JOURNAL OF GEOPHYSICAL RESEARCH: SPACE PHYSICS, VOL. 118, 3846–3855, doi:10.1002/jgra.50354, 2013
Effect of trends of middle atmosphere gases on the mesosphere
and thermosphere
Liying Qian,1 Daniel Marsh,2 Aimee Merkel,3 Stanley C. Solomon,1 and Raymond G. Roble 1
Received 2 February 2013; revised 19 April 2013; accepted 22 May 2013; published 21 June 2013.
[1] We conducted model simulations to examine how changes in concentration of radiatively
active trace gases in the middle atmosphere affect long-term changes in the upper atmosphere.
We focused our model study on the impact of increases in carbon dioxide (CO2), methane
(CH4), and water vapor (H2O), and decreases in ozone (O3) between 1983 and 2003. We
used both the National Center for Atmospheric Research Whole Atmosphere Community
Climate Model and the Thermosphere-Ionosphere-Mesosphere-Electrodynamics General
Circulation Model, global mean version, in this study. The model simulations indicate
that CO2 is the main forcing mechanism of long-term changes in the thermsophere,
with minor influences from O3, CH4, and H2O. At 400 km altitude, global mean
thermospheric neutral density decreased by ~4.5% due to CO2 forcing alone,
whereas it decreased by ~4.8% due to the combined forcing from all four gases. O3
depletion caused cooling in the stratosphere and mesosphere (maximum decrease of
0.5 K) due to reduced absorption of solar ultraviolet radiation, but had nearly no
cooling effect in the thermosphere. However, due to thermal contraction in the
stratosphere and mesosphere, O3 depletion caused a small decrease in thermospheric
neutral density of ~0.25%. Increases in both CH4 and H2O may slightly warm the
upper mesosphere and thermosphere due to increased chemical heating and absorption
of solar ultraviolet radiation.
Citation: Qian, L., D. Marsh, A. Merkel, S. C. Solomon, and R. G. Roble (2013), Effect of trends of middle atmosphere
gases on the mesosphere and thermosphere, J. Geophys. Res. Space Physics, 118, 3846–3855, doi:10.1002/jgra.50354.
1.
Introduction
[2] CO2 is considered the primary greenhouse gas that
drives global changes in the upper atmosphere (mesosphere
and thermosphere) [e.g., La
stovi
cka et al., 2006; Akmaev
and Fomichev, 2000; Qian et al., 2011]. Modeling studies
have focused on the effects of CO2 increases on long-term
changes in the upper atmosphere [e.g., Roble and
Dickinson, 1989; Rishbeth and Roble, 1992; Akmaev and
Fomichev, 2000; Qian et al., 2006, 2008, 2009]. CO2 is a
nonreactive species and is not important in the chemistry
of the middle atmosphere. However, it is a major greenhouse gas and plays a critical role in the thermal budget of
the atmosphere, warming the troposphere, but causing a
net cooling effect in the stratosphere and above with the
1
High Altitude Observatory, National Center for Atmospheric Research,
Boulder, Colorado, USA.
2
Atmospheric Chemistry Division, National Center for Atmospheric
Research, Boulder, Colorado, USA.
3
Laboratory for Atmospheric and Space Physics, University of Colorado
Boulder, Boulder, Colorado, USA.
Corresponding author: L. Qian, High Altitude Observatory, National Center
for Atmospheric Research, 3080 Center Green Dr., Boulder, CO 80301, USA.
([email protected])
©2013. American Geophysical Union. All Rights Reserved.
2169-9380/13/10.1002/jgra.50354
exception of the cold summer mesopause region [Akmaev
et al., 2006].
[3] A recent study by Walsh and Oliver [2011] suggests
that the long-term cooling in the thermosphere may be
largely due to O3 depletion, rather than changes in CO2.
Using an ion temperature data set spanning 1966 to 1987
over Saint Santin, France at 200–400 km, they reported a
cooling trend that accelerated abruptly around 1979 (a
breakpoint). Walsh and Oliver [2011] further demonstrated
that 1979 was also a breakpoint year in O3 column density,
but did not coincide with large changes in CO2 concentration.
La
stovi
cka [2012] questioned this hypothesis based on analysis of other long-term data sets. He pointed out that ozone
depletion only continued through the late 1990s; therefore,
if O3 is the main driver of the upper atmosphere cooling, then
there should be a second breakpoint in upper atmospheric
trends around that time. La
stovi
cka [2012] demonstrated that
long-term data sets of thermospheric neutral density [e.g.,
Emmert et al., 2008] and ion temperature [Zhang et al.,
2011] exhibit no second breakpoint. Akmaev [2012] also
questioned the Wash and Oliver [2011] hypothesis,
concluding that cooling of the stratosphere due to O3 depletion was unlikely to contribute to substantial negative trends
in the thermosphere.
[4] O3 concentration is partly determined by pure oxygen
chemistry, which involves photolysis of O3 and O2, collisions of excited atomic oxygen (O1D) with N2 and O2, and
O and O2 three-body recombination to produce O3. In
3846
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
Figure 1. Global mean profiles of gas concentrations simulated by WACCM for the 2003 control case
(scenario #2, red) and the CO2 variation case (scenario #3, black). Vertical axis on the right shows
geometric heights of model pressure surfaces. (a) CO2, (b) CH4, (c) H2O, and (d) O3.
addition, O3 concentration is determined by interactions
between hydrogen, nitrogen, and chlorine compounds.
Furthermore, it is influenced by transport, particularly in the
lower stratosphere where it is a long-lived species. The most
important process that causes O3 depletion is the catalytic
destruction of ozone by atomic halogens, mainly atomic
chlorine. The main source of these chlorine atoms in the
stratosphere is the photodissociation of anthropogenic
halocarbon refrigerants. CFC-11 (CFCl3) and CFC-12
(CF2Cl2) are two such compounds with high ozone depletion
potential. Production of these two compounds was phased
out during the 1990s, following adoption of the Montreal
Protocol in 1989.
[5] CFC-11 and CFC-12 are fairly stable in the troposphere and can therefore be transported to the stratosphere,
where they are eventually dissociated by ultraviolet (UV)
radiation and produce chlorine atoms. They can also react
with hydrogen radical OH and produce chlorine atoms.
When a chlorine atom is produced in the stratosphere, it
reacts with O3 to produce chlorine monoxide. Chlorine
monoxide reacts with O, destroys odd oxygen, and produces a chlorine atom. These two reactions constitute an
important catalytic cycle that destroys O3. In addition,
chlorine monoxide reacts with NO and also reforms
atomic chlorine. These reformed chlorine atoms again catalytically destroy O3. Stratospheric O3 depletion then cools
the upper atmosphere through reduced absorption of
UV radiation.
[6] Other heteronuclear molecules are potential agents of
change in the upper atmosphere energy budget, including
methane (CH4) and water vapor (H2O). CH4 is the most
abundant organic molecule in the Earth’s atmosphere and
plays important roles in both the radiative energy budget
and global atmospheric chemistry [Brasseur et al., 1999].
CH4 concentration has increased in the industrial era, causing
a direct warming effect in the troposphere [Forster et al.,
2007]. CH4 is produced at ground level both by natural
sources in the biosphere and lithosphere, and by anthropogenic production through industrial, agriculture, and
mining activities. It is then transported into the
atmosphere until it is oxidized. CH4 is oxidized by either
an OH radical, excited oxygen (O1D), or chlorine (Cl).
The end products of methane oxidation are H2O, CO,
and CO2. Therefore, an increase in CH4 contributes to
radiative forcing indirectly, through interactions with
chlorine that influence ozone concentrations, increase of
stratospheric water vapor, and generation of a small additional source of CO2.
[7] H2O in the stratosphere is due in part from tropospherestratosphere exchanges. In the middle and upper stratosphere, H2O is formed by methane oxidation process. The
increase in H2O abundance in the stratosphere can be
explained by an increase of CH4 concentration in the industrial era [Brasseur and Solomon, 1995]. In the stratosphere
and lower mesosphere, H2O reacts with O(1D) and produces
OH radicals. In the thermosphere and upper mesosphere,
H2O is photolyzed by Lyman-alpha, producing hydrogen
free radicals.
[8] A modeling study by Akmaev et al. [2006] investigated
the impact of changes in CO2, O3, and H2O from 1980 to
3847
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
(a)
(b)
Figure 2. Global mean profiles of neutral temperature (TN)
changes between the 2003 control case (scenario #2) and the
CO2 variation case (scenario #3) (scenario #2 minus scenario
#3). Vertical axis on the right shows geometric heights of
model pressure surfaces. (a) Simulated by the 1-D model
and (b) simulated by WACCM.
2000 on the long-term changes in the mesosphere and lower
thermosphere, similar to this study. They found that O3
depletion caused cooling in the mesosphere comparable to
that of CO2 through reduced daytime heating in the stratosphere. In addition, the O3 depletion caused a warming effect
in the lower thermosphere due to thermal contraction of the
underlying atmosphere if changes are computed at fixed
altitude levels, due to the fact that temperature has a large
positive vertical gradient in the lower thermosphere
[Akmaev, 2012]. However, because the upper boundary of
their model was 200 km, they could not infer the impact in
the middle and upper thermosphere. Therefore, our modeling
study expands on the work by Akmaev et al. [2006] and investigates how radiatively active gases, including CO2,
O3, CH4, and H2O, impact long-term changes in the mesosphere and thermosphere. Our focus is on the global mean
impact of these gases on the long-term changes in the mesosphere and thermosphere.
2.
Method
[9] For our investigation, we use a combination of simulations from two types of atmospheric models. We use an
upper atmosphere model called the National Center for
Atmospheric Research (NCAR) Thermosphere-IonosphereMesosphere-Electrodynamics General Circulation Model
(TIME-GCM) (35–600 km) [Roble and Ridley, 1994] and a
coupled chemistry global circulation model called the
Whole Atmosphere Community Climate Model (WACCM)
[Garcia et al., 2007; Marsh et al., 2007]. Specifically, since
our purpose is to investigate the global mean impact of
CO2, O3, CH4, and H2O, we use a 1-dimensional (1-D) version of the TIME-GCM. The 1-D model is a single-column
representation of the aeronomic processes in the 3-D
NCAR TIME-GCM. It is a self-consistent global mean
model of the mesosphere, thermosphere, and ionosphere
[Roble et al., 1987; Roble, 1995]. We will simply refer
to this 1-D version of the TIME-GCM as the 1-D model
from now on. The concentration of chemical constituents in the mesosphere and lower thermosphere in the
1-D model are specified from the global mean of
WACCM simulation output. This allows us to study
the impacts of these gases from the chemistry climate
model on the upper atmosphere using an upper atmospheric model, which will be discussed in detail in the
following paragraphs.
[10] Model investigations of the effects of CO2 on the
upper atmosphere are relatively straightforward due to
the fact that CO2 is not chemically active in the
troposphere, stratosphere, and mesosphere. Therefore,
CO2 concentrations measured at the Earth’s surface can
simply be applied at the lower boundary of an upper atmospheric model to adequately conduct CO2 investigations in the upper atmosphere [e.g., Akmaev and
Fomichev, 1998; Qian et al., 2006]. However, unlike
CO2, the other three gases in our study (O3, CH4, and
H2O) are chemically active below the thermosphere.
An upper atmosphere model such as the TIME-GCM
is not sufficient for these investigations alone. A global
circulation model, which extends from the surface to
the upper atmosphere and is coupled with a detailed
chemical transport model, is required for these investigations. WACCM extends from the surface to the lower
thermosphere (140 km) and includes a fully interactive
chemical transport model [Marsh et al., 2007].
However, it does not cover the middle and upper thermosphere. Therefore, investigations of the effects of
these gases on the entire upper atmosphere require the
use of WACCM as well as the 1-D model.
[11] The WACCM climate simulations consist of five scenarios. These scenarios are (1) a control case for 1983 in
which WACCM was run for 10 years to reach a steady state
with gas concentrations kept at their 1983 levels; (2) similarly, a control case for 2003; (3) a variation from the 2003
control case in which CO2 concentration was scaled back to
the 1983 level, all other gases were kept at 2003 levels; (4)
a similar case to case (3) but chlorofluorocarbon compounds
(CFCs) were scaled back to 1983 levels; and (5) similar to
case (3) but CH4 was at its 1983 level. Since CFC
compounds are responsible for ozone destruction, scenario
#4 examines effects of O3 depletion. Scenario #5 examines
the effects of both CH4 and H2O, since variations of CH4
concentration will change H2O concentrations through CH4
oxidation. For each scenario, WACCM was run for 10 years
to reach steady state. The profiles from the last 5 years’ runs
3848
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
Figure 3. Global mean profiles of gas concentrations simulated by WACCM for the 2003 control case
(scenario #2, red) and the CFC variation case (scenario #4, black). Increase of CFCs from 1983 to 2003
caused O3 depletion. Vertical axis on the right shows geometric heights of model pressure surfaces. (a)
CFC-11, (b) CFC-12, and (c) O3.
were averaged to obtain global mean profiles for each month.
The September global mean profiles were used as climatological profiles for each scenario in this study and will be referred to as global mean profiles in the rest of the paper. The
September profiles were chosen because O3 shows larger
depletion in springtime over the Earth’s polar regions, and
the most pronounced O3 depletion has been observed in the
Antarctic region [Solomon et al., 1987]. The global mean
WACCM profiles were used to specify the concentration of
chemical constituents in the mesosphere and lower thermosphere in the 1-D model.
[12] Radiative cooling rates by CO2 (15 um) and O3
(9.6 um) in the mesosphere and lower thermosphere are calculated using the radiative transfer algorithm developed by
Fomichev et al. [1998] in both WACCM and TIME-GCM.
Radiative cooling from CH4 and H2O in the mesosphere,
which does not exceed 10% of the total cooling [Fomichev
et al., 2004], is not included in either WACCM or TIMEGCM. Various solar heating terms are included in both
WACCM [Marsh et al., 2007] and TIME-GCM [Roble
et al., 1987]. These heating terms are the following:
absorption of solar extreme ultraviolet by CO2, O3, CH4,
and H2O; absorption of Schuman-Lunge continuum and
bands by CO2, O3, and H2O; absorption of Lyman-a by
CO2, O3, and CH4; absorption of Herzberg, Hartley,
Huggins, and Chappuis bands by H2O and O3; and absorption in the near-infrared by CO2 (between 1.05 and 4.3 mm).
[13] The vertical coordinates of both the 1-D model and
WACCM are on pressure surfaces but are defined differently.
Therefore, the WACCM profiles were linearly interpolated to
the 1-D pressure coordinates. The pressure interfaces of the
1-D model are defined as lev = ln(p0/p), where P0 is a reference pressure at 5 10 7 hPa. The 1-D model has 49 pressure surfaces covering the altitude range from ~35 km to
~600 km, with lev ranging from 17 to 7 and a vertical
resolution of one half scale height. The scale height is
~5 km near 100 km and increases to ~50 km in the upper
thermosphere. WACCM uses a hybrid-sigma pressure
coordinate [Janjic et al., 2010], which combines sigmadenominated layers at the bottom (following terrain)
with isobaric (fixed pressure) levels aloft, with a total of 66
vertical levels [Collins et al., 2006].
3849
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
(a)
reflect upper atmospheric states, including neutral temperature and density, were then analyzed to examine how the
upper atmosphere responds to the climatological gas profiles
from each WACCM scenario.
3.
(b)
Figure 4. Global mean profiles of neutral temperature (TN)
changes between the 2003 control case (scenario #2) and the
CFC variation case (scenario #4) (scenario #2 minus scenario
#4). Vertical axis on the right shows geometric heights of
model pressure surfaces. (a) Simulated by the 1-D model
and (b) simulated by WACCM.
[14] There are five chemical compounds that are important
in determining the concentrations of CH4, H2O, and O3 in the
middle atmosphere: carbon, hydrogen, oxygen, chlorine, and
nitrogen compounds [Brasseur and Solomon, 1995]. These
five chemical compounds react with CH4, H2O, and O3 and
change their concentrations. The chemical species considered in the 1-D model that involves the four gases [CO2,
CH4, H2O, O3] and the five chemical compounds are the
following: CO2, CO; O3, Cl, and ClO; CH4, H2O, H2, H,
H2O2, HO2, and OH; N2O; and NO. We produced climatological global mean profiles for each of these species
from the defined WACCM scenarios. These profiles were
then used in the 1-D model for the atmosphere below
lev = 5 (~105 km); above lev = 0 (~130 km), gas profiles
calculated by the 1-D model were used. Between these pressure levels, the WACCM profiles were gradually nudged to
the 1-D model profiles through linear interpretation in the
pressure space. The 1-D model was run to steady state with
the WACCM global mean profiles for each WACCM
scenario defined above. For each simulation, the 14 gases
were enforced at each time step below 105 km, under solar
activity F10.7 = 100 and geomagnetically quiet conditions.
Parameters from both the 1-D model and WACCM that
Results and Discussions
3.1. Effect of CO2 Increase
[15] Comparisons between the CO2 variation case
(scenario # 3) and the 2003 control case (scenario # 2)
show the impact of CO2 forcing on the upper atmosphere. Figure 1 shows the global mean profiles for
CO2, CH4, H2O, and O3, for scenarios #2 and #3, simulated by WACCM. CO2 concentration changed from
344 ppmv in 1983 to 375 ppmv in 2003, which is a 9%
increase (Figure 1a). Figure 1a shows that CO2 volume
mixing ratio starts to decrease with altitudes at ~77 km.
This is consistent with the CO2 altitude profile used in
Akmaev and Fomichev [1998], where CO2 starts to decrease with altitudes at ~80 km [Fomichev et al., 1998].
Even though CH4, H2O, and O3 were kept at the same
levels as those in 2003 in scenario #3, their concentrations are slightly different from those for the 2003 control case, due to small changes in temperature-dependent
chemical reaction rates.
[16] In the upper mesosphere and lower thermosphere,
CO2 bend-stretch vibrational mode is excited by thermal
collisions with atomic oxygen, followed by radiative deexcitation at 15 mm infrared emission [López-Puertas et al.,
1986a, 1986b; Wintersteiner et al., 1992]. The 1-D model
uses a CO2 radiative transfer algorithm developed by
Fomichev et al. [1998]. The collisional excitation rate
constant is set at k = 1.5 10 12cm3s 1 [Khvorostovskaya
et al., 2002].
[17] The global mean profiles shown in Figure 1 were used
in the 1-D model using the method that was described in the
last paragraph of section 2. Figure 2 shows the temperature
differences driven by the CO2 change, simulated by the 1-D
model (Figure 2a) and WACCM (Figure 2b). The 1-D model
results indicate that the temperature decrease due to the CO2
forcing reaches the maximum of ~5.5 K at ~110 km. In the
middle and upper thermosphere, the temperature decrease is
nearly a constant of ~4.4 K. In the mesosphere, the temperature decrease is between 0 and 5 K. The WACCM simulations also obtained a maximum temperature decrease of
~5.8 K near 110 km, consistent with the 1-D model profiles.
Above 110 km, the WACCM results start to diverge from
the 1-D model results. This is due to the fact that this altitude
range is close to the upper boundary of WACCM, and
therefore, the results at these altitudes should be used with
caution. In the mesosphere, the WACCM results show a
temperature decrease between ~0 and 5 K, which is
consistent with the 1-D model results. However, the
WACCM simulations exhibit a slight warming of ~0.5 K
at an altitude range in the mid-mesosphere (Figure 2b),
which is likely due to absorption in the near-infrared by
CO2 (between 1.05 and 4.3 mm) [Marsh et al., 2007].
Other modeling studies found net cooling effects in the
mesosphere driven by CO2 increase [e.g., Akmaev and
Fomichev, 1998], which is consistent with the 1-D
model results.
3850
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
Figure 5. Global mean profiles of gas concentrations simulated by WACCM for the 2003 control case
(scenario #2, red) and the CH4 variation case (scenario #5, black). Increase of CH4 from 1983 to 2003
caused increase H2O. Vertical axis on the right shows geometric heights of model pressure surfaces. (a)
CH4, (b) H2O, (c) O3, and (d) CO2.
3.2. Effect of O3 Depletion
[18] Figure 3 shows the global mean profiles of the CFC
variation case (scenario #4) and the 2003 control case
(scenario #2), for CFC-11 and CFC-12, and O3, simulated
by WACCM. The increases of CFC-11 (Figure 3a) and
CFC-12 (Figure 3b) from 1983 to 2003 resulted in a ~3%
global mean O3 depletion at its peak altitude in the stratosphere during this period.
[19] Stratospheric O3 absorbs UV radiation and thus heats
the atmosphere. O3 depletion causes cooling in the stratosphere and mesosphere through this reduced solar heating.
The global mean gas profiles shown in Figure 3 were used
in the 1-D model using the method described in the last paragraph of section 2. Figure 4a shows the global mean neutral
temperature change driven by the O3 depletion simulated by
the 1-D model. The 1-D model results indicate that the O3
depletion causes 0–0.8 K cooling in the upper stratosphere
and mesosphere, up to an altitude of about 100 km. O3
depletion has nearly no effect above this altitude. It is important to point out that Figure 4 shows the temperature
changes on constant pressure levels. On fixed altitude levels,
the O3 depletion causes ~1.2 K warming effect in the lower
thermosphere around 110 km, due to the cooling and contraction of the underlying atmosphere, as well as the large positive temperature gradient in this altitude range (not shown).
This is consistent with the results of Akmaev et al. [2006].
The WACCM simulations also produced a ~0–0.8 K cooling
below 100 km (Figure 4b), which is consistent with the 1-D
model results.
3.3. Effect of CH4/H2O Increase
[20] Figure 5 shows the global mean profiles for the CH4
variation case (scenario #5) and the 2003 control case simulated by WACCM. CH4 increased from 1.6 ppmv in 1983 to
1.8 ppmv in 2003, which is a 12.5% increase (Figure 5a).
The resulting H2O increase in the stratosphere and mesosphere is shown in Figure 5b. O3 concentration increased
slightly, due to the reaction of CH4 with chlorine
(Figure 5c). The effect of CH4 increase on CO2 concentration is negligible (Figure 5d).
[21] The global mean gas profiles shown in Figure 5 were
used in the 1-D model using the method described in the last
paragraph of section 2 to examine the impact of CH4/H2O
forcing on the upper atmosphere. The 1-D model simulations
show that the increases of CH4/H2O caused a global
mean temperature increase in the mesosphere ~0–0.7 K,
and a 0.5 K warming in the thermosphere (Figure 6a), due to
increased chemical heating. The WACCM results indicated a
global mean warming (~ 0–1.2 K) above ~97 km (Figure 6b),
which is qualitatively consistent with the 1-D model results.
The WACCM results show a cooling (~0–0.75 K) between
~50 and ~90 km, due to reduced chemical heating at these
altitudes. Below ~50 km, the WACCM results again exhibit
a warming effect, which is due to increased solar heating. It
3851
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
Figure 6. Global mean profiles of neutral temperature (TN)
changes between the 2003 control case (scenario #2) and the
CH4 variation case (scenario #5) (scenario #2 minus scenario
#5). Vertical axis on the right shows geometric heights of
model pressure surfaces. (a) Simulated by the 1-D model
and (b) simulated by WACCM.
is important to point out that neither of the two models
considers H2O and CH4 infrared cooling. It is not clear how
much effect of the infrared cooling will have on the temperature. We do know that concentrations of CH4 and H2O, as
well as the changes of their concentrations from 1983 to
2003, are negligible above ~80 km, so the local cooling effect
from CH4 and H2O infrared cooling should be very small at
these altitudes.
3.4. Effect From All Gases
[22] We compared global mean concentrations of the four
radiatively active gases from the WACCM 1983 control case
(scenario #1) to the WACCM 2003 control case (scenario
#2). From 1983 to 2003, CO2 concentration increased by
9% (Figure 7a), CH4 concentration increased by 11.6%,
H2O increased by 4%, whereas O3 concentration decreased
by 2%. Figure 8 shows the total effect from these changes
of concentrations. The 1-D model and WACCM produced
consistent results that the global mean temperature decreased
by 0.5–2 K in the mesosphere. In addition, both the models
indicated a maximum cooling near pressure surface 6
(in the lower thermosphere) of about 5–5.5 K. Above this
pressure level, WACCM is too close to its upper boundary.
The 1-D model predicted a cooling of ~4.2 K in the upper
thermosphere from 1983 to 2003.
[23] The temperature changes from Figures 2a, 4a, 6a, and
8a, which shows the global mean temperature changes
between 1983 and 2003 due to contributions from CO2, O3,
CH4/H2O, and the total temperature changes due to all these
gases, are displayed together in Figure 9. In the mesosphere,
CO2 forcing caused a cooling in the range of 0.5–2 K, O3
forcing caused a cooling between 0 and 0.8 K, whereas CH4/
H2O forcing caused a warming between ~0 and 1.2 K. In the
thermosphere, CO2 forcing caused a cooling between 2 and
5.5 K, the effect of O3 depletion is negligible, whereas CH4/
H2O forcing caused a slight warming of ~0.4 K. Recall, however, that the missing infrared cooling by CH4 and H2O may
cancel some of the warming effect, and therefore, the warming
effect may be smaller. The total temperature change (black
line) between the 1983 and 2003 control cases is not the sum
of the temperature changes from individual contributions of
the four gases. This indicates that the effects from these gases
are not linear. The gases affect each other through chemical
reactions, as well as temperature change. Temperature change
can affect temperature-dependent chemical reaction rates.
[24] Figure 10 shows the global mean altitude distributions
of neutral density changes between 1983 and 2003 due to
contributions from CO2, O3, CH4/H2O, and the total due to
all these gases, simulated by the 1-D model using the
WACCM climatological gas profiles below the thermosphere. The neutral density decrease at 400 km due to all
the gases (black) is 4.8% in this 20 years. This is in excellent
agreement with results from long-term satellite drag data
analysis [e.g., Marcos et al., 2005; Emmert et al., 2008], which
range from 1.7%/decade to 3.0%/decade at 400 km, for the
past three to four decades [Qian et al., 2011]. CO2 forcing
caused a neutral density decrease of 4.5% at 400 km. O3 depletion caused a maximum density decrease of ~2% near 100 km.
Although the temperature change due to O3 depletion is negligible in the middle and upper thermosphere, O3 depletion caused
a density decrease of ~0.25% at these altitudes. This highlights a
fact that temperature change is a local effect whereas density
change at a fixed altitude is cumulative. Cooling in the
stratosphere and mesosphere due to O3 depletion causes a
density decrease in the thermosphere through thermal contraction. CH4/H2O increases caused a density increase in the
mesosphere and thermosphere, with a 0.7% density increase at
400 km from 1983 to 2003. Effects of CO2 increase, O3 depletion, and CH4/H2O increases are comparable and important in
the mesosphere, whereas in the thermosphere, especially in
the middle and upper thermosphere, CO2 is the main driver
of the global cooling. At 400 km, density decrease due to
CO2 alone is ~4.5% compared to the total density decrease
of ~4.8% due to all gases.
4.
Conclusions
[25] In this paper, we conducted model simulations to examine how changes in concentrations of CO2, O3, CH4, and
H2O affect long-term changes in the upper atmosphere.
Concentrations of both CO2 and CH4 have increased consistently in the industrial era due to human activities. Increased
concentration of CH4 has increased H2O concentration in the
stratosphere through oxidation of CH4. Halocarbon
3852
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
Figure 7. Global mean profiles of gas concentrations simulated by WACCM for the 2003 control case
(scenario #2, red) and the 1983 control case (scenario #1, black). Vertical axis on the right shows geometric
heights of model pressure surfaces. (a) CO2, (b) CH4, (b) H2O, and (c) O3.
refrigerants, especially CFC-11 and CFC-12, have also increased in the industrial era. Halocarbon compounds produce
chlorine atoms that catalytically destroy O3 and have caused
stratospheric O3 depletion in the past several decades.
[26] Our model simulations demonstrated that O3 causes
cooling in the mesosphere (maximum decrease of 0.5 K from
1983 to 2003) that is comparable to the cooling caused by
CO2 forcing, and a warming effect around 110 km due to
large positive temperature gradient in this altitude region
and the cooling and contraction of the underlying atmosphere. This is consistent with the results of Akmaev et al.
[2006]. In addition, the model simulations show that O3
depletion has nearly no effect above the lower thermosphere,
but it causes ~0.25% neutral density decrease in the thermosphere from 1983 to 2003, due to thermal contraction in the
stratosphere and mesosphere. WACCM simulated O3 depletion is consistent with various observational data sets [Garcia
et al., 2007]. Even if the actual O3 depletion from 1983 to
2003 were twice as large as the WACCM calculation, the
O3 effect on the thermosphere would still be minor compared
to CO2 forcing. Therefore, our conclusion is that O3 depletion is not a main driver of the thermosphere cooling,
contradicting the speculation of Walsh and Oliver [2011].
[27] Long-term changes in the thermosphere are mainly
caused by CO2, with minor contributions from O3, CH4,
and H2O. Thermosphere neutral density decreases by 4.5%
at 400 km from 1983 to 2003 due to CO2 forcing alone,
whereas it decreases by 4.8% due to effects from all four
gases. The modeling results indicate that the increase of
CH4 and H2O slightly warms the upper mesosphere and thermosphere, due to increased chemical heating and absorption
of solar ultraviolet radiation. The model simulations show
that temperature effects from the four gases are nonlinear.
The contribution from each gas does not add up to their total
temperature effect, because the gases affect each other
through chemical reactions. In addition, temperature change
can affect chemical reaction rates.
[28] In this study, we used the 1-D version of TIME-GCM
to conduct the simulations. The upper atmosphere cooling
due to CO2 forcing simulated by the 1-D version of TIMEGCM [Roble and Dickinson, 1989] is consistent with that
simulated by the 3-D version of TIME-GCM [Qian et al.,
2011]. Similarly, we expect that the effects of O3, CH4, and
H2O forcing on the upper atmosphere simulated by the 1-D
model will be consistent with those simulated by the 3-D version of TIME-GCM.
[29] Future modeling studies should aim to use one general
circulation model that extends from the Earth’s surface to the
top of the thermosphere, coupled to a detailed chemical transport model of the troposphere and the middle atmosphere.
However, this more comprehensive approach should be in
qualitative agreement with the results shown here. In addition,
with the recovery from ozone depletion, the mesospheric
cooling trend is expected to gradually decrease, while the thermospheric cooling trend will be largely unaffected.
[30] Acknowledgments. This research was supported by NASA
grants NNX10AF21G and NNX09AJ60G to the National Center for
Atmospheric Research. We would also like to acknowledge the Center for
Integrated Space Weather Modeling (CISM), which is funded by the
National Science Foundation’s STC program under agreement number
ATM-0120950. NCAR is sponsored by the National Science Foundation.
[31] Robert Lysak thanks Uwe Berger and another reviewer for their assistance in evaluating this paper.
3853
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
(a)
(b)
Figure 8. Global mean profiles of neutral temperature (TN)
changes between the 2003 control case (scenario #2) and the
1983 control case (scenario #1) (scenario #2 minus scenario
#1). Vertical axis on the right shows geometric heights of model
pressure surfaces. (a) Simulated by the 1-D model and (b) simulated by WACCM.
Figure 9. Global mean profiles of neutral temperature (TN)
changes from 1983 to 2003 due to CO2 increase (scenario #2 minus scenario #3), O3 depletion (scenario #2 minus scenario #4),
CH4/H2O increase (scenario #2 minus scenario #5), and concentration changes of all these gases (scenario #2 minus scenario #1)
simulated by the 1-D model. Vertical axis on the right shows geometric heights of model pressure surfaces.
Figure 10. Global mean profiles of neutral density
changes from 1983 to 2003 due to CO2 increase (scenario
#2 minus scenario #3), O3 depletion (scenario #2 minus
scenario #4), CH4/H2O increase (scenario #2 minus
scenario #5), and concentration changes of all these gases
(scenario #2 minus scenario #1). The y axis values are
geometric heights. The star symbol gives global average
neutral density trend from 1967 to 2007 at 400 km from
Emmert et al. [2008], which is derived from orbits of
5000 near-Earth objects from 1967 to 2007. The triangle
symbol shows average neutral density trend from 1970
to 2000 at 400 km from Marcos et al. [2005], which is
derived from orbits of five satellites with moderately
eccentric orbits from 1970 to 2000.
References
Akmaev, R. A. (2012), On estimation and attribution of long-term temperature trends in the thermosphere, J. Geophys. Res., 117, A09321,
doi:10.1029/2012JA018058.
Akmaev, R. A., and V. I. Fomichev (1998), Cooling of the mesosphere and
lower thermosphere due to doubling of CO2, Ann. Geophys., 16,
1501–1512.
Akmaev, R. A., and V. I. Fomichev (2000), A model estimate of cooling in
the mesosphere and lower thermosphere due to the CO2 increase over the
last 3–4 decades, Geophys. Res. Lett., 27, 2113–2116.
Akmaev, R. A., V. I. Fomichev, and X. Zhu (2006), Impact of middleatmospheric composition changes on greenhouse cooling in the upper
atmosphere, J. Atmos. Sol. Terr. Phys., 68, 1879–1889, doi:10.1016/
j.jastp.2006.03.008.
Brasseur, G., and S. Solomon (1995), Aeronomy of the Middle Atmosphere,
2nd ed., D. Reidel, Springer, The Netherlands.
Brasseur, G., J. Orlando, and G. Tyndall (Eds.) (1999), Atmospheric
Chemistry and Global Change, 654 pp. Oxford Univ. Press, New York.
Collins, W. D., et al. (2006), The Community Climate System Model version
3 (CCSM3), J. Clim., 19, 2122–2143.
Emmert, J. T., J. M. Picone, and R. R. Meier (2008), Thermospheric
global average density trends 1967–2007, derived from orbits of
5000 near-Earth objects, Geophys. Res. Lett., 35, L05101,
doi:10.1029/2007GL032809.
Fomichev, V. I., J.-P. Blanchet, and D. S. Turner (1998), Matrix parameterization of the 15-mm CO2 band cooling in the middle and upper atmosphere for
variable CO2 concentration, J. Geophys. Res., 103, 11,505–11,528.
Fomichev, V. I., C. Fu, J. de Grandpré, S. R. Beagley, V. P. Ogibalov, and
J. C. McConnell (2004), Model thermal response to minor radiative energy
sources and sinks in the middle atmosphere, J. Geophys. Res., 109,
D19107, doi:10.1029/2004JD004892.
Forster, P., et al. (2007), Changes in Atmospheric constituents and in radiative forcing, in Climate Change 2007: The Physical Science Basis.
Contribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change, edited by S. Solomon
et al., pp. 130–234, Cambridge Univ. Press, Cambridge, U. K.
3854
QIAN ET AL.: TRACE GASES IMPACT ON UPPER ATMOSPHERE
Garcia, R. R., D. R. Marsh, D. E. Kinnison, B. A. Boville, and F. Sassi (2007),
Simulation of secular trends in the middle atmosphere, 1950–2003,
J. Geophys. Res., 112, D09301, doi:10.1029/2006JD007485.
Janjic, Z., R. Gall, and M. E. Pyle (2010), Scientific Documentation for the
NMM Solver, NCAR Tech. Note NCAR/TN-477+STR, doi:10.5065/
D6MW2F3Z.
Khvorostovskaya, L. E., I. Y. Potekhin, G. M. Shved, V. P. Ogibalov, and
T. V. Uzyukova (2002), Measurement of the rate constant for quenching
1
CO2(01 0) by atomic oxygen at low temperatures: Reassessment of the
rate of cooling by the 15-mm emission in the lower thermosphere,
Atmos. Ocean. Phys., 38, 694–706.
Lastovicka, J. (2012), On the role of ozone in long-term trends in the
upper atmosphere–ionosphere system, Ann. Geophys., 30, 811–816,
doi:10.5194/angeo-30-811-2012.
Lastovicka, J., R. A. Akmaev, G. Beig, J. Bremer, and J. T. Emmert (2006),
Global change in the upper atmosphere, Science, 314, 1253–1254.
López-Puertas, M., R. Rodrigo, A. Molina, and F. W. Taylor (1986a), A nonLTE radiative transfer model for infrared bands in the middle atmosphere,
I, Theoretical basis and application to CO2 15mm bands, J. Atmos. Terr.
Phys., 48, 729–748.
López-Puertas, M., R. Rodrigo, J. J. López-Moreno, and F. W. Taylor
(1986b), A non-LTE radiative transfer model for infrared bands in the
middle atmosphere, II, CO2 (2.7 m and 4.3 m and water vapour (6.3 m bands
and N2(1) and O2(1) vibrational levels, J. Atmos. Terr. Phys., 48, 749–764.
Marcos, F. A., J. O. Wise, M. J. Kendra, N. J. Grossbard, and B. R. Bowman
(2005), Detection of a long-term decrease in thermospheric neutral
density, Geophys. Res. Lett., 32, L04103, doi:10.1029/2004GL021269.
Marsh, D. R., R. R. Garcia, D. E. Kinnison, B. A. Boville, F. Sassi, and
S. C. Solomon (2007), Modeling the whole atmosphere response to solar
cycle changes in radiative and geomagnetic forcing, J. Geophys. Res.,
112, D23306, doi:10.1029/2006JD008306.
Qian, L., R. G. Roble, S. C. Solomon, and T. J. Kane (2006), Calculated and
observed climate change in the thermosphere, and a prediction for solar
cycle 24, Geophys. Res. Lett., 33, L23705, doi:10.1029/2006GL027185.
Qian, L., S. C. Solomon, R. G. Roble, and T. J. Kane (2008), Model simulations of global change in the ionosphere, Geophys. Res. Lett., 35, L07811,
doi:10.1029/2007GL033156.
Qian, L., A. G. Burns, S. C. Solomon, and R. G. Roble (2009), The effect of
carbon dioxide cooling on trends in the F2-layer ionosphere, J. Atmos. Sol.
Terr. Phys., 71, 1592–1601, doi:10.1016/j.jastp.2009.03.006.
Qian, L., J. Lastovicka, R. G. Roble, and S. C. Solomon (2011),
Progress in observations and simulations of global change in the
upper atmosphere, J. Geophys. Res., 116, A00H03, doi:10.1029/
2010JA016317.
Rishbeth, H., and R. G. Roble (1992), Cooling of the upper atmosphere by
enhanced greenhouse gases - Modelling of thermospheric and ionospheric
effects, Planet. Space Sci., 40, 1011–1026.
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., and R. E. Dickinson (1989), How will changes in carbon dioxide and methane modify the mean structure of the mesosphere and lower
thermosphere?, Geophys. Res. Lett., 16, 1441–1444.
Roble, R. G., and E. C. Ridley (1994), A thermosphere-ionospheremesosphere-electrodynamics general circulation model (TIME-GCM):
Equinox solar cycle minimum simulations (30–500 km), Geophys. Res.
Lett., 21, 417–420.
Roble, R. G., E. C. Ridley, and R. E. Dickinson (1987), On the global mean
structure of the thermosphere, J. Geophys. Res., 92, 8745–8758.
Solomon, P. M., B. Connor, R. L. de Zafra, A. Parrish, J. Barrett, and
M. Jaramillo (1987), High concentrations of chlorine monoxide at low
altitudes in the Antarctic spring stratosphere: Secular variation, Nature,
328(6129), 411–3, doi:10.1038/328411a0.
Walsh, P. L., and W. L. Oliver (2011), Is thermospheric long-term cooling
due to CO2 or O3?, Ann. Geophys., 29, 1779–1782, doi:10.5194/angeo29-1779-2011.
Wintersteiner, P. P., R. H. Picard, R. D. Sharma, J. R. Winick, and
R. A. Joseph (1992), Line-by-line radiative excitation model for the
nonequilibrium atmosphere: Application to CO2 15 mm emission,
J. Geophys. Res., 97, 18083–18117.
Zhang, S.-R., J. M. Holt, and J. Kurdzo (2011), Millstone Hill ISR observations of upper atmospheric long-term changes: Height dependency,
J. Geophys. Res., 116, A00H05, doi:10.1029/2010JA016414.
3855