Stratospheric transport in global circulation models and chemistry–climate
models is an important component in simulating the recovery of the ozone
layer as well as changes in the climate system. The Brewer–Dobson
circulation is not well constrained by observations and further investigation
is required to resolve uncertainties related to the mechanisms driving the
circulation. This study has assessed the specified dynamics mode of the
Canadian Middle Atmosphere Model (CMAM30) by comparing to the Atmospheric
Chemistry Experiment Fourier transform spectrometer (ACE-FTS) profile
measurements of CFC-11 (CCl3F), CFC-12 (CCl2F2), and N2O.
In the CMAM30 specified dynamics simulation, the meteorological fields are
nudged using the ERA-Interim reanalysis and a specified tracer was employed
for each species, with hemispherically defined surface measurements used as
the boundary condition. A comprehensive sampling technique along the line of
sight of the ACE-FTS measurements has been utilized to allow for direct
comparisons between the simulated and measured tracer concentrations. The
model consistently overpredicts tracer concentrations of CFC-11, CFC-12, and
N2O in the lower stratosphere, particularly in the northern hemispheric
winter and spring seasons. The three mixing barriers investigated, including
the polar vortex, the extratropical tropopause, and the tropical pipe, show
that there are significant inconsistencies between the measurements and the
simulations. In particular, the CMAM30 simulation underpredicts mixing
efficiency in the tropical lower stratosphere during the
June–July–August season.
Introduction
As highlighted by , interest in stratospheric
transport has increased over the last 20 years as a result of significant
developments in stratosphere-resolving general circulation models (GCMs)
e.g. and chemistry–climate models
(CCMs) e.g.. Accurate projections of
stratospheric ozone and climate rely on the ability of these models to
simulate stratospheric transport and chemistry. The distribution of
long-lived trace gases in the stratosphere is primarily controlled by the
Brewer–Dobson circulation (BDC), which is generally characterized by
tropospheric air entering the stratosphere in the tropics, poleward
transport, and descent in the midlatitude and polar regions of the winter
hemisphere e.g.and references
therein. The BDC describes the primary
features of stratospheric circulation, based upon a conceptual model proposed
to explain observations of ozone and water vapour
. Over the last decade, the
response of stratospheric circulation to changes in anthropogenic climate
forcing has been studied . It is now understood that
the residual circulation and quasi-isentropic mixing are key factors to
understanding the structure of the BDC
.
and state that two-way mixing
between the tropics and extratropics and the rapid stirring of air parcels
are important components of stratospheric transport. The influence of
planetary waves on stirring is predominately in the winter midlatitude surf
zone but synoptic-scale wave
activity occurs throughout the year in the subtropical lower stratosphere and
its influence can extend upwards to 25 km .
argued that the greenhouse gas induced warming of the
climate system has led to an upward displacement of the critical layers for
wave breaking. Subsequently, it has been suggested that the BDC changes are
characterized more by a vertical lifting of rather than an acceleration of
the meridional circulation .
and identified two distinct
pathways within the BDC. These are the “deep branch”, defined as the
poleward transport in the winter hemisphere extending into the middle and
upper stratosphere, and the “shallow branch(es)”, defined as multiple
pathways of faster poleward transport that are observed in both hemispheres
throughout the year and are generally restricted to the lower-to-middle
stratosphere. It is likely that the shallow branches are driven by
Rossby wave pumping on a synoptic scale .
As part of the Climate Chemistry Model Validation (CCMVal) project,
investigated simulated changes in the BDC by considering
three branches separately: the transition, shallow, and deep branches. They
found that changes in the transition and shallow branches of the BDC were
consistent with the increase of greenhouse gas concentrations and the trends
were associated with changes in subtropical jets and tropical upper
tropospheric temperatures, which is also consistent with the mechanism
described by . The acceleration of the deep branch is
consistent with that of the transition and shallow branches but is seasonally
modulated by changes in ozone concentrations with the exact mechanisms yet to
be determined . The mechanisms that lead to an acceleration
or deceleration in the deep branch remain unresolved
. Observational evidence seems to indicate
a deceleration in the deep branch .
However, there is a fair degree of confidence that the acceleration of the
shallow branch is likely since it is driven by the vertical lifting mechanism
proposed by , which is related to large-scale changes
and also has been diagnosed from changes in stratospheric constituent
distributions .
The BDC is well characterized in models but remains poorly constrained by
observations . A direct comparison to determine how
the BDC and quasi-horizontal mixing combine to produce the distribution of
long-lived tracers with tropospheric sources is not possible. Therefore, a
number of observational techniques have been used to investigate
stratospheric transport characteristics, such as age of air diagnostics
, tropical lower stratosphere ascent
rates , and descent rates in the Antarctic
polar vortex and in the
Arctic polar vortex
. Except
for , these observations do not provide global
seasonally resolved quantitative estimates of the BDC
. Thus, identified that it is
difficult to deduce changes in the strength of the BDC using available
measurements
e.g..
Recently, determined that, for the BDC, a period of 30 years is required for a trend to be identified from noise due to natural
variability. They also found that dynamic variability can obscure a trend in
the BDC if it is based on less than 12 years of data .
It is clear that the transport of chemical tracers will be impacted by
changes in the BDC, which will in turn influence ozone recovery projections,
lifetimes of ozone depleting gases, and mass exchange between the troposphere
and stratosphere . Understanding how the structure of
the BDC will change depends greatly upon the ability to simulate its current
behaviour. This is typically assessed by investigating how capable a model is
at simulating tracer concentrations
– in particular,
assessing the characteristics of the simulated tracers such as concentrations
and behaviour at the mixing barriers in the stratosphere (i.e. the polar
vortex edge, the extratropical tropopause, and the tropical pipe).
In this study, measurements of the long-lived chlorofluorocarbons, CCl3F
(CFC-11) and CCl2F2 (CFC-12), and N2O from the Atmospheric Chemistry
Experiment Fourier transform spectrometer (ACE-FTS) are used to evaluate the
specified dynamics simulation mode of the Canadian Middle Atmosphere Model
(CMAM30). Using these global measurements, areas in which simulated tracers
agree with observations and where improvements are needed have been
investigated. Since ACE-FTS measurements are vertically resolved, these data
are useful for testing model simulations
e.g.; however, care
must be taken in the methods used for these comparisons. In addition to
comparison methods, external tools are useful for the interpretation of
differences between CMAM30 and ACE-FTS. An idealized stratospheric model, the
tropical leaky pipe (TLP) model, described by has been
used to test factors contributing to the differences observed between ACE-FTS
and CMAM30.
Many limb-viewing satellite missions, including ACE, are contributing data to
the SPARC Data Initiative, whose purpose is to compile and assess a
repository of climatologies for comparison with model output
. Some of
the instruments involved in the initiative, such as ACE-FTS, do not cover all
latitudes and altitudes each month. This matter has been the subject of
recent studies . They have
shown that the impacts of sampling patterns of various instruments must be
considered when performing comparisons because the model simulates data
points that are evenly spaced throughout each latitude range in the zonal
mean at each pressure level while the measurements can represent a subset of
that domain.
This work addresses issues related to climatological comparisons in several
ways: first, by using a nudged version of the CMAM; second, by breaking out
the species of interest from the halocarbon family arrangement in the model;
and third, by sampling the model output along the individual measurement
profile pathways through the atmosphere. The latter addresses sampling issues
identified by , , and
. By isolating the model output in this way, the transport
and chemical processes in CMAM can be evaluated. This type of assessment of
CMAM simulations has not been possible until recently because of the
typically free-running nature of the model simulations. The CMAM specified
dynamics simulation has been investigated in a few recent studies:
evaluated the polar cap mesospheric transport and
midlatitude mean zonal winds, and long term observational records of water
vapour and ozone were also used to evaluate the CMAM30 run by
and , respectively.
Additionally, investigated the CMAM30 polar regions
using satellite data, including ACE-FTS. Comparisons of the CMAM30
simulations to observations remain limited in their extent, a gap which
this work attempts to fill.
This paper is structured as follows: Sect. 2 describes the tools used in
this study including measurements from ACE-FTS, CMAM30 simulations, and
TLP model simulations. Section 3 examines methods of sampling and comparison
techniques and considers the impact of the sampling of ACE-FTS. Section 4
examines the measured and simulated zonally averaged morphologies of CFC-11,
CFC-12, and N2O. Section 5 investigates the three barriers to mixing: the
polar vortex, the extratropical tropopause, and the tropical pipe. In Sect. 6, the TLP model is used to investigate changes in tropical upwelling and
quasi-isentropic mixing required to allow the model to more effectively
simulate the lower stratosphere. Finally, the results are summarized and
discussed in Sect. 7.
ToolsACE-FTS
The ACE mission on board the Canadian satellite SCISAT was launched to
investigate the distribution of upper tropospheric and stratospheric ozone
with the goal to further our understanding of the chemical and dynamical
processes that influence its behaviour . ACE entered a
circular low-earth orbit (650 km, 74∘ inclination), on 12 August 2003,
to observe the earth's atmosphere using the solar occultation technique
. The high-resolution Fourier transform spectrometer
ACE-FTS is the primary instrument on SCISAT. It has a high spectral
resolution (0.02 cm-1) and a spectral range of 2.2–13.3 µm
(750–4400 cm-1) . ACE-FTS does not
require filters for its operation, which allows it to measure solar
absorption spectra for dozens of atmospheric constituents simultaneously.
ACE-FTS is ideal for studying the vertical structure of constituent gases
from cloud tops to 100 km. The retrieved profiles are particularly useful in
the upper troposphere and lower stratosphere, where the vertical resolution is
approximately 3 km . The ACE-FTS products are derived
from the solar absorption spectra measured and include the vertical profiles
of temperature, pressure, and the concentration expressed as a volume-mixing
ratio (VMR) for several dozen molecules of atmospheric interest over
latitudes from 85∘ N to 85∘ S . These
data products are useful for the development of climatologies
e.g, trends
e.g., and lifetimes e.g.,
among other applications
e.g..
The version 3.0 ACE-FTS retrievals of CFC-11 (CCl3F), CFC-12
(CCl2F2), and N2O based on spectra recorded between June 2004 and
May 2010 have been used throughout this work. A description of the retrieval
process used for ACE-FTS is provided by .
The earlier work details the retrieval process for version 2.2 of the data
while the latter describes improvements that have been implemented for the
recent versions. The CFC-11 retrieval ranges from 5 km to a maximum of
28 km in the tropical latitudes but is limited in the stratosphere at higher
latitudes due to low concentrations. CFC-12 is retrieved between 5 and
36 km, while the N2O retrieval covers 5 to 95 km. Due to the vertical
limitation of the CFC-11 retrieval, this work focuses on the upper
troposphere and lower stratosphere (5 to 30 km).
The validity of these measurements has been investigated in several studies.
compared both the ACE-FTS CFC-11 and CFC-12 v2.2
products to the FIRS-2 and MkIV Interferometer measurements. They found
ACE-FTS to be approximately 10 % lower than FIRS-2 between 12 and 16 km
in the case of CFC-11. Using a non-coincident technique,
found agreement to be better than 20 % between
ACE-FTS and MkIV over the range of 17 to 24 km. The CFC-12 comparisons
made by and show a consistent
difference with ACE-FTS approximately 10 % lower than MkIV. Using a
climatological validation approach within the SPARC Data Initiative to
compare the CFC-11 and CFC-12 products of HIRDLS, MIPAS, and ACE-FTS,
found excellent agreement in the lower stratosphere
(up to 50 hPa) and increasing positive deviations above this level to around
20 % from the multi-instrument mean. provided an
extensive validation of the ACE-FTS v2.2 N2O product using satellite,
aircraft, balloon, and ground-based FTIR measurements. Differences observed
were typically within +15 % . The work of
was consistent with the results of
. compared the previously
validated CFC-11 and CFC-12 v2.2 products with the v3.0 products used in this
work and found that around 15 km there is a slight increase in CFC-11 in the
new product, bringing it closer to the correlative measurements used in the
studies described above. Similarly for CFC-12, found
an increase of 2–5 % between 6 and 22 km. They also found an
approximately 10 % decrease in the concentration of N2O above 35 km in
v3.0, bringing the differences found in to within
approximately +5 %.
A set of quality control flags for the ACE-FTS products on the 1 km
retrieval grid is available . The version 1.1 flags
were applied to the data used in this work by removing profiles that
contained a flag between 4 and 7, as recommended by . The
method rejects a maximum of 6 % of data over all the species retrieved. For
CFC-11, CFC-12, and N2O, 1.7, 2.1, and 4.3 % of the data are
rejected, respectively . In addition to the quality control flags,
derived meteorological products from GEOS 5.2.0, based on the techniques
described in , are used here along with the geographic
location information to account for the geographic extent and meteorological
context of ACE-FTS profiles.
Canadian Middle Atmosphere ModelThe model
CMAM is a freely running CCM based on an upwardly extended version of the
Canadian Centre for Climate Modelling and Analysis (CCCma) third-generation
Atmospheric General Circulation Model .
The chemistry includes the Ox, HOx, NOx, ClOx, and BrOx
catalytic cycles that control ozone in the stratosphere; the chemistry of
N2O, CH4, and seven long-lived halocarbon species; and a representation
of heterogeneous chemistry on background stratospheric sulfate aerosols and
on polar stratospheric clouds . CMAM
has been used extensively to investigate the middle atmosphere and to study
complex processes of the climate system
e.g..
Results from the CMAM have also been assessed during two phases of the CCMVal
project and, more recently, the Chemistry-Climate Model Initiative (CCMI).
The extensive investigation of dynamical and chemical processes in CCMs that
took place during CCMVal-2 are detailed in the SPARC report
. For the simulations used here, CMAM was run at a T47
spectral resolution, equivalent to approximately 3.75∘× 3.75∘, with 71 vertical levels topping out at 0.08 Pa (approximately
equivalent to 95 km in altitude).
Due to the chaotic nature of atmospheric circulation, free-running models are
unable to reproduce the day-to-day evolution of the atmosphere. Therefore,
simulated fields, such as tracer concentrations, cannot be compared to
observations directly or on a day-to-day basis. There has recently been an
effort to circumvent this limitation by constraining the evolution of the
circulation and temperatures in CCMs with fields from reanalysis datasets
through the use of Newtonian relaxation
e.g., known colloquially as
“nudging”. The ability to constrain the dynamical fields to follow the
reanalysis, the best approximation of reality, enables direct
model–measurement comparisons of chemical tracers in the model by eliminating
the internal variability in the simulated circulation. This type of specified
dynamics simulation has been used here to allow for space and time matched
comparisons of output from CMAM with ACE-FTS observations over the June 2004
to May 2010 period.
The specified dynamics simulation
CMAM30, the specified dynamics version of the CMAM, uses meteorological
fields from the ERA-Interim reanalysis to constrain the
dynamical fields while the chemical fields are allowed to freely evolve. The
model horizontal winds and temperature are nudged towards the ERA-Interim
fields from the surface to 1 hPa. The 6-hourly reanalysis data are linearly
interpolated in time to produce fields for nudging at intermediate time
steps. The relaxation is applied at only the synoptic scales and larger by
constraining only wave numbers up to 21. The application of nudging on only
the large scales and the use of a relaxation time constant of 24 h has
been found to produce root-mean-square differences between the CMAM30 and the
reanalysis comparable to those found between different reanalysis datasets
for fields such as temperature and vorticity .
examined improvements in transient features of the
model. In addition, some minor adjustments were made to the global average
temperature in the ERA-Interim fields at 5 hPa and above to remove
discontinuities associated with changes in the observing system as described
by . Sea surface temperatures and sea ice were
specified using the HadISST dataset . The nudging helped
correct some large-scale biases in CMAM and a comparison of more transient
circulation features like sudden stratospheric warmings against independent
observations was improved by nudging above 10 hPa.
As noted, tracers in CMAM30 evolve freely subject to advection by the
resolved circulation and vertical redistribution by physical
parameterizations. Advection of tracers is calculated using spectral
advection, which is inherently mass conservative though not necessarily
positive definite. The generation of negative concentrations upon
transformation from spectral to physical space is corrected through “hole
filling” with any artificially added mass to remove negatives tracked and
corrected for in the global average. The tracers analysed here are long lived
and smoothly varying, resulting in spatial distributions that are well
represented in spectral space and produce minimal problems with the
generation of negative concentrations. No nudging of surface pressure is
performed and the global average surface pressure is continually corrected
back to a predefined constant value in the CMAM30 simulation, in the exact
same manner as is done in free-running simulations. While mass conservation
in the CMAM30 simulation has not been analysed specifically, no significant
differences with free-running simulations have been seen for diagnostics such
as the evolution of total stratospheric chlorine.
The standard CMAM chemical mechanism uses a lumping approach for the
halocarbon tracers to reduce the number of chemical species that must be
transported by the model. A limited number of halocarbons are explicitly
treated by the chemical scheme and the remaining long-lived halocarbons are
combined into the model species based on their “fractional release values”
. The concentration specified as a lower boundary
condition is increased so that the total amount of organic chlorine (or
bromine) of all halocarbons represented by the model species is conserved.
For example, the model explicitly treats the chemistry of CFC-12
(CCl2F2), but the concentration of CFC-12 was increased to account for
the additional chlorine carried by CFC-113. Because of the time-varying
contribution of the individual halocarbons to the tropospheric concentration
of the model species, the numerous assumptions that would be required to
rescale the model species concentration that could be compared with
observations would introduce significant uncertainties. Therefore, to
directly compare the model halocarbon concentrations with observations, a
parallel set of halocarbons was added to the model that explicitly represents
individual halocarbon species. These parallel species undergo the appropriate
chemical reactions using the photolysis rates and concentrations calculated
by the full model chemical mechanism, though the reactions of the parallel
species do not feed into the concentration of other model species. The CMAM30
simulation including the additional explicit halocarbon species will be
referred to as CMAM30HR for the remainder of this work.
Influence of the surface boundary conditions
In CMAM, global average concentrations are typically applied as the lower
boundary condition for long-lived species, such as N2O. To capture the
interhemispheric differences in tropospheric concentrations of the
halocarbon tracers, the parallel species have separate northern and southern
hemispheric surface mixing ratios imposed as lower boundary conditions. The
application of hemispherically defined lower boundary conditions based on
observations is consistent with the proposed approach for the upcoming sixth
phase of the Coupled Climate Model Intercomparison Project (CMIP6)
. The lower boundary conditions were derived from the
annual average hemispheric mixing ratios from the National Oceanographic and
Atmospheric Administration's Halocarbon and other Atmospheric Trace Species
(HATS) program . The annual average
values were linearly interpolated in time to calculate an instantaneous
surface-layer mixing ratio for the model. The model mixing ratio in the
lowest six model layers (approximately the lowest 1 km) was relaxed towards
the specified concentration with a time constant that increased from 25 days
near the Equator to 12 h at 25∘ of latitude. In CMAM30HR, all
important losses for the species of interest have been considered. The
photolysis rates and reaction rates have been updated to the values from
JPL-2010 . The chemical losses of CFC-11 are dominated
by reactions with O(1D) and photolysis in the mid-to-lower stratosphere,
particularly in the tropical region. The chemical losses of CFC-12 and N2O
are similar to that of CFC-11, except they generally occur at a higher
altitude in the stratosphere. CFC-11, CFC-12, and N2O losses are
insignificant in the troposphere.
Comparison of CMAM30HR simulations of CFC-11 (blue), CFC-12 (green),
and N2O (black) to the HATS surface flask network of measurements at
various locations around the world. Locations of measurement sites are
indicated by latitude. Relative differences are calculated as the difference
between the concentration at the surface site and the lowest model layer of
the nearest neighbour grid box to the site in the CMAM30HR output, divided by
the measured concentrations. The relative differences were calculated based
on the monthly averaged observations and simulations. Shown here are the mean
of the relative differences between May 2004 and June 2010 and the error bars
indicate 1 standard deviation of the mean of the relative differences over
the time period.
To ensure the consistency of the boundary conditions applied to the CMAM30HR
simulation, the model output was compared to measurements at surface
monitoring sites using data from the NOAA HATS program. The monthly mean
measurements of N2O, CFC-11, and CFC-12 have been compared to the monthly
mean in the CMAM30HR output between 2004 and 2010. Because of the way the
surface boundary condition was imposed, each site was compared to the lowest
model level of the closest grid point in the CMAM30HR output. The relative
differences have been calculated by subtracting the measurement from the
simulation and dividing by the measurement for each month in the time series.
The differences over time were compared and no trend in the differences was
observed. The comparisons of each trace gas at each site are summarized in
Fig. , ordered by latitudinal location. Data included here are
averaged over the 2004 to 2010 period and error bars (±) indicate 1
standard deviation.
Both CFC-11 and CFC-12 simulations have reasonably small differences,
generally less than 1 %, from the surface measurements. Over the time
period compared, CMAM30HR appears to underpredict CFC-11 at all HATS sites
while the CFC-12 comparisons are not significantly different from zero within
1 standard deviation for all but two sites in the Southern Hemisphere. The
N2O comparisons show that the model consistently overpredicts the
concentrations at 11 of the 13 sites. There appears to be some latitudinal
dependence in the comparisons of N2O which may be caused by the
application of a globally averaged boundary condition in the run. Generally
the differences for all the species shown are within ±1 %, which is to
be expected since the model boundary conditions were derived from these
measurements.
Influence of nudging on the age of stratospheric air
The mean age of stratospheric air, i.e. the average time elapsed since the
last time an air parcel was in the troposphere, can provide a diagnostic for
determining differences in isentropic transport and mixing between different
model runs . The stratospheric age of air in
the model is derived from an idealized SF6 tracer whose lower boundary
condition linearly increases over time. In this section, the CMAM30, rather
than CMAM30HR, mean age of air is used since no transport changes were made
between the two runs of the model. Averaged between 2004 and 2010, Fig. a shows the zonal distribution of the CMAM30 mean age subtracted
from the mean age from an identical, but freely running, version of the CMAM
using the same specified sea surface temperatures and sea-ice data. The
differences in age range from approximately -1.25 to +1 years, where
the positive differences indicate areas where the air in CMAM is older than
that in CMAM30, and the negative differences indicate areas in which the
CMAM30 age is older than in the CMAM age. In general, the nudging of CMAM
appears to affect the Southern Hemisphere more than the Northern Hemisphere.
Below 50 hPa in the tropics and midlatitude regions, the difference between
the age in the two versions of the model is close to zero. Above 50 hPa in
the tropics and midlatitudes and above 150 hPa in the polar regions, air in
CMAM is older than that in CMAM30, with peaks occurring around the surf zones
in the stratosphere. This implies that for the majority of the lower
stratosphere the process of nudging leads to an apparent decrease in the CMAM
age of air. The cause of this has not been fully explained in the literature
at this time; however, it can be speculated that nudging the model to the
reanalysis could be a source of artificial drag that drives the BDC to be
more rapid than in the free-running version of the model.
A comparison of the age of stratospheric air in the free-running
CMAM and CMAM30 averaged between 2004 and 2010. (a) The zonal mean
difference of the CMAM30 mean age subtracted from the free-running mean age
(years) and the monthly time series difference (b) at
80∘ S and (c) at 80∘ N.
In the polar regions of Fig. a, the comparisons exhibit a different
behaviour in the lowermost stratosphere relative to the rest of the
stratosphere. The differences are close to zero at the lowest pressure level
shown (400 hPa). Above approximately 300 hPa, the CMAM30 air is older than
the CMAM air and younger above approximately 150 hPa. These differences,
while strongest at the latitudes poleward of 60∘, can extend to
approximately 40∘ latitude in both hemispheres. Figure b and
c show the monthly evolution at 80∘ S and
80∘ N, respectively. Below 150 hPa, the CMAM30 air appears to
be older than the air in the free run and this tends to be pronounced during
the respective summer months in each hemisphere. At 80∘ S, the
pattern of differences in the age of air change in altitude over time. At
approximately 150 hPa in May, the air in CMAM30 is older than the air in
CMAM. This difference appears at approximately 100 hPa by October, with a
larger magnitude. In November and December, during austral spring, the age
difference remains such that the CMAM30 air is older than the air in CMAM. In
Fig. c, the evolution of the differences in age of air at
80∘ N appears restricted to the same altitudes but the seasonal
timing of the pattern is similar. The difference peaks in spring–summer and
dissipates through the fall and winter. The prevalence of the older air in
the polar lowermost stratosphere in the nudged run is significant because,
throughout the stratosphere, air in CMAM30 is younger than that in CMAM. It
is known that the freely running CMAM has a cold bias inside the Antarctic
vortex. suggest that there may be missing gravity
wave drag (GWD) in the Southern Hemisphere based on comparisons of the free-running model simulations and reanalysis data. By effectively adding this
missing GWD through the nudging to reanalysis data, downwelling between
70 and 90∘ S is increased, leading to higher
temperatures – a reduction in the cold bias – during September and
October. The increased downwelling pushes the older air deeper into the
lowermost stratosphere, causing the observed differences in age between the
two versions of CMAM.
In general, synoptic-scale waves are filtered out close to the tropopause and
only planetary-scale waves can propagate further up into the stratosphere
. showed that synoptic-scale wave
drag drives the lower branch of the BDC, while the drag that drives the
deeper parts of the BDC are associated with planetary wave drag. CMAM30
appears to reproduce the upper troposphere simulated in CMAM and, to some
degree, the lower branch of the BDC as well. This is evidenced by the
near-zero differences in Fig. a between 50∘ S and
50∘ N and between 100 and 50 hPa; however,
the absolute ages in this region tend to be quite small. Understanding the
impact of the nudging on the age of air provides the basis for an
interpretation of the isentropic transport and mixing differences between the
two model runs. While it is difficult to quantify the extent to which the
differences in age of air would change tracer concentrations at a given
location, it is necessary to consider these results when considering
implications for the free-running model, particularly for the deep branch of
the BDC. Based on Fig. , CMAM30 clearly has older air in the
extratropical lowermost stratosphere. It is potentially caused by either
stronger downwelling of the older air from above, consistent with a stronger
BDC, or reduced isentropic mixing of tropospheric air from lower latitudes
e.g.. Therefore, the differences in age appear to
suggest a slower shallow branch or a faster deep branch of the BDC.
TLP model
A modified TLP model is used to interpret the differences
between the CMAM30HR simulations and the ACE-FTS measurements. The modified
TLP is based on a set of three coupled one-dimensional equations relating
transport between the tropics and each hemispheric extratropical region
. The model includes advection,
vertical diffusion, and horizontal mixing between the extratropics and the
tropics. Significant changes to the modified version of the TLP model include
common pressure coordinates in all regions and the addition of particle
trajectories with photochemistry. The modification was done to allow for
direct comparisons between TLP output and other models and/or measurements.
The Lagrangian approach is described by . The tropical
boundaries in the TLP model averages were chosen based on observational
estimates of the upwelling region . The model was run with
a vertical resolution of 200 m and a maximum altitude of 40 km above
tropopause; however, the results included here are limited to 30 km in
altitude above the tropopause. To ensure the effectiveness of the TLP as an
interpretation tool, established that the TLP could
accurately simulate the CMAM30HR output with its mean circulation and a
TLP-derived mixing parameter as an input. The mixing parameter was derived
from a suite of simulations conducted with the TLP at varying amounts of
mixing. The resultant best match to the averaged 2004–2010 CMAM30HR
CFC-11, CFC-12, and age of air profiles was the mixing efficiency selected to
initiate the simulations. The TLP model is used here to identify the changes
to the CMAM30HR tropical upwelling and effective mixing that may improve the
comparisons between ACE-FTS and CMAM30HR by testing a range of tropical
upwelling and mixing efficiency settings.
Diagnosing the biases in the model stratospheric circulation requires a
complete separation of the effects of the strength of the BDC and mixing. A
simplified model, such as the TLP, is useful to interpret differences between
measurements and CCMs because of the complexity of wave activity contributing
to stratospheric mean circulation and mixing. It would not be prudent to
adjust model parameterizations in CMAM to modify wave breaking because many
aspects of the model climatology would be impacted with no way of separating
the effects . While the TLP-based analysis does not
identify a specific mechanism, it can separate the contributions of model
biases in the BDC and in mixing to biases in the resulting distribution of
species.
Comparison methods and sampling considerationsMeasurement–model comparison techniques
Two of the comparison techniques used in this study are described here: the
first is the comparison of zonal means and the second is the computation of
joint probability density functions.
Zonal mean comparison technique
To assess the transport and chemistry in CMAM30HR, measurements of N2O,
CFC-12, and CFC-11 are compared with simulated concentrations in
latitude-pressure coordinates. A common method of visualizing the
distribution of long-lived trace gases is the zonal mean cross section. In
this work, data from the ACE-FTS profiles, sampled CMAM30HR profiles, and the
relative difference profiles were averaged in 5∘ latitude bins and
over 18 pressure levels (equally distributed in the log of pressure from 450 to 10 hPa), corresponding to altitude ranges from approximately 5 to
30 km. In these plots, colour contours indicate the VMR of the species. The
comparison of the ACE-FTS measurements and the subsampled CMAM30HR output is
shown as the average of the differences, defined as CMAM30HR minus ACE-FTS
divided by the ACE-FTS measurement. The altitude of the average thermal
tropopauses is typically indicated by a black line. All measurements and
subsampled model output between June 2004 and May 2010 have been included,
representing an average of 6 years of observations and simulations.
Joint probability density functions
Tracer–tracer correlations have been used in a number of studies to identify
transport and mixing characteristics in the stratosphere and to derive
climatologies from sparse data
e.g..
showed that long-lived species exhibit compact
correlations even with varying meteorological conditions, minimizing
discrepancies resulting from sampling and daily variations; thus, sparse
measurements, such as those from aircraft, can be useful in model assessment
studies e.g.. The correlations used here are
N2O–CFC-11 and include only the stratospheric data available (data
located at altitudes above 2 km above the thermal tropopause). The
correlations produced from both the ACE-FTS measurements and CMAM30HR
simulations exhibit compact relationships that tend to be densely populated.
Determining whether the model can capture the clustering in addition to the
overall shape is important to understanding whether the stratosphere is
well-reproduced by the model. Understanding the density distribution of a
dataset is particularly useful for tracer–tracer relationships with compact
correlations. Following the methods of and
, normalized joint probability
density functions (JPDFs) have
been calculated for the ACE-FTS and CMAM30HR correlations described above.
JPDFs are two-dimensional histograms that reveal the clustering of data
and can be used to test how well a model captures the
behaviour of trace gases in the stratosphere. Hegglin and Shepherd (2007)
have shown the impact of ACE-FTS measurement uncertainties in JPDFs by
comparing the full model output, subsampled model output, and ACE-FTS
measurements. They found that there was larger variability in the ACE-FTS
JPDFs compared to those of the subsampled CMAM output.
The influence of beta angle
ACE-FTS records a series of spectra along a slanted path line of sight during
each occultation. The length of this slanted path is different for each
occultation. Each ACE-FTS occultation
is assigned a latitude and longitude at the 30 km tangent point,
geometrically calculated . A sample year of
the geometric 30 km tangent point latitudes is provided in
Fig. (black circles), showing the annual repeating
latitudinal coverage. The beta angle parameter, a measure of the angle
between the solar vector and the satellite orbit plane, has also been
included in Fig. (red circles). The beta angle is an
important parameter to consider because as it changes, so does the geographic
distance between each spectrum acquired through the profile of any given
occultation. The distance is greatest at high beta angles (both positive and
negative), which occur when ACE is in view of the sun for longer periods.
Since the FTS instrument measurement frequency is held at a constant 2 s
interval, more measurements per profile and longer ground-paths of the
retrieved profile occur at high beta angles.
The ACE-FTS sampling pattern for the year 2005. Each black circle is
the latitude of the 30 km tangent height of an occultation and each red
circle is the corresponding beta angle of the occultation.
Considering the impact of observation sampling is a critical step when
comparing measurements with model output. The work of ,
, and illustrate the necessity for
considering the sampling patterns resulting from different measurement
techniques and satellite orbits. The ground path length of a profile is
considered because a single profile can be representative of more than one
geographic region, typically varying more over latitude than longitude. A
refraction model is used to determine the geographic locations along the
slant path of the ACE-FTS profiles . At the
30 km tangent altitude, it has been found that for 98 % of the ACE-FTS
occultations, the difference between the geometric latitude and the
refraction calculation is less than 0.2∘. A useful marker of a nominal
occultation length is at a beta angle of 53∘, corresponding to an
occultation duration of 3 min. Occultations longer than 3 min, at beta angles larger than 53∘, measure across large spatial
distances and represent approximately 12 % of the ACE-FTS data used in this
work. Both horizontal and vertical variations within the CMAM30HR output will
impact the comparison to ACE-FTS measurements. The CMAM30HR fields are output
on a grid with a spatial resolution of approximately 400 km. While most
ACE-FTS occultations have a shorter horizontal extent than the CMAM30HR grid
point footprints, there are some occultations that fall outside of a single
grid point range in the upper troposphere and lower stratosphere. For
example, between 5 and 30 km, 15 % of occultations extend across more
than one CMAM30HR grid point footprint, of which 82 % are at beta angles
greater than 53∘ and 80 % are at latitudes poleward of 30∘.
Various model sampling techniques have been investigated since occultations
span multiple grid point footprints in both latitude and longitude, as well
as vertically.
The comparisons of the three sampling methods described in the text
using N2O simulations in CMAM30HR. The relative differences (%) defined
as the difference between the advanced sampling and (a) the full
model output, (b) basic sampling, and (c) intermediate
sampling, relative to the advanced sampling. Note the different colour scales
in each panel.
Comparison of sampling techniques
To determine the impact of sampling the model output at varying levels of
detail, three methods were tested by sampling the full model output (the
CMAM30HR output at all latitudes and longitudes for each 5∘ latitude
bin) between June 2004 and May 2010. All three methods began with identifying
the temporally coincident three-dimensional CMAM30HR output (latitude,
longitude, pressure) for each ACE-FTS profile; the output within 3 h of
the occultation was selected with no temporal interpolation. The
three-dimensional output was interpolated in the vertical dimension to the
ACE-FTS profile pressure grid, which is different for each occultation since
the retrievals are provided on an altitude scale. The “basic” sampling method
involved selecting a vertical column based on the nearest neighbour grid
point to the 30 km tangent point location with no interpolation and no
consideration of the vertical extent of the profile. The “intermediate” level
of sampling extracted the vertical column based on a bilinear interpolation
of the four closest grid points to the 30 km geometric tangent point but with
no consideration of the variation in geographical location of the tangent
points above or below 30 km. The “advanced” sampling method improves on the
intermediate level of sampling by performing the bilinear interpolation at
each level of the ACE-FTS profile using the distinct geographic locations,
derived from the refraction model , for the
respective level. Therefore, at each vertical level in the ACE-FTS profile, a
spatial bilinear interpolation including the four geographically closest grid
points was computed to determine the comparable CMAM30HR VMR. To illustrate
the sampling effect between 450 and 10 hPa (5–30 km), the differences,
relative to the advanced technique, in the zonal mean of N2O over the
observation period (June 2004–May 2010) are compared in Fig. . The advanced method is compared to the full output of the
model (Fig. a), the basic model sampling
(Fig. b),
and the intermediate sampling (Fig. c).
The comparison of the advanced sampling and full model output in the
stratosphere is dominated by the influence of the polar vortex in both
hemispheres. Generally, there is good agreement throughout the troposphere,
with less than 5 % differences. For the long-lived tracers investigated
in this work, the free troposphere is well mixed such that there is minimal
influence of the ACE sampling pattern. In the stratosphere, however, there
are pronounced differences on the order of 20 %. Air in the polar vortex
is typically composed of older air brought down from higher altitudes.
Therefore, tracer concentrations within the vortex and vortex edge tend to be
significantly different from those in the midlatitude surf zone during the
winter in each hemisphere. The differences seen in Fig. a occur
because comparing the full output to measurement-like samples of the output
accounts for neither the variability of the vortex edge in both longitude and
latitude nor the differences in spatially and temporally sampled large-scale
downwelling of air within the vortex compared to a zonal mean average that
includes the model simulation at all longitudes and time periods. At the edge
of the vortex, tracer concentration gradients are strong, so comparing
measurements to the full output of the model will tend to smear the influence
of the vortex on tracer concentrations. The differences are not symmetric
latitudinally due to different dynamical conditions in each hemisphere. For
example, in the Antarctic stratosphere in September there is a strong
decrease in the geographic extent of the polar vortex with height such that
the vortex is much wider geographically at 100 hPa
than at 10 hPa. A similar phenomenon occurs in the Arctic but it is much
more variable both spatially and vertically.
The advanced sampling technique is compared to the basic sampling in Fig. b. The distinction between Fig. a and
b is that rather than using the full model output, the nearest
grid point is selected based on the geographic location of the 30 km tangent
point of the ACE-FTS measurements. Even this basic level of sampling improves
the comparison in the stratosphere substantially, bringing the range of
differences down to ±10 %. It is worth noting that the stratospheric
differences in the midlatitudes are on the same order of magnitude with a
similar latitudinal pattern but of opposite sign. The differences of 5–10 %
are primarily negative in the stratosphere in the Southern Hemisphere and positive
in the Northern Hemisphere. This pattern occurs because each ACE-FTS profile
is tilted such that the top of the profile is always further north than the
bottom of the profile, leading to a directional bias. In the Northern
Hemisphere, profiles tend to “point” toward the North Pole, and therefore
measurements in this hemisphere are subject to a poleward bias. In the
Southern Hemisphere, profiles point toward the Equator, leading to an
equatorward bias in sampling. Therefore, the choice of “closest” grid point
likely biases the comparisons, leading to the differences in Fig. b.
Figure illustrates the average latitudinal extent of
occultations between the 5 and 30 km tangent
altitudes (where, for an individual occultation, the latitude at 5 km is subtracted from the latitude at 30 km), showing the two
directionalities for each 5∘ latitude bin included in
Fig. b, with error bars indicating 1 standard deviation from
the mean latitudinal extent. A poleward bias implies that the 30 km tangent
point is located poleward of the 5 km tangent point and an equatorward bias
reflects when the 30 km tangent point is located equatorward of the 5 km
tangent point.
In the midlatitude region of the Northern Hemisphere, the average latitudinal
extent of occultations exhibits a primarily poleward bias while the
occultations in the southern hemispheric midlatitudes exhibit a primarily
equatorward bias. The northern hemispheric poleward bias in Fig. corresponds to the positive differences in the
northern hemispheric midlatitude stratosphere in Fig. b and the southern
hemispheric equatorward bias corresponds to the negative differences in the
southern hemispheric midlatitude stratosphere. The comparison to the advanced
technique in Fig. b reflects the combined influence of the
horizontal interpolation and, to a lesser degree, the geographical extent of
the ACE-FTS profiles. In the Northern Hemisphere, contributions from sampling
the model at lower latitudes (which tend to have a higher concentration) lead
to positive differences between the two sampling techniques; in contrast, in the
Southern Hemisphere, contributions from sampling the model at higher
latitudes (lower concentrations) lead to negative differences between the
advanced and basic sampling techniques. This nuance in the sampling pattern
highlights the importance of considering the sampling pattern of the ACE-FTS
occultations when comparing measurements to model output.
Average latitude ranges covered by ACE-FTS occultations in a given
5∘ latitude bin, separated by an equatorward bias (black) and a
poleward bias (red) as defined in the text. The error bars indicate 1
standard deviation from the mean latitudinal extent for the given 5∘
latitude bin.
Zonally averaged annual-mean latitude-altitude distributions of
N2O: (a) CMAM30HR (ppbv), (b) ACE-FTS (ppbv),
and (c) the mean relative difference (in %) between sampled model and
ACE-FTS profiles, divided by the ACE-FTS profiles
(100 × (CMAM30HR - ACE-FTS) / ACE-FTS). The black line
indicates the location of the thermally defined tropopause.
Figure c compares the advanced and the intermediate sampling.
With approximately ±2 % differences between the two techniques, it is
clear that the intermediate sampling technique can account for much of the
geographic extent of the ACE-FTS profiles at this model resolution. However,
if comparing to a model with a finer resolution or if a larger vertical
extent is considered, accounting for the full geographic extent of the
profile will become more important. The more detail that is included in the
sampling of the model, the more comparable the output is to the observations.
The advanced method of sampling provides the most appropriate model profiles
for direct comparison between ACE-FTS and CMAM30HR. Therefore, all comparison
results shown in this study utilize CMAM30HR output that has been sampled
using the advanced technique.
Zonally averaged tracer morphologiesGeneral features of tracer morphology comparisons
Same as Fig. c for (a) N2O,
(b) CFC-12, and (c) CFC-11. The grey regions indicate where
no data are available.
The zonally averaged annual-mean distribution of N2O is presented in Fig. .
The N2O simulated by CMAM30HR is shown in Fig. a, the ACE-FTS measurements are shown in Fig. b,
and the average of the profile differences within each 5∘ latitude bin
is shown in Fig. c. Both the ACE-FTS measurements and the
CMAM30HR distribution of N2O in Fig. show many of the
features that are expected of a long-lived tracer with a tropospheric source
and chemical losses that occur primarily in the stratosphere. The
distributions show a decrease in concentration of N2O with altitude at all
latitudes and also moving from the Equator poleward at each pressure level
and in each hemisphere. There is a hemispheric asymmetry in the decrease with
altitude beyond the tropical region. The southern extratropical and Antarctic
concentrations of N2O tend to decrease with altitude more rapidly than
those in the Northern Hemisphere. This asymmetry is likely driven by
differences in the isolation of the polar vortex in each hemisphere and the
large-scale downwelling that is largely dependent on this isolation. By
visual comparison, the lowest concentrations observed and simulated appear to
be in the Antarctic region between 30 hPa and 10 hPa and the Arctic
stratosphere above 20 hPa. The quantitative comparison between the ACE-FTS
and CMAM30HR zonal mean N2O distributions in the bottom panel of Fig. reveals significant differences throughout the lower
stratosphere, with the largest differences in the northern polar region.
CMAM30HR simulates larger concentrations of N2O in the lower stratosphere.
Upwelling in the tropics, descent in the extratropics, and mixing in the surf
zone define the transport controls on the distributions in the stratosphere.
The differences observed in Fig. c are influenced by the
combined effects of these features on the measured and the simulated
concentrations of N2O. Therefore, if there were no issues in the simulated
stratospheric transport, the differences would be of similar magnitude to the
upper troposphere comparisons (less than ±5 %), unless there was a
significant flaw in the chemical losses in the model.
Investigating measurement–model comparisons using more than one trace gas
leverages the varying lifetimes of and chemical processes of each gas. The
comparisons of CMAM30HR and ACE-FTS, equivalent to the bottom panel of Fig. , are shown in Fig. a–c for N2O, CFC-12, and
CFC-11, respectively. Each of the panels shows the differences as a
percentage. All three species show good measurement–model agreement (within
approximately 5 %) in the well-mixed troposphere. In the tropics, the VMRs of
these three species remain relatively constant up into the lower stratosphere
where chemical loss processes begin to break down the compounds. Above 70 hPa
in the tropics, the CFC-12 and N2O comparisons show similar agreement (on
the order of 5 %). However, above 50 hPa in the tropics, CFC-11 exhibits
both positive and negative differences between the measurements and model
simulations. These differences in CFC-11 are also observed outside the
tropics above 70 hPa and are much higher (on the order of 50 %). In the
northern hemispheric extratropics, the differences are primarily positive but
become more variable closer to the northern polar region. Very small
concentrations above 70 hPa, which occur because of the significant
photolytic losses in the tropical lower stratosphere, lead to the large
magnitude of the differences in CFC-11. The irregular pattern in the CFC-11
differences is driven by the variability in the measurements as ACE-FTS
reaches its detection limit.
The relative mean of individual ACE-FTS profiles subtracted from
CMAM30HR profiles of N2O, divided by the ACE-FTS measurements for each
season: (a) December–January–February (DJF),
(b) March–April–May (MAM),
(c) June–July–August (JJA), and
(d) September–October–November (SON). The grey regions
indicate where no data are available.
Seasonality of the tracer morphology comparisons
The structure and intensity of the BDC varies seasonally. In general, the BDC
is strongest in the northern hemispheric winter because of wave-driven
enhancements initiated by topography and because, during this time of
year, climatological westerlies facilitate wave propagation into the
stratosphere e.g.. It is well known
that tropical upwelling is stronger in the summer hemisphere; therefore,
during the December–January–February (DJF) season the upwelling is strongest
in the Southern Hemisphere e.g.. Investigating the
comparisons between the CMAM30HR simulations and the ACE-FTS observations in
a seasonal context helps to determine whether the differences observed
earlier are related to the behaviour of the BDC. If the differences observed
in Fig. are driven by the simulation of the BDC in the model,
it would be expected that the morphology of the seasonal differences would
appear to follow the behaviour of the BDC.
For each season, Fig. identifies the differences between the
simulation and the measurements. The seasonal composites shown here do not
fully represent the seasons because of the sampling pattern of the ACE-FTS
(recall Fig. ). However, the comparisons are relevant since
the CMAM30HR output has been subsampled, as previously described. The most
obvious features across all seasons in Fig. are the same as
those of the differences shown in Fig. a. There is good
agreement in the lower stratosphere at all latitudes and in the tropics up to
about 50 hPa. In the mid-stratosphere, CMAM30HR simulates higher
concentrations of N2O than those measured by ACE-FTS. Some of the largest
differences occur at the high northern latitudes during boreal winter and
spring, presumably in the region of downwelling within the polar vortex.
The large disagreements in the north polar region during winter and spring
indicate that the downwelling portion of the BDC across the different seasons
is not well characterized by CMAM30HR. Meanwhile, the shifting of the
agreement in the tropical region through the seasons indicates that the
simulation is consistent with the spatial distribution of the observations in
this region. For example, the difference in the southern tropical latitudes
appears small (close to 0 %) up to 50 hPa and to approximately 40∘ S
in DJF, but the agreement diminishes in this region in the March–April–May
(MAM) season, presumably when the tropical upwelling begins to decline in
strength and shift toward the Equator. A similar pattern is observed in the
Northern Hemisphere during austral winter where the differences in the
northern tropical latitudes appear to be small up to 50 hPa and to
approximately 40∘ N in the June–July–August (JJA) and
September–October–November (SON) seasons. These results support the
understanding that the most rapid tropical upwelling occurs in the summer
hemisphere as first reported by .
In the comparisons shown in Fig. , the northern high latitudes
measurement–model differences are significantly different compared to those
in the southern high latitudes. There are negative differences at high
southern latitudes in MAM, JJA, and part of SON. The differences seem quite
asymmetric when compared with the results for the Northern Hemisphere. The
negative differences at high southern latitudes appear to descend between MAM
and JJA and begin to weaken in SON with the vortex breakdown. There is also some asymmetry in the
differences between 30 and 10 hPa between the Northern and Southern
hemispheres, particularly in winter for each hemisphere. These differences
are likely due to the behaviour of the polar vortex in each hemisphere. In
particular, the differences may be related to the model's (either CMAM30HR or
the ERA-Interim reanalysis used for the nudging or both) ability to represent
transport processes in the strong, cold, quiescent Antarctic vortex versus
the warmer and more variable Arctic vortex.
Since the model run compared here has been nudged to the ERA-Interim
meteorology, it cannot be simply concluded that the differences are due to
the variable nature of the vortex. The vertical migration of the negative
differences in the southern polar region across MAM, JJA, and SON suggests
the vortex variability physical or chemical mechanism as the cause. The
negative differences mean that there is an underprediction in the CMAM30HR
simulation, which could happen if air that is too old is brought down into
the vortex. As the vortex forms in fall, the negative differences appear and
descend through the winter, reaching a maximum latitudinal extent. The
appearance of the negative differences in the comparison of N2O between
the observations and the model is conspicuous because elsewhere N2O is
higher in CMAM30HR throughout the lower and middle stratosphere.
In Fig. b and d, the large positive
differences in the northern hemispheric stratosphere may be caused by too much
orographic wave driving in CMAM30HR (and CMAM30). This would lead to air moving from the tropical region into the extratropics and polar
regions too quickly, thereby simulating higher than expected
concentrations of N2O in the northern hemispheric stratosphere.
Comparison of mixing barriers
It is well understood that quasi-horizontal mixing flattens tracer isopleths
in mixing regions and sharpens gradients at mixing barriers
e.g.. However, it can be very difficult to separate
the effects of mixing barriers from the residual circulation when looking at
zonal mean comparisons between measurements and models. Therefore, it is
necessary to scrutinize mixing barriers individually.
A monthly time series of the average relative differences between
the model and the measurements (CMAM30HR minus ACE-FTS, divided by ACE-FTS),
as in Fig. c for (a) N2O and (b) CFC-12 in
the Northern Hemisphere (60–90∘ N) between June 2004 and May
2010. The grey regions indicate where no data are available.
The polar vortex
Consideration of the behaviour of the polar vortex in both hemispheres is
necessary as they have atmospheric processes that affect their behaviour
differently over time. For this purpose, the monthly mean differences between
ACE-FTS and CMAM30HR over the time period of the study have been determined
for the stratospheric abundances of N2O and CFC-12. CFC-11 has been
excluded from this comparison because the limited vertical extent of the
sensitivity of the measurement results in too few data in the stratosphere.
All comparisons shown here are profiles located poleward of 60∘ and
show the mean of the difference between the ACE-FTS and CMAM30HR profiles,
relative to the ACE-FTS profile. The comparisons here extend the work of
, who discussed the polar region simulations in
CMAM30 extensively by comparing temperature, ozone, methane, and water vapour
up to 0.001 hPa with a variety of satellite instruments including ACE-FTS.
All the species investigated in have much shorter
lifetimes than those of N2O and CFC-12. The advantage of using species
with long lifetimes is that at least some of the parcels of air that are
sampled have been through the deep branch of the BDC. By restricting
comparisons to the polar stratosphere, it is primarily air from the deep
branch of the BDC that is being investigated. Tracers with a stratospheric
sink are mostly depleted from the deep branch because they have had the most
time for chemical loss to occur since they entered the stratosphere.
The comparison of the N2O and CFC-12 difference time series (Fig. ) demonstrates that there is interannual variability that is
consistent between the two gases in the Arctic. While the two species shown
follow similar patterns over time, there appear to be larger differences in
the CFC-12 comparison than in the N2O comparison. There are two possible
(and related) reasons for this difference: the range in the concentrations of
CFC-12 is much larger than that of N2O, and there are differences in their
respective chemical losses. For example, the photolysis loss of CFC-12 is
faster than that of N2O throughout much of the stratosphere, as evidenced
by the differences in their lifetimes (102 years for CFC-12 and 123 years for
N2O). Generally, it appears that the model simulates higher concentrations
of both species compared to ACE-FTS measurements through much of the
stratosphere, with the largest differences occurring above 30 hPa. When the
concentration of either tracer becomes very small (typically air that has
descended from the upper stratosphere or mesosphere), the relative
differences between ACE-FTS and CMAM30HR can be enhanced. These differences
are most clear in the N2O comparisons during the autumns of 2004 and 2009
and the springs of 2007 and 2010; in contrast, in the CFC-12 comparisons, the
springs of 2005, 2006, and 2008 exhibit additional occurrences of this
feature.
A monthly time series of the average relative differences between
measurements and the model (CMAM30HR minus ACE-FTS, divided by ACE-FTS), as
in Fig. c for (a) N2O and (b) CFC-12 in
the Southern Hemisphere (60–90∘ S) between June 2004 and May
2010. The grey regions indicate where no data are available.
During each of these periods, ACE-FTS observed much lower concentrations
compared to CMAM30HR. Both the speed and structure of the residual
circulation within the CMAM30HR run can contribute to the observed
differences. It is possible that the BDC in the CMAM30HR simulation is
drawing air through the deep branch of the BDC too rapidly. The vertical
structure of the differences observed in Fig. , particularly
between 70 and 30 hPa, may be caused by ACE-FTS measuring a descent in
the air mass that CMAM30HR does not simulate. It is more likely that the
model circulation is not moving enough air through the loss region of these
tracers and through to the polar vortex. For photolytic tracers, the
structure of the circulation is more important than the speed of the residual
circulation because photolysis rates are so fast. Above a certain level in
the stratosphere, the tracer is completely destroyed when air passes through
the region. The distribution of photolytic species is a mixture of air that
passed through the region of rapid loss and the air that by-passed the loss
processes. This result is consistent with the work of
, where they found large differences in temperature
and ozone between satellite observations and CMAM30.
There is less interannual variability in the Southern Hemisphere comparisons
(Fig. ) than in the Northern Hemisphere (Fig. ). This is expected since the variability of the southern polar
stratospheric dynamics is much less than that in the northern polar
stratosphere. However, the magnitude of the differences between the
measurements and simulations is larger in the Antarctic stratosphere than in
the Arctic stratosphere. Moreover, while the patterns of the differences in
N2O and CFC-12 are quite similar in Fig. a and
b, the magnitude is more pronounced in the CFC-12 comparisons.
The largest differences occur above 30 hPa where the concentrations of CFC-12
are extremely low. The peak in the magnitude of the CFC-12 differences
appears to increase in vertical extent through the austral springtime. The
largest differences tend to occur during summer (December) at around 40 hPa
for both tracers. The differences in CFC-12 are established at the top of the
vortex in July and propagate down until vortex breakup in December. However,
this propagation does not occur to the same extent in the N2O comparisons,
which may be a reflection of the differences in the chemistry of the two
tracers. For example, a source of N2O in the lower thermosphere has
recently been identified in ACE-FTS measurements by .
The N2O source descends into the mesosphere and stratosphere, thereby
influencing air that is circulated in the BDC . The
transport of enhanced N2O downwards from the upper atmosphere has also
been detected by . The CMAM30HR does not
include this source of N2O. The results presented here are consistent with
the methane comparisons discussed in . Generally,
the descent of the model's high bias is observed in both hemispheres for all
three trace gas species. The results of indicate
that the high bias is consistent with a fast BDC and that the downward
propagation of the bias is a problem with the parameterizations in the model
above 10 hPa.
The extratropical tropopause
The transport barrier at the extratropical tropopause can be permeable,
allowing the exchange of air between the troposphere and the stratosphere.
Understanding how CMAM30HR simulates this exchange assists in the
interpretation of mixing effectiveness in the model and the impact of its
finite resolution on the comparisons. Since ACE-FTS predominantly samples the
polar regions and has fewer samples of extratropical latitudes, interpreting
latitudinal or seasonal dependence of the exchange of air across the
tropopause in the full atmosphere using these measurements must be considered
from a tropopause coordinate perspective . In this study,
a diagnostic of the tropopause barrier has been developed for comparison
between the simulations and the measurements. The tropopause height, used in
this analysis to define tropospheric air and stratospheric air, is the
thermally defined tropopause based on the derived meteorological products for
ACE-FTS and based on sampled temperature profiles from
CMAM30 output for the CMAM30HR simulations.
The frequency of intrusion across the tropopause for both CMAM30HR
(black circles) and ACE-FTS (red squares) below the 420 K potential
temperature isotherm in (a) the southern hemispheric and
(b) the northern hemispheric extratropical region between
20 and 60∘ N/S, respectively. For each 5∘ latitude bin,
the frequency of intrusion events are separated by stratospheric intrusions
(solid lines), where stratospheric-like air is found in the troposphere, and
tropospheric intrusions (dashed lines), where tropospheric-like air is found
in the stratosphere.
Since CFC-11 has a strong vertical gradient in concentration in the
stratosphere, it can be used as a proxy for determining the exchange of air
across the tropopause. The diagnostic developed for this analysis is the
frequency of intrusions, which signifies the frequency of stratospheric
(tropospheric) air penetrating into the troposphere (stratosphere). A data
point is defined as an intrusion based on two criteria: the physical location
of data point and the concentration relative to a tropospheric concentration
threshold and only data below the 420 K potential temperature layer of
the atmosphere have been considered. The tropospheric threshold was defined
separately for each hemisphere as the tropospheric mean minus 1.5 times the
tropospheric standard deviation. An intrusion frequency metric has been
developed for comparison between the simulated and observed concentrations. A
tropospheric intrusion is identified when a measurement is physically located
in the stratosphere but its concentration is larger than the tropospheric
threshold. A stratospheric intrusion is identified when a measurement is
physically located in the troposphere but its concentration is smaller than
the tropospheric threshold. For each 5∘ latitude bin between
20 and 60∘ latitude (N and S), the frequency of tropospheric
and stratospheric intrusions have been determined. This technique was used to
calculate frequencies for both the ACE-FTS measurements and CMAM30HR
profiles.
The comparison of ACE-FTS and CMAM30HR tropospheric and stratospheric
intrusions within the southern and northern extratropical latitudes is shown
in Fig. a and b, respectively. There appears to be
better agreement for the stratospheric intrusion comparisons and at some
latitudes there is very good agreement between ACE-FTS (red) and CMAM30HR
(black). However, there are some latitudes, such as 30–40∘ S
and 20–25∘ N, where the stratospheric intrusions identified
in the simulation are a factor of 2 fewer than those from the ACE-FTS
measurements. In general, isentropes that are in the extratropical lowermost
stratosphere are in the troposphere in the tropics
e.g.. Fast isentropic transport occurs because wave
motions cause air to rapidly change latitude, leading to the transport of
stratospheric air into the troposphere. Since the disagreement in measured
stratospheric exchanges is largest at latitudes where this isentropic
transport tends to occur implies that CMAM30HR is not capturing this
mechanism well or simply lacks the resolution to fully resolve the
stratospheric intrusions features.
Based on Fig. , tropospheric intrusions occur more frequently in
the ACE-FTS data than stratospheric intrusions and vary more significantly in
number across the latitudes in both hemispheres. Additionally, the
differences between the measurements and simulations are larger than for the
stratospheric intrusions. It is possible that there is more tropospheric air
found in the stratosphere with this method than stratospheric air found in
the troposphere because tropospheric CFC-11 is more easily distinguishable.
The manner in which the intrusions have been defined here does not rule out
stratospheric air being identified as tropospheric if there is rapid,
poleward transport out of the lower tropical stratosphere.
suggest that it is easier to distinguish tropospheric
intrusions into the stratosphere using tropospheric tracers and
stratospheric intrusions into the troposphere using stratospheric tracers.
Tropospheric CFC-11 concentrations found in the stratosphere at extratropical
latitudes indicate air that likely has not cycled through the BDC because the
air parcels have not experienced any loss processes that can only happen in
the stratosphere, while stratospheric air in the troposphere could be
representative of a range of aged air. For example, this air could have
cycled through the BDC very quickly by re-entering the troposphere in the
subtropics or it could have gone through the deep branch, allowing it to be
more depleted and therefore more identifiable as stratospheric-like air.
A seasonal representation of the intrusions depicted in
Fig. : (a) tropospheric intrusions in the Southern
Hemisphere, (b) tropospheric intrusions in the Northern Hemisphere,
(c) stratospheric intrusions in the Southern Hemisphere, and
(d) stratospheric intrusions in the Northern Hemisphere. Each season
is an average of the extratropical intrusion frequency with error bars
indicating 1 standard deviation of the seasonal mean. ACE-FTS intrusions
are in red squares and CMAM30HR intrusions are in black circles. Note the
seasonality represented may be impacted by the sampling of ACE-FTS and may
not representative for the full atmosphere.
A mechanism for extratropical tropospheric air to be uplifted into the
stratosphere has been identified recently. Building on the work of
, showed that the occurrence of double
tropopauses is associated with the strength of the tropopause inversion layer
(TIL), as well as Rossby wave breaking. They also showed that as the strength
of the TIL increases, cyclonic circulation in the upper troposphere switches
to anticyclonic circulation, thereby driving an increase in the upward
motion. Based on the double tropopause calculations of ,
it is likely that ACE-FTS observes this phenomenon of upward motion, leading
to a higher frequency of tropospheric intrusions across the extratropical
regions. CMAM30HR appears to inadequately simulate this mechanism since it
does not have a sharply defined tropopause compared to reality
. It is possible that both the spatial and vertical
resolutions of the simulations performed limit the model's ability to capture
this synoptic-scale activity.
JPDFs, as described in the text, of N2O–CFC-12 for
(a) ACE-FTS and (b) CMAM30HR, and N2O–CFC-11 for
(c) ACE-FTS and (d) CMAM30HR. All stratospheric ACE-FTS
observations and subsampled model output in the Northern Hemisphere are
included.
Seasonal averages of the tropospheric and stratospheric intrusions are shown
in Fig. , where Fig. a and b are the tropospheric
intrusion frequencies for the southern and northern extratropics,
respectively, and Fig. c and d are the stratospheric intrusion
frequencies for each hemisphere. For this case, the data shown in
Fig. were averaged seasonally in each hemisphere and the error
bars represent the latitudinal variability defined as 1 standard deviation of
the mean. Recall that since the intrusion frequencies in Fig.
are affected by the sampling pattern of ACE-FTS, the seasonality shown may
not be representative of the actual seasonality of the atmosphere. The
stratospheric intrusion comparisons are remarkably good; there does not
appear to be a significant difference between the simulations and the
measurements. Therefore, the differences are primarily due to the finite
horizontal and vertical resolution of the model, which leads to differences
in the representation of stratosphere–troposphere intrusion events. The
comparisons of tropospheric intrusions exhibit similar behaviour between the
two hemispheres. The measurement–model differences appear to be largest
during JJA and smallest during MAM.
The consistency of the increase during SON in tropospheric intrusion
frequency between the two hemispheres may be the result of convective
overspill into the stratosphere across the subtropical tropopause barrier if
the majority of the tropospheric intrusions are driven by convection in the
tropics. This convective influence could extend across the extratropical
latitudes via isentropic transport from the tropics. The strength of the
convection during JJA and SON and its influence on mixing across the
tropopause barrier is observed by the largest variability in the
stratospheric intrusions.
The tropical pipe
In this section, JPDFs are used to investigate the tropical pipe barrier.
Figure illustrates the N2O–CFC JPDFs of the entire
northern hemispheric stratosphere, beginning at 2 km above the tropopause to
30 km (up to 10 hPa), for ACE-FTS (Fig. a, c) and CMAM30HR (Fig. b, d) for CFC-12
(Fig. a, b) and CFC-11 (Fig. c, d). have highlighted the use of this
technique for comparing CMAM and ACE-FTS measurements. The
N2O–CFC-12 JPDFs (Fig. a and b) exhibit a
quasi-linear relationship in both the measurements and the simulations. The
loss rates of CFC-12 and N2O are very similar in the upper troposphere and
lower stratosphere, leading to a linear relationship in the JPDFs. The model
JPDFs tend to peak at the higher concentrations of N2O and CFC-12 but are
more evenly distributed throughout the range of concentrations. These
differences between the measurements and simulations are likely due to the
overly rapid BDC in the model simulation, leading to higher concentrations of
the trace gases in the simulation, which is consistent with the zonal mean
comparisons discussed in Sect. 4. The ACE-FTS data in Fig. a
are much more scattered than the CMAM30HR in Fig. b, where the
simulated tracers are highly correlated and the JPDF is highly compact. The
differences in the spread of the correlations are related to differences in
mixing and chemistry but are also influenced by the precision of the ACE-FTS
measurements and the constraint of the boundary conditions in the
simulations. The CMAM30HR JPDF is very compact because of the similarity of
the chemical losses of the two tracers in the model. Additionally, the
surface boundary conditions applied do not represent the variability observed
in the atmosphere. It is the atmospheric variability that contributes to the
variability observed in the ACE-FTS JPDF around 150–200 ppbv of
N2O .
Tropical JPDFs of N2O–CFC-11, as described in the text,
separated by season. Only stratospheric ACE-FTS observations (left column)
and CMAM30HR simulations (right column) within the height-dependent tropical
turnaround latitudes have been included.
The N2O–CFC-11 JPDFs show two segments of linear correlations in
both the measurements and model results (shown in Fig. c and
d, respectively) that would have otherwise been overlooked in
data-dense tracer–tracer correlation plots (also see
). The presence of a bimodal correlation between
N2O and CFC-11 has been previously observed by . The
separation is caused by differences in local chemistry in the tropical pipe
region. The tropics are somewhat isolated from the midlatitudes so that the
steeper slope is a signature of the local chemistry, or the relative loss
rates of CFC-11 and N2O in the tropical lower stratosphere. This
relationship is observed because the photochemical lifetime of CFC-11 is
shorter than the timescale for mixing by horizontal eddy transport timescale
. According to , slope equilibrium
conditions that define the linear relationship seen in the
N2O–CFC-12 JPDF are only satisfied if the photochemical lifetime of
a species is much greater than the horizontal eddy transport timescale. When
this condition is satisfied, the slopes of the isopleths are only a function
of atmospheric circulation.
Isolating the measurements and simulations in the tropical region allows for
the characteristics of the tropical pipe to be investigated. A JPDF
comparison is provided in Fig. for both the ACE-FTS
measurements and the CMAM30HR simulations in the tropics during the DJF, MAM,
JJA, and SON seasons. The data were selected from the tropical latitude
region using estimates of the turnaround latitude, the height-dependent
latitude where the tropical upwelling is zero, determined from CMAM30HR
monthly mean vertical velocities. Bimodal behaviour is observed in each
season and in both the measured and simulated JPDFs. In general, the maximum
of the JPDF appears to be positioned towards higher concentrations in the
simulation compared to the measurements, where the maximum in the probability
tends to extend throughout the shallower segment. As was observed in the
N2O–CFC-12 JPDFs in Fig. , the probabilities tend to be
weighted towards the higher concentrations, implying that there is younger air
(of tropospheric origin) in the simulated stratosphere than the measured
stratosphere. Of note is that the length and the width of the shallower
segment are consistently larger in the measurements. The longer length –
extending to low concentrations – indicates the presence of older air not
found in the model simulations. The larger width also coincides with the
degree of separation between the primary and secondary segment in each
season, where the simulated JPDFs appear to have a much greater separation
than the measurements. These features are likely dependent on the amount of
quasi-horizontal mixing influencing the JPDFs, implying that there is not
enough quasi-horizontal mixing occurring in the simulation. However, the
differences between the measurements and simulations are primarily in the
steepness of this segment of the JPDF, which is an indication of insufficient
mixing into the tropics rather than too rapid tropical ascent. If it was
just too fast ascent, both tracers would be affected in the same way. But
because the midlatitude N2O–CFC-11 relationship is less steep, it
is the mixing in of this air that makes the tropics have shallower slope.
Then it becomes a question of whether mixing is underestimated because the
ascent is too fast or whether the model does not simulate the structure of the
pathways of the BDC correctly.
There is an evolution of the characteristics of the JPDFs shown in Fig. . During DJF, there is minimal separation between the two
segments. The position of the maximum in the JPDF (the red region) in Fig. a does not extend below 200 ppbv of N2O, while the location
of the maximum of the simulated JPDF in Fig. b is limited to
above 250 ppbv of N2O with the maximum primarily located above the
separation of the two segments. During MAM, the maximum probability of the
ACE-FTS JPDF in Fig. c extends throughout the shallow and steep
segments. This implies that there is significant mixing occurring during this
time period. In the simulated MAM, Fig. d, the maximum of the
probability is restricted to the higher concentrations of N2O, prior to
the separation of the two segments. This indicates that while mixing is
occurring, it does not occur frequently enough to simulate the atmosphere
well.
During JJA, the two segments begin to separate in both the measured and
simulated JPDFs. The shallower segment in the simulated JPDF is the shortest
during JJA of all the seasons with the maximum of the probability residing at
the highest concentrations. This indicates that the upwelling during this
season is too strong and it may also be too isolated since there appears to
be more mixing in the measured JPDF. During SON, the separation between the
segments is most prominent, indicating that this time of year exhibits the
least mixing. Based on these comparisons, it is still difficult to discern
the relative contributions of the tropical upwelling and quasi-horizontal
mixing to the differences in the JPDFs of the model and measurements. To
interpret these differences, the results of the TLP model simulations
described previously are used. These simulations provide a basis for
determining how much the residual circulation needs to slow down within
CMAM30HR and what impact that may have on mixing of air between the tropical
and extratropical regions.
Using a TLP model to interpret CMAM30HR
An analysis of the effects of the strength of the BDC and the mixing in
CMAM30HR was achieved through a suite of simulations computed with the TLP to
test a range of mean circulation strengths and mixing efficiencies. The TLP
runs began with the CMAM30HR best fit to the TLP model and then the best
combination was selected to match the ACE-FTS measurement profiles of CFC-11
and CFC-12, as well as an age-of-air estimate derived from balloon-borne
measurements . There were 480 simulations initialized with
different combinations of w∗ (velocity of tropical upwelling) and
ϵ (the mixing efficiency) settings. The fraction of the CMAM30
w∗ used to initialize the TLP model ranged from 0.20 to 1.24 and the
ϵ ranged from 0.18 to 1.50. In each TLP simulation run, the
relationship between mean circulation and mixing is constrained by the
vertically averaged mixing efficiency . The mixing
efficiency in the TLP model is defined as ϵ=α/λτ, where α is the ratio of tropical to extratropical
mass, λ is the rate of the mean circulation influence, specifically
the mass flux out of the tropics due to mean circulation, and τ is the
mixing time or timescale for mass flux between the tropics and extratropics
. Therefore, the mixing efficiency is dependent
on both the mean circulation and the horizontal mixing mass flux
. The same suite of TLP model experiments
describe are also used here but the analysis has been
modified to investigate the behaviour of the tropical pipe in CMAM30HR in
further detail.
found that the CMAM30HR simulations best match the ACE-FTS
measurements when the w∗ is between 0.27 and 0.32 mm s-1 (a reduction from the fitting estimate of
0.4 mm s-1) and ϵ ranges from 0.7 to 1.2 (an increase from
the fitting estimate of 0.55). found that since ϵ
is inversely proportional to both λ and τ, there is a
compensating effect with changes in w∗ (λ) or τ. For the
CMAM30HR changes derived, w∗ needed to be slowed down significantly
below 20 km, and above 24 km. For constant ϵ that would result in
larger τ (less mixing). However, found that
ϵ also needed to be increased so there needed to be more mixing than
would result from slower w∗ and constant τ, but not enough of an
increase in ϵ so that the mixing times were less (more mixing) than
CMAM30HR has currently. concluded that with the increase
in ϵ, mixing times are reduced but still longer than the current
CMAM30HR mixing times.
Examples of the TLP comparisons described in the text. The grey
scale on each panel indicates the comparison between ACE-FTS and the TLP
simulations over various w∗ (y axes) and ϵ cases
(x axes). The white lines identify the agreement with ACE-FTS (grey colour
scale). The red marker indicates the CMAM30HR estimated values of w∗
and ϵ. The examples shown are (a) CFC-11 tropics during the
DJF season, (b) CFC-11 Northern Hemisphere during the MAM season,
(c) CFC-12 Northern Hemisphere during the JJA season, and
(d) CFC-11 Southern Hemisphere during the SON season.
In this study, the TLP simulations are used in a spatial and seasonal context
to determine the changes required in the CMAM30HR simulations to match the
ACE-FTS observations. The three regions investigated (the tropics and the
northern and southern extratropics) were defined by the turnaround latitude
of the tropical upwelling and by exclusion of the polar vortex in each
hemisphere using a 1.2 × 10-4 s-1 scaled potential vorticity
threshold e.g.. The vortex is excluded in these
comparisons because the TLP model does not simulate its complexity. For each
of the 480 simulations of the TLP model and the ACE-FTS measurements,
profiles for CFC-11 and CFC-12 were averaged over 16–27 km in the
extratropics and 16–29 km in the tropics. To capture the monthly
coverage of the ACE-FTS measurements, a weighting function for each region
and season was applied based on the relative contribution of the occultations
observed in each month of the particular season and region (the tropics and
northern/southern extratropics). The vertically averaged and monthly weighted
measurements were compared to the individual simulations of the TLP model
that represent varying levels of tropical upwelling and mixing efficiencies.
For each of the comparisons, the absolute value of the differences is used
and scaled to the maximum of the range of all differences so that each of the
scales range from zero to one.
Figure shows four examples of these comparisons plotted by the
seasonally averaged, regionally specific modelled values of mixing,
ϵ, and upwelling, w∗, for each simulation. The x-axis and
y-axis values do not represent the setting used to run the simulation –
they are the values simulated by the TLP model and lead to the curvature seen
in the plots. The shading of each plot indicates the level of agreement,
where the darker regions indicate the minimum differences between ACE-FTS and
the TLP simulations. The white-to-black shading is reinforced by white
contours of the same quantity to illustrate the comparison more clearly. The
relative CMAM30HR position is shown by the red marker and the error bars
represent the estimated range of uncertainty based on the optimization
exercise with the TLP model. These comparisons all indicate that the CMAM30HR
values of ϵ and w∗ would require some change to bring the
simulations closer to agreement with the ACE-FTS measurements (i.e. bringing
the red marker towards the dark shaded region). The dependence of w∗
on ϵ and vice versa is not the same across regions and seasons for
either CFC-11 or CFC-12. There are often limited ranges in which agreement
between the TLP simulations and ACE-FTS can be assessed but it is clear that
changes could lead to improvements in the agreement (such as in Fig. b). There are some scenarios where the mixing is required to
change much more than the tropical upwelling (see Fig. c for an
example) or where there can be a range of mixing efficiency and upwelling
values that lead to better agreement (as in Fig. d).
A summary of the TLP comparisons by season and region:
panel (a) indicates the changes required in tropical upwelling,
w∗, while panel (b) indicates the changes in the mixing,
ϵ, required for the CMAM30HR simulation to agree with the ACE-FTS
observations. The “-” and “+” symbols at the top of the figure
indicate the direction of mass transport (downwelling or upwelling,
respectively) for each season and region (green for Northern Hemisphere,
black for tropics, and red for the Southern Hemisphere). The changes required
based on CFC-11 are shown in cyan and the changes required based on CFC-12
are shown in blue.
The changes that would bring CMAM30HR into agreement with ACE-FTS based on
the TLP simulation comparisons for CFC-11 and CFC-12 have been calculated.
This was done by finding where the differences between ACE-FTS and the TLP
simulation are below a certain threshold (0.2 for all regions and seasons)
and determining the ranges of w∗ and ϵ over which this
occurs. The average values (cyan “x” markers for CFC-11 and blue square
markers for CFC-12) and ranges (error bars) of the changes to w∗ and
ϵ calculated are shown in Fig. . The changes in
w∗ were calculated based on the absolute value of w∗ so
that the interpretation of the differences calculated was not dependent on
the sign of w∗. This means that a positive change is always an
acceleration of the BDC, even when w∗ is typically negative, such as
in the extratropics. Figure a shows these changes for each
region and across the four seasons. The sign of w∗ values is
indicated above Fig. a, where the green symbols are for the
northern hemispheric extratropics, the black symbols are for the tropics, and
the red symbols are for the southern hemispheric extratropics. In some cases,
the agreement between ACE-FTS and the TLP simulations did not meet the 0.2
threshold so there is no result available for Fig. .
The changes required to w∗ in the tropical region are significant in
the CFC-11 comparisons in two seasons, DJF and JJA, indicating that the
w∗ strength should be reduced by 0.05 to 0.10 mm s-1 in DJF and by 0.03 to 0.12 mm s-1 in JJA. The lack of
a significant change during MAM and SON indicates that there is no argument
for changes to the tropical upwelling in CMAM30HR. The Southern Hemisphere
w∗ changes are not significantly different from zero during any
season, suggesting that changes in the w∗ in this region would not
improve the comparisons. However, there are significant changes identified in
three of the four seasons in the extratropical region of the Northern
Hemisphere. In DJF, an increase in w∗ is required in the Northern
Hemisphere by the CFC-12 comparisons, implying that strength of the
downwelling during this season should increase by up to 0.08 mm s-1.
However, the CFC-11 comparisons indicate that w∗ should decrease by
approximately 0.05 mm s-1. This contradiction necessitates the
conclusion that no change is recommended in w∗ in the northern
hemispheric extratropical region during DJF. During MAM in the Northern
Hemisphere, CFC-11 and CFC-12 comparisons suggest an increase in the
downwelling by approximately 0.05 mm s-1. By summer in the Northern
Hemisphere (JJA), the tropical pipe has shifted poleward such that the
northern hemispheric extratropical w∗ values are positive and require
a decrease in value as compared to the measurements. While this season tends
to have the most active tropical upwelling, these results suggest that the
upwelling in CMAM30HR may be too rapid or, since the southern extratropics
appear to have increased downwelling during JJA, may be displaced in
latitude.
The calculations of changes to the mixing parameter, ϵ, shown in
Fig. b, indicate that changes could be made in the tropics and
extratropics differently across the seasons. The results indicate, based on
CFC-11 and CFC-12 comparisons, that mixing efficiency in the tropics and in
the northern hemispheric extratropics needs to increase during JJA. Based on
the CFC-12 comparison, mixing efficiency in the southern hemispheric
extratropics should also be increased during JJA. Significant changes are
indicated from the CFC-11 comparisons for the remaining seasons in the
extratropical regions. In the Northern Hemisphere and Southern Hemisphere
during DJF, decreased mixing efficiency is suggested. During MAM, CFC-12
indicates there may be an increase suggested but the result is not supported
by CFC-11 comparisons. In the Southern Hemisphere, CFC-11 comparisons show
that a decrease in mixing efficiency is necessary. During SON, CFC-11
comparisons also indicate a decrease in mixing efficiency in the Northern
Hemisphere only. Generally, in all seasons except JJA, mixing efficiency in
the extratropics is suggested to decrease.
Combining the ϵ parameter results with the decrease in w∗
suggested in Fig. a, an increase in mixing timescales between
the tropics and northern hemispheric extratropics is required during JJA. It
follows that if w∗ needs to be reduced in the model then a reduction
in wave activity is required. The specific waves that break in the lower,
middle, and upper stratosphere that could be investigated for possible
sources of increased w∗. For mixing changes, the background state of
the winds and corresponding critical layers for wave breaking could be
investigated for critical layers that extend too far into the tropics.
Conclusions
In this work, ACE-FTS measurements of CFC-11, CFC-12, and N2O have been
used to assess the CMAM30HR simulations of these tracers and, thereby,
indirectly the transport processes in the lower stratosphere in the model. By
treating each tracer in the specified dynamics simulation explicitly, the
CMAM30HR run allows for the direct comparison of the measurements to model
output. The advanced sampling technique employed here allows for detailed
interpretation of the comparisons. Of the species investigated, it was found
that CMAM30HR consistently overpredicts tracer concentrations in the lower to
mid-stratosphere. The largest and most widespread overpredictions occur in
the northern hemispheric winter and spring, when the BDC is most active in
that hemisphere.
The investigation of simulated mixing barriers identified a number of issues
in the CMAM30HR simulations. The polar vortex comparisons reveal issues in
both the timing and strength of the downwelling portion of the deep branch,
which is related to the too-rapid BDC in CMAM30 simulations observed in the
zonal mean comparisons. The extratropical tropopause barrier in the model
appears to represent stratospheric intrusion events well, as evidenced by
Fig. c and d. However, tropospheric intrusions
are poorly simulated in most seasons (Fig. a and
b), with the largest discrepancies occurring during JJA in both
hemispheres. The tropical pipe mixing barrier analysis suggests that while
the strength of the simulated BDC (i.e. upwelling in the tropics and
downwelling in the extratropics) may partially explain the too young air
found in the mid-stratosphere, mixing efficiency may play at least as
prominent a role and seems to be underestimated, particularly in the JJA
season in all regions.
The analysis presented here highlights the importance of scrutinizing the
mixing efficiency in CCMs and GCMs since it may be related to the mechanisms
driving the projected trends in stratospheric circulation, thereby
influencing the simulations of stratospheric ozone recovery and climate
change. The techniques used in this work, including the advanced sampling and
use of the tropical leaky pipe model, have proven illuminating. It is
suggested that other CCMs and GCMs investigate the use of these techniques in
future studies.
The ACE-FTS Level 2 data used in this study can be obtained
via the ACE-FTS website (registration required,
http://www.ace.uwaterloo.ca). The HATS combined CFC data can be
acquired via NOAA's ftp site, ftp://ftp.cmdl.noaa.gov/hats/cfcs/. The
CMAM30 data set can be downloaded via Environment and Climate Change Canada's
climate modelling website,
http://climate-modelling.canada.ca/climatemodeldata/cmam/output/.
Additional fields specific to the CMAM30HR simulation are available available
from David Plummer (david.plummer@canada.ca). The TLP data are available from
Eric Ray (eric.ray@noaa.gov).
The authors declare that they have no conflict of
interest.
Acknowledgements
This project was funded by grants from the Canadian Space Agency (CSA) and
the Natural Sciences and Engineering Research Council of Canada (NSERC). The
Atmospheric Chemistry Experiment (ACE), also known as SCISAT, is a
Canadian-led mission mainly supported by the CSA and the NSERC. The
development of the CMAM30 data set was funded by the CSA. We thank Peter Bernath for his leadership of the ACE mission. We also thank Ted Shepherd,
Dylan Jones, and John Scinocca for their leadership and support of the CMAM30
Project. The HATS measurements were funded in part by the Atmospheric
Chemistry Project of the National Oceanographic and Atmospheric
Administration (NOAA) Climate and Global Change Program. The authors thank
Geoff Dutton, Debra Mondeel, and Steve Montzka for their work on measurements
of these important trace gases in the HATS program. The authors also thank
the anonymous reviewers, Matthew Toohey, and Chris McLinden for their
extensive and thoughtful feedback on this paper.
Edited by: Martin Dameris
Reviewed by: Matthew Toohey and two anonymous referees
References
Abalos, M., Legras, B., Ploeger, F., and Randel, W. J.: Evaluating the
advective Brewer-Dobson circulation in three reanalyses for the period
1979–2012, J. Geophys. Res., 120, 7534–7554, 2015.Abrams, M. C., Manney, G. L., Gunson, M. R., Abbas, M. M., Chang, A. Y.,
Goldman, A., Irion, F. W., Michelsen, H. A., Newchurch, M. J., Rinsland,
C. P., Salawitch, R. J., Stiller, G. P., and Zander, R.: Trace gas transport
in the Arctic Vortex inferred from ATMOS ATLAS-2 observations during April
1993, Geophys. Res. Lett., 23, 2345–2348, 10.1029/96GL00704,
1996a.Abrams, M. C., Manney, G. L., Gunson, M. R., Abbas, M. M., Chang, A. Y.,
Goldman, A., Irion, F. W., Michelsen, H. A., Newchurch, M. J., Rinsland,
C. P., Salawitch, R. J., Stiller, G. P., and Zander, R.: ATMOS/ATLAS3
observations of long-lived tracers and descent in the Antarctic Vortex in
November 1994, Geophys. Res. Lett., 23, 2341–2344, 10.1029/96GL00705,
1996b.Allen, D. R., Stanford, J. L., Nakamura, N., López Valverde, M. A.,
López-Puertas, M., Taylor, F. W., and Remedios, J. J.: Antarctic polar
descent and planetary wave activity observed in ISAMS CO from April to July
1992, Geophys. Res. Lett., 27, 665–668, 10.1029/1999GL010888, 2000.Allen, N. D. C., Bernath, P. F., Boone, C. D., Chipperfield, M. P., Fu, D.,
Manney, G. L., Oram, D. E., Toon, G. C., and Weisenstein, D. K.: Global
carbon tetrachloride distributions obtained from the Atmospheric Chemistry
Experiment (ACE), Atmos. Chem. Phys., 9, 7449–7459,
10.5194/acp-9-7449-2009, 2009.Austin, J., Shindell, D., Beagley, S. R., Brühl, C., Dameris, M.,
Manzini, E., Nagashima, T., Newman, P., Pawson, S., Pitari, G., Rozanov, E.,
Schnadt, C., and Shepherd, T. G.: Uncertainties and assessments of
chemistry-climate models of the stratosphere, Atmos. Chem. Phys., 3, 1–27,
10.5194/acp-3-1-2003, 2003.Beagley, S. R., de Grandpré, J., Koshyk, J. N., McFarlane, N. A., and
Shepherd, T. G.: Radiative-dynamical climatology of the first-generation
Canadian middle atmosphere model, Atmos. Ocean, 35, 293–331,
10.1080/07055900.1997.9649595, 1997.Bernath, P. F.: The Atmospheric Chemistry Experiment (ACE), J. Quant.
Spectrosc. Ra., 186, 3–16, 10.1016/j.jqsrt.2016.04.006,
2017.Bernath, P. F., McElroy, C. T., Abrams, M. C., Boone, C. D., Butler, M.,
Camy-Peyret, C., Carleer, M., Clerbaux, C., Coheur, P. F., Colin, R., DeCola,
P., DeMazière, M., Drummond, J. R., Dufour, D., Evans, W. F. J., Fast,
H., Fussen, D., Gilbert, K., Jennings, D. E., Llewellyn, E. J., Lowe, R. P.,
Mahieu, E., McConnell, J. C., McHugh, M., McLeod, S. D., Michaud, R.,
Midwinter, C., Nassar, R., Nichitiu, F., Nowlan, C., Rinsland, C. P., Rochon,
Y. J., Rowlands, N., Semeniuk, K., Simon, P., Skelton, R., Sloan, J. J.,
Soucy, M. A., Strong, K., Tremblay, P., Turnbull, D., Walker, K. A., Walkty,
I., Wardle, D. A., Wehrle, V., Zander, R., and Zou, J.: Atmospheric
Chemistry Experiment (ACE): Mission overview, Geophys. Res. Lett., 32, L15S01,
10.1029/2005GL022386, 2005.Birner, T. and Bönisch, H.: Residual circulation trajectories and transit
times into the extratropical lowermost stratosphere, Atmos. Chem. Phys., 11,
817–827, 10.5194/acp-11-817-2011, 2011.Boone, C. D., Nassar, R., Walker, K. A., Rochon, Y., McLeod, S. D., Rinsland,
C. P., and Bernath, P. F.: Retrievals for the atmospheric chemistry
experiment Fourier-transform spectrometer, Appl. Optics, 44, 7218–7231,
10.1364/AO.44.007218, 2005.
Boone, C. D., Walker, K. A., and Bernath, P. F.: Version 3 retrievals for
the
atmospheric chemistry experiment Fourier transform spectrometer (ACE-FTS),
The Atmospheric Chemistry Experiment ACE at 10: A Solar Occultation
Anthology, A. Deepak Publishing, Hampton, Va., USA, 2013.Brewer, A. W.: Evidence for a world circulation provided by the measurements
of helium and water vapour distribution in the stratosphere, Q. J. Roy.
Meteor. Soc., 75, 351–363, 10.1002/qj.49707532603, 1949.Brown, A. T., Chipperfield, M. P., Boone, C., Wilson, C., Walker, K. A., and
Bernath, P. F.: Trends in atmospheric halogen containing gases since 2004,
J. Quant. Spectrosc. Ra., 112, 2552–2566,
10.1016/j.jqsrt.2011.07.005, 2011.Brown, A. T., Volk, C. M., Schoeberl, M. R., Boone, C. D., and Bernath, P.
F.: Stratospheric lifetimes of CFC-12, CCl4, CH4, CH3Cl and
N2O from measurements made by the Atmospheric Chemistry
Experiment-Fourier Transform Spectrometer (ACE-FTS), Atmos. Chem. Phys., 13,
6921–6950, 10.5194/acp-13-6921-2013, 2013.Brown, A. T., Chipperfield, M. P., Richards, N. A. D., Boone, C., and
Bernath, P. F.: Global stratospheric fluorine inventory for 2004–2009 from
Atmospheric Chemistry Experiment Fourier Transform Spectrometer (ACE-FTS)
measurements and SLIMCAT model simulations, Atmos. Chem. Phys., 14, 267–282,
10.5194/acp-14-267-2014, 2014.Butchart, N.: The Brewer-Dobson circulation, Rev. Geophys., 52, 157–184,
10.1002/2013RG000448, 2014.Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P.,
Kobayashi,
S., Andrae, U., Balmaseda, M. A., Balsamo, G., Bauer, P., Bechtold, P.,
Beljaars, A. C. M., van de Berg, L., Bidlot, J., Bormann, N., Delsol, C.,
Dragani, R., Fuentes, M., Geer, A. J., Haimberger, L., Healy, S. B.,
Hersbach, H., Hólm, E. V., Isaksen, L., Kållberg, P., Köhler, M.,
Matricardi, M., McNally, A. P., Monge Sanz, B. M., Morcrette, J. J., Park,
B. K., Peubey, C., de Rosnay, P., Tavolato, C., Thépaut, J. N., and
Vitart, F.: The ERA-Interim reanalysis: configuration and performance of the
data assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553–597,
10.1002/qj.828, 2011.deGrandpré, J., Beagley, S. R., Fomichev, V. I., Griffioen, E.,
McConnell,
J. C., Medvedev, A. S., and Shepherd, T. G.: Ozone climatology using
interactive chemistry: Results from the Canadian Middle Atmosphere Model, J.
Geophys. Res., 105, 26475–26491, 10.1029/2000JD900427, 2000.Diallo, M., Legras, B., and Chédin, A.: Age of stratospheric air in the
ERA-Interim, Atmos. Chem. Phys., 12, 12133–12154,
10.5194/acp-12-12133-2012, 2012.Dickinson, R. E.: Vertical propagation of planetary rossby waves through an
atmosphere with Newtonian cooling, J. Geophys. Res., 74, 929–938,
10.1029/JB074i004p00929, 1969.Dobson, G. M. B.: Origin and Distribution of the Polyatomic Molecules in the
Atmosphere, P. Roy. Soc. A-Math. Phy., 236, 187–193, 10.1098/rspa.1956.0127,
1956.Dobson, G. M. B., Harrison, D. N., and Lawrence, J.: Measurements of the
Amount
of Ozone in the Earth's Atmosphere and Its Relation to Other
Geophysical Conditions. Part III, P. Roy. Soc. A-Math. Phy., 122, 456–486,
10.1098/rspa.1929.0034, 1929.Elkins, J. W., Thompson, T. M., Swanson, T. H., Butler, J. H., Hall, B. D.,
Cummings, S. O., Fisher, D. A., and Raffo, A. G.: Decrease in the growth
rates of atmospheric chlorofluorocarbons 11 and 12, Nature, 364, 780–783,
10.1038/364780a0, 1993.Engel, A., Möbius, T., Bönisch, H., Schmidt, U., Heinz, R., Levin,
I.,
Atlas, E., Aoki, S., Nakazawa, T., Sugawara, S., Moore, F., Hurst, D.,
Elkins, J., Schauffler, S., Andrews, A., and Boering, K.: Age of
stratospheric air unchanged within uncertainties over the past 30 years,
Nat. Geosci., 2, 28–31, 10.1038/NGEO388, 2009.Eyring, V., Harris, N. R. P., Rex, M., Shepherd, T. G., Fahey, D. W.,
Amanatidis, G. T., Austin, J., Chipperfield, M. P., Dameris, M., Forster, P.
M. D. F., Gettelman, A., Graf, H. F., Nagashima, T., Newman, P. A., Pawson,
S., Prather, M. J., Pyle, J. A., Salawitch, R. J., Santer, B. D., and Waugh,
D. W.: A Strategy for Process-Oriented Validation of Coupled
Chemistry-Climate Models, B. Am. Meteorol. Soc., 86, 1117–1133,
10.1175/BAMS-86-8-1117, 2005.Funke, B., García-Comas, M., López-Puertas, M., Glatthor, N.,
Stiller, G. P., von Clarmann, T., Semeniuk, K., and McConnell, J. C.:
Enhancement of N2O during the October–November 2003 solar proton
events, Atmos. Chem. Phys., 8, 3805–3815,
10.5194/acp-8-3805-2008, 2008a.Funke, B., López-Puertas, M., Garcia-Comas, M., Stiller, G. P., von
Clarmann, T., and Glatthor, N.: Mesospheric N2O enhancements as observed
by MIPAS on Envisat during the polar winters in 2002–2004, Atmos. Chem.
Phys., 8, 5787–5800, 10.5194/acp-8-5787-2008, 2008b.Garny, H., Birner, T., Bönisch, H., and Bunzel, F.: The effects of mixing
on age of air, J. Geophys. Res., 119, 7015–7034, 10.1002/2013JD021417,
2014.Gerber, E. P.: Stratospheric versus Tropospheric Control of the Strength and
Structure of the Brewer-Dobson Circulation, J. Atmos. Sci., 69, 2857–2877,
10.1175/JAS-D-11-0341.1, 2012.Greenblatt, J. B.: Correction to “Tracer-based determination of vortex
descent
in the 1999/2000 Arctic winter” by J. B. Greenblatt et al., J. Geophys.
Res., 107, SOL 50-1–SOL 50-2 , 10.1029/2002JD001597, 2003.Greenblatt, J. B., Jost, H. J., Loewenstein, M., Podolske, J. R., Hurst,
D. F.,
Elkins, J. W., Schauffler, S. M., Atlas, E. L., Herman, R. L., Webster,
C. R., Bui, T. P., Moore, F. L., Ray, E. A., Oltmans, S., Vömel, H.,
Blavier, J. F., Sen, B., Stachnik, R. A., Toon, G. C., Engel, A., Müller,
M., Schmidt, U., Bremer, H., Pierce, R. B., Sinnhuber, B. M., Chipperfield,
M., and Lefèvre, F.: Tracer-based determination of vortex descent in the
1999/2000 Arctic winter, J. Geophys. Res., 107, SOL 22-1–SOL 22-21, 10.1029/2001JD000937,
2002.Haenel, F. J., Stiller, G. P., von Clarmann, T., Funke, B., Eckert, E.,
Glatthor, N., Grabowski, U., Kellmann, S., Kiefer, M., Linden, A., and
Reddmann, T.: Reassessment of MIPAS age of air trends and variability, Atmos.
Chem. Phys., 15, 13161–13176, 10.5194/acp-15-13161-2015,
2015.Hall, T. M. and Plumb, R. A.: Age as a diagnostic of stratospheric
transport, J. Geophys. Res., 99, 1059–1070, 10.1029/93JD03192, 1994.Hall, T. M. and Waugh, D. W.: Stratospheric residence time and its
relationship to mean age, J. Geophys. Res., 105, 6773–6782,
10.1029/1999JD901096, 2000.Hardiman, S. C., Lin, P., Scaife, A. A., Dunstone, N. J., and Ren, H.-L.: The
influence of dynamical variability on the observed Brewer-Dobson circulation
trend, Geophys. Res. Lett., 44, 2885–2892, 10.1002/2017GL072706, 2017.Haynes, P. and Shuckburgh, E.: Effective diffusivity as a diagnostic of
atmospheric transport: 2. Troposphere and lower stratosphere, J. Geophys.
Res., 105, 22795–22810, 10.1029/2000JD900092, 2000.Hegglin, M. I. and Shepherd, T. G.: O3-N2O correlations from the
Atmospheric Chemistry Experiment: Revisiting a diagnostic of transport and
chemistry in the stratosphere, J. Geophys. Res., 112, D19301,
10.1029/2006JD008281, 2007.Hegglin, M. I., Boone, C. D., Manney, G. L., Shepherd, T. G., Walker, K. A.,
Bernath, P. F., Daffer, W. H., Hoor, P., and Schiller, C.: Validation of
ACE-FTS satellite data in the upper troposphere/lower stratosphere (UTLS)
using non-coincident measurements, Atmos. Chem. Phys., 8, 1483–1499,
10.5194/acp-8-1483-2008, 2008.Hegglin, M. I., Boone, C. D., Manney, G. L., and Walker, K. A.: A global view
of the extratropical tropopause transition layer from Atmospheric Chemistry
Experiment Fourier Transform Spectrometer O3, H2O, and CO, J. Geophys.
Res., 114, D00B11, 10.1029/2008JD009984, 2009.Hegglin, M. I., Tegtmeier, S., Anderson, J., Froidevaux, L., Fuller, R.,
Funke,
B., Jones, A., Lingenfelser, G., Lumpe, J., Pendlebury, D., Remsberg, E.,
Rozanov, A., Toohey, M., Urban, J., von Clarmann, T., Walker, K. A., Wang,
R., and Weigel, K.: SPARC Data Initiative: Comparison of water vapor
climatologies from international satellite limb sounders, J. Geophys. Res.,
118, 11824–11846, 10.1002/jgrd.50752, 2013.Hegglin, M. I., Plummer, D. A., Shepherd, T. G., Scinocca, J. F., Anderson,
J.,
Froidevaux, L., Funke, B., Hurst, D., Rozanov, A., Urban, J., von Clarmann,
T., Walker, K. A., Wang, H. J., Tegtmeier, S., and Weigel, K.: Vertical
structure of stratospheric water vapour trends derived from merged satellite
data, Nat. Geosci., 7, 768–776, 10.1038/NGEO2236, 2014.Hoffmann, L., Hoppe, C. M., Müller, R., Dutton, G. S., Gille, J. C.,
Griessbach, S., Jones, A., Meyer, C. I., Spang, R., Volk, C. M., and Walker,
K. A.: Stratospheric lifetime ratio of CFC-11 and CFC-12 from satellite and
model climatologies, Atmos. Chem. Phys., 14, 12479–12497,
10.5194/acp-14-12479-2014, 2014.Holton, J. R., Haynes, P. H., McIntyre, M. E., Douglass, A. R., Rood, R. B.,
and Pfister, L.: Stratosphere-troposphere exchange, Rev. Geophys., 33,
403–439, 10.1029/95RG02097, 1995.Jin, J. J., Semeniuk, K., Jonsson, A. I., Beagley, S. R., McConnell, J. C.,
Boone, C. D., Walker, K. A., Bernath, P. F., Rinsland, C. P., Dupuy, E.,
Ricaud, P., De La Noë, J., Urban, J., and Murtagh, D.: Co-located ACE-FTS
and Odin/SMR stratospheric-mesospheric CO 2004 measurements and comparison
with a GCM, Geophys. Res. Lett., 32, L15S03, 10.1029/2005GL022433,
2005.Jones, A., Walker, K. A., Jin, J. J., Taylor, J. R., Boone, C. D., Bernath,
P. F., Brohede, S., Manney, G. L., McLeod, S., Hughes, R., and Daffer, W. H.:
Technical Note: A trace gas climatology derived from the Atmospheric
Chemistry Experiment Fourier Transform Spectrometer (ACE-FTS) data set,
Atmos. Chem. Phys., 12, 5207–5220, 10.5194/acp-12-5207-2012,
2012.Jonsson, A. I., de Grandpré, J., Fomichev, V. I., McConnell, J. C., and
Beagley, S. R.: Doubled CO2-induced cooling in the middle atmosphere:
Photochemical analysis of the ozone radiative feedback, J. Geophys. Res.,
109, D24103, 10.1029/2004JD005093, 2004.Kawamoto, N. and Shiotani, M.: Interannual variability of the vertical
descent
rate in the Antarctic polar vortex, J. Geophys. Res., 105, 11935–11946,
10.1029/2000JD900076, 2000.Koo, J.-H., Walker, K. A., Jones, A., Sheese, P. E., Boone, C. D., Bernath,
P. F., and Manney, G. L.: Global climatology based on the ACE-FTS version 3.5
dataset: Addition of mesospheric levels and carbon-containing species in the
UTLS, J. Quant. Spectrosc. Ra., 186, 52–62,
10.1016/j.jqsrt.2016.07.003, 2017.Lin, P. and Fu, Q.: Changes in various branches of the Brewer-Dobson
circulation from an ensemble of chemistry climate models, J. Geophys. Res.,
118, 73–84, 10.1029/2012JD018813, 2013.Mahieu, E., Duchatelet, P., Demoulin, P., Walker, K. A., Dupuy, E.,
Froidevaux, L., Randall, C., Catoire, V., Strong, K., Boone, C. D., Bernath,
P. F., Blavier, J.-F., Blumenstock, T., Coffey, M., De Mazière, M.,
Griffith, D., Hannigan, J., Hase, F., Jones, N., Jucks, K. W., Kagawa, A.,
Kasai, Y., Mebarki, Y., Mikuteit, S., Nassar, R., Notholt, J., Rinsland, C.
P., Robert, C., Schrems, O., Senten, C., Smale, D., Taylor, J., Tétard,
C., Toon, G. C., Warneke, T., Wood, S. W., Zander, R., and Servais, C.:
Validation of ACE-FTS v2.2 measurements of HCl, HF, CCl3F and
CCl2F2 using space-, balloon- and ground-based instrument
observations, Atmos. Chem. Phys., 8, 6199–6221,
10.5194/acp-8-6199-2008, 2008.Manney, G. L., Daffer, W. H., Zawodny, J. M., Bernath, P. F., Hoppel, K. W.,
Walker, K. A., Knosp, B. W., Boone, C., Remsberg, E. E., Santee, M. L.,
Harvey, V. L., Pawson, S., Jackson, D. R., Deaver, L., McElroy, C. T.,
McLinden, C. A., Drummond, J. R., Pumphrey, H. C., Lambert, A., Schwartz,
M. J., Froidevaux, L., McLeod, S., Takacs, L. L., Suarez, M. J., Trepte,
C. R., Cuddy, D. C., Livesey, N. J., Harwood, R. S., and Waters, J. W.:
Solar occultation satellite data and derived meteorological products:
Sampling issues and comparisons with Aura Microwave Limb Sounder, J.
Geophys. Res., 112, D24S50, 10.1029/2007JD008709, 2007.Manney, G. L., Harwood, R. S., MacKenzie, I. A., Minschwaner, K., Allen, D.
R., Santee, M. L., Walker, K. A., Hegglin, M. I., Lambert, A., Pumphrey, H.
C., Bernath, P. F., Boone, C. D., Schwartz, M. J., Livesey, N. J., Daffer, W.
H., and Fuller, R. A.: Satellite observations and modeling of transport in
the upper troposphere through the lower mesosphere during the 2006 major
stratospheric sudden warming, Atmos. Chem. Phys., 9, 4775–4795,
10.5194/acp-9-4775-2009, 2009.McIntyre, M. E. and Palmer, T. N.: Breaking planetary waves in the
stratosphere, Nature, 305, 593–600, 10.1038/305593a0, 1983.McIntyre, M. E. and Palmer, T. N.: The “surf zone” in the stratosphere,
J. Atmos. Terr. Phys., 46, 825–849, 10.1016/0021-9169(84)90063-1,
1984.McLandress, C., Shepherd, T. G., Scinocca, J. F., Plummer, D. A., Sigmond,
M.,
Jonsson, A. I., and Reader, M. C.: Separating the Dynamical Effects of
Climate Change and Ozone Depletion. Part II: Southern Hemisphere
Troposphere, J. Climate, 24, 1850–1868, 10.1175/2010JCLI3958.1, 2011.McLandress, C., Shepherd, T. G., Polavarapu, S., and Beagley, S. R.: Is
Missing
Orographic Gravity Wave Drag near 60∘ S the Cause of the Stratospheric
Zonal Wind Biases in Chemistry-Climate Models?, J. Atmos. Sci., 69, 802–818,
2012.McLandress, C., Scinocca, J. F., Shepherd, T. G., Reader, M. C., and Manney,
G. L.: Dynamical Control of the Mesosphere by Orographic and Nonorographic
Gravity Wave Drag during the Extended Northern Winters of 2006 and 2009, J.
Atmos. Sci., 70, 5449–5463, 10.1175/JAS-D-12-0297.1, 2013.McLandress, C., Shepherd, T. G., Reader, M. C., Plummer, D. A., and Shine,
K. P.: The Climate Impact of Past Changes in Halocarbons and CO2 in the
Tropical UTLS Region, J. Climate, 27, 8646–8660,
10.1175/JCLI-D-14-00232.1, 2014.Meinshausen, M., Vogel, E., Nauels, A., Lorbacher, K., Meinshausen, N.,
Etheridge, D. M., Fraser, P. J., Montzka, S. A., Rayner, P. J., Trudinger, C.
M., Krummel, P. B., Beyerle, U., Canadell, J. G., Daniel, J. S., Enting, I.
G., Law, R. M., Lunder, C. R., O'Doherty, S., Prinn, R. G., Reimann, S.,
Rubino, M., Velders, G. J. M., Vollmer, M. K., Wang, R. H. J., and Weiss, R.:
Historical greenhouse gas concentrations for climate modelling (CMIP6),
Geosci. Model Dev., 10, 2057–2116, 10.5194/gmd-10-2057-2017,
2017.Merrifield, M. A., Thompson, P. R., and Lander, M.: Multidecadal sea level
anomalies and trends in the western tropical Pacific, Geophys. Res. Lett.,
39, L13602, 10.1029/2012GL052032, 2012.Millán, L. F., Livesey, N. J., Santee, M. L., Neu, J. L., Manney, G. L.,
and Fuller, R. A.: Case studies of the impact of orbital sampling on
stratospheric trend detection and derivation of tropical vertical velocities:
solar occultation vs. limb emission sounding, Atmos. Chem. Phys., 16,
11521–11534, 10.5194/acp-16-11521-2016, 2016.Montzka, S. A., Butler, J. H., Myers, R. C., Thompson, T. M., Swanson, T. H.,
Clarke, A. D., Lock, L. T., and Elkins, J. W.: Decline in the Tropospheric
Abundance of Halogen from Halocarbons: Implications for Stratospheric Ozone
Depletion, Science, 272, 1318–1322, 10.1126/science.272.5266.1318,
1996.Mote, P. W., Rosenlof, K. H., McIntyre, M. E., Carr, E. S., Gille, J. C.,
Holton, J. R., Kinnersley, J. S., Pumphrey, H. C., Russell III, J. M., and
Waters, J. W.: An atmospheric tape recorder: The imprint of tropical
tropopause temperatures on stratospheric water vapor, J. Geophys. Res., 101,
3989–4006, 10.1029/95JD03422, 1996.Neu, J. L. and Plumb, R. A.: Age of air in a “leaky pipe” model of
stratospheric transport, J. Geophys. Res., 104, 19243–19255,
10.1029/1999JD900251, 1999.Neu, J. L., Hegglin, M. I., Tegtmeier, S., Bourassa, A., Degenstein, D.,
Froidevaux, L., Fuller, R., Funke, B., Gille, J., Jones, A., Rozanov, A.,
Toohey, M., von Clarmann, T., Walker, K. A., and Worden, J. R.: The SPARC
Data Initiative: Comparison of upper troposphere/lower stratosphere ozone
climatologies from limb-viewing instruments and the nadir-viewing
Tropospheric Emission Spectrometer, J. Geophys. Res., 119, 6971–6990,
10.1002/2013JD020822, 2014.Niwano, M., Yamazaki, K., and Shiotani, M.: Seasonal and QBO variations of
ascent rate in the tropical lower stratosphere as inferred from UARS HALOE
trace gas data, J. Geophys. Res., 108, 4794, 10.1029/2003JD003871,
2003.Oberländer-Hayn, S., Gerber, E., Abalichin, J., Akiyoshi, H.,
Kerschbaumer,
A., Kubin, A., Kunze, M., Langematz, U., Meul, S., Michou, M., Morgenstern,
O., and Oman, L. D.: Is the Brewer-Dobson circulation increasing or moving
upward?, Geophys. Res. Lett., 43, 1772–1779, 10.1002/2015GL067545,
2016.Pan, L. L., Randel, W. J., Gille, J. C., Hall, W. D., Nardi, B., Massie, S.,
Yudin, V., Khosravi, R., Konopka, P., and Tarasick, D.: Tropospheric
intrusions associated with the secondary tropopause, J. Geophys. Res.,
114, D10302, 10.1029/2008JD011374, 2009.Park, M., Randel, W. J., Kinnison, D. E., Emmons, L. K., Bernath, P. F.,
Walker, K. A., Boone, C. D., and Livesey, N. J.: Hydrocarbons in the upper
troposphere and lower stratosphere observed from ACE-FTS and comparisons with
WACCM, J. Geophys. Res., 118, 1964–1980, 10.1029/2012JD018327,
2013.Pawson, S., Kodera, K., Hamilton, K., Shepherd, T. G., Beagley, S. R.,
Boville,
B. A., Farrara, J. D., Fairlie, T. D. A., Kitoh, A., Lahoz, W. A., Langematz,
U., Manzini, E., Rind, D. H., Scaife, A. A., Shibata, K., Simon, P.,
Swinbank, R., Takacs, L., Wilson, R. J., Al-Saadi, J. A., Amodei, M., Chiba,
M., Coy, L., de Grandpré, J., Eckman, R. S., Fiorino, M., Grose, W. L.,
Koide, H., Koshyk, J. N., Li, D., Lerner, J., Mahlman, J. D., McFarlane,
N. A., Mechoso, C. R., Molod, A., O'Neill, A., Pierce, R. B., Randel, W. J.,
Rood, R. B., and Wu, F.: The GCM-reality intercomparison project for SPARC
(GRIPS): Scientific issues and initial results, B. Am. Meteorol. Soc., 81,
781–796, 10.1175/1520-0477(2000)081<0781:TGIPFS>2.3.CO;2, 2000.Peevey, T. R., Gille, J. C., Homeyer, C. R., and Manney, G. L.: The double
tropopause and its dynamical relationship to the tropopause inversion layer
in storm track regions, J. Geophys. Res., 119, 10194–10212,
10.1002/2014JD021808, 2014.Pendlebury, D., Plummer, D., Scinocca, J., Sheese, P., Strong, K., Walker,
K., and Degenstein, D.: Comparison of the CMAM30 data set with ACE-FTS and
OSIRIS: polar regions, Atmos. Chem. Phys., 15, 12465–12485,
10.5194/acp-15-12465-2015, 2015.Ploeger, F., Abalos, M., Birner, T., Konopka, P., Legras, B., Müller, R.,
and
Riese, M.: Quantifying the effects of mixing and residual circulation on
trends of stratospheric mean age of air, Geophys. Res. Lett., 42, 2047–2054,
10.1002/2014GL062927, 2015a.Ploeger, F., Riese, M., Haenel, F., Konopka, P., Müller, R., and Stiller,
G.:
Variability of stratospheric mean age of air and of the local effects of
residual circulation and eddy mixing, J. Geophys. Res., 120, 716–733,
10.1002/2014JD022468, 2015b.Plumb, R. A.: A “tropical pipe” model of stratospheric transport, J.
Geophys. Res., 101, 3957–3972, 10.1029/95JD03002, 1996.Plumb, R. A.: Stratospheric Transport, J. Meteorol. Soc. Jpn., 80,
793–809, 10.2151/jmsj.80.793, 2002.Plumb, R. A. and Eluszkiewicz, J.: The Brewer-Dobson Circulation: Dynamics
of
the Tropical Upwelling, J. Atmos. Sci., 56, 868–890,
10.1175/1520-0469(1999)056<0868:TBDCDO>2.0.CO;2, 1999.Plumb, R. A. and Ko, M. K. W.: Interrelationships between mixing ratios of
long-lived stratospheric constituents, J. Geophys. Res., 97,
10145–10156, 10.1029/92JD00450, 1992.Plummer, D. A., Scinocca, J. F., Shepherd, T. G., Reader, M. C., and Jonsson,
A. I.: Quantifying the contributions to stratospheric ozone changes from
ozone depleting substances and greenhouse gases, Atmos. Chem. Phys., 10,
8803–8820, 10.5194/acp-10-8803-2010, 2010.Ray, E. A., Moore, F. L., Elkins, J. W., Hurst, D. F., Romashkin, P. A.,
Dutton, G. S., and Fahey, D. W.: Descent and mixing in the 1999–2000
northern polar vortex inferred from in situ tracer measurements, J. Geophys.
Res., 107, 8285, 10.1029/2001JD000961, 2002.Ray, E. A., Moore, F. L., Rosenlof, K. H., Davis, S. M., Sweeney, C., Tans,
P.,
Wang, T., Elkins, J. W., Bönisch, H., Engel, A., Sugawara, S., Nakazawa, T.,
and Aoki, S.: Improving stratospheric transport trend analysis based on
SF6 and CO2 measurements, J. Geophys. Res., 119, 14110–14128,
10.1002/2014JD021802, 2014.Ray, E. A., Moore, F. L., Rosenlof, K. H., Plummer, D. A., Kolonjari, F., and
Walker, K. A.: An idealized stratospheric model useful for understanding
differences between long-lived trace gas measurements and global
chemistry-climate model output, J. Geophys. Res., 121, 5356–5367,
10.1002/2015JD024447, 2016.Rayner, N. A., Parker, D. E., Horton, E. B., Folland, C. K., Alexander,
L. V.,
Rowell, D. P., Kent, E. C., and Kaplan, A.: Global analyses of sea surface
temperature, sea ice, and night marine air temperature since the late
nineteenth century, J. Geophys. Res., 108, 4407, 10.1029/2002JD002670, 2003.Rosenlof, K. H.: Seasonal cycle of the residual mean meridional circulation
in the stratosphere, J. Geophys. Res., 100, 5173–5191,
10.1029/94JD03122, 1995.
Sander, S. P., Friedl, R. R., Abbatt, J. P. D., Barker, J. R., Burkholder,
J. B., Golden, D. M., Kolb, C. E., Kurylo, M. J., Moortgat, G. K., Wine,
P. H., Huie, R. E., and Orkin, V. L.: Chemical Kinetics and Photochemical
Data for Use in Atmospheric Modeling, Evaluation Number 17, JPL Publication
10-6, Jet Propulsion Laboratory, Pasadena, USA, 2011.Sankey, D. and Shepherd, T. G.: Correlations of long-lived chemical species
in a middle atmosphere general circulation model, J. Geophys. Res., 108,
4494, 10.1029/2002JD002799, 2003.Schauffler, S. M., Atlas, E. L., Donnelly, S. G., Andrews, A., Montzka,
S. A.,
Elkins, J. W., Hurst, D. F., Romashkin, P. A., Dutton, G. S., and Stroud, V.:
Chlorine budget and partitioning during the Stratospheric Aerosol and Gas
Experiment (SAGE) III Ozone Loss and Validation Experiment (SOLVE), J.
Geophys. Res., 108, 4173, 10.1029/2001JD002040, 2003.Schwartz, M. J., Manney, G. L., Hegglin, M. I., Livesey, N. J., Santee,
M. L.,
and Daffer, W. H.: Climatology and variability of trace gases in
extratropical double-tropopause regions from MLS, HIRDLS, and ACE-FTS
measurements, J. Geophys. Res., 120, 843–867, 10.1002/2014JD021964,
2015.Scinocca, J. F., McFarlane, N. A., Lazare, M., Li, J., and Plummer, D.:
Technical Note: The CCCma third generation AGCM and its extension into the
middle atmosphere, Atmos. Chem. Phys., 8, 7055–7074,
10.5194/acp-8-7055-2008, 2008.
Seviour, W. J. M., Butchart, N., and Hardiman, S. C.: The Brewer-Dobson
circulation inferred from ERA-Interim, Q. J. Roy. Meteorol. Soc., 138,
878–888, 2012.Sheese, P. E., Boone, C. D., and Walker, K. A.: Detecting physically
unrealistic outliers in ACE-FTS atmospheric measurements, Atmos. Meas. Tech.,
8, 741–750, 10.5194/amt-8-741-2015, 2015.Sheese, P. E., Walker, K. A., Boone, C. D., Bernath, P. F., and Funke, B.:
Nitrous oxide in the atmosphere: First measurements of a lower thermospheric
source, Geophys. Res. Lett., 43, 2866–2872, 10.1002/2015GL067353,
2016.Shepherd, T. G.: Transport in the Middle Atmosphere, J. Meteorol. Soc. Jpn.,
85B, 165–191, 10.2151/jmsj.85B.165, 2007.Shepherd, T. G. and McLandress, C.: A Robust Mechanism for Strengthening of
the Brewer-Dobson Circulation in Response to Climate Change: Critical-Layer
Control of Subtropical Wave Breaking, J. Atmos. Sci., 68, 784–797,
10.1175/2010JAS3608.1, 2011.Shepherd, T. G., Plummer, D. A., Scinocca, J. F., Hegglin, M. I., Fioletov,
V. E., Reader, M. C., Remsberg, E., von Clarmann, T., and Wang, H. J.:
Reconciliation of halogen-induced ozone loss with the total-column ozone
record, Nat. Geosci., 7, 443–449, 10.1038/NGEO2155, 2014.SPARC-CCMVal (Stratospheric Processes And their Role in Climate): Report on
the Evaluation of Chemistry-Climate Models, SPARC report No. 5, WCRP-132,
WMO/TD-No. 1526, available at:
http://www.sparc-climate.org/publications/sparc-reports/ (last access:
8 May 2018), 2010.Sparling, L. C.: Statistical perspectives on stratospheric transport, Rev.
Geophys., 38, 417–436, 10.1029/1999RG000070, 2000.Stiller, G. P., von Clarmann, T., Höpfner, M., Glatthor, N., Grabowski,
U., Kellmann, S., Kleinert, A., Linden, A., Milz, M., Reddmann, T., Steck,
T., Fischer, H., Funke, B., López-Puertas, M., and Engel, A.: Global
distribution of mean age of stratospheric air from MIPAS SF6 measurements,
Atmos. Chem. Phys., 8, 677–695, 10.5194/acp-8-677-2008,
2008.Stiller, G. P., von Clarmann, T., Haenel, F., Funke, B., Glatthor, N.,
Grabowski, U., Kellmann, S., Kiefer, M., Linden, A., Lossow, S., and
López-Puertas, M.: Observed temporal evolution of global mean age of
stratospheric air for the 2002 to 2010 period, Atmos. Chem. Phys., 12,
3311–3331, 10.5194/acp-12-3311-2012, 2012.Strahan, S. E., Douglass, A. R., Stolarski, R. S., Akiyoshi, H., Bekki, S.,
Braesicke, P., Butchart, N., Chipperfield, M. P., Cugnet, D., Dhomse, S.,
Frith, S. M., Gettelman, A., Hardiman, S. C., Kinnison, D. E., Lamarque,
J. F., Mancini, E., Marchand, M., Michou, M., Morgenstern, O., Nakamura, T.,
Olivié, D., Pawson, S., Pitari, G., Plummer, D. A., Pyle, J. A.,
Scinocca, J. F., Shepherd, T. G., Shibata, K., Smale, D., Teyssèdre, H.,
Tian, W., and Yamashita, Y.: Using transport diagnostics to understand
chemistry climate model ozone simulations, J. Geophys. Res., 116, D17302,
10.1029/2010JD015360, 2011.Strong, K., Wolff, M. A., Kerzenmacher, T. E., Walker, K. A., Bernath, P. F.,
Blumenstock, T., Boone, C., Catoire, V., Coffey, M., De Mazière, M.,
Demoulin, P., Duchatelet, P., Dupuy, E., Hannigan, J., Höpfner, M.,
Glatthor, N., Griffith, D. W. T., Jin, J. J., Jones, N., Jucks, K.,
Kuellmann, H., Kuttippurath, J., Lambert, A., Mahieu, E., McConnell, J. C.,
Mellqvist, J., Mikuteit, S., Murtagh, D. P., Notholt, J., Piccolo, C.,
Raspollini, P., Ridolfi, M., Robert, C., Schneider, M., Schrems, O.,
Semeniuk, K., Senten, C., Stiller, G. P., Strandberg, A., Taylor, J.,
Tétard, C., Toohey, M., Urban, J., Warneke, T., and Wood, S.: Validation
of ACE-FTS N2O measurements, Atmos. Chem. Phys., 8, 4759–4786,
10.5194/acp-8-4759-2008, 2008.Tegtmeier, S., Hegglin, M. I., Anderson, J., Bourassa, A., Brohede, S.,
Degenstein, D., Froidevaux, L., Fuller, R., Funke, B., Gille, J., Jones, A.,
Kasai, Y., Krüger, K., Kyrölä, E., Lingenfelser, G., Lumpe, J., Nardi, B.,
Neu, J., Pendlebury, D., Remsberg, E., Rozanov, A., Smith, L., Toohey, M.,
Urban, J., von Clarmann, T., Walker, K. A., and Wang, R. H. J.: SPARC Data
Initiative: A comparison of ozone climatologies from international satellite
limb sounders, J. Geophys. Res., 118, 12229–12247,
10.1002/2013JD019877, 2013.Tegtmeier, S., Hegglin, M. I., Anderson, J., Funke, B., Gille, J., Jones, A.,
Smith, L., von Clarmann, T., and Walker, K. A.: The SPARC Data Initiative:
comparisons of CFC-11, CFC-12, HF and SF6 climatologies from
international satellite limb sounders, Earth Syst. Sci. Data, 8, 61–78,
10.5194/essd-8-61-2016, 2016.
Toohey, M. and von Clarmann, T.: Climatologies from satellite measurements:
the impact of orbital sampling on the standard error of the mean, Atmos.
Meas. Tech., 6, 937–948, 10.5194/amt-6-937-2013, 2013.Toohey, M., Hegglin, M. I., Tegtmeier, S., Anderson, J., Añel, J. A.,
Bourassa, A., Brohede, S., Degenstein, D., Froidevaux, L., Fuller, R., Funke,
B., Gille, J., Jones, A., Kasai, Y., Krüger, K., Kyrölä, E., Neu, J. L.,
Rozanov, A., Smith, L., Urban, J., von Clarmann, T., Walker, K. A., and Wang,
R. H. J.: Characterizing sampling biases in the trace gas climatologies of
the SPARC Data Initiative, J. Geophys. Res., 118, 11847–11862,
10.1002/jgrd.50874, 2013.Toon, G. C., Blavier, J. F., Sen, B., Margitan, J. J., Webster, C. R., May,
R. D., Fahey, D., Gao, R., Del Negro, L., Proffitt, M., Elkins, J.,
Romashkin, P. A., Hurst, D. F., Oltmans, S., Atlas, E., Schauffler, S.,
Flocke, F., Bui, T. P., Stimpfle, R. M., Bonne, G. P., Voss, P. B., and
Cohen, R. C.: Comparison of MkIV balloon and ER-2 aircraft measurements of
atmospheric trace gases, J. Geophys. Res., 104, 26779–26790,
10.1029/1999JD900379, 1999.Velazco, V. A., Toon, G. C., Blavier, J.-F. L., Kleinböhl, A., Manney,
G. L., Daffer, W. H., Bernath, P. F., Walker, K. A., and Boone, C.:
Validation of the Atmospheric Chemistry Experiment by noncoincident MkIV
balloon profiles, J. Geophys. Res., 116, D06306,
10.1029/2010JD014928, 2011.Vyushin, D. I., Fioletov, V. E., and Shepherd, T. G.: Impact of long-range
correlations on trend detection in total ozone, J. Geophys. Res., 112, D14307,
10.1029/2006JD008168, 2007.Waugh, D. and Hall, T.: Age of Stratospheric Air: Theory, Observations, and
Models, Rev. Geophys., 40, 1-1–1-26, 10.1029/2000RG000101, 2002.Waymark, C., Walker, K. A., Boone, C. D., and Bernath, P. F.: ACE-FTS
version 3.0 data set: validation and data processing update, Ann. Geophys.,
56, Fast Track–1, 10.4401/ag-6339, 2013.Yulaeva, E., Holton, J. R., and Wallace, J. M.: On the Cause of the Annual
Cycle in Tropical Lower-Stratospheric Temperatures, J. Atmos. Sci., 51,
169–174, 10.1175/1520-0469(1994)051<0169:OTCOTA>2.0.CO;2, 1994.