assuming the exponential decaying nature of densities for heights above the peak height as
mentioned above. The dotted (red colored) line in Figure 2a shows the IEC values in the units of 1016
electrons/m2. The daytime optical airglow data obtained from Hyderabad from where neutral GW informa-
tion is available are used to compare with the GW behavior obtained from radio measurements. Thus, for the
sake of consistency, similar method of analysis as used for optical is applied to that of the radio data as well.
The solid (blue colored) line shows the 3‐hr running average of the IEC data. The dash‐dotted (black colored)
line shows the residual obtained by subtracting the running average data from the IEC data. Figure 2b shows
the Lomb‐Scargle periodogram obtained for this residual data. It can be seen that in contrast to the dayglow
emissions, on this day, the IEC data do not show presence of any time periods that are greater than the FAL
(black colored dashed line). This difference in the existence of GW features in optical versus radio data is not
surprising and can be understood if one considers the fact that the IEC from the radio data is averaged over a
larger height range of 100–800 km, while the daytime OI 630.0‐nm emission layer semithickness is around
100 km. Similarly, the GPS‐derived TEC data over Ahmedabad have been analyzed (not shown here), which
too do not show any significant periodicities within 3 hr. As discussed earlier, TEC corresponds to integrated
contribution from both the ionosphere and the plasmasphere, and therefore, oscillatory features of GW
regime are smoothed out. When perturbations occur in the ionosphere as a whole, they do get reflected in
the spectral analysis of IEC measurements, and therefore, large‐scale features, such as traveling ionospheric
disturbances, have been conventionally investigated using data obtained from digisondes. However, in case
of GWs, as they do not perturb the ionosphere at all the heights simultaneously with the same phase, it is not
possible to unambiguously determine the GW periodicity in the IEC variations.
Similar analysis methodology has been followed for peak F2 layer height (hmF2) variation and peak fre-
quency of F2 layer (foF2) variation, the results of which are shown in Figures 2c–2e. These variations of
hmF2 and foF2 when subjected to periodogram analysis show the presence of periodicities of 2.68 and 2.63
hr, respectively. Conventionally, these time periods in hmF2 and foF2 variation are attributed to be of GW
origin (Rama Rao et al., 2006; Tulasi Ram et al., 2012). But these wave‐like fluctuations can be caused by
either one or combined effects of variations in electric field, wind, temperature, and wave activity.

MANDAL ET AL. 4
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

Figure 2. (a) Integrated electron content (IEC; red colored dotted line) obtained from the digisonde measurements along
with 3‐hr running average (blue colored solid line) and its residual (black colored dash‐dotted line). (b) Normalized power
spectral density of IEC, showing that there are no periodicities in the gravity wave range that are significant, as their power
is lower than the 90% FAL (dashed line). (c, d) Same as in (a, b) but for height of peak F2 layer and (e, f) same as in (a, b) but
for critical frequency of F2 layer. FAL = false alarm limit; PSD = power spectral density.

Therefore, these time periods cannot be stated unambiguously to be of GW origin. It can be appreciated that
as the variations in IEC, hmF2, and foF2 are either integrated effect or variation at a particular height of
ionosphere, they do not yield any information in terms of propagation characteristics of GWs. Depending
on the strength of the source, GWs propagate both horizontally and vertically and thus alter the
ionospheric densities at different heights at different times (phases). Therefore, it is required that height
variations of different electron densities be investigated independently. In order to do this, all the
individual ionograms have been meticulously scaled and analyzed for obtaining information on the true
heights of the ionosphere at different frequencies of transmissions. Results of such analyses are shown in
Figure 3a for the same day as in Figure 2. The true height variation corresponding to a fixed transmission
frequency (i.e., a particular value of ionospheric electron density) with time is known as an isoelectron
density contour. In Figure 3a, the height variations of isoelectron density contours from 8 to 12 MHz of
operational frequencies of digisonde at an interval of 1 MHz are shown as a function of time. It may be
noted that there exists a clear time difference in the modulations of ionospheric heights which are earlier
at greater transmission frequencies (and higher altitudes) as compared to the smaller ones (lower
altitudes). To aid the eye, this is also shown as dashed (black colored) lines at three different times in
Figure 3a. From Hines (1960) GW theory, it is known that the downward phase movement is associated
with upward propagation of GWs. If this downward phase movement is indeed due to GW propagation,
then it should be verifiable by spectral analysis of variations in the heights of different isoelectron
densities. Similar to the optical data shown above (Figure 1), residual has been obtained in the variation
of heights of isoelectron density contours by subtracting it with a 3‐hr running average. This residual has

MANDAL ET AL. 5
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

Figure 3. (a) Variations of true heights of isoelectron density contours at five transmission frequencies (8, 9, 10, 11, and 12
MHz) on 18 May 2015 obtained after meticulous scaling of the ionograms. (b) Lomb‐Scargle periodogram of the residuals
obtained by subtracting the 3‐hr running average values from original data, showing the time period of 2.2 hr (as indicated
by a black colored arrow) to be significant at all the heights corresponding to all the transmission frequencies. PSD =
power spectral density.

been subjected to Lomb‐Scargle periodogram analysis, the result of which is shown in Figure 3b. It may be
noted that a common time period of 2.20 ± 0.08 hr is present in the height variations of isoelectron densities
corresponding to all of the frequencies of transmission (shown by an arrow in Figure 3b). The presence of
common periodicity and downward phase propagation in the time variation of optical emissions from
different altitudes provided an unambiguous signature of GW propagation from lower to higher altitudes
in the mesosphere‐lower thermosphere region (Singh & Pallamraju, 2016). It can be noted in the present
case as well that the height variations of isoelectron density contours show a common time period and a
downward phase movement (Figure 3a), which clearly indicate propagation of GWs from lower to higher
altitudes in the ionosphere.
3.2.1. Estimation of Vertical Phase Speeds (cz) of GW Propagation
It is apparent that the ionospheric behavior is a linear superposition of GWs of different periodicities that
propagate with different speeds and directions. As seen in Figures 3a and 3b on 18 May 2015, each of the iso-
electron density contours also shows the presence of periodicities other than the common time period and so
they will also exhibit independent behavior. We have analyzed the GW characteristics at each of the com-
mon time periods present at all the height variations of isoelectron density contours. For that, we have con-
sidered one significant common time period at a time and removed the contributions of all other time
periods in the Fourier domain. In order to do that, only the power corresponding to a particular dominant
time period (power spectral density, PSD > FAL) is considered by using a band‐pass filter with a width of
peak time period ± Brunt‐Väisälä time period of 0.25 hr (Brunt‐Väisälä frequency, N = (2g*/5H)1/2;
Shiokawa et al., 2009, where H is the neutral scale height [an average value of 54 km is considered for this
duration] and g* is the acceleration due to gravity at a particular altitude), while the powers of the rest of the
spectra are equated to zero. Thus, by such a treatment (Hocke & Kämpfer, 2009), the power of the common
GW time period corresponding to the fluctuations in isoelectron density contour at a given transmission fre-
quency is isolated in Fourier (frequency) domain. The inverse Fourier transformation of this spectral series
yields information of GW fluctuations in time domain. As shown in Figure 4a, such analysis has been carried
out for the GW time period of 2.2 hr. Such a method was successfully used in an earlier study to obtain the
first three‐dimensional GW characteristics in the daytime through optical measurement technique (e.g.,
Pallamraju et al., 2016). In the present case the results obtained for dominant time period of 2.2 hr are shown
in Figure 4a, wherein an average height corresponding to each of the transmission frequencies is added to
the relative variations to obtain information on temporal variations due to one GW time period only. As
can be seen in this figure, clear wave‐like fluctuations of a given time period are apparent with troughs
and crests at a given altitude. It can also be noted that the crests and troughs at lower heights occur at a
slightly later time than those at higher altitudes showing clear downward phase propagation in the height

MANDAL ET AL. 6
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

Figure 4. (a) Variations in the heights of isoelectron density contours corresponding to transmission frequencies of 8, 9,
10, 11, and 12 MHz on 18 May 2015 for the gravity wave (GW) time period of 2.2 hr are shown. Clear downward phase
movement can be seen. The maximum and minimum phase propagation speeds and the corresponding vertical scale sizes
of the GW as obtained are also indicated. (b) The vertical scale sizes (blue colored squares) obtained for the duration of 16–
21 May 2015. (c) Distribution of GW vertical scale sizes with altitude as seen from different measurements, boxes A–D
from Yamanaka and Fukao (1994), box E from Oliver et al. (1997), and box F is the results obtained through the current
method.

variations of isoelectron density contours, indicating the characteristic signature of upward propagating
GWs (Hines, 1960). The vertical phase speed (cz) of the GW can be calculated with the knowledge of the
difference in heights (Δh) and the phase offset times (Δt), by using the relation, cz = Δh/Δt. The time
offsets between the crests/troughs of the height variations of consecutive isoelectron density contours
have been linearly fitted to obtain a consolidated picture. On this day (Figure 4a) vertical phase speeds
vary from 37.74 ± 5.40 to 39.81 ± 5.62 m/s for the time period of 2.2 hr; however, these are not distinct as
they are within the uncertainties associated with each of these values. In all these analyses, it is assumed
that the phase speeds do not change as a function of altitude.
3.2.2. Estimation of Vertical Scale Sizes of GWs
In Figure 4a, the phase variations that correspond to the GW time period of 2.2 hr have been shown. On such
occasions, from the common period (τ) obtained and the derived phase speeds (cz), the vertical scale sizes (λz)
of the GW were estimated as follows: cz × τ, which on this day range from 298.92 ± 44.21 to 315.35 ± 46.07
km. Such analysis, as shown in Figures 3a and 4a have been carried out for the duration of 16–20 May 2015.
Phase speeds obtained in this duration range from 30.06 ± 4.35 to 45.69 ± 11.84 m/s. On 17 May 2015, no
signatures of wave propagation are seen. Considering the uncertainties in the measurement of height by
the digisonde and the data cadence, the uncertainties in phase speeds are estimated which turn out to be
in the range of 14–26%. Similar analyses as described above have been successfully applied to airglow inten-
sity variations originating at multiple altitudes to characterize the effect of tropical cyclone‐generated GW in
the lower atmosphere on the upper atmosphere at mesospheric heights (Singh & Pallamraju, 2016).

MANDAL ET AL. 7
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

Figure 4b shows all the vertical scale sizes (blue squares) of GWs as
obtained from digisonde for this duration. It is apparent that the magni-
tudes of the vertical scale sizes of GWs vary from one day to another.
Further, as there are waves of different periodicities that travel with differ-
ent speeds of propagation, multiple vertical scale sizes of GWs can exist in
the same day. In this duration, however, all the vertical scale sizes
obtained in a day vary within the measured uncertainties (which vary
from 39.23 to 66.38 km). During this observational window, the range of
vertical scale sizes obtained is around 183.21 ± 39.23 to 393.07 ± 66.38
km. These values of vertical scale sizes do agree well with the theoretical
works and values reported earlier (Oliver et al., 1997; Yamanaka &
Fukao, 1994).
Figure 4c is adapted from Oliver et al. (1997) and redrawn with the
updated values that resulted from the present work. The x and y axes of
this figure represent the vertical scale sizes and the altitudes at which they
are calculated, respectively. The boxes A–D are from observations using
different measurements as reported in Yamanaka and Fukao (1994).
Box E depicts the results obtained from MU radar observations (Oliver
et al., 1997), and box F is the vertical scale sizes estimated from digisonde
measurements as described in this work. Considering a scale height of 50
km in the relation of vertical GW wave number m as a function of height h
given by Yamanaka and Fukao (1994), the resulting vertical GW scale
sizes are plotted in Figure 4c (black colored solid line).

m ¼ ð2π=1:5 kmÞ: expð−z=50 kmÞ (2)

It may be noted that the present λz measurements at altitude correspond-


ing to ~270 km as reported in box F agree well with the theoretical esti-
mate as given in equation (2). To the best of our knowledge,
Figure 5. (a) Gravity wave (GW) periodicities obtained by optical (red measurements at heights corresponding to ~270 km (box F) as depicted in
colored triangle) and radio (blue colored circle) measurements during 16– Figure 4c is being reported for the first time.
21 May 2015 are shown. These are the GW time periods at which vertical
propagation of GWs was seen in the digisonde data at different frequencies.
The dotted and solid circles around the time periods indicate the common 3.3. Intercomparison of GW Periodicities From Simultaneous
time periods between the optical and radio measurements. The solid and Radio and Optical Measurements
dashed circles show in phase (b) and out of phase (c) variations between the
The optical data duration is from 7 to 17 hr. For the digisonde data, the
GW behavior of simultaneous dayglow and radio measurements. (b) Phase
evolution comparison of OI 630.0‐nm dayglow and isoelectron density con- duration of presence of the echo corresponding to the transmission fre-
tour corresponding to transmission frequency of 8 MHz for 19 May 2015. (c) quency of 12 MHz has been considered, which is typically from 11 to 20
Same as (b) but for 20 May 2015. The OI 630.0‐nm dayglow data are available hr. The data duration for other transmission frequencies have been
from 7 to 17 hr, and data for height variation of isoelectron density contour truncated accordingly so as to keep the same data range for the sake
corresponding to transmission frequency of 8 MHz are typically available
of consistency in analysis. Nevertheless, there is a considerable overlap
from 11 hr onward.
in these two data sets of around 6 hr (from 11 to 17 hr) to
enable intercomparison.
Similar to the method described in Figure 1, time series analyses have been carried out for optical measure-
ments for all clear‐sky days during 16–20 May 2015. Figure 5a shows the comparison of the time periods
obtained from optical measurements (in red colored triangles) and from digisonde (in blue colored circles).
A comparison can be made between the optical and radio measurements as there is considerable overlap ~6
hr (11–17 hr) of these data sets, and the data duration of each of the sets is ~10 hr, which is much greater than
the GW time periods (less than 3 hr) that are being considered here. Each blue circle corresponds to the value
of the common time period present at all the height variations of isoelectron density contours, which also
show phase propagation features at all the heights of isoelectron density contours. In this way it is ensured
that they are of GW origin. As mentioned above, no wave propagation features were seen on 17 May 2015 in
the variation of isoelectron density contours. Optical measurements are not available on 16 May 2015 due to
unfavorable sky condition.

MANDAL ET AL. 8
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

It can be noted that the GW periodicities present in the height variations of isoelectron densities correspond-
ing to all transmission frequencies match with the daytime optical data on 19 and 20 May 2015 (encircled
values in Figure 5a). The solid and dashed circles represent the matching of time periods corresponding to
a similar/dissimilar phase evolution between the two as described below. While matching, time periods
within the limits of the Brunt‐Väisälä period of around 0.25 hr at these heights were considered to be similar.
Also, while there are similarities between the GW time periods obtained by optical and radio measurements,
not all of them match exactly with one another. This can be understood to be due to the fact that (i) the alti-
tude regions of integration for optical and radio measurements are not exactly similar and (ii) although the
data are obtained simultaneously, the instruments are not collocated in the present study; therefore, depend-
ing on the ambient wind conditions and temperature gradients, which could be different at different lati-
tudes, these waves could evolve and propagate differently.
These similar values of GW time periods could be due to either the presence of same wave structures at these
two locations or common source of fluctuations. To investigate the reason for the existence of similar GW
time periods measured from both methods, the phase evolution obtained by these two measurements is com-
pared. For such comparison, height variation of isoelectron density contour corresponding to the transmis-
sion frequency of 8 MHz is chosen as its height is closest to the peak emission altitude of OI 630.0‐nm
dayglow emission (~230 km), and the times shown in Figures 5b and 5c correspond to the local time for
Ahmedabad. The same method to derive the variation in heights of isoelectron density contour for a parti-
cular period as described above (as depicted in Figure 4a) has been applied to the optical OI 630.0‐nm day-
glow emission data analysis as well to obtain the variation in emission intensity for a particular GW time
period. On 19 May 2015, the variations in these respective parameters corresponding to periods of 2.52
and 2.63 hr for optical (red colored dashed line in Figure 5b) and radio (black colored solid line in
Figure 5b) measurements do match well with each other (within the limit of Brunt‐Väisälä time period;
Figure 5b). In spite of large separation (~1,000 km) between the two observational locations, such a match
between the optical and radio data is remarkable. Therefore, on days like 19 May 2015, when GW time per-
iod and their phase evolution matches, it is possible that these two locations are influenced by a similar
broad source (e.g., equatorial electric field) which can simultaneously affect the spatially separated locations
along the same longitude. On 20 May, however, the phase evolution corresponding to time periods of 1.51
and 1.47 hr obtained in a similar method do not match with each other (Figure 5c). In such cases, the plau-
sible reasons for such observations are that these two locations are being influenced by independent
sources/causes. Such comparison between the GW time period and phase at these two spatially separated
locations provides a broader regional scale picture of the GW characteristics in the daytime thermosphere
over low latitudes.

4. Discussion and Conclusion


This work provides a detailed description of an approach for obtaining the GW characteristics using
digisonde. Conventionally, established method for deriving GW periodicities, scale sizes, and propaga-
tion speeds has been through the analysis of temporal variation in optical airglow emission intensities.
Ground‐based airglow emission measurements yield column integrated values; however, as they ema-
nate from a finite thickness, they retain the information of fluctuations of smaller timescales as the reac-
tants responsible for the airglow emissions do get affected by the passage of any wave‐like fluctuations
in the medium (e.g., Karan & Pallamraju, 2017, 2018; Laskar et al., 2015; Pallamraju et al., 2010,
2014, 2016).
As the GPS and other satellite‐based transmitters provide information on the total column integrated elec-
tron densities, it is possible to obtain wave information over large time periods (tidal scales and greater;
e.g., Hunsucker, 1982; Mendillo, 2006). However, in such methods, the smaller timescale fluctuations are
usually averaged out, unlike in the optical measurements. Also, due to the integrated nature of such mea-
surements, vertical propagations of waves, if any, are also averaged out. Thus, for insights on the GW propa-
gation characteristics, it is required that investigations be carried out by monitoring the individual height
variations at different isoelectron densities as they respond to any wave propagation. This approach also pre-
vents any averaging of wave‐like features in space and time in the process of their propagation. The techni-
que of ionospheric sounding by radio waves provides one such effective method.

MANDAL ET AL. 9
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

The sources of GWs are different, and they propagate in varying directions
away from the source. If the same source perturbs the ionosphere at differ-
ent altitudes, the ionospheric behavior at all the altitudes should show the
presence of common time periods as demonstrated in Figure 3b. Further,
as they propagate upward, corresponding phase variation should also be
reflected in the ionosphere as pointed out in Figure 3a, and such propaga-
tion exists not only on this chosen day but also on several other days as
evidenced in the common time periods present in the height variations
of isoelectron density contours as shown in Figure 5a. Based on the verti-
cal propagation speeds and time periods, vertical scale sizes of GWs are
obtained (Figure 4b) which are well in agreement within the estimates
of GW theory (Figure 4c). Therefore, the data obtained by the present
radio method do actually reflect the neutral GW behavior as also con-
firmed by its comparison with time period and phase of the simultaneous
measurements of optical emissions in the daytime (Figures 5b and 5c).

4.1. Estimation of Horizontal Scale Sizes of GWs


It has been shown above that GW time periods, their vertical phase propa-
gation speeds, and vertical scale sizes have been derived using digisonde
measurements. Using some of these derived parameters, an attempt has
been made to estimate the horizontal scale sizes of GWs. The GW disper-
sion relation (Fritts & Alexander, 2003) used is given as follows:
  
k 2 þ l2 N 2 −ω2 1
m2 ¼  2 2 − 2 (3)
ω −f 4H

In this GW dispersion relation, k, l, and m are the wave numbers in x, y


Figure 6. (a) Simulation of gravity wave (GW) vertical scale sizes (λz) for (horizontal), and z (vertical) directions, respectively; N is the Brunt‐
horizontal scale size (λH) of 1,000 km using wind values ranging from 0 to Väisälä frequency; f = 2Ωsinϕ is the Coriolis parameter (Ω is the Earth's
100 m/s and the relative angles (θ = θu − θc) of 0°, 45°, 90°, 145°, and 180° ! !
rotation rate, and ϕ is the latitude); ω = ω− k :!
u − l :!
v , is the intrinsic
between the direction of propagations of wave (θc) and wind (θu) for GW
periods of 1.5 hr (solid lines) and 2.5 hr (dashed lines) obtained using GW GW frequency, that is, the frequency value if measured in the reference
frame moving with the background wind (!
dispersion relation. (b) Simulation of λH obtained for range of values of λz, at u , zonal, and !v , meridional);
θ = 90° for GW time periods of 1.0 (red), 1.5 (orange), 2.0 (cyan), 2.5 (blue),ω is the GW frequency as observed from ground; and H is the neutral scale
and 3.0 hr (purple) is shown to enable a first‐order estimation of horizontal
height. N is given as (2g*/5H)1/2 (Pallamraju et al., 2016; Shiokawa et al.,
GW scale sizes in the absence of information on winds and GW propagation
direction. 2009), where g* is acceleration due to gravity at a given altitude. In this
relation (3), we use vertical scale size (λz ¼ 2πm ) values derived through
the radio method and model NRLMSIS‐00 neutral atmosphere as inputs
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
to calculate the horizontal scale sizes (λH ¼ k2πH , kH ¼ k 2 þ l2 ) of GWs. Simulations are carried out to esti-
mate the horizontal scale sizes (λH) for different vertical scale sizes (λz) for wind speeds (U) varying from 0 to
100 m/s. As the relative direction of winds with respect to that of propagation of GWs (Δθ = θu − θc; θu is the
wind propagation direction, and θc is the direction of wave propagation) has a bearing on the GWs charac-
teristics, several representative angles (Δθ) of 0°, 45°, 90°, 145°, and 180° have been considered. The results
for two different GW time periods (1.5 and 2.5 hr) are shown in Figure 6a as solid and dashed lines. If the
ambient wind is in the same direction as the GW propagation (Δθ = 0°), then horizontal scale size grows
at the expense of vertical scales. Conversely, if the wind in the opposite direction to GW propagation (Δθ
= 180°), then the vertical scale increases as the ambient wind speed increases. Thus, as Δθ increases form
0° to 180°, the same value of horizontal scale size of 1,000 km corresponds to higher values of vertical scale
sizes. It may be noted that when the winds flow in orthogonal direction (Δθ = 90°) to that of the wave pro-
pagation, the magnitudes of ambient wind speeds do not have any effect on the vertical and horizontal scale
sizes. Therefore, in the absence of knowledge of wind magnitudes and directions at those altitudes and lim-
itation of information on the wave propagation direction, we have considered orthogonality in wind flows
with respect to GW propagation direction (Δθ = 90°), to get an estimate of λH values for the measured time
periods and λz values of GWs as shown in Figure 6b. The calculations are carried out for different GW time

MANDAL ET AL. 10
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

periods of 1.0 hr (red colored dotted line), 1.5 hr (yellow colored dashed line), 2.0 hr (cyan colored dash‐
dotted line), 2.5 hr (blue colored dash‐dot‐dot‐dot line), and 3.0 hr (purple colored solid line) and are pre-
sented in Figure 6b. The simulation results given in this figure can be considered to get a first‐order estimate
of the horizontal scale sizes of a GW, given its τ and λz. For example, from this figure, for τ ≅ 2 hr and λz of
300 km, λH turns out to be ~2,500 km, whereas for τ ≅ 1.5 hr and λz of 300 km, the value of λH is ~1,400 km.
These values are consistent with the simulations (Hocke & Schlegel, 1996; Richmond, 1978; Vadas, 2007)
which have considered the effects of ambient temperature and wave dissipations as a function of altitudes.
As mentioned above, these are values for no wind conditions; nevertheless, they provide a first‐order esti-
mate of the horizontal scale sizes of GWs for the observed duration. However, to remove ambiguity with
respect to knowledge on the wave propagation direction, it is planned to carry out simultaneous and collo-
cated radio and optical measurements in the near future which will help to converge on more accurate
estimates.
Another aspect which emerges is that this technique is applicable for daytime conditions only when the
thickness of ionosphere is much larger than in the nighttime (typically, in the daytime echoes can be
obtained until 12 MHz for ~10‐hr duration, while in the nighttime it is 5 MHz for the same duration).
This can be understood that for realistic estimates of vertical scale size, information on phase propagations
from large spatial extents is required, which is possible in the daytime.
This possibility of deriving information on GW time periods, vertical propagation speeds, and vertical scale
sizes using digisondes opens up new avenues in the understanding of the neutral wave dynamics in the
upper atmosphere. The radio measurements can be interspersed between two optical observational locations
to fill in the much needed gap in obtaining a larger‐scale picture of wave dynamical interactions in the ther-
mosphere. Further, as the radio measurements are not affected by cloud cover, these measurements can fill
in the information with regard to the understanding of waves during monsoon period as well.

5. Summary
Results from the combined investigations of radio and optical airglow emission data in daytime are pre-
sented with a specific aim to extract information on GW characteristics. A different method of approach
of radio measurements is described wherein waves of GW regime can be retrieved and various wave charac-
teristics can be obtained from them. Digisonde measurements from Ahmedabad are used, and the iono-
grams so obtained are carefully scaled to obtain the height variation in the isoelectron density contours,
which are used to clearly identify the phase propagation speeds of GWs. Spectral analysis of ionospheric
height variation of each of the isoelectron density contours showed common time periods which indicated
that the same wave had propagated upward. Using this information, vertical scale sizes of GWs were
obtained. The vertical scale sizes obtained using this method do agree well with the theoretical estimates
of an exponential increase of vertical scale sizes with altitude as well as with earlier observations of GW scale
sizes at different altitudes. Further, using the derived GW time periods and their phase speeds, vertical GW
scale sizes are calculated. The GW time periods, phase speeds, and vertical wavelengths during 16–20 May
2015 over a low‐latitude location, Ahmedabad, were found to be in the range of 1.47 ± 0.05 to 2.64 ± 0.07
hr, 30.06 ± 4.35 to 45.69 ± 11.84 m/s, and 183.21 ± 39.23 to 393.07 ± 66.38 km, respectively. Simulations have
been carried out using these experimentally measured parameters and neutral atmosphere model as inputs
to the Hines (1960) GW theory to estimate to the first order, the plausible horizontal scale sizes of GWs, in
the absence of information on neutral winds and the relative directions of propagation between the winds
Acknowledgments
We would like to thank T. Vijaya and GWs. This new technique has great potential in yielding information on the neutral GW characteristics
Lakshmi and her team at Jawaharlal and the ensuing dynamics in all weather conditions.
Nehru Technological University,
Hyderabad, India, for the logistics sup-
port for carrying out the optical experi-
ments. The data obtained are property
References
of the Physical Research Laboratory. Alexander, M. J. (1996). A simulated spectrum of convectively generated gravity waves: Propagation from the tropopause to the mesopause
Text files containing data needed to and effects on the middle atmosphere. Journal of Geophysical Research, 101(D1), 1571–1588. https://doi.org/10.1029/95JD02046
reproduce the figures in this paper are Bretthorst, L. (1988). Bayesian spectrum analysis and parameter estimation. Berlin: Springer. https://doi.org/10.1007/978‐1‐4684‐9399‐3
provided in the supporting information. Chakrabarty, D., Sekar, R., Sastri, J. H., & Ravindran, S. (2008). Distinctive effects of interplanetary electric field and substorm on nighttime
This work is supported by the equatorial F layer: A case study. Geophysical Research Letters, 35, L19108. https://doi.org/10.1029/2008GL035415
Department of Space, Government of Forbes, J. M., Bruinsma, S. L., Miyoshi, Y., & Fujiwara, H. (2008). A solar terminator wave in thermosphere neutral densities measured by
India. the CHAMP satellite. Geophysical Research Letters, 35, L14802. https://doi.org/10.1029/2008GL034075

MANDAL ET AL. 11
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

Fritts, D. C., & Alexander, M. J. (2003). Gravity wave dynamics and effects in the middle atmosphere. Reviews of Geophysics, 41(1), 1003.
https://doi.org/10.1029/2001RG000106
Fritts, D. C., & Vadas, S. L. (2008). Gravity wave penetration into the thermosphere: Sensitivity to solar cycle variations and mean winds.
Annales Geophysicae, 26(12), 3841–3861. https://doi.org/10.5194/angeo‐26‐3841‐2008
Hines, C. O. (1960). Internal atmospheric gravity waves at ionospheric heights. Canadian Journal of Physics, 38(11), 1441–1481. https://doi.
org/10.1139/p60‐150
Hocke, K., & Kämpfer, N. (2009). Gap filling and noise reduction of unevenly sampled data by means of the Lomb‐Scargle periodogram.
Atmospheric Chemistry and Physics, 9(12), 4197–4206. https://doi.org/10.5194/acp‐9‐4197‐2009
Hocke, K., & Schlegel, K. (1996). A review of atmospheric gravity waves and travelling ionospheric disturbances: 1982–1995. Annales
Geophysicae, 14(9), 917–940. https://doi.org/10.1007/s00585‐996‐0917‐6
Hunsucker, R. D. (1982). Atmospheric gravity waves generated in the high‐latitude ionosphere: A review. Reviews of Geophysics, 20(2),
293–315. https://doi.org/10.1029/RG020i002p00293
Karan, D. K., & Pallamraju, D. (2017). Small‐scale longitudinal variations in the daytime equatorial thermospheric wave dynamics as
inferred from oxygen dayglow emissions. Journal of Geophysical Research: Space Physics, 122, 6528–6542. https://doi.org/10.1002/
2017JA023891
Karan, D. K., & Pallamraju, D. (2018). Effect of geomagnetic storms on the daytime low‐latitude thermospheric wave dynamics. Journal of
Atmospheric and Solar ‐ Terrestrial Physics, 170, 35–47. https://doi.org/10.1016/j.jastp.2018.02.003
Klausner, V., Fagundes, P. R., Sahai, Y., Wrasse, C. M., Pillat, V. G., & Becker‐Guedes, F. (2009). Observations of GW/TID oscillations in the
F2 layer at low latitude during high and low solar activity, geomagnetic quiet and disturbed periods. Journal of Geophysical Research, 114,
A02313. https://doi.org/10.1029/2008JA013448
Lakshmi Narayanan, V., Gurubaran, S., & Emperumal, K. (2010). Airglow imaging observations of small‐scale structures driven by con-
vective instability in the upper mesosphere over Tirunelveli (8.7°N). Journal of Geophysical Research, 115, D19119. https://doi.org/
10.1029/2009JD012937
Laskar, F. I., Pallamraju, D., Veenadhari, B., Vijaya Lakshmi, T., Anji Reddy, M., & Chakrabarti, S. (2015). Gravity waves in the thermo-
sphere: Solar activity dependence. Advances in Space Research, 55(6), 1651–1659. https://doi.org/10.1016/j.asr.2014.12.040
Lin, C. H., Wang, W., Hagan, M. E., Hsiao, C. C., Immel, T. J., Hsu, M. L., et al. (2007). Plausible effect of atmospheric tides on the equatorial
ionosphere observed by the FORMOSAT‐3/COSMIC: Three‐dimensional electron density structures. Geophysical Research Letters, 34,
L11112. https://doi.org/10.1029/2007GL029265
Lomb, N. R. (1976). Least‐squares frequency analysis of unequally spaced data. Astrophysics and Space Science, 39(2), 447–462. https://doi.
org/10.1007/BF00648343
Mendillo, M. (2006). Storms in the ionosphere: Patterns and processes for total electron content. Reviews of Geophysics, 44, RG4001. https://
doi.org/10.1029/2005RG000193
Oliver, W. L., Otsuka, Y., Sato, M., Takami, T., & Fukao, S. (1997). A climatology of F region gravity wave propagation over the middle and
upper atmosphere radar. Journal of Geophysical Research, 102(A7), 14,499–14,512. https://doi.org/10.1029/97JA00491
Pallam Raju, D., Sridharan, R., Gurubaran, S., & Raghavarao, R. (1996). First results from ground‐based daytime optical investigation of the
development of the equatorial ionization anomaly. Annales Geophysicae, 14(2), 238–245. https://doi.org/10.1007/s00585‐996‐0238‐9
Pallamraju, D., Baumgardner, J., Singh, R. P., Laskar, F. I., Mendillo, C., Cook, T., et al. (2014). Daytime wave characteristics in the
mesosphere lower thermosphere region: Results from the balloon‐borne investigations of regional‐atmospheric dynamics experiment.
Journal of Geophysical Research: Space Physics, 119, 2229–2242. https://doi.org/10.1002/2013JA019368
Pallamraju, D., Chakrabarti, S., & Valladares, C. (2004). Magnetic storm‐induced enhancement in neutral composition at low latitudes as
inferred by O(1D) dayglow measurements from Chile. Annales Geophysicae, 22(9), 3241–3250. https://doi.org/10.5194/angeo‐22‐3241‐
2004
Pallamraju, D., Das, U., & Chakrabarti, S. (2010). Short‐ and long‐timescale thermospheric variability as observed from OI 630.0 nm
dayglow emissions from low latitudes. Journal of Geophysical Research, 115, A06312. https://doi.org/10.1029/2009JA015042
Pallamraju, D., Karan, D. K., & Phadke, K. A. (2016). First three dimensional wave characteristics in the daytime upper atmosphere derived
from ground‐based multiwavelength oxygen dayglow emission measurements. Geophysical Research Letters, 43, 5545–5553. https://doi.
org/10.1002/2016GL069074
Pallamraju, D., Laskar, F. I., Singh, R. P., Baumgardner, J., & Chakrabarti, S. (2013). MISE: A multiwavelength imaging spectrograph using
echelle grating for daytime optical aeronomy investigations. Journal of Atmospheric and Solar ‐ Terrestrial Physics, 103, 176–183. https://
doi.org/10.1016/j.jastp.2012.12.003
Pramitha, M., Ratnam, M. V., Taori, A., Krishna Murthy, B. V., Pallamraju, D., & Vijaya Bhaskar Rao, S. (2015). Evidence for tropospheric
wind shear excitation of high phase speed gravity waves reaching the mesosphere using the ray tracing technique. Atmospheric
Chemistry and Physics, 15(5), 2709–2721. https://doi.org/10.5194/acp1527092015
Raghavarao, R., Sridharan, R., Sastri, J. H., Agashe, V. V., Rao, B. C. N., Rao, P. B., & Somayajulu, V. V. (1988). The equatorial ionosphere;
WITS hand‐ book. World Ionosphere/Thermosphere Study (Vol. 1, pp. 48–93, SCOSTEP Secretariat). Urbana: University of Illinois.
Rama Rao, P. V. S., Gopi Krishna, S., Niranjan, K., & Prasad, D. S. V. V. D. (2006). Temporal and spatial variations in TEC using simul-
taneous measurements from the Indian GPS network of receivers during the low solar activity period of 2004–2005. Annales
Geophysicae, 24(12), 3279–3292. https://doi.org/10.5194/angeo‐24‐3279‐2006.
Rastogi, R. G., & Klobuchar, J. A. (1990). Ionospheric electron content within the equatorial F2‐layer anomaly belt. Journal of Geophysical
Research, 95(A11), 1355–1363. https://doi.org/10.1029/JA095iA11p19045
Ravindran, S., G. Manju, G, C. V. Devasia, R. Sridharan, S. Thampi, S. Pradeep, et al.. (2014). Plasmaspheric electron content variation in
the magnetic equatorial region during space weather events: Results from the CRABEX (Coherent Radio Beacon Experiment) using the
beacon onboard the Indian geostationary satellite (GSAT ‐ 2).
Richmond, A. D. (1978). Gravity wave generation, propagation, and dissipation in the thermosphere. Journal of Geophysical Research,
83(A9), 4131–4145. https://doi.org/10.1029/JA083iA09p04131
Rishbeth, H. (1967). The effects of winds on the ionospheric F2 peak. Journal of Atmospheric and Terrestrial Physics, 29(3), 225–238. https://
doi.org/10.1016/0021‐9169(67)90192‐4
Scargle, J. D. (1982). Studies in astronomical time series analysis. II. Statistical aspects of spectral analysis of unevenly spaced data.
Astronomy Journal, 263(2), 835–853. https://doi.org/10.1086/160554
Shiokawa, K., Otsuka, Y., & Ogawa, T. (2009). Propagation characteristics of nighttime mesospheric and thermospheric waves observed by
optical mesosphere thermosphere imagers at middle and low latitudes. Earth Planets Space, 61(4), 479–491. https://doi.org/10.1186/
BF03353165

MANDAL ET AL. 12
Journal of Geophysical Research: Space Physics 10.1029/2019JA026723

Singh, R. P., & Pallamraju, D. (2016). Effect of cyclone Nilofar on mesospheric wave dynamics as inferred from optical nightglow obser-
vations from Mt. Abu, India. Journal of Geophysical Research: Space Physics, 121, 5856–5867. https://doi.org/10.1002/2016JA022412
Tinsley, B. A., Rohrbaugh, R. P., Hanson, W. B., & Broadfoot, A. L. (1997). Images of transequatorial F region bubbles in 630‐ and 777‐nm
emissions compared with satellite measurements. Journal of Geophysical Research, 102(A2), 2057–2077. https://doi.org/10.1029/
95JA01398
Tulasi Ram, S., Yamamoto, M., Tsunoda, R. T., & Thampi, S. V. (2012). On the application of differential phase measurements to study the
zonal large scale wave structure (LSWS) in the ionospheric electron content. Radio Science, 47, RS2001. https://doi.org/10.1029/
2011RS004870
Vadas, S. L. (2007). Horizontal and vertical propagation and dissipation of gravity waves in the thermosphere from lower atmospheric and
thermospheric sources. Journal of Geophysical Research, 112, A06305. https://doi.org/10.1029/2006JA011845
Yadav, S., & Pallamraju, D. (2015). On the coupled interactions between Ring current intensity and high‐latitude ionospheric electron
density variations. Journal of Atmospheric and Solar ‐ Terrestrial Physics, 125‐126, 50–58. https://doi.org/10.1016/j.jastp.2015.02.006
Yamanaka, M. D., & Fukao, S. (1994). A simple model of gravity‐wave momentum and energy fluxes transferred through the middle
atmosphere to the upper atmosphere. Journal of Atmospheric and Terrestrial Physics, 56(10), 1375–1385. https://doi.org/10.1016/0021‐
9169(94)90075‐2
Zhang, S. P., & Shepherd, G. G. (2004). Solar influence on the O(1D) dayglow emission rate: Global‐scale measurements by WINDII on
UARS. Geophysical Research Letters, 31, L07804. https://doi.org/10.1029/2004GL019447
Zhang, S. P., & Shepherd, G. G. (2005). On the response of the O(1S) dayglow emission rate to the Sun's energy input: An empirical model
deduced from WINDII/UARS global measurements. Journal of Geophysical Research, 110, A03304. https://doi.org/10.1029/
2004JA010887

MANDAL ET AL. 13

S-ar putea să vă placă și