Indian Journal of Radio & Space Physics
Vol 42, October 2013, pp 340-355
Migrating and non-migrating diurnal and semi-diurnal tides over a
tropical and an equatorial station
K N Uma$,*, Karanam Kishore Kumar & Siddarth Shankar Das
Space Physics Laboratory, Vikram Sarabhai Space Centre, ISRO-PO, Thiruvananthapuram 695 022, India
$
E-mail: [email protected]
Received 22 January 2013; revised 13 May 2013; accepted 16 May 2013
VHF radar measurements of hourly winds in the lower atmosphere are utilized to study the diurnal and semi-diurnal tides
over two key geographical locations, Gadanki (13.45°E, 79.18°E), India and Kotatabang (0.2°S, 100.2°E), Indonesia. The
analysis shows that tidal amplitudes exhibit maxima in upper troposphere lower stratosphere (UTLS) region during
June-September over Gadanki and during March and September over Kotatabang. The vertical wavelength is 3-5 km at
Gadanki and 25-30 km at Kotatabang in UTLS region, which reveal the existence of non-migrating and migrating tides,
respectively. The analysis shows the maximum rainfall during June-September over Gadanki, whereas uniform rainfall
throughout the year over Kotatabang. The brightness temperature shows deep (at Gadanki) and shallow (at Kotatabang)
convective clouds during these periods. The release of latent heat due to deep convective cloud is found to be the main
source mechanism for non-migrating tides at Gadanki. The significance of the present study lies in bringing out the
important differences in tidal characteristics over equatorial and low latitude stations.
Keywords: Migrating tides, Non-migrating tides, Rainfall, Convective clouds, Brightness temperature
PACS Nos: 92.60.hh; 92.60.Nv; 92.60.jf
1 Introduction
Atmospheric tidal oscillations are global scale
oscillations with periods equal to the harmonics of
one solar day. The absorption of solar radiation by
water vapour in the troposphere and ozone in the
stratosphere sets up a thermal forcing which sets the
atmosphere in to oscillations with a period of integral
fractions of a day1. These tides dictate the energetics
and dynamics of the middle and upper atmosphere.
The classical theory of tides by Chapman & Lindzen2
extensively reviews the characteristics of tides and
they have shown that the general characteristics of
tides in the upper atmosphere agree well with the
theoretical predictions. The diurnal tides propagate
vertically before saturation or dissipation in the
mesosphere and lower thermosphere region. Tidal
variability can be observed in many atmospheric
parameters ranging from surface pressure3,4 to tropical
convection5 to upper atmospheric winds6. Two modes
of tides are generally observed in the atmosphere:
(i) migrating tides which are sun synchronous with
their phase propagating to the west traversing the
globe in one solar day; and (ii) non-migrating which
are non synchronous with the sun having zonal
numbers different than one. Generally, non-migrating
tides are generated due to the asymmetry in the
distribution of water vapour and ozone. This
asymmetry is caused by the distribution of continents
and oceans as well as the local topography7-9. Nonmigrating tides may also be generated by latent heat
release by convective clouds. The migrating
components have high phase speed than the nonmigrating mode and hence, they are less susceptible to
critical interactions with the background wind; and
any seasonal variability observed in the lower
stratosphere is due to critical wave filtering of nonmigrating components10.
Extensive studies have been made in the past on
diurnal and semi-diurnal tides in the mesosphere and
lower thermosphere but only few studies were carried
out in the troposphere using radiosondes, radars,
models and satellites. Wallace & Tadd11 detected the
distribution of diurnal tides and found significant
regional deviations using 12 h radiosonde
observations. McKenzie12 accounted the diurnal
amplitudes below 30 km by simplifying the heat
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
sources in both latitude and longitude in the numerical
model. Using gravity wave f-plane model, Forbes &
Groves8 have found that the influence of nonmigrating components can change the tidal
amplitudes in the lower thermospheric region by
15-20%. Tsuda & Kato9 developed a numerical model
for diurnal non-migrating tides and found that the
upward heat flux in the planetary boundary layer is
the main source for non-migrating tides. Miyahara
et al.13, using the Kyushu University general
circulation model, found that the forcing for diurnal
tide is due to global distribution of solar insolation
heating, latent heating, convection heating and eddy
conduction heating. Ekanayake et al.14 have made use
of two dimensional steady state model and
characterized the non-migrating tides both at low and
high latitudes. The infra-red cloud brightness
temperature was also used to study the tidal
variability15-18. Tsuda et al.19,20 carried out intensive
campaigns over the equatorial region with radiosonde
and radar measurements to study the tidal oscillation
on wind components in the lower and middle
atmosphere. They found that the overall structure of
the observed wind velocity agree well with the
models above 20 km. The complicated structures and
observed discrepancy in the wind velocity below
20 km is attributed to the effects of local topography.
The tidal studies over China using radiosonde were
found to be dominated by diurnal non-migrating
components21. Using Arecibo radar, studies on diurnal
and semi-diurnal winds in the lower stratosphere were
carried out by Fukao et al.22. They found that the tidal
periods show strong downward phase propagation and
also their amplitudes of both diurnal and semi-diurnal
winds were found to be larger than the values
predicted by classical tidal theory. Whiteman &
Bian23, using Doppler radars at three locations, found
that the semi-diurnal tide in the boundary layer
compared well with the model but above the tidal
amplitudes were more than a factor of two compared
with that of the model. Williams & Avery15 studied
the diurnal winds in the tropical troposphere using
network of wind profiles across the equatorial pacific.
They found that the observations agree with the
model winds off the equator than the observations
near the equator. They attributed this to forcing
caused by solar absorption of water vapour. Riggin
et al.10 studied the seasonal structure of diurnal tides
using VHF radar at Jicamarca. They found that
341
vertical wavelengths of about 10 km were found
above the tropopause with standing wave structure
below.
Only limited studies have been made in the past
over the observational site Gadanki (13.45°N and
79.18°E). Sasi et al.24 studied the diurnal tidal
oscillations
using
Mesosphere
Stratosphere
Troposphere (MST) radar located at Gadanki. They
found that the non-migrating modes are dominant
over this region and they also found that the observed
amplitudes of the non-migrating modes compared
well with that of the model by Tsuda & Kato9.
Sasi et al.25 have also simulated the amplitudes and
phases of the diurnal tide using classical tidal theory
by incorporating different sources and short wave
solar radiation absorption by clouds. The above
studies used wind data collected over a certain period,
that is, during September-October 1995. Dutta
et al.26 have also compared the tidal amplitudes and
phases with the global scale wave model. However,
these studies were limited to certain seasons of the
year and there was no attempt to study the annual
variation of tidal amplitudes. In order to characterize
the tidal oscillations over this low latitude site, long
term data accounting for the monthly variability is
required. Thus, the main objective of the present
study is to establish the monthly variability of both
diurnal and semi-diurnal tides using four years of
observations from Indian MST Radar (IMSTR)
located at a low latitude station Gadanki (13.45°N,
79.18°E). The second objective is to compare the
amplitudes and phases of both diurnal and semidiurnal tides at Gadanki with that of the equatorial
station Kotatabang (0.2°S, 100.2°E), Indonesia, where
the observations were utilized using Equatorial
Atmosphere Radar (EAR). The observations of
diurnal winds over the Indian region are compared
with that of Kotatabang to better understand the
differences and similarities in the low latitude and
equatorial station diurnal tides, if any.
2 Data and System description
The present analysis is based on measurements of
diurnal winds made by two powerful VHF wind
profilers, IMSTR and EAR, over four years
(2002-2005) and one year (2003), respectively.
Complete details of the IMSTR and EAR are given by
Rao et al.27 and Fukao et al.28, respectively. The
diurnal cycle (24 h) data collected with the IMSTR
and EAR are used for the present study. However, the
observation strategies are different at both locations.
342
INDIAN J RADIO & SPACE PHYS, OCTOBER 2013
At Gadanki, the radar was operated in diurnal cycle
mode for typically two days in a month, while
continuous measurements were available for more
than 25 days per month in most of the months at
Kotatabang. So, only one year of EAR data has been
used to study the diurnal wind pattern as the available
data proved to be sufficient to study such a pattern.
Quality checks were made on the radar observations
to eliminate the data associated with weak echoes.
The quality controlled data have been used to
construct hourly averages.
The data utilized in the present study consists of
32 days (diurnal cycles) of IMSTR observations
collected during January 2002-December 2005.
During this period, the IMSTR has been operated with
16 µs pulse width and 1 µs baud length providing a
range resolution of 150 m. The data have been
recorded from 3.6 km to 30 km. However, useful
returns were obtained only up to a height of ~20 km.
The IMSTR main beam can be positioned anywhere
in a 20° cone in two orthogonal directions (EW and
NS) with 1° resolution. However, during the
observational period, IMSTR has been operated with
6 beams (two in zenith and four in 10° off-zenith
directions, E10, W10, Zx, Zy, N10, and S10)
providing a temporal resolution of 2 minutes.
Spurious data points have been removed from the
analysis using the methodology given by Uma &
Rao29. The noise level is found to vary with season
from –17 and –19 dB with a 2σ (σ is standard
deviation) level of 3 dB. The data is considered only
when its corresponding signal-to-noise ratio (SNR) is
greater than the above threshold. If there is a data gap
of more than ~5 h in a 24 h time series, then the data
corresponding to that height is discarded; and if the
missing data points are less than 5 h, then the gaps are
filled using cubic spline interpolation. Since the data
collection procedures are different over both Gadanki
and Kotatabang, around ~320 days of observations
over EAR were available. Before estimation of winds
from IMSTR, a thorough quality check has been done
for the signal detectability. Data have been discarded
if the SNR is less than -15 dB for IMSTR for the
estimation of winds. Winds are estimated using
adaptive and least square analysis method (details
presented by Anandan et al.30). On the other hand,
over EAR winds data, downloaded from www.
rish.kyoto-u.ac.jp/ear/data/index.html,
has
been
already quality controlled and the quality check of
signal detectability is presented by Fukao et al.28.
Figure 1 shows the height profiles of percentage of
wind data after quality check used for the present
study for both IMSTR and EAR for each month along
with the mean profile. It is observed that for IMSTR,
>90% of data has been used below 10 km, >50% data
has been used between 12-16 km and above that
>70% data is used. Similar pattern is observed over
Kotatabang, however, the percentage of data utilized
is found to vary for each month below 10 km.
3 Results and Discussion
Figure 2(a) shows time-height cross sections of
monthly averaged zonal winds from 3.6 to 20 km,
with a vertical resolution of 150 m from January 2002
- December 2005. The diurnal data available in each
hour have been averaged and considered as a
representative of that hour. These hourly profiles are
then averaged for all the days available in that
particular month. It is seen from the figure that above
6 km, the winds are westerlies during the winter
(December-February) and pre monsoon months
(March-May). However, the magnitude of zonal
winds decreases in pre-monsoon as compared to
winter season. The maximum magnitude of the zonal
winds during winter season (December-February)
goes up to 15 ms-1. Below 6 km, generally, easterlies
are observed in the winter, post monsoon (OctoberNovember) and pre monsoon (March-May) seasons.
During monsoon season (June-September), the local
weather is dominated by the Asian summer monsoon
and hence, strong easterlies, popularly known as
tropical easterly jet (TEJ), are clearly seen from the
figure at a height of 12-17 km and the magnitude of
these jet decreases during post monsoon season.
During monsoon, the winds are westerlies below 4 km
and are known as low level jet (LLJ). During the
monsoon season, strong winds of nearly 35 ms-1, on
an average, exist around 16 km height. The easterlies
in the post monsoon season are relatively smaller in
magnitude, with a maximum of 20 ms-1. Monsoon
season is more prone for tropical convection and
intense amount of latent heat is released during this
season which forms the main source for generation of
gravity waves and non-migrating tides.
Figure 2(b) shows time-height cross sections of
monthly averaged meridional winds from 3.6 to
20 km, with a vertical resolution of 150 m during
January 2002-December 2005. Generally, above
7 km, the meridional winds are southerlies in the
winter, pre monsoon, and post monsoon seasons, with
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
343
Fig. 1 — Percentage of occurrence of the data base at: (a) Gadanki and (b) Kotatabang
Fig. 2 — Monthly mean derived from MST radar measurements over Gadanki of: (a) zonal wind, (b) meridional wind, and
(c) vertical wind
344
INDIAN J RADIO & SPACE PHYS, OCTOBER 2013
maximum magnitudes of 5 and 2 ms-1 in the monsoon
and post monsoon seasons, respectively. Around 1213 km, a jet-like wind pattern with strong southerlies
is seen during post monsoon, winter and pre monsoon
with varying core thickness. The maximum
magnitudes of 6–7 ms-1 is observed. The height of
these jet-like winds is lower than that of the tropical
easterlies in zonal winds. Figure 2c shows time-height
cross sections of monthly averaged vertical winds
from 3.6 to 20 km, with a vertical resolution of 150 m
during January 2002 - December 2005. The positive
and negative magnitudes represent upward and
downward motion, respectively. Below 5 km, the
vertical wind is downward during pre monsoon,
monsoon and post monsoon seasons. During monsoon
and post monsoon season, strong upward motion is
seen in the upper troposphere. Earlier studies have
shown similar observations in vertical velocities over
Gadanki using 10 years of data29.
In order to understand the diurnal and semi-diurnal
tidal variability, the diurnal and semi-diurnal
components are extracted using least square analysis.
For each day, the diurnal and semi-diurnal amplitudes
are determined from the hourly averaged diurnal data
using least square fitting to the time series at each
height levels. The reconstructed time height section of
zonal, meridional and vertical wind for both semidiurnal and diurnal components are shown in
Figs 3(a-f), respectively. The phase fronts of the wave
can be easily identified from the figures. The
downward (upward) phase propagation indicates the
presence of an upward (downward) propagating tide.
Figs 3(a-b) show the reconstructed signal for zonal
semi-diurnal and diurnal tides. The amplitude is found
to be 4-5 ms-1, and the figures clearly show the
upward and downward phase propagation of tides.
One interesting feature to be noted is that both the
upward propagating tides from the troposphere and
downward propagating tides from the stratosphere
interact at the tropopause. However, the present
interest is to study the amplitudes and phases of these
propagating tides and their sources. Similar kind of
Fig. 3 — Reconstructed height-time intensity map of: (a) zonal, (b) meridional, and (c) vertical winds, obtained using least square fitting
during a typical day on 30-31 July 2005
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
propagation can be seen in the meridional wind also
[Figs 3(c-d)], but the amplitudes are relatively less
than that of zonal wind. Figures 3(e-f) show the
reconstructed signal for vertical wind and one can
observe the magnitudes are an order of magnitude less
than that of the zonal and meridional wind. The
detailed study of the diurnal variability of vertical
wind is studied by Uma & Rao29. They have observed
the semi-diurnal component to be dominant in the
middle/upper troposphere as well as in the lower
stratosphere over the present observational site.
Figure 4(a-b) shows the monthly variation of
amplitude and phases of the diurnal tides in the zonal
wind for all the months. The amplitudes show an
increasing trend with height for all the months. In the
upper troposphere and lower stratosphere (UTLS)
(16-20 km), the amplitudes are high during June to
September as compared to other months. The
amplitudes ranged 4-6 ms-1 during the above months
with maximum in the month of June. The amplitudes
345
are showing maxima near the tropopause. The
secondary maximum is observed in the middle
troposphere (8-12 km) during March and September. In
the lower troposphere (4-5 km), amplitudes are
relatively high during the months of April and May.
The amplitudes reported here are comparable with the
magnitudes obtained by Sasi et al.24 for a single day in
the month of September. The corresponding phase
profiles are given besides the amplitude for each month
in Fig. 4. The height profile of phase shows large
variations in all the months. The phase profile in the
UTLS region remains constant with less variations.
Semi-diurnal amplitudes and phases in the zonal wind
are shown in Figs 4(c-d). The semi-diurnal amplitude
shows similar variation as that of diurnal tidal
amplitudes but their magnitudes are less compared to
diurnal tides. The variability in the semi-diurnal tidal
phase structure is more than the diurnal tide.
Figure 5(a-b) shows the amplitude and phases of
the diurnal tide in meridional winds for all the months
Fig. 4 — Monthly height variation of: (a) amplitude and (b) phase, of diurnal tide in zonal wind; (c) amplitude and (d) phase, of
semi-diurnal tides in zonal wind; over Gadanki
346
INDIAN J RADIO & SPACE PHYS, OCTOBER 2013
similar to Fig. 4. In the UTLS region, the amplitudes
are high during June - September, with maximum in
the month of June and minimum in the month of
August. One interesting feature is that the amplitudes
show dominance throughout the troposphere during
the month of March, which is not observed in the
zonal wind. In the lower troposphere, as similar to
zonal wind, the amplitudes show maxima during
March-May. During October-November, a peak is
observed in the height region of 14-16 km. The
amplitudes observed are relatively less compared to
that of zonal diurnal tides in the UTLS region. The
observed phase profiles exhibit irregular structures
except in the UTLS region where propagation is
observed. But one should note that if these irregular
features are removed, the phase profiles appear to be
constant as in the phase profile of zonal diurnal tides.
Figures 5(c-d) show the amplitude and phase profiles
of semi-diurnal component of meridional wind for all
the months similar to Figs 4(c-d). The amplitudes
show increasing trend with altitude for all the months
as similar to zonal tide amplitudes. The amplitude
shows maxima in the UTLS region during JuneSeptember. In the middle troposphere (8-12 km), the
amplitudes are dominant from March-November, and
in the lower troposphere (4-6 km), the amplitudes
show similar pattern like zonal semi-diurnal tide. The
phase profile remains almost constant with irregular
structures throughout all the months, but propagation
is observed in the UTLS region.
Figures 6(a-d) show amplitudes and phases of the
diurnal and semi-diurnal tides in the vertical wind. It
is very interesting to note that during the month of
June and July, the amplitudes show significant peak
from middle troposphere and extend to lower
stratosphere. This dominant amplitude is seen in the
UTLS region during August and September. The
amplitudes are, in general, an order of magnitude less
Fig. 5 — Monthly height variation of: (a) amplitude and (b) phase, of diurnal tide in meridional wind; (c) amplitude and (d) phase, of
semi-diurnal tides in meridional wind; over Gadanki
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
347
Fig. 6 — Monthly height variation of: (a) amplitude and (b) phase, of diurnal tide in vertical wind; (c) amplitude and (d) phase, of
semi-diurnal tides in vertical wind; over Gadanki
than that of the zonal and meridional diurnal and
semi-diurnal tide amplitudes. This is because the
magnitude of the vertical wind is itself an order of
magnitude less than that of zonal and meridional
winds29. The phase profiles are almost constant during
all the months except for the month of August, where
some variability is seen in the lower troposphere.
Using few days of data, Dutta et al.26 have shown the
variation of diurnal, semi-diurnal and ter-diurnal tides
in the zonal and meridional wind over Gadanki using
MST radar and compared with the Global Scale Wave
Model (GSWM). They have also observed similar
phase structure in different seasons and have reported
that no apparent variations of phases can be traced
with seasons. The observed phases also found to be in
excellent agreement with the modeled phases
especially in the meridional wind. Now as a next step,
the vertical wavelength is estimated from the phase
profiles for both diurnal and semi-diurnal tides in
zonal, meridional and vertical wind, respectively and
a typical example is presented in Fig. 7. The vertical
wavelength is estimated in the UTLS region where a
clear amplitude enhancement is seen in the above
mentioned region. Although, the phase propagation
characteristics are not clearly seen in the lower
troposphere, one can see a clear propagation in the
UTLS region from the typical example shown in
Fig. 7. The vertical wavelength from this example is
found to be ~3 km for both diurnal and semi-diurnal
tides in the zonal, meridional and vertical winds.
Generally, as mentioned above, in the earlier studies,
the presence of tides are more clearly observed above
the troposphere as convective motions, waves from
different sources dominate the lower troposphere.
Riggin et al.10 have also mentioned that the diurnal
tide observed to be propagated above the tropopause,
as below this height the dominant motion were
convective motions that exhibited little phase
348
INDIAN J RADIO & SPACE PHYS, OCTOBER 2013
Fig. 7 — Phase profile for: (a) zonal, meridional and vertical semi-diurnal tide; and (b) zonal, meridional and vertical diurnal tide
progression with time. Over Gadanki, the tropopause
is situated around ~16 km (Ref. 31), and in the present
study the phase propagation above the tropopause is
observed. Liberman & Leovy32 investigated the
response of the middle atmosphere to longitudinally
dependent radiative, sensible and latent heat profiles.
They found that the tides generated from sensible heat
have higher amplitudes in the lower atmosphere as
compared to the middle atmosphere. But on the other
hand, they found that the tides triggered from solar
heating and latent heat release contribute less in the
troposphere. Williams & Avery15 have also found that
diurnally migrating precipitating clouds generate
non-migrating modes whose presence is more
observed in the mesosphere where they deposit
enormous momentum. The vertical wavelength
calculated for all the days during June - September in
the present study varied ~3-5 km in the zonal,
meridional and vertical wind. The observed
amplitudes in the UTLS region are about 3-4 ms-1.
Tsuda & Kato9 showed through numerical model that
non-migrating components in the horizontal winds in
the UTLS region are characterized by short vertical
wavelengths (2-5 km) and fluctuating amplitudes with
height (1-3 ms-1). Observations made by VHF radars
have clearly shown the presence of large amplitude
and short vertical wavelength non-migrating tidal
oscillations in the lower stratosphere22,33.
From the above results, one can observe that
diurnal and semi-diurnal tides are prominent in the
UTLS region during June-August. In addition to it,
their presence is also felt in the lower troposphere
during March as well as in September, but their
amplitudes are less compared to the UTLS region. At
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
first sight, it gives an impression that the
non-migrating tides are dominant over this region
(Gadanki), as they are observed during the monsoon
season. The amplitudes of migrating tides, generally,
show maxima in the equinoctial period (March and
September)34. During June-September, the Indian
region is dominated by the monsoon convection
activity29,35 and hence as a resultant, large amount of
latent heat is released, which is the main source for
non-migrating tides. Sasi et al.25 have compared the
amplitudes and phases of diurnal tides over Gadanki
with the theoretical model of Forbes & Gorves8 and
observed that these tidal oscillations are manifestation
of non-migrating modes of diurnal oscillations. Dutta
et al.26 have compared the diurnal, semi-diurnal and
ter-diurnal tidal oscillations with GSWM over
Gadanki and observed that during June-August
(convective season), the observed amplitudes are
found to be much larger than the model indicating that
the non-migrating modes are dominant over this study
region. Sasi et al.24 and Jani et al.36 have also observed
similar amplitudes and vertical wavelengths over
Gadanki. The vertical wavelengths inferred from
amplitudes and phases varied 3-9 km in a study by
Dutta et al.26 over the present observational site. All
the above studies have attributed that the observed
characteristics of tidal activity over Gadanki are that
of non-migrating tides and they have concluded that
the latent heat release by deep convective clouds have
short vertical wavelength and large amplitudes. So,
the present results very well agree with the previous
studies over this location, which clearly indicates that
there is dominance of non-migrating tides in the
UTLS region during monsoon season. But the earlier
studies were for a limited period and the present study
spans over four years, as long term observations are
very much essential to have a meaningful
understanding of the tidal variability. For example,
the presence of migrating tides could be seen in the
present study over Gadanki, which could be
evidenced from the enhanced amplitudes (secondary
maxima) during March and September in 8-12 km
altitude. But the amplitudes are less compared to that
observed during June-August. An attempt is made to
calculate the vertical wavelength during March and
September (figures not shown). The phase profiles
were highly variable and the vertical wavelengths
varied in the range 12-18 km in the UTLS region. The
range of vertical wavelengths observed during March
and September indicates that they might be generated
349
by solar short wave heating of water vapour/clouds.
This has to be investigated in detail and will be taken
up in near future.
The above analysis indicates that non-migrating
tides are dominant over this low latitude site during
June-August. Also, the variation of tidal activity over
an equatorial station, Kotatabang is studied to
delineate differences/similarities with that of Gadanki,
if any. The observational altitude over Kotatabang
using EAR is from 2 km to 19.05 km. The same
exercise is repeated by extracting the diurnal and
semi-diurnal oscillations using least square analysis in
the zonal, meridional and vertical wind. Figure 8(a-d)
shows the monthly variation of amplitude and phases
of the diurnal tides for zonal diurnal tide for all the
months. One can observe from the figure that the
diurnal and semi-diurnal oscillations are having
maximum amplitude during March and September in
the middle troposphere and in the UTLS region,
which is in contrast to Gadanki where the dominance
is observed during June-August. The amplitudes are
in the range of 5-6 ms-1 and it is slightly greater than
that observed over Gadanki. Figure 9(a-d) shows the
monthly variation of amplitude and phase of the
diurnal and semi-diurnal tide in the meridional wind.
The features observed are similar to zonal wind but
during the month of March, the amplitude maxima is
observed throughout the depth of the troposphere.
Figure 10(a-d) shows diurnal and semi-diurnal
amplitudes in the vertical wind. The dominant
amplitude is seen only in the month of March in the
middle troposphere, unlike zonal and meridional
winds, where it is observed in September also. The
phase profiles given besides the amplitudes of the
diurnal and semi-diurnal tide in zonal, meridional and
vertical wind, respectively do not show much
irregular structure at Gadanki and they remain
constant throughout the observational altitude in a
month. The vertical wavelength calculated from the
phase profiles varied ~25-30 km. The wavelength,
that is observed, is consistent with that of the
migrating diurnal component8. Similar constant phase
structure is also observed during the coupling process
in the equatorial atmosphere (CPEA) campaign using
radiosondes over Indonesia36. Forbes et al.16 have
shown that large vertical wavelengths are generally
excited by the migrating mode (1,1) and higher order
propagating modes have much shorter vertical
wavelengths and hence, are severely damped by
the turbulent dissipation. The above studies indicate
350
INDIAN J RADIO & SPACE PHYS, OCTOBER 2013
Fig. 8 — Monthly height variation of: (a) amplitude and (b) phase of diurnal tide in zonal wind; (c) amplitude and (d) phase of
semi-diurnal tide in zonal wind; over Kotatabang
Fig. 9 — Monthly height variation of: (a) amplitude and (b) phase of diurnal tide in meridional wind; (c) amplitude and (d) phase of
semi-diurnal tide in meridional wind; over Kotatabang
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
351
Fig. 10 — Monthly height variation of: (a) amplitude and (b) phase of diurnal tide in vertical wind; (c) amplitude and (d) phase of
semi-diurnal tide in vertical wind; over Kotatabang
that the tidal structures observed over Kotatabang
must be that of the migrating mode as they peak
during the equinoctial months and the observed
vertical wavelengths are large compared to the
non-migrating modes.
From the above results, it gives an indication that
non-migrating tides are dominant over Gadanki and
migrating tides over Kotatabang. Earlier studies
clearly indicate that heating due to water vapour and
ozone gives rise to migrating tidal modes and latent
heat release due to convection gives rise to nonmigrating tidal modes. Generally, precipitation can be
used as a proxy for latent heat release as it involves
phase transformation of water vapour. Hence, the
rainfall measurements are used by surface optical rain
gauge (ORG) over Gadanki and Kotatabang. ORG
over Gadanki provides high resolution (1 minute)
rainfall rate data with good accuracy (95% on
accumulation data). The continuous measurements of
ORG are available from 1999 over Gadanki and rain
rate is obtained over Kotatabang during the CPEA
campaign during 2002-2006. The monthly
accumulated rain rate is calculated during 2002-2005
and is shown in Fig. 11(a and b) for Gadanki and
Kotatabang, respectively along with their mean and
standard deviation. It is clearly observed from the
figure that during June-September, which is generally
the monsoon season over India, the accumulated rain
rate is maximum compared to the other months over
Gadanki. During the year 2002, the rainfall was less
as it was a drought over the Indian region and hence,
the mean value is slightly shifted to lower value. The
mean rain rate has been shifted toward maximum
value in the month of October during the year 2005,
which is due to intense cyclonic activity during that
period38. But on an average, the rain rate is maximum
during June-September and one can expect that
release of latent heat will also be more during these
months. This may be the plausible reason for the
dominance of non-migrating modes in the UTLS
region over Gadanki. On the other hand, the rainfall
pattern over Kotatabang [Fig. 11(b)] clearly exhibits
that it rains throughout the year over this region.
There is little variation in the observed rainfall with
352
INDIAN J RADIO & SPACE PHYS, OCTOBER 2013
water vapour to water (latent heat of condensation),
whereas in deep clouds (cold clouds), phase
transformation of water to ice (latent heat of freezing)
is also involved, thus releasing large latent heat.
Fig. 11 —Monthly mean of accumulated rain rate during
2002-2005 along with mean and standard deviation for:
(a) Gadanki, and (b) Kotatabang
respect to months, but the variability seems to be not
significant. So, from the rainfall pattern, one can
expect that the non-migrating modes are dominant
over this region throughout the year, which is not the
case as one can observe from Figs (8-10) where the
maximum tidal activity is seen only during March and
September. Even though the rainfall, an indicator of
convection, was observed throughout the year over
Kotatabang, surprisingly one could not observe nonmigrating tides in the UTLS region, which raises a
question that whether this rainfall is caused by deep
convective clouds or shallow convective clouds. On
an average, deep and shallow convective clouds have
vertical extent of 10-12 km and 3-4 km, respectively.
The vertical extent of the cloud is very important in
estimating the latent heat release. Shallow clouds
(warm clouds) involve only phase transformation of
The next attempt has been to identify whether
clouds formed over Gadanki and Kotatabang are deep
or shallow convective clouds. To explore this, the
cloud top brightness temperature (Tbb) is processed,
which is the intensity of the radiation emitted by a
black body and sensed by radiometers. The data used
for the present study is derived from the infrared
channels (IR1) of the Geostationary Meteorological
Satellite (GMS5 and GMS6) and Geostationary
Operational Environmental Satellite (GOES 9)
(Ref. 39). The three hourly data during 2003-2005 is
used, which have a high resolution of 0.05°×0.05° to
represent the extent of clouds over Gadanki and
Indonesia. Methods to identify tropical deep
convective clouds using infrared measurements are
generally based on thresholds of cloud top
temperatures, which have been defined with different
values. Mapes & Houze40 and Zeudima41 have
classified deep convective clouds when Tbb is less
than 235 K. The clouds with cloud top temperatures
below 240 K include deep high-top precipitating
cloud, anvil with stratiform precipitating cloud and
deep convective precipitating cloud. So, the same
criteria is followed to define the convective clouds
over both the study regions and the monthly
variability of Tbb is given in Fig. 12(a) for Gadanki
and Fig. 12(b) for Kotatabang. The figure shows
mean Tbb profile for 2002-2005 (thick) along with
standard deviations. Over Gadanki during DecemberMarch, Tbb is very high ~280 K and during
April-May, it varies around ~250-270 K. During
monsoon (June-August), the Tbb is less than 240 K.
During September, the value of Tbb is 242 K just
above the threshold. During October-November, it
varied from ~250 to 260 K. So from this figure, it is
clearly evident that during June–September, deep
convective clouds prevail over Gadanki, which
releases large amount of latent heat, thereby,
triggering the non-migrating tidal modes over this
location. On the other hand, over Kotatabang, the
value of Tbb is always greater than 260 K. During
January-April, it varies from ~275 to 280 K and from
May-December, the variability is very less and it is
around 263 K. This clearly reveals that the rainfall
over Kotatabang is mostly due to the shallow
convective clouds and so the possibility of source
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
353
analysis for extracting diurnal and semi-diurnal
amplitudes and phases. The amplitudes were
dominant in the UTLS region over Gadanki during
June-September, whereas over Kotatabang, it was
during March and September. The phase profiles
revealed a vertical wavelength of ~3-4 km over
Gadanki during large tidal amplitudes (JuneSeptember) and ~25-30 km over Kotatabang. Further,
the brightness temperature analysis over these
locations revealed that deep clouds dominated during
the monsoon season over Gadanki, whereas over
Kotatabang most of the time shallow clouds were
observed. The analysis showed that maximum rainfall
occurred during June-September over Gadanki,
whereas uniform rainfall was observed throughout the
year over Kotatabang. The present results revealed
that non-migrating tides are dominant over Gadanki
and migrating tides are dominant over Kotatabang.
The presence of deep convective clouds over Gadanki
is found to be the main reason for the dominance of
non-migrating tides. Even though, the rainfall persists
throughout the year over Kotatabang, the migrating
mode is found dominant over Kotatabang. This is
attributed to the shallow convection taking place over
this region, which are less efficient in releasing latent
heat as compared to deep convection. Thus, the
present observations revealed the tidal characteristics
over two geographical locations within the tropics,
which will have implications on modeling of tides and
their variability.
Fig. 12 — Monthly mean of brightness temperature during
2003-2005 along with standard deviation for: (a) Gadanki, and
(b) Kotatabang
(latent heat) triggering the non-migrating tidal modes
is expected to be lesser compared to Gadanki. The
presence of shallow convective clouds may be one of
the plausible reasons for the absence of non-migrating
modes; thereby, the dominance of migrating
components is observed over Kotatabang. The present
study, thus, clearly reveals the variability of tidal
characteristics over low latitude and equator, which
further needs to be substantiated with global hourly
observations of geophysical parameters, which will be
the next exploration on tides.
Acknowledgement
The Gadanki VHF radar belongs to National
Atmospheric Research Laboratory (NARL), under
Department of Space, Government of India. The
authors would like to thank NARL Director and
technical staff for their support in conducting the
radar experiments. The authors thank Dr M V Ratnam
and Dr Y Shibagaki for providing high resolution
brightness temperature datasets. They also thank
Equatorial Atmosphere Radar (EAR) scientists and
engineers for providing the radar data over
Kotatabang.
References
4 Conclusions
VHF radar observations have been utilized to study
the diurnal and semi-diurnal tides over Gadanki and
Kotatabang. The VHF radar measured hourly winds
over both these locations are subjected to least square
1
2
3
Lindzen R S, Thermally driven diurnal tide in the
atmosphere, Q J R Meteorol Soc (UK), 93 (1967) pp 18-42.
Chapman S & Lindzen R S, Atmospheric tides (D Reidel,
Norwell, Mass), 1970.
Haurwitz B & Cowley A D, The diurnal and semi-diurnal
barometric oscillations: Global distribution and annual
354
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
INDIAN J RADIO & SPACE PHYS, OCTOBER 2013
variation, Pure Appl Geophys (Switzerland), 103 (1973)
pp 193-222.
Hamilton K, Latent heat release as a possible forcing
mechanism for atmospheric tides, Mon Weather Rev (USA),
109 (1981) pp 3-17.
Salby M L, Hendon H H, Woodberry K W & Tanaka K,
Analysis of global cloud imagery from multiple satellites,
Mon Weather Rev (USA), 72 (1991) pp 467-480.
Mason A H, Meek C E, Teitelbaum H, Vail F, Schminder R,
Kurschner D, Smith G D, Fraser G J & Clark R R,
Climatologies of semi-diurnal and diurnal tides in the middle
atmosphere (70-110 km) at middle latitudes (40-55o),
J Atmos Terr Phys (UK), 51 (1989) pp 579-593.
Kato S, Tsuda T & Wantanabe F, Thermal excitation of
non-migrating tides, J Atmos Terr Phys (UK), 44 (1982)
pp 131-146.
Forbes J M & Groves G V, Diurnal propagating tides in the
low latitude middle atmosphere, J Atmos Terr Phys (UK),
49 (1987) pp 153-164.
Tsuda T & Kato S, Diurnal non-migrating tides excited by a
differential heating due to land-sea distribution, J Meteorol
Soc Jpn (Japan), 67 (1989) pp 43-54.
Riggin D M, Kudeki E, Feng Z, Sarango M F & Liberman
R S, Jicamarca radar observations of the diurnal and
semi-diurnal tide in the troposphere and lower stratosphere,
J Geophys Res (USA), 107 (D8) (2002) 4062, doi:
10.1029/2001JD001216.
Wallace J M & Tadd R F, Some further results concerning
the vertical structure of atmospheric tidal motions within the
lowest 30 kilometers, Mon Weather Rev (USA), 102 (1974)
pp 795-803.
Mc Kenzie, The diurnal atmospheric tide with Newtonian
cooling and longitudinally dependent drices, Ph D thesis,
University of Washington, Seattle, 1968, 139.
Miyahara S, Yoshida Y & Miyoshi Y, Dynamical coupling
between the lower atmosphere by tides and gravity waves,
J Atmos Terr Phys (UK), 55 (1993) pp 1039-1053.
Ekanayake E M P, Aso T & Miyahara S, Background wind
effect on propagation of non-migrating tides in the
middle atmosphere, J Atmos Sol-Terr Phys (UK), 59 (1997)
pp 401-429.
Williams C R & Avery S K, Diurnal non-migrating tidal
oscillations forced by deep convective clouds, J Geophys Res
(USA), 101 (1996) pp 4079-4091.
Forbes J M, Hagan M E, Zhang X & Hamilton K, Upper
atmospheric tidal oscillations due to latent heat release in
the tropical troposphere, Ann Geophys (France), 15 (1997)
pp 1165-175.
Hagan M E & Forbes J M, Migrating and non-migrating
diurnal tides in the middle and upper atmosphere excited by
tropospheric latent heat release, J Geophys Res (USA), 107
(D24) (2002) 4754, doi: 10.1029/2001JD001236.
Lieberman R S, Ortland D A & Yarosh E S, Climatology and
inter annual variability of diurnal water vapor heating,
J Geophys Res (USA), 108 (D3) (2003) 4123, doi:
10.1029/2002JD002308.
Tsuda T, Murayama Y, Wiryosumarto H, Harijono S W B &
Kato S, Radiosonde observations of equatorial atmosphere
dynamics over Indonesia Pt 1: Equatorial waves and diurnal
tides, J Geophys Res (USA0, 99 (1994) pp 10,491-10,505.
20 Tsuda T, Nakamura T, Shimizhu A, Yoshino T, Harijono
S W B, Sribimawati T & Wiryosumarto H, Observations of
diurnal oscillations with a meteor wind radar and radiosondes
in Indonesia, J Geophys Res (USA), 102 (D22) (1997)
pp 26,217-26,224.
21 Huang C M, Zhang S D & Yi F, Intensive radiosonde
observations of the diurnal tide and planetary waves in the
lower atmosphere over Yichang (111°180'E, 30°42'N),
China, Ann Geophys (Germany), 27 (2009) pp 1079-1095.
22 Fukao S, Kato S, Yokoi S, Harper R M, Woodman R F &
Gordon W E, One full-day radar measurement of lower
stratospheric winds over Jicamarca, J Atmos Terr Phys (UK),
40 (1978) pp 1331-1337.
23 Whiteman C D & Bian X, Radar wind profiler observations
of solar semi diurnal atmospheric tides, Geophys Res Lett
(USA), 22 (1995) pp 901-904.
24 Sasi M N, Ramkumar G & Deepa V, Non-migrating diurnal
tides in the troposphere and lower stratosphere over Gadanki
(13.5°N, 79.2°E), J Geophys Res (USA), 103 (1998) pp
19,485-19,494.
25 Sasi M N, Ramkumar G & Deepa V, Tidal wind oscillations
in the tropical lower atmosphere as observed by Indian MST
radar, Ann Geophys (Germany), 19 (2001) pp 991-999.
26 Dutta G, Bapiraju B, Balasubrahmanyam P, Aleem Siddique
M A & Aleem Basha H, Seasonal variation of solar tides in
the troposphere and lower stratosphere over Gadnki:
Comparisons with the global scale wave model, Radio Sci
(USA), 37 (2002), doi: 10.1029/2000RS002571.
27 Rao P B, Jain A R, Kishore P, Balamuralidhar P, Damle S H
& Viswanathan G, Indian MST radar Pt 1: System
description and sample vector wind measurements in ST
mode, Radio Sci (USA), 30 (1995) pp 1125-1138.
28 Fukao S, Hashiguchi H, Yamamoto M, Tsuda T, Nakamura
T, Yamamoto M K, Sato T, Hagio M & Yabugaki Y,
Equatorial Atmosphere Radar (EAR): System description
and first results, Radio Sci (USA), 38 (3) (2003) 1053, doi:
10.1029/2002RS002767.
29 Uma K N & Rao T N, Diurnal variation in vertical air motion
over a tropical station, Gadanki (13.5°N, 79.2°E), and its
effect on the estimation of mean vertical air motion,
J Geophys Res (USA), 114 (2009) D20106, doi:
10.1029/2009JD012560, 2009.
30 Anandan V K, Balamuralidhar P, Rao P B, Jain A R &
Pan C J, An adaptive moments estimation technique applied
to MST radar echoes, J Atmos Ocean Technol (USA),
22 (2005) pp 396-408.
31 Das S S, Jain A R, Kumar K K & Rao D N, Diurnal
variability of the tropical tropopause: Significance of MST
radar measurements, Radio Sci (USA), 43 (2008) RS6003,
doi: 10.1029/2008RS003824.
32 Lieberman R S & Leovy C B, A numerical model of nonmigrating diurnal tides between the surface and 65 km,
J Atmos Sci (USA), 52 (1995) pp 389-409.
33 Maekawa Y, Fukao S, HirotaI, Sulzer M P & Kato S, Some
further results on long term mesospheric and lower
thermospheric wind observations by the Arecibo UHF radar,
J Atmos Terr Phys (UK), 49 (1987) pp 63-71.
34 Forbes J M & Garrett H B, Theoretical studies of
atmospheric tides, Rev Geophys Space Phys (USA), 17
(1979) pp 1951-1981.
UMA et al.: MIGRATING AND NON-MIGRATING DIURNAL AND SEMI-DIURNAL TIDES
35 Kumar K K, Jain A R & Rao D N, VHF/UHF radar
observations of tropical mesoscale convective systems
over southern India, Ann Geophys (Germany), 23 (2005)
pp 1673-1683.
36 Jani Y N, Joshi H P & Iyer K N, Seasonal differences of nonmigrating tides in the troposphere and lower stratosphere
over Gadanki (13.5°N, 79.2°E), Indian J Radio Space Phys,
29 (2000) pp 210-221.
37 Alexander S P & Tsuda T, Observations of the diurnal tide
during seven intensive radiosonde campaigns in Australia
and Indonesia, J Geophys Res (USA), 113 (2008), doi:
10.1029/2007JD008717.
38 Das S S, Patra A K & Rao D N, VHF radar echoes
in the vicinity of tropopause during the passage of tropical
355
cyclone: First observations from the Gadanki MST radar,
J Geophys Res (USA), 113 (2008) D09113, doi:
10.1029/2007JD009014.
39 Miyakawa T & Satomura T, Seasonal variation and
environmental properties of southward propagating
mesoscale convective systems over the Bay of Bengal,
Sci Online Lett Atmos (Japan), 2 (2006) pp 088-091,
doi: 10.2151/sola.2006-023.
40 Mapes B E & Houze Jr R A, Cloud clusters and super
clusters over the oceanic warm pool, Mon Weather Rev
(USA), 121 (1993) pp 1398-1415.
41 Zuidema P, Convective clouds over the Bay of Bengal, Mon
Weather Rev (USA), 131 (2003) pp 780-798.