Geophysical Research Letters
RESEARCH LETTER
10.1002/2014GL060698
Key Points:
• Dissipating tides induce a net
meridional and vertical transport
of oxygen
• Net tidal transport is comparable to
tidal-induced net advective transport
• The migrating diurnal tide is primarily
responsible for net oxygen changes
Correspondence to:
M. Jones Jr.,
[email protected]
Citation:
Jones, M., Jr., J. M. Forbes, and M.
E. Hagan (2014), Tidal-induced net
transport effects on the oxygen
distribution in the thermosphere,
Geophys. Res. Lett., 41, 5272–5279,
doi:10.1002/2014GL060698.
Received 28 MAY 2014
Accepted 9 JUL 2014
Accepted article online 15 JUL 2014
Published online 24 JUL 2014
Tidal-induced net transport effects on the oxygen
distribution in the thermosphere
M. Jones Jr.1 , J. M. Forbes1 , and M. E. Hagan2
1 Department of Aerospace Engineering Sciences, University of Colorado Boulder, Boulder, Colorado, USA, 2 High Altitude
Observatory, National Center for Atmospheric Research, Boulder, Colorado, USA
Abstract Through a series of numerical experiments performed with the National Center for
Atmospheric Research Thermosphere-Ionosphere-Mesosphere-Electrodynamics General Circulation Model,
we evaluate a new mechanism by which the dissipation of vertically propagating tides acts to change the
O distribution in the thermosphere. Jones et al. (2014) proposed that the tides induced a net transport
of constituents themselves, in addition to the transport provided by the mean circulation induced by the
dissipation of tides. Through diagnosis of the continuity equation for [O], our results show that the net
meridional and vertical transport of O induced by the tides appreciably contributes to [O] changes in the
lower thermosphere. Combined with recombination, these transport mechanisms drive a net reduction
in [O] of ∼25% that is transmitted to higher altitudes by molecular diffusion. The migrating diurnal tide
appears to be the main driver of the [O] variations during September.
1. Introduction
Atomic oxygen (O) is fundamental to the aeronomy of the ionosphere-thermosphere system. The distribution of O in the upper thermosphere is controlled by the transport, photochemical, and diffusion processes
in the upper mesosphere and lower thermosphere (MLT). The MLT region (approximately 80–120 km) is subject to regular and repeatable forcing from the troposphere and stratosphere via a spectrum of vertically
propagating waves. Diurnal (24 h) and semidiurnal (12 h) thermal tides are one class of these waves that are
generated by the absorption of solar radiation and by latent heat release due to deep tropical convection.
The dissipation of atmospheric tides by eddy and molecular diffusion in the MLT leads to a deposition of
energy and momentum into the mean flow, thereby affecting the transport, photochemical, and diffusion
processes that determine the distribution of O near and above the turbopause.
The effects of eddy mixing, vertical winds, and atmospheric tides on the O distribution in the mesosphere and thermosphere have been studied since the 1960s [e.g., Colegrove et al., 1966; Shimazaki, 1968,
and references therein]. For example, an increase in eddy diffusivity will act to enhance the downward
transport of O to altitudes where recombination is strong causing a decrease in O density in the thermosphere [Colegrove et al., 1966; Shimazaki, 1968; Akmaev and Shved, 1980; Angelats i Coll and Forbes,
1998; Vlasov and Kelley, 2010, and references therein]. The downward transport and loss of O in the thermosphere is accompanied by an upward flow of and subsequent photodissociation of O2 . Akmaev and
Shved [1980], Angelats i Coll and Forbes [1998], Vlasov and Kelley [2010], and references therein have all
shown that the advection of O resulting from downward (upward) vertical winds in the lower thermosphere acts to change O in a similar manner to that of increased (decreased) eddy diffusion. Also, Roble
and Shepherd [1997] showed that the overall oxygen number density, [O], was considerably less for strong
versus weak diurnal tidal forcing at the lower boundary of the National Center for Atmospheric Research
(NCAR) Thermosphere-Ionosphere-Mesosphere-Electrodynamics General Circulation Model (TIME-GCM),
and attributed this to the dynamical effects associated with the diurnal tide. Using a one-dimensional
model, Akmaev and Shved [1980] proposed a mechanism by which atmospheric tides affect the net vertical
transport of O in the MLT. They reported that vertical motions due to the diurnal tide cause vertically oscillating O parcels to dip into the recombination region, where they experience a net loss, analogous to that
of increased eddy mixing. Using the nomenclature defined in Gardner and Liu [2010], the Akmaev and Shved
[1980] mechanism may be defined as wave-induced chemical transport, where a wave-induced oscillation of
a reactive species chemically interacts to modify the constituent profile. Recently, Yamazaki and Richmond
[2013] used the Thermosphere-Ionosphere-Electrodynamics General Circulation Model (TIE-GCM) with
JONES ET AL.
©2014. American Geophysical Union. All Rights Reserved.
5272
Geophysical Research Letters
10.1002/2014GL060698
specified tidal lower boundary conditions for the migrating (i.e., sun-synchronous) diurnal tide (DW1) and
showed that the Akmaev and Shved [1980] tidal mixing mechanism was inefficient due to the relatively slow
recombination time compared with tidal period. Yamazaki and Richmond [2013] concluded that the downward vertical transport of O comes from the mean meridional circulation induced by the dissipating DW1,
which acts to redistribute thermospheric O through advective transport.
A substantial reduction in [O] also appeared in the TIE-GCM simulations of Jones et al. [2014]. Similarities
between the seasonal-latitudinal distribution of this reduction and that of the migrating semidiurnal tide
(SW2) led these authors to suggest that SW2 was playing a significant role. They also suggested another
mechanism by which the dissipation of vertically propagating tides can act to alter the oxygen distribution
in the lower thermosphere, similar to the wave-induced dynamical transport of sodium shown to be produced by dissipating gravity waves [Gardner and Liu, 2010]. Hence, the purpose of this study is to verify the
claim provided by Jones et al. [2014], i.e., the tides induce a net transport of O themselves and an advective
transport of O via the mean meridional circulation generated by tidal dissipation. We test the above hypothesis using a set of numerical experiments from the TIME-GCM in order to quantify the relative contributions
of the different tidal-induced mechanisms which act to alter the [O] presented in Jones et al. [2014]. The
results presented herein are important because changes in the neutral constituents can have a significant
effect on the plasma density in the F region [Forbes et al., 1993; Yamazaki and Richmond, 2013; Siskind et al.,
2014; Jones et al., 2014; Yue and Wang, 2014].
2. Thermosphere GCM Simulations
The NCAR Thermosphere General Circulation Models or TGCMs (i.e., both the TIE- and TIME-GCM) are
three-dimensional time-dependent general circulation models designed to self-consistently simulate the
circulation, temperature, composition, and electrodynamics of the ionosphere-thermosphere system. The
main difference between the TIE- and TIME-GCM is the altitudes of their respective lower boundaries, which
is ∼97 km for the TIE-GCM and ∼30 km for the TIME-GCM. Therefore, the TIME-GCM includes additional
physics and chemistry in order to simulate the composition and dynamics of the mesosphere. For a more
detailed description of the TGCMs, the reader is referred to Roble and Ridley [1994] and Roble [1995, 1996].
The TGCM simulations in the present study were performed using a 2.5◦ × 2.5◦ horizontal resolution and
four grid points per scale height in the vertical. We invoked a 10.7 cm solar radio flux (F10.7 ) value of 120 solar
flux units, a hemispheric power value [after Evans, 1987] of 8 GW, and a cross cap potential drop of 30 kV.
The TIE-GCM simulations reported herein are detailed in Jones et al. [2014]. We ran the TIME-GCM until it
reached a diurnally reproducible state for day of year 258, in order to simulate geomagnetically quiet, solar
medium September conditions. The TIME-GCM inherently accounts for atmospheric tides generated in the
middle and upper atmosphere due to the absorption of ultraviolet and extreme ultraviolet solar radiation.
Atmospheric tides of tropospheric or lower stratospheric origin must be introduced as lower boundary conditions (i.e., 1 kPa or ∼30 km), and we do this using the September climatologies from the Global Scale Wave
Model 2009 (GSWM-09) [Zhang et al., 2010a, 2010b]. To elucidate the different tropospheric tidally induced
transport mechanisms of O in the MLT region, difference fields were calculated between TIME-GCM simulations that included and excluded lower boundary atmospheric tidal forcing from the GSMW-09 (hereafter
referred to as “with/without TBCs”).
3. Results and Discussion
Figure 1 illustrates the percent change in zonally and diurnally averaged [O] when tides are included at the
lower boundary of the NCAR TGCMs. Figure 1a shows the percent change in [O] due to the tropospheric and
stratomesospheric tides as a function of month and latitude at 130 km, while Figure 1b shows the percent
change in [O] as a function of latitude and height during the month of September. Both Figures 1a and 1b
are from the TIE-GCM simulations discussed in Jones et al. [2014], with Figure 1a being identical to Figure 6a
in Jones et al. [2014] and Figure 1b being analogous to Figure 6b in Jones et al. [2014], except for the month
of September. Figure 1a clearly shows a reduction in [O] at 130 km and low latitudes during all times of the
year with a maximum decrease of 13% occurring during the equinoxes. During September, this reduction in
[O] occurs in the E region and is transmitted to higher altitudes via molecular diffusion resulting in at least
an ∼8% decrease in [O] around 300 km (Figure 1b).
JONES ET AL.
©2014. American Geophysical Union. All Rights Reserved.
5273
Geophysical Research Letters
10.1002/2014GL060698
Since the lower boundary of the TIE-GCM
is near the turbopause and the constituent
distributions within the first one or two
scale heights are controlled by the fixed
constituent lower boundary condition, it is
difficult to assess the mechanisms responsible for altering the O distribution in the
MLT. Furthermore, Jones et al. [2014] showed
that the largest tidal components entering the thermosphere around 97 km were
DW1, SW2, as well as the nonmigrating
diurnal eastward propagating tide with
zonal wave number 3 (DE3), and that the
TIE-GCM was capable of producing comparable tidal amplitude structures to those
produced in the TIME-GCM. Therefore, we
utilize the TIME-GCM (i.e., model lower
boundary ∼30 km) to further investigate
the tidal-induced mechanisms that act to
modify the O distribution in the height
region around the TIE-GCM lower boundary. Figure 1c depicts the percent change in
[O] from the TIME-GCM during the month
of September extending from 80 to 300
km, with dashed line (at 100 km) approximately indicating the lower boundary of the
TIE-GCM. Figure 1c clearly shows a reduction
in [O] of ∼25% at 100 km, which extends
to higher altitudes with percent decreases
double that is shown in Figure 1b from the
TIE-GCM. The smaller percent changes in
[O] above about 150 km shown in Figure 1b
compared to Figure 1c are partially due to
an increase in temperature when tides are
included in the TIE-GCM, which causes an
increase in O scale height (i.e., a less rapid
decrease in O with height). Temperature
Figure 1. Percent differences in [O] between TIE-GCM simulations
with/without TBCs (a) as a function of month and latitude at 130 km decreases cause the opposite behavior in
and (b) as a function of latitude and altitude during September. Per- the TIME-GCM (Figure 1c). Also depicted in
cent differences in [O] between TIME-GCM simulations with/without Figure 1c is a large increase in [O] (∼65%)
TBCs (c) as a function of latitude and height during September.
between 85 and 95 km. Since the main
Percent differences are contoured every ±4%.
source of SW2 excitation (insolation absorption by stratospheric ozone) is the same in
both TIME-GCM simulations, the above variations in [O] can be explained mainly by the dissipation of DW1,
with SW2 and DE3 playing secondary roles.
The largest [O] differences displayed in Figure 1c occur in the altitude range of 85 to 105 km, where transport, chemical, eddy, and molecular diffusion processes can be of equal importance. In order to diagnose
which one of the above mechanisms is of utmost importance in this region, one can refer to their respective
time constants, which are calculated using the following:
JONES ET AL.
𝜏v,trans =
L
vTidal
𝜏w,trans =
HO
wTidal
©2014. American Geophysical Union. All Rights Reserved.
5274
Geophysical Research Letters
10.1002/2014GL060698
𝜏chem =
1
2k[O] [N2 ]
2
𝜏eddy =
H
Kzz
2
𝜏mole =
HO
D(O, N2 )
where the terms with overbars represent the zonal and diurnal means calculated at a constant altitude,
L is the characteristic meridional length (we assumed it to be ∼10◦ in latitude or 1000 km), HO is the
diffusive-equilibrium scale height of O, vTidal (wTidal ) are the effective meridional (vertical) net tidal transport velocities that are explicitly defined below, k is the rate constant of three-body recombination of O,
9.59 × 10−34 exp( 480
) in cm6 s−1 [after Allen et al., 1984], H is the mean scale height, Kzz is the eddy diffusion
T
coefficient from the TIME-GCM, and D(O, N2 ) is the mutual molecular diffusion coefficient from Colegrove
et al. [1966]. The lifetime of O due to meridional and vertical tidal transport ranges from ∼10 to 100 days
near the [O] peak around 100 km and remains relative constant with increasing altitude. Both 𝜏v,trans and
𝜏w,trans are comparable to or faster than chemistry, which is near ∼100 days at 100 km and increases with
height. Due to the long chemical lifetime of O, compared to tidal period (hours) and the time scales of eddy
(∼10 days at 100 km) and molecular diffusion (∼20 days at 100 km), the aforementioned Akmaev and Shved
[1980] mechanism is not likely to cause the changes in [O] shown in Figure 1, as concluded by Yamazaki
and Richmond [2013]. Also, near the [O] peak, the tidal transport and eddy diffusion time constants (e.g.,
∼1 to 30 days) are of comparable importance and similar to the time constant of molecular diffusion in the
TIME-GCM. Above an altitude of ∼125 km, molecular diffusion dominates over tidal transport and eddy
diffusion. Thus, the net effects of meridional and vertical transport associated with the dissipation of vertically propagating tides in the altitude range of 80 to ∼125 km (see Figure 1) are playing a significant role in
determining the vertical structure of [O], which in turn affect higher altitudes via molecular diffusion.
In order to separate the tidally induced advective and net transport of [O] in the MLT region, we derive the
continuity equation for [O], assuming that the dependent variables (i.e., meridional wind (v ), vertical wind
(w), and [O]) can be written as follows:
′
⎡ v(𝜆, 𝜃, z, t) ⎤ ⎡ v(𝜃, z) + v (𝜆, 𝜃, z, t) ⎤
⎢ w(𝜆, 𝜃, z, t) ⎥ = ⎢ w(𝜃, z) + w′ (𝜆, 𝜃, z, t) ⎥
⎥
⎢
⎥ ⎢
⎣ [O](𝜆, 𝜃, z, t) ⎦ ⎣ [O](𝜃, z) + [O]′ (𝜆, 𝜃, z, t) ⎦
(1)
where 𝜆 = longitude, 𝜃 = latitude, z = height, t = time; as previously mentioned above, the terms with overbars represent the zonal and diurnal means; and the prime terms represent their residuals. If we use (1) in
the continuity equation, neglect chemical production and loss, and take the zonal and diurnal average of
that equation, we get the following:
𝜕[O]
𝜕
𝜕 ′ ′
𝜕
𝜕
= − v[O] −
v [O] − w[O] − w′ [O]′
𝜕t
𝜕𝜃
𝜕𝜃
𝜕z
𝜕z
(2)
The first and third terms on the right-hand side of (2) represent the meridional and vertical [O] flux divergences due to the zonally and diurnally averaged meridional and vertical winds, respectively. The transport
of [O] due to the tidally induced mean winds is what we refer to as the advective transport of [O] or the
prominent transport mechanism proposed by Yamazaki and Richmond [2013]. The remaining two terms
on the right-hand side of (2) represent the divergences of the net transport fluxes of [O] due to the tides
themselves, i.e., the mechanism proposed by Jones et al. [2014]. Following the derivation of the effective
dynamical transport velocity shown by Gardner and Liu [2010], we derive effective meridional and vertical
net tidal transport velocities of [O], which are given by the following:
v ′ [O]′ ≈ vTidal [O]
(3a)
w′ [O]′ ≈ wTidal [O]
(3b)
We note that dynamical transport velocity as defined by Gardner and Liu [2010] is the same for all species,
which are assumed to be chemically inert. This is not the case for O, which is why we use the term net tidal
JONES ET AL.
©2014. American Geophysical Union. All Rights Reserved.
5275
Geophysical Research Letters
10.1002/2014GL060698
Figure 2. Meridional tidal and advective transport velocities as a function of latitude and altitude from TIME-GCM
simulations during September. (a and b) From TIME-GCM simulations, vTidal and vAdv velocities including GSWM-09
lower boundary tidal forcing are shown, respectively. (c and d) Differences in vTidal and vAdv velocities from TIME-GCM
simulations with/without TBCs are shown, respectively. Velocities are contoured every 2 m s−1 .
transport velocity throughout the remainder of this paper. If vTidal and wTidal are of comparable magnitude
to meridional and vertical advective velocities (vAdv and wAdv ), we deduce that they are also playing an
important role in transporting O in the MLT region.
Figure 2 (Figure 3) shows TIME-GCM September vTidal and vAdv (wTidal and wAdv ) results with GSWM-09 tides
at the lower boundary in the top row, and with/without TBCs in the bottom row. Comparison between
Figures 2a and 2b shows that vTidal is larger than vAdv between 80 to 95 km with values ranging between
±8 m s−1 , while vAdv is larger than vTidal above 95 km with values ranging from −10 to 6 m s−1 . Further comparison between difference fields calculated in Figures 2c and 2d reveals a meridional structure of vTidal
that is relatively unchanged, whereas there is a reduction in vAdv . The similarity between Figures 2a and 2c
implies that the tidally induced meridional wind is mainly coming from the tropospherically forced DW1
Figure 3. Same as Figure 2, except for vertical transport velocities wTidal and wAdv . Velocities are contoured
every 1 cm s−1 .
JONES ET AL.
©2014. American Geophysical Union. All Rights Reserved.
5276
Geophysical Research Letters
10.1002/2014GL060698
with the stratomesospherically forced DW1 and SW2, as well as the tropospherically forced DE3, playing
secondary roles.
A discussion similar to the one above can be for the vertical transport velocities, wTidal and wAdv (Figure 3).
Below 100 km, wTidal is downward at low and middle latitudes with a maximum downward velocity of
2 cm s−1 near the equator (Figure 3a). Above 100 km, wTidal is upward with maximum velocities of 3 cm s−1
centered about the equator. The differences between wTidal with/without TBCs show that below (above)
100 km, the downward (upward) wTidal velocities remain relatively unchanged (reduced) due to the dissipation of tides. At low latitudes, wAdv has a clear three-cell structure that is upward at the equator (5 cm s−1 )
and downward just poleward of the equator (−4 cm s−1 ) at 100 km (Figure 3b). This three-cell vertical
circulation structure is evidently generated by the dissipation of the tides because it is present in Figure 3d.
It is clear from Figures 2 and 3 that the magnitudes of the net tidal transport velocities are comparable to
the magnitudes of the advective transport velocities. We now consider the flux divergences due to each one
of the different transport velocities and their respective difference fields in order to describe the [O] difference fields shown in Figure 1. Equation (2) can be rewritten in terms of difference fields between TIME-GCM
simulations with/without TBCs, which results in the following equation:
𝜕
𝜕
𝜕
𝜕 ′ ′
𝜕
Δ[O] ≈ − (v 1 − v 0 )[O]1 − (w1 − w0 )[O]1 −
v [O] − w′ [O]′
𝜕t
𝜕𝜃
𝜕z
𝜕𝜃 1 1 𝜕z 1 1
(4)
where subscript zero (one) refers to TIME-GCM simulations excluding (including) GSWM-09 lower boundary
tidal forcing, and Δ[O] = [O]1 − [O]0 . Equation (4) also assumes [O]1 ≈ [O]0 (i.e., only changing by 10–20%;
see Figure 1) and w0′ [O]′0 ≪ w1′ [O]′1 .
Figure 4 illustrates both the individual terms on the right-hand side of (4) and the sum of these terms as
function of latitude and height from the September TIME-GCM simulations in order to facilitate the interpretation of the [O] differences (Figure 4h). Figures 4a, 4c, and 4e depict the [O] flux divergences due to tidally
induced net transport, whereas Figures 4b, 4d, and 4f portray the [O] flux divergences due to tidal-induced
advective transport. The meridional component of the [O] flux divergence due to net (advective) tidal transport of [O] in Figures 4a (4b) results in a net source (sink) of [O] centered close to the equator near 100 km
on the order of 25 × 105 cm−3 s−1 (−25 × 105 cm−3 s−1 ). Also, in Figures 4a (4b) are two net sink (source)
regions of [O] just poleward of source (sink) regions near 100 km on the order of −10 × 105 cm−3 s−1
(15 × 105 cm−3 s−1 ). The vertical component of the [O] flux divergences due to tidally induced net transport reveals a net sink in [O] between ∼90 and ∼105 km over the low and middle latitudes, while below
∼90 km there is a source region of [O] over the low and middle latitudes (Figure 4c). These source (sink) maxima (minima) are 6 × 105 cm−3 s−1 (−5 × 105 cm−3 s−1 ), which are an order of magnitude smaller than the
meridional [O] flux divergences. Vertical [O] flux divergences due to advective transport show antisymmetric source and sink regions centered about 100 km and between ±30◦ in latitude (Figure 4d). Below (above)
100 km, there is a net sink (source) of [O] at equator flanked by two source (sink) regions with magnitudes
on the order of ±12 × 105 cm−3 s−1 . The sum of the meridional and vertical [O] flux divergences due to tidally
induced net (advective) transport shown in Figures 4e (4f ) reveals the same structure as the meridional
flux divergences shown in Figures 4a (4b), with a source (sink) of [O] at the equator, flanked on each side
by a sink (source) of [O] near 100 km. The sums of all of the flux divergence terms are shown in Figure 4g.
The large percent increase (i.e., source) in [O] between 85 and 95 km and close to the equator shown in
Figure 4h appears to be consistent with the net meridional divergence term due to the tides, which is then
extended poleward in both hemispheres by the advective part of the meridional and vertical divergence
terms. This percent increase in [O] and could possibly be connected with the effects of tides on the mesospheric odd-oxygen chemistry. Also, depicted in Figure 4h is a percent decrease (i.e., sink) in [O] of ∼25% at
100 km and tropical latitudes that appears to be consistent with the advective meridional divergence term
due to the tides, which is then extended downward and poleward in both hemispheres due to the net tidal
portion of the meridional divergence term. Hence, both the tidally induced net and advective transport
of [O] in the MLT region are equally contributing to the [O] distribution at these altitudes, with the meridional component playing the primary role and the vertical component playing a secondary role. We note
here that the above analysis of the continuity equation was done using Eulerian mean quantities, which
is different than Lagrangian mean quantities (i.e., mean motion following an air parcel). A complementary
Lagrangian analysis is beyond the purview of this report.
JONES ET AL.
©2014. American Geophysical Union. All Rights Reserved.
5277
Geophysical Research Letters
10.1002/2014GL060698
Figure 4. Meridional and vertical derivatives in the [O] fluxes from the continuity equation (4) calculated from TIME-GCM
simulations during September. (a) Third term on the right-hand side of (4), (b) first term on the right-hand side of (4),
(c) fourth term on the right-hand side of (4), (d) second term on the right-hand side of (4), (e) total divergence due to
tidal transport, (f ) total divergence due to advective transport, (g) the full right-hand side of (4), and (h) same as Figure 1
except from 80 to 150 km.
It is also important to note that since the TIME-GCM does not resolve gravity waves (i.e., gravity wave effects
are parameterized after Lindzen [1981]), the primed terms in equations (1), (2), and (4) only include effects
associated with the tides. Although we cannot unequivocally separate tidal and gravity wave effects in
the TIME-GCM, an analysis of Kzz shows that DW1 is causing diurnal variations in the eddy diffusivities in a
manner similar to that demonstrated by Fritts and Vincent [1987]. These diurnal variations could in principle affect the diurnal mean [O] distribution through some nonlinear process. Having said that, differences
between zonally and diurnally averaged Kzz values from TIME-GCM simulations with/without TBCs are small
around 100 km, suggesting that gravity wave effects are playing a comparatively minor role in driving the
percent changes in [O] shown in Figure 1c.
4. Summary and Conclusions
As stated in section 1, there are a number of different mechanisms that can act to change the O concentration in the MLT region and overlying thermosphere. We have revealed through a series of TIME-GCM
JONES ET AL.
©2014. American Geophysical Union. All Rights Reserved.
5278
Geophysical Research Letters
10.1002/2014GL060698
simulations a new mechanism by which the tides can act to change [O] throughout the thermosphere, i.e.,
a net meridional and vertical transport of [O] is induced by the tides themselves, in addition to the advective transport of [O] due to the mean circulation induced by the dissipation of the tides. The largest terms
in the continuity equation are calculated for between 80 and 105 km. Due to its large meridional and vertical wind amplitudes, we conclude that the DW1 is the primary tidal component responsible for driving
constituent changes in the TIME-GCM during September. DW1 tends to be the dominant tidal component
contributing to constituent changes in the MLT region around the equinoxes; however, during the solstices,
SW2 plays a greater role (not shown), comparable to that of DW1. We do note here that the inherent presence of stratomesospheric tidal components (i.e., DW1 and SW2) in our TIME-GCM simulation excluding
GSWM-09 tidal perturbations does not affect the overall conclusions presented above because they tend to
dissipate at altitudes (approximately 130 km) where the time scale of molecular diffusion dominates over
tidal transport.
Although the main purpose of this paper was to quantify the transport mechanisms acting on O due to dissipating tides, recombination also acts to translate transport effects into [O] changes. Evidence for chemical
recombination effects are embodied in the net ∼10% increase in [O2 ] produced by the presence of dissipating tides (not shown here, but see Jones et al. [2014]). In addition, we determined that net heat fluxes
ranging from 20 × 10−4 W m−2 at 95 km to 1.5 × 10−4 W m−2 at 120 km are produced by dissipating
tides. The implications of these and other effects (e.g., changes in solar heating rates, adiabatic heating and
cooling, ozone chemistry, and radiative cooling) warrant future investigation.
Acknowledgments
The authors thank Art Richmond for
providing valuable comments on our
paper prior to submission. TIE- and
TIME-GCM outputs in NetCDF format
are archived on the NCAR High
Performance Storage System and are
available upon request. This work
was supported under a Predoctoral
Fellowship from The Ford Foundation
and the University of Colorado Glenn
Murphy Endowed Chair. This material is based upon work supported
by the National Science Foundation
under Cooperative Support Agreement
AGS-0856145. Any opinions, findings,
and conclusions or recommendations
expressed in this material are those
of the author(s) and do not necessarily reflect the views of the National
Science Foundation.
The Editor thanks Gordon Shepherd
and an anonymous reviewer for their
assistance in evaluating this paper.
JONES ET AL.
References
Akmaev, R. A., and G. M. Shved (1980), Modeling of the composition of the lower thermosphere taking into account of the dynamics
with applications to tidal variations of the [OI] 5577 Å airglow, J. Atmos. Terr. Phys., 42, 705–716.
Allen, M., J. I. Lunine, and Y. L. Yung (1984), The vertical distribution of ozone in the mesosphere and lower thermosphere, J. Geophys.
Res., 89(D3), 4841–4872, doi:10.1029/JD089iD03p04841.
Angelats i Coll, M., and J. M. Forbes (1998), Dynamical influences on atomic oxygen and 5577 Å emission rates in the lower
thermosphere, Geophys. Res. Lett., 25, 461–464.
Colegrove, F. D., F. S. Johnson, and W. B. Hanson (1966), Atmospheric composition in the lower thermosphere, J. Geophys. Res., 71(9),
2227–2236, doi:10.1029/JZ071i009p02227.
Evans, D. S. (1987), Global statistical patterns of auroral phenomena, in Proceedings of the Symposium on Quantitative Modeling of
Magnetospheric-Ionospheric Coupling Processes, edited by Y. Kamide and R. A. Wolf, pp. 325–330, Kyoto Sangyo Univ., Kyoto, Japan.
Forbes, J. M., R. G. Roble, and C. G. Fesen (1993), Acceleration, heating, and compositional mixing of the thermosphere due to upward
propagating tides, J. Geophys. Res., 98(A1), 311–321, doi:10.1029/92JA00442.
Fritts, D. C., and R. A. Vincent (1987), Mesospheric momentum flux studies at Adelaide, Australia: Observations and a gravity wave-tidal
interaction model, J. Atmos. Sci., 44, 605–619.
Gardner, C. S., and A. Z. Liu (2010), Wave-induced transport of atmospheric constituents and its effect on mesospheric Na layer,
J. Geophys. Res., 115, D20302, doi:10.1029/2010JD014140.
Jones, M., Jr., J. M. Forbes, M. E. Hagan, and A. Maute (2014), Impacts of vertically propagating tides on the mean state of the
ionosphere-thermosphere system, J. Geophys. Res. Space Physics, 119, 2197–2213, doi:10.1002/2013JA019744.
Lindzen, R. S. (1981), Turbulence and stress owing to gravity wave and tidal breakdown, J. Geophys. Res., 86, 9707–9714.
Roble, R. G. (1995), Energetics of the mesosphere and thermosphere, in The Upper Mesosphere and Lower Thermosphere: A Review of
Experiment and Theory, Geophys. Monogr. Ser., vol. 87, edited by R. M. Johnson and T. L. Killeen, pp. 1–21, AGU, Washington, D. C.
Roble, R. G. (1996), The NCAR thermosphere-ionosphere-mesosphere-electrodynamics general circulation model (TIME-GCM), in STEP
Handbook on Ionospheric Models, edited by R. W. Schunk, pp. 281–288, Utah State Univ., Logan, Utah.
Roble, R. G., and E. C. Ridley (1994), A thermosphere-ionosphere-mesosphere-electrodynamics general circulation model (TIME-GCM):
Equinox solar cycle minimum simulations (30–500 km), Geophys. Res. Lett., 21(6), 417–420, doi:10.1029/93GL03391.
Roble, R. G., and G. G. Shepherd (1997), An analysis of wind imaging interferometer observations of O (1 S) equatorial emission rates
using the thermosphere-ionosphere-mesosphere-electrodynamics general circulation model, J. Geophys. Res., 102(A2), 2467–2474,
doi:10.1029/96JA02930.
Shimazaki, T. (1968), Dynamic effects on height distributions of neutral constituents in the Earth’s upper atmosphere: A calculation of
atmospheric model between 70 km and 500 km, J. Atmos. Terr. Phys., 30, 1279–1292.
Siskind, D. E., D. P. Drob, K. F. Dymond, and J. P. McCormack (2014), Simulations of the effects of vertical transport on the thermosphere
and ionosphere using two coupled models, J. Geophys. Res. Space Physics, 119, 1172–1185, doi:10.1002/2013JA019116.
Vlasov, M. N., and M. C. Kelley (2010), Estimates of eddy turbulence consistent with seasonal variations of atomic oxygen and its possible
role in the seasonal cycle of mesopause temperature, Ann. Geophys., 28, 2103–2110.
Yamazaki, Y., and A. D. Richmond (2013), A theory of ionospheric response to upward propagating tides: Electrodynamic effects and tidal
mixing effects, J. Geophys. Res. Space Physics, 118, 5891–5905, doi:10.1002/jgra.50487.
Yue, J., and W. Wang (2014), Changes of thermospheric composition and ionospheric density caused by quasi 2-day wave dissipation,
J. Geophys. Res. Space Physics, 119, 2069–2078, doi:10.1002/2013JA019725.
Zhang, X., J. M. Forbes, and M. E. Hagan (2010a), Longitudinal variation of tides in the MLT region: 1. Tides driven by tropospheric net
radiative heating, J. Geophys. Res., 115, A06316, doi:10.1029/2009JA014897.
Zhang, X., J. M. Forbes, and M. E. Hagan (2010b), Longitudinal variation of tides in the MLT region: 2. Relative effects of solar radiative
and latent heating, J. Geophys. Res., 115, A06317, doi:10.1029/2009JA014898.
©2014. American Geophysical Union. All Rights Reserved.
5279