We present a comparison of trends in total column ozone
from ten two-dimensional and four three-dimensional models and SBUV/2
satellite observations from the period 1979-2003. Trends for the past
(1979-2000), the recent seven years (1996-2003), and the future
(2000-2050) are compared. We have analyzed the data using both simple
linear trends and linear trends derived with a hockey stick method
including a turnaround point in 1996. If the last seven years,
1996-2003, are analyzed in isolation, the SBUV/2 observations show no
increase in ozone and most of the models predict continued depletion,
although at a lesser rate. In sharp contrast to this the recent data
show positive trends for the northern and the southern hemisphere if
the hockey stick method with a turnaround point in 1996 is employed for
the models and observations. The analysis shows that the observed
positive trends in both hemispheres in the recent seven- year period
are much larger than what is predicted by the models. The trends
derived with the hockey stick method are very dependent on the values
just before the turnaround point. The analysis of the recent data
therefore depends greatly on these years being representative of the
overall trend. Most models underestimate the past trends at mid- and
high latitudes. This is particularly pronounced in the northern
hemisphere. Quantitatively there is much disagreement among the models
concerning future trends. However, the models agree that future trends
are expected to be positive and less than half the magnitude of the
past downward trends. Examination of the model projections shows that
there is virtually no correlation between the past and future trends
from the individual models.
Recent observational studies [1,2] uncovered a layer of strongly enhanced static stability that is on average about 2-3 km thick and resides just above the local extratropical tropopause. These observational studies are based on vertically high--resolved (~ 50 m) radiosonde data. The question arises to what extent this layer exists in current climate models. Here we adress this question using output from the Canadian Middle Atmosphere Model (CMAM). General aspects of the UTLS thermal and wind structure in the CMAM will be presented as well.
[1] Birner, T., A. D{\"o}rnbrack, U. Schumann (2002), How sharp is
the tropopause at the midlatitudes?, {\it Geophys. Res. Lett., 29}, 1700, doi: 10.1029/2002GL015,142.
[2] Birner, T. (2005), The fine--scale structure of the extratropical
tropopause region, {\it submitted to J. Geophys. Res.}
Within the EU integrated project SCOUT-O3 different
chemistry-climate models will be used to verify and predict past and
future changes in climate and composition of the upper troposphere and
lower stratosphere. A central workpackage within SCOUT-O3 is the
characterisation and intercomparison of existing and upcoming new
integrations for the near past and future using these chemistry-climate
models, requiring the definition and application of standard diagnostic
tools. Here, we will use singular value decomposition of geopotential
heights to obtain coupled modes from the participating models. We will
present an intercomparison of the detected coupled modes in conjunction
with their impacts on ozone. In addition, we will contrast the model
results with observational evidence of coupled modes obtained from
ERA-40 data.
A pronounced ENSO cycle occurred in the second half of
the 1980s, with an El Niño during the 1986/87 winter and a relatively
strong La Niña in the winter 1988/89. Distinct anomalies in global
climate, stratospheric circulation, and total ozone were observed
during these winters, which makes this ENSO cycle an interesting case
for comparing and validating climate models. In this paper we compare
the two winters 1986/87 and 1988/89 in ensemble simulations that were
forced with observed SSTs. We analyse the output from two different
models and compare the results with observations. The first model is
the AGCM MRF9, which was run at a resolution of T40, with 18 sigma
levels (model top at 50 hPa). Two ensembles of 45 members are available
for the two winters (November to March). The second model is the CCM
SOCOL, which consists of MA-ECHAM4 and chemistry model MEZON. The model
was run with a horizontal resolution of T30 and 39 levels (model top at
0.01 hPa). With this model we performed 11 simulations for each winter.
Results from the two models are compared with respect to differences in
late winter (January to March) 1987 and 1989 for temperature and
pressure at the surface and in the lower stratosphere. The agreement
with each other and with the observations is good, and many features
obtained from the observations also appear in the results of both
models with a high level of confidence. For the SOCOL model, more
specific comparisons with observations can be performed with respect to
stratospheric dynamics and chemistry. We present comparisons for middle
stratospheric temperatures, EP flux, and total ozone. In addition, we
analyse ozone in an equivalent latitude–potential temperature framework
and compare the differences to a corresponding assimilated data set
that is based on TOMS/GOME/SBUV observations.
One of the key dynamical processes identified at the first CCMval
meeting is the stratospheric response to planetary wave drag.
Mid-winter stratospheric sudden warming (SSW) events are the
preeminent manifestation of this process. Therefore, we examine the
characteristics of SSWs in a small number of current generation GCMs
specifically designed to resolve the stratospheric circulations, and
we contrast the modeling results with both NCAR/NCEP and ERA-40
reanalysis datasets. Most models are found to have fewer sudden
warmings than are present in the observations, with some models
showing a remarkable lack of sudden warming activity. Several key,
process-based diagnostic benchmarks of the type, amplitude,
distribution and precursors to SSWs are defined. Comparing these
benchmarks between the reanalysis datasets and the GCMs is helpful in
understanding the reasons for the relative lack of SSW activity in GCMs.
The representation of the quasi-biennial oscillation in
atmospheric GCMs is still challenging. In the context of CCMVal the
inability of a GCM to simulate the QBO is indicative of the biased or
incomplete representation of the processes driving the QBO that has to
be addressed eventually. By the dynamical influence of the QBO, a
missing representation of the QBO is not just disturbing but is also
considered as a cause of biases in the simulation of transport or
mixing of tracers in the middle atmosphere, or in the vortex dynamics.
The assimilation of the QBO by relaxation schemes has been employed
therefore in some CCM integrations to overcome these secondary problems
related to the missing ability to simulate the QBO. This study presents
an overview of the QBOs in a number of CCMs which have been integrated
over the 1980 to 1999 period. The average QBO structure is compared
between the models and ERA-40, and the resulting signals in temperature
and chemical compounds are discussed. Further aspects of this study are
the annual mean effect of the QBO on the general circulation of the
tropical middle atmosphere, as for example on the tropical upwelling,
the atmospheric tape recorder or the SAO, and the question if these
effects can be reproduced in models using externally forced QBOs.
We have run two 20-year simulations of GSFC General Circulation Model (GCM, GEOS-4 Version) with mean ozone distributions derived from our CTM, corresponding to 1978-1980 and 1998-2000. These were input to the radiation code in the GCM to evaluate the dynamical and radiative response of the model to ozone change. Several significant changes occur with decreasing ozone. For the 2000 simulation, there is about three weeks of delay in break-up of southern hemispheric polar vortex. Over the Antarctic region, there is a decrease in temperature in the lower stratosphere with an increase in the middle stratosphere. On an average, the Antarctic tropopause height is several hundred meters higher in the 2000 simulation as compared to the 1980 simulation. There is no significant change in the annual extratropical stratosphere-troposphere mass exchange as well as the mass of the extratropical lowermost stratosphere between the tropopause and 380 K surface.
These model simulations represent a first step in running fully interactive coupled simulations of stratospheric ozone. Presently, two 20-year time slice simulations of the coupled model for 1980 and 2000 are underway. We will also present results from these simulations and compare them against those from non-interactive simulations.
The response to the 11 year solar cycle is investigated using a 3-D mechanistic stratosphere-mesosphere model, in particular how the level of planetary wave activity changes the effect of the solar cycle, and the zonal asymmetry in the solar signal.
Planetary waves are initiated at the lowest boundary level of the model, which corresponds to the tropopause height. Model simulations are carried out in pairs, with one simulation using solar forcing corresponding to solar minimum and the other to solar maximum. The level of lower boundary planetary wave forcing is varied between pairs of model simulations.
The results illustrate the crucial role played by the planetary wave forcing in the solar cycle temperature signal. The solar cycle temperature signal in the tropics and subtropics is about 1K for all values of wave forcing. However, in the extra-tropics the solar signal varies critically with wave forcing, giving a solar signal as strong as 16K for intermediate values of wave forcing. Despite some spatial differences, the simulations with a specific wave forcing show good qualitative agreements with observational results from rocketsondes, lidars and satellites.
Above a critical level of wave activity, the non-linear interaction with the mean flow induces a stratospheric warming and strong temperature change. The critical wave-forcing amplitude necessary to produce such an event is very sensitive to the initial state of the atmosphere and a small change of the mean wind, due for example to an enhancement of the solar forcing, can generate a large difference in temperature, depending on the level of the wave forcing. The numerical simulations presented here suggest a mechanism by which a small change induced by the solar forcing can generate a large atmospheric response.
Model results are also compared to the solar signal seen from five rocket-sonde sites and one lidar site, all located at mid-latitude, but at different longitudes. Although there are some significant differences between the model results and the observations, the model results help to explain why there are significant variations in the solar signal in the observations made at different longitudinal positions. Such longitudinal dependence obviously does not appear in zonal mean results, hence this raises questions about the meaning of comparing zonal mean model results with results from local observations.
Further simulations will be carried out examining these themes in the CCM model LMDz-REPRO (coupling of the GCM LMDz and the
chemistry module of the CTM REPROBUS).
Using chemistry-climate models (CCMs) to investigate whether recovery of Antarctic ozone is likely to be delayed or accelerated by climate change is complicated by the fact that there is significant disagreement between different models. Temperature biases within CCMs, and their parameterizations dependent on temperature (e.g. PSC formation processes) result in inaccurate representation of the Antarctic ozone hole. In this study we first quantify the sensitivity of Antarctic ozone depletion to mid-latitude planetary wave activity and South Pole temperature anomalies using a simple statistical regression model based on observations (Huck et al., 2005). We then apply the regression model coefficients to wave activity and temperature anomalies calculated from geopotential height, meridional wind velocity, and temperature fields from CCMs for which these data are available. By using only the year-to-year temperature and wave activity anomalies from a CCM run as the explanatory variables in our regression model, many of the complicating factors are avoided.
The required geopotential height, meridional wind velocity, and temperature fields don’t necessarily need to come from CCM model output. Another possibility is to use output from the large series of new climate change simulations with coupled ocean-atmosphere climate models from about 19 modelling centers that have been made available for the IPCC fourth assessment report (available through PCMDI). However, it has been shown that in the northern hemisphere when chemistry is switched off in UMETRAC, the model predicts a future increase in overall generation of planetary waves (Gillet et al., 2002), while UMETRAC with chemistry switched on shows a decrease in mid-latitude planetary wave activity over the period 1975 to 2020 (Austin et al., 2003). This suggests that feedbacks between chemistry and the dynamics may be important for understanding future changes in planetary wave activity and its potential influence on the evolution of Antarctic ozone depletion.
Ozone mini-holes, transient and localized events of
severe total ozone reduction, have been sometimes observed during
winter seasons in both hemispheres. In this study, dynamical
contributions to ozone mini-hole formation are examined for 22 events
in both hemispheres during the period from 1994 to 2003, on the basis
of daily distributions of ozone mixing ratio and isentropic potential
vorticity. In case of mini-holes in the Northern Hemisphere, about a
half of the dynamical reduction of total ozone is due to isentropic
transport of ozone-poor air, which consists of a poleward direction
around the tropopause and equatorward one in the middle stratosphere.
The remaining half is explained by thinning of ozone-rich isentropic
layers caused by vertical uplift of isentropes in the lower
stratosphere. On the other hand, mini-holes in the Southern Hemisphere
could be classified into two types: One is a similar type to that in
the Northern Hemisphere. The other is a type of mini-holes caused by
displacement and/or stretching of the Antarctic ozone hole toward
mid-latitudes, where the total ozone is reduced throughout the
stratosphere by equatorward transport of ozone-depleted polar air,
while the local uplift of isentropes due to associated cold temperature
advection also contributes to the reduction. Finally the interannual
change of their appearance frequency will be discussed.
New measurements from the Microwave Limb Sounder and
other sensors associated with NASA's "A-Train" are providing
unprecedented new prospects for characterizing the hydrological
processes in the upper troposphere, such as the structure and evolution
of cloud amount, cloud ice content, particle size, temperature, and
relative humidity. When combined with dynamical quantities, such data
present an altogether new opportunity to describe and understand cloud
microphysics and its connections to large-scale dynamics. We'll present
examples of using these new data resources, assessing their consistency
in describing upper-tropospheric cloud processes, and using this
information to help evaluate and improve model representations of
convection, clouds and microphysical processes in regional and global
atmospheric models.
Correlations of ozone (O3) and nitrous oxide (N2O) have been suggested as a tool for validating photochemical models and as a reference for estimating high latitude ozone loss. However, so far no data set is available that provides a good temporal coverage throughout all seasons. Here, we combine measurements from the Improved Limb Atmospheric Spectrometers (ILAS/ILAS-II) to derive an O3/N2O climatology for the high latitude regions in the Northern and Southern Hemisphere for each month of the year, thus providing a complete seasonal cycle. ILAS operated for 8 months in 1996/1997 and ILAS-II for 7 month in 2003. From the ILAS-II measurements the months that are not available from ILAS are covered. The ILAS/ILAS-II correlations of ozone vs. nitrous oxide are organized monthly in both hemispheres by partitioning these data into equal bins of altitude or potential temperature. The resulting families of curves allow to separate ozone changes due to photochemistry from those due to transport. The combined ILAS/ILAS-II data set corroborates earlier findings that the families of O3/N2O curves are separated and generally do not cross and further that the separation is much clearer for the potential temperature binning than for the altitude binning.
The seasonal cycle of O3/N2O distributions in the
Northern and Southern Hemisphere high latitudes is found to be rather
different. In the Southern Hemisphere O3/N2O distributions are
influenced by the strong chemical ozone loss in the Antarctic vortex
and a much longer duration of the polar vortex. In the Northern
Hemisphere, diabatic descent is much more pronounced. Solely during the
setup phase of the polar vortex the N2O/O3 distributions in the two
hemispheres are rather similar.The data set derived here provides a
tool which can be used for the validation of simulations of the
seasonal cylce of ozone high latitudes by photochemical models.
The tropical tropopause layer is a key region for
climate change. To explain past changes and to predict the future
development in the tropical upper troposphere and lower stratosphere we
have to understand the relevant processes in the tropical tropopause
layer. These include convection, cirrus formation, dehydration, aerosol
microphysics and chemical processes. The goal is to construct climate
prediction models with accurate descriptions of these processes. As a
first step the ability of current chemistry-climate models (CCMs) to
represent the observed tropical mean state and variability has to be
assessed. In this talk we present an intercomparison of climate change
simulations for the period 1980-1999 performed by several European
CCMs. The model results are validated against different observational
datasets.
Ensemble simulations of the GFDL coupled
chemistry-climate model are performed for the period of 1960-2100. The
simulations consist of two time-slice runs with fixed forcing at 1960
and 2000 respectively, three past transient runs (1960-2005) and three
future runs (1990-2100). This paper focuses on the dynamical
performance of the model with an emphasis on the polar stratosphere.
The polar stratosphere temperature is essential for PSC generation and
ozone depletion. It's well known that the polar stratosphere
temperature is controlled by the strength of the Brewer-Dobson
circulation, which is induced by wave dissipation (planetary and
gravity waves) in the stratosphere. Relationships between heat flux and
temperature correlations previously explored by Newman et al. (2001)
and Austin et al. (2003) are investigated and compared with
observations. The trends in wave activity will be investigated using
the future run results. Possible impacts of the simulated wave activity
trend on polar ozone recovery will be discussed.
Recent results from re-analysis and an ensemble of
simulations from a middle atmosphere general circulation model report
the emergence of a distinct stratospheric response to ENSO events,
characterized by a time-height evolution of anomalies. Specifically, a
polar warming has been found in late winter, in the lower stratosphere
of the Northern hemisphere. In order to evaluate the generality of
these results and quantify their implications for ozone climate
interactions, the role of variations in sea surface temperatures
associated with ENSO events is evaluated in relevant simulations
performed with the chemistry climate models available to CCMVal.
Results will be reported for the simulated anomalies of meteorological
and chemical fields, constructed from time series of an ENSO index
(such as sea surface temperature, SST, anomalies in the NINO 3 or
NINO3.4 region) and following the methodology used in Manzini et al
(2005).
Two related diagnostics are used to evaluate the
performance of NCAR CTM (MOZART3) and CCM (WACCM) in the extratropical
upper tropospheric and lower stratospheric region. The first one
compares the observed and modeled trace gas profiles (O3, CO and H2O)
in a relative altitude coordinate (altitude relative to the
tropopause). The second one compares the observed and modeled tracer
relationships. Together, they characterize the model’s ability in
reproducing observed chemical transition across the extratropical
tropopause. The ER-2 in situ observations of CO, ozone and water vapor
and satellite data from AIRS on Aqua and MLS on Aura are used in this
analysis. The results show that the chemical transition across the
extratropical tropopause is well reproduced in NCAR WACCM and MOZART3
driven by the ECMWF met field. The gradients of change across the
tropopause, however, are not as sharp as that in the observations.
Using the tracer-tracer correlation method, we show that the fully
coupled CCM (WACCM2) run compares better with the observations than the
results of MOZART3 driven by the ECMWF winds.
The GEOS-4 GCM has been coupled with the middle
atmospheric chemistry modules from a CTM and used for interactive
chemistry-climate simulations. It was run with a resolution of 2.5
degrees longitude by 2 degrees latitude, with 55 levels extending to
0.01hPa. Aspects of the validation of this CCM are presented by
Douglass et al. and Gupta et al. (at this meeting). This poster will
focus on simulations of the past period 1950-2000, which will be
continued into the future, 2000-2100. For the past, sea-surface
temperatures and ice (SST/ice) distributions will be imposed from
observations (Hadley) and established emissions scenarios for
radiatively active and ozone-depleting gases will be used. For the
future, these values will be merged with SST/ice data from a coupled
ocean-atmosphere model prediction and gaseous emissions will follow
established scenarios. The poster will focus on the experiment design,
the realism of results for the past period (by evaluation against
observations), and preliminary estimates of changes into the future. By
the time of the CCMVal meeting, the simulation should have reached
about 2050. We will examine in detail the impacts of strategies for
merging past (observed) SST/ice with future (predicted) values. Results
will focus on long-term changes in aspects of the dynamics (e.g.,
behaviour of polar vortices) and transport (e.g., age-of-air and
changes in long-lived species), as well as an assessment of the ozone
evolution.
In the stratosphere, the Northern and Southern Hemisphere annular
modes exhibit the main modes of dynamic variability. We will compare the autocorrelation function of the annular modes between observations and
MA-GCMs. We will determine the time scale of the annular modes as measured by the time (days) for the autocorrelation function to drop to
1/e (~0.378) (Baldwin et al., 2003). We suggest that this time-scale analysis is a key diagnostics for evaluating chemistry –climate models because it provides insight into the persistence of the stratospheric polar vortex.
The ozone QBO is known to precede the zonal wind QBO by
about a quarter-cycle in the lower stratosphere but to be retarded by
similar time-lag above about 15 hPa (28km). Chemistry-climate model
(CCM) of Meteorological Research Institute is used to investigate this
phase relationship between chemical species including ozone and
dynamical quantities. The CCM is T42L68 with vertical spacing of 500 m
in the QBO altitude range. Two runs were prepared for control realistic
QBO simulations. One with non-interactive ozone reproduces a 27-month
period; the other with interactive ozone yields a 31-month period. Two
experiment runs were made by switching on and off the ozone radiative
feedback for the non-interactive ozone run and the interactive ozone
run. The switched-on run prolonged the QBO period from 27 to 48 months,
while the switched-off run shortened it from 31 to 20 months. All the
runs are found to approximately reproduce the observed phase
relationship between the ozone and zonal wind QBOs, if the timescale is
normalized by their own QBO periods. Phase relations for other chemical
species are also investigated.
We have compared NASA/TIMED satellite data with 3D model
output to quantify the dynamical coupling between the stratosphere and
mesosphere. The data are from the Sounding of the Atmosphere with
Broadband Emission Radiometry (SABER) experiment on TIMED. The model is
a new high altitude version of the Navy’s operational forecast system,
called NOGAPS-ALPHA (Advanced Level Physics High Altitude). Of specific
interest are periods during polar winter when the middle atmosphere was
dynamically disturbed. During August, 2002 for example, SABER shows a
clear signatures of a mesospheric cooling in concert with a wave 1
minor stratospheric warming. NOGAPS-ALPHA hindcast analyses of the same
period show good agreement in the data in both the vertical and
temporal extent of the temperature perturbation. Somewhat suprsingly,
the vertical extent of the mesospheric cooling is shown to be less than
previously thought. A strong anticorrelation between stratospheric and
mesospheric temperatures persists up to .01 mb, but breaks down at
higher altitudes. This susgests that airglow measurements, typically
made at 83-90 km, may not be representative of the entire mesosphere.
Simulations with Rayleigh friction suggest the importance of gravity
wave drag in governing the vertical extent and depth of the mesospheric
cooling. Finally, an examination of 3D EP-flux vectors during August
2002, suggests that the planetary wave responsible for the
warming/cooling event originated from a horizontally localized region
of the troposphere.
Results from a large intercomparison of tropospheric
chemistry models will be presented, focussing on tropospheric ozone,
and in particular the influence of the stratosphere on tropospheric
ozone. The ACCENT intercomparison was split into two experiments: the
first considered 1860-2000-2100, and included changes in stratospheric
composition and climate; the second considered 2000-2030, with various
2030 emissions scenarios, and also the influence of climate change in
2030. Results indicate that there are at least two major influences of
climate change on global tropospheric ozone, comparable in magnitude to
the effects of emissions changes. The first is a negative feedback that
involves increased levels of water vapour: this mechanism generally
promotes ozone destruction, particularly in relatively unpolluted
regions. The second is a positive feedback driven by an enhanced
Brewer-Dobson circulation, and increased stratosphere-troposphere
exchange. Current chemistry-climate models exhibit a wide range of
behaviour, with no consensus on which of these two climate feedbacks
dominate globally. Model responses to emissions changes are more
consistent. The largest inter-model differences in response occur in
the tropics and the upper troposphere.
In the Arctic lower stratosphere, perturbations of ozone
concentration with the zonal wavenumber of one or two propagate
westward in summer. To examine their interannual variations, we
calculated ozone concentration in the Arctic stratosphere of the period
of 1978 - 2003 using the CCSR/NIES nudging chemical transport model.
The experiment shows that ozone perturbations are dominant around the
isentrope of 650K in every summer. The perturbations are due to the
large latitudinal gradient of ozone concentration formed by the NOx
chemistry. However, zonal wavenumber of ozone perturbations in each
summer varies depending on the dominant zonal wavenumber of potential
vorticity. Phasespeed of ozone perturbations is also similar to that of
PV anomaly. Thus, it is likely that dynamical processes are important
for interannual variations in ozone perturbations. In the height region
where ozone perturbations are dominant, mean ozone concentration in the
northern high latitude is positively correlated with the amplitude of
ozone perturbations in each summer. This result suggests that ozone
perturbations have a role to induce interannual variations of ozone
concentrations in the northern high latitude in summer. The relation of
the perturbations with QBO is also discussed.
Recently has emerged a need to improve the representation of shortwave (SW) radiative processes in CCMs (coupled climate-chemistry models), focusing on the following aspects:
the implementation of new UV bands is of relevance to the upper stratosphere for heating rate calculations (including O2 and O3 absorption); there should be consistency between heating rates and photolysis rates estimations; detailed clouds informations are essential for calculating surface UV quantities.
In order to contribute to these aspects results will be presented concerning new developments of radiation schemes within the ECHAM model family and comparison with more detailed schemes, in terms of solar heating rates, SW fluxes (upward, downward and net) and photolysis rates for standard atmospheric profiles with specified zenith angles and surface albedo for both clear-sky and cloudy-sky conditions.
A tropospheric/stratospheric chemistry scheme has been incorporated into the GFDL climate model which will be run for the period 1960 to 2100. The past runs (1960-2005) are forced with observed SSTs, the 'future' runs (1990-2100) are forced with GFDL model SSTs. The model has high spatial resolution: 2 deg. latitude by 2.5 deg. longitude with 48 levels extending to the upper mesosphere. A three member ensemble is being run, and at the time of writing, the past simulations have been completed and the future simulations have reached 2017, 2026 and 2032, each requiring approximately one week wallclock time per decade of simulation. For the past the impact of volcanic aerosol is clearly apparent in globally averaged ozone, leading to an increase following the Agung eruption, and a decrease following the Mt. Pinatubo eruption. After the El Chichon eruption, ozone is approximately constant for a few years due to a balance between chlorine and nitrogen impacts as well as complications due to the solar cycle. Minimum global ozone is attained in about the year 2015. Thereafter, slight recovery in the model values is noted out to the year 2032. Polar ozone by contrast shows faster recovery with a
minimum occurring near 2005. In all the diagnostics
computed to date, all three ensemble members show the same features.
However, for the future runs, volcanic aerosol has been kept constant
at background levels and any future volcanic eruptions over the period
until about 2025 could be significant for ozone.
The NIWA assimilated total column ozone data base, described on another paper submitted to CCMVal2005, has been used to generate a set of indicators describing attributes of the Antarctic ozone hole for the period 1979 to 2004, including:
i) daily measures of the area over Antarctica where ozone levels are below 150 DU, below 220 DU, more than 30% below 1979 to 1981 norms, and more than 50% below 1979 to 1981 norms,
ii) the date of disappearance of 150 DU ozone values, 220 DU ozone values, values 30% below 1979 to 1981 norms, and values 50% below 1979 to 1981 norms, for each year,
iii) daily minimum total column ozone values over Antarctica, and
iv) daily values of the ozone mass deficit based on a O3<220 DU threshold.
Similar indices will be calculated based on output from all available CCM runs contributing to CCMVal (existing runs and REF1, REF2, SCN1 and SCN2 if available) and compared with those based on measurements to assess the extent to which these models can track past changes in the behaviour of the ozone hole.
We will also examine the CCM based indicators for signs of Antarctic ozone hole recovery over the coming decades. Recent publications (Alvarez-Madrigal et al., 2005; Bodeker et al., 2005) have shown that the mean size of the Antarctic ozone hole during the month of November has decreased significantly over the period 1999 to 2003. In a preliminary analyses, based on ozone profiles from an earlier UMETRAC run (Struthers et al., 2004), monthly, seasonal and annual averages of partial ozone columns over 4 altitude ranges (100-200 hPa, 50-100 hPa, 10-50 hPa and 50-150 hPa), and over 4 latitude ranges (60-70S, 70-80S, 80-90S and 60-90S) have been calculated and examined for which most clearly signals recovery of the Antarctic ozone hole. It was found that the region showing the largest change in trend before and after 2000, was November mean partial columns from 50-150 hPa between 60S and 70S, consistent with the two studies cited above. We will extend this analysis to include as many as possible of the CCM run results available through CCMVal, and subject the model output to more robust statistical analyses of the change in trends (Reinsel et al., 2002), to see if this result is consistent across all of the models. Finally we will examine 50-150 hPa November mean partial ozone columns from Syowa (69.0S, 39.58E) to see if indeed a statistically significant change in trends is observable.
Two assimilated data bases have been generated at NIWA for validation of chemistry-climate models:
1) The NIWA assimilated total column ozone data base (Bodeker et al., 2005), and
2) The NIWA combined vertical trace gas profile data base.
The NIWA assimilated total column ozone data base combines satellite-based total column ozone measurements from 4 Total Ozone Mapping Spectrometer (TOMS) instruments, 3 different retrievals from the Global Ozone Monitoring Experiment (GOME), and data from 4 Solar Backscatter Ultra-Violet (SBUV) instruments. Comparisons with the global ground-based Dobson spectrophotometer network are used to remove offsets and drifts between the different data sets to produce a global homogeneous data set that combines the advantages of good spatial coverage of satellite data with good long-term stability of ground-based measurements. This data set, available from November 1978 to December 2004 provides daily total column ozone fields which can be used to validate chemistry-climate model ozone fields. The data have also been regridded by equivalent latitude based on potential vorticity fields on the 330, 450, 550 and 650 K isentropic levels.
The NIWA combined trace gas profile data base is still in development. At present it includes solar occultation measurements of:
• O3, NO2, H2O, aerosol extinctions at 385, 453, 525, 1020 nm, aerosol surface area density and aerosol effective radius from the Stratospheric Aerosol and Gas Experiment (SAGE II) instrument,
• measurements of O3, and aerosol extinction at 352.3, 441.6, 601.4, 781, 921 and 1060 nm from the Polar Ozone and Aerosol Measurement (POAM II) instrument,
• O3, NO2, H2O, aerosol extinctions at 354, 442.2, 603, 779, 922.4, and 1018 nm from the Polar Ozone and Aerosol Measurement (POAM III) instrument.
The data base is provided in three different formats:
1) Indexed by latitude, longitude, altitude and time.
2) Indexed by latitude, longitude, pressure and time.
3) Indexed by equivalent latitude (calculated on the 300, 315, 330, 350, 400, 450, 550 and 650 K isentropic levels), potential temperature and time.
In the near future we plan to add to the data base measurements from the Halogen Occultation Experiment (HALOE) and ozone measurements from ozonesondes for a number of locations. This data base should also be useful for validation of chemistry-climate model output.
The paper will discuss the construction of both data bases in greater detail, and will provide some examples of how they can be used to validate chemistry-climate model output.
Results of a so-called transient simulation with the CCM
E39/C have been investigated which covers the 60-year time period
between 1960 and 2020. Natural and anthropogenic forcing as precise as
possible has been considered: Based on observations and model
estimates, sea surface temperatures (SSTs) and sea-ice cover are
prescribed. The increase of greenhouse gas concentrations as well as of
chloroflurocarbons and nitrogen oxide emissions is taken into account.
In addition, the 11-year solar cycle is taken into account for the
calculation of heating rates and photolysis of chemical species. The
quasi-biennial oscillation (QBO) is introduced using observed lower
stratospheric tropical zonal winds by a linear regression method, i.e.,
so-called nudging. The three major volcanic eruptions which happened
during that time (Agung, 1963; El Chichon, 1982; Pinatubo, 1991) are
considered, i.e., additional heating rates and sulphate aerosol
surfaces are used which are derived from model estimates and
measurements, respectively. The general agreement between model results
and observations (1960 until 2000) is quite satisfactory, e.g., the
mean state of dynamic and chemical values and parameters as well as
seasonal and inter-annual variability; long-term changes are reasonable
reproduced by E39/C (Dameris et al., 2005). This is a solid basis for
an assessment of future changes in atmospheric composition and climate,
e.g., the behaviour of the stratospheric ozone layer within the coming
years. The investigation focuses on the beginning of ozone recovery.
The GMES PROMOTE ozone monitoring service is partly a response to the demand for consistent 3D ozone analyses and ozone records. Therefore, the focus is to derive a multi-year ozone record by combining satellite observations of ozone and related species, meteorological reanalyses and a CTM. Hence, all MIPAS and SCIAMACHY limb observations are assimilated using the chemistry-transport model ROSE/DLR in order to derive consistent global chemical analysis of the stratosphere. MIPAS observations of O3, H2O, HNO3, CH4, N2O and NO2 are considered as well as O3, NO2 and BrO of SCIAMACHY. Sequential assimilation is performed using an optimum interpolation scheme with error propagation. Optimized assimilation parameters are derived using Chi2 diagnostics. Results are analyzed using observation minus first-guess error (OMF) statistics and are additionally compared to UARS/HALOE data. The resulting ozone record will contribute in fulfilling the general objective to better evaluate the three-dimensional representation of ozone in CCMs in time and space.
The use of data services of various variables and constituents derived from satellite data, ground-based measurements, and models is extremely important for CCM validation to better understand the representation of various physical and chemical processes and long-term changes of the atmosphere. An improved understanding of these processes and, more generally, of the interaction between chemistry and climate is needed if credible predictions of the future levels of stratospheric ozone and surface UV radiation are to be made
SPARC CCMVal has joined in the ESA Project PROMOTE consortium as Core User for the evaluation of products related to Stratospheric Ozone.
Atmospheric concentrations of radiatively active
(greenhouse) gases and ozone destroying compounds have increased
significantly in the last 50 years, but the impact of these changes has
not been fully assessed. To investigate this question we use the Whole
Atmosphere Community Climate Model, v.3 (WACCM3) to study trends in
atmospheric composition and temperature over the second half of the
20th century. WACCM3 includes a complete chemical scheme for the middle
atmosphere that is fully interactive with radiation and dynamics. The
model is run with specified boundary conditions and external forcings
for greenhouse gases, chlorine and bromine compounds, sea-surface
temperatures, volcanic aerosols, and solar inputs taken from
observations for 1950-2000. Linear trends for temperature and key
chemical species (ozone, water vapor, etc.) are calculated and compared
with trends obtained from ground-based and satellite observations. The
statistical significance of computed trends, and their stability, is
assessed. Wherever possible, physical mechanisms responsible for
calculated trends are identified.
In ECHAM5/MESSy a comprehensive set of modules (Modular Earth Submodel System) for atmospheric chemistry in the troposphere, stratosphere and mesosphere has been interactively coupled to the general circulation model ECHAM5. ECHAM5/MESSy can be used with nudged meteorology or in free running mode.
We compare results of an integration applying the middle atmosphere configuration in a relatively high resolution (T42, 90 layers, resolution in UTLS about 600m) with UARS, ENVISAT and TOMS data. We focus on ozone depletion at high latitudes (including PSC-chemistry), the relations of ozone and long-lived source gases at the polar and subtropical barriers (including standard tests like PDFs), the Brewer Dobson circulation and exchange at the tropopause. We present also examples of how photolysis of NO2 and O3 in the lower stratosphere is effected by tropospheric clouds, based on the online photolysis scheme of the model.
Motivated by the need to study the climatic impact of
aerosol-related cirrus cloud changes, a physically-based
parameterization scheme of ice initiation and initial growth of ice
crystals in young cirrus clouds has been developed. The scheme tracks
the number density and size of nucleated ice crystals as a function of
vertical wind speed, temperature, ice saturation ratio, aerosol number
size distributions, and preexisting cloud ice, allowing for competition
between heterogeneous ice nuclei and liquid aerosol particles during
freezing. Its implementation in a general circulation model is briefly
outlined, and examples from pure homogeneous freezing and idealized
heterogeneous ice nucleation simulations are presented. This new scheme
establishes a flexible framework for a comprehensive assessment of
indirect aerosol effects on and properties of cirrus clouds in global
climate, chemistry transport, and weather forecast models.
The Modular Earth Submodel System (MESSy) developed at the Max-Planck-Institute for Chemistry in Mainz, Germany, has been used to investigate the influence of different model resolutions on the distribution of stratospheric tracers like N2O as well as chemically active substances like Ozone and NO2. MESSy consist of a atmospheric base model (ECHAM) and different submodules, combined through an online interface structure which controls the data exchange between the base model and the submodels and their interconnections. It provides a possibility to study feedback mechanisms between chemical, physical and biological processes (by two-way coupling) and includes through switchable submodules a very easy way to adapt the model on the own research interests.
We use version 0.9 of MESSy in the 39 layer version
which covers the atmosphere from the surface up to 1 Pa. As our
interest focus are the examination of the middle atmosphere only that
submodules of MESSy has been used which are necessary to simulate the
stratospheric and mesospheric processes: MECCA for chemistry, CONVECT
for convection, JVAL as photolysis modul, PSC, H2O and OFFLEM for the
ground emissions. Thereby we simulate one year and three years periods.
The results of the model runs with different model resolutions (T31,T42
and T63) show, almost independent of the resolution, realistic
distribution of the examined substances. The ozone loss inside the
Antarctic vortex is simulated with MESSy version 0.9, but the extent of
ozone depletion with respect of the size of the ozone hole as well as
the strength of the ozone loss is slightly underestimated.
The Network for Detection of Stratospheric Change (NDSC) is a set of remote-sounding research stations for observing and understanding the physical and chemical state of the stratosphere. The NDSC is a well established global network that provides high quality time series of trace gas measurements. Long-term time series of measurements are an important resource for model validation, in particular for testing model response to changing forcings over decadal time scales.
As part of the NDSC, measurements of NO2 (nitrogen dioxide) have been made using zenith-sky UV-visible absorption spectroscopy at Lauder, New Zealand (45S, 170E), and at Arrival Heights, Antarctica (78S, 167E). The observations span a time period of 25 and 23 years, respectively. Also as part of the NDSC, measurements of HNO3 from direct-sun infrared absorption spectra using high resolution Fourier transform spectrometers have been made at these two sites, spanning 14 and 8 years, respectively.
In a previous analysis, trends in NO2 derived from a 45 year integration of a UMETRAC run were compared with the long-term NO2 time series at both sites [Struthers et al., 2004]. The observed trends in NO2 at both sites exceed the modelled trends in N2O, the primary source gas for stratospheric NO2. This suggests that the processes driving the NO2 trend are not solely dictated by changes in N2O but are coupled to global atmospheric change, either chemically or dynamically or both. The model results also suggest an anti-correlation between the trend in NO2 and the trend in HNO3 at Lauder. This study will compare measured trends in NO2 and HNO3 with results from the UMETRAC model to demonstrate the use of NDSC ground based trace gas measurements for validation of CCMs.
References
Struthers, H., K. Kreher, J. Austin, R. Schofield, G.E. Bodeker, P.V. Johnston, H. Shiona, and A. Thomas, Past and future simulations of NO2 from a coupled chemistry-climate model in comparison with observations, Atmospheric Chemistry and Physics, 4, 2227–2239, 2004.
Rinsland, C.P. et al., Long-term trends of inorganic chlorine from ground-based infrared solar spectra: Past increases and evidence for stabilization, J. Geophys. Res., 108(D8), 4252, doi: 10.1029/2002JD003001, 2003.
The prediction of Arctic polar ozone with chemistry climate models (CCMs) is
problematic given the complex interactions of climate change, stratospheric
temperature and ozone depletion. Additionally, amongst an ensemble of CCMs,
differences in numerics, boundary conditions,
detail of included processes, and extent of
parameterisation have lead to a wide range of predictions for the timing of
possible ozone recovery in the Arctic stratosphere.
We examine results from the 2015 time slice experiment performed with
the coupled chemistry climate model E39C, which showed the earliest
signs of ozone recovery in the last WMO ozone assessment.
For a more realistic representation of stratospheric
chemistry, we employ the chemistry transport model CLaMS in an off-line mode on
seasonal polar winter meteorology from the 20-year CCM time slice.
Compared to the chemistry climate model, the chemistry transport simulations
show much lower spring Arctic ozone columns. For the winter with the lowest
Arctic ozone column, the difference amounts to 40 DU, lowering the predicted minimum
Arctic ozone column from 290 to 250 DU.
With this updated calculation, there is no indication for ozone recovery by 2015.
Calculations of chemical ozone loss with the tracer-tracer
correlation technique in both the CCM and the CTM show that chemical ozone loss
is greatly underestimated by the CCM: while the vortex average 90-day chemical
ozone loss in the CCM is 36 DU, the CTM simulates 76 DU. Of these, 30% are due
to bromine catalytic cycles which are not implemented in the CCM.
The simulations show that around 2015 northern hemispheric ozone
depletion may be as severe as observed in medium to cold 1990s winters.
The disparate results for ozone between the CCM and CTM simulations for the same
meteorological and chemical boundary conditions call for a thorough assessment
of the implementation of chemical processes.
Specifically, within the framework of process-oriented validation of CCMs,
seasonal chemistry transport model calculations should be used to validate
ozone loss in cold Arctic winter simulations.
The tracer-tracer correlation technique has been widely
employed to infer chemical ozone loss from observational data. Yet,
its applicability to global circulation model data has been
disputed. We report the successful application of the tracer-tracer
correlation technique on climate model data to derive chemical ozone
loss. By comparison with chemical ozone loss calculated with the
passive tracer method in a Lagrangian transport model (CLaMS), we
quantify effects of anomalous internal mixing and cross vortex
boundary mixing on ozone loss derived with the correlation technique.
As a test case, we use predicted results for a 2015 cold Arctic winter
from the coupled chemistry climate model E39C, where features typical
for polar winter conditions are simulated, for example, sufficient
polar stratospheric cloud formation potential, denitrification and
dehydration, and intermittent and final stratospheric warming events.
For calculations using the correlation technique we find an average
underestimation of ozone loss with respect to the passive tracer
method on the order of -100 ppm in the lower stratosphere, or 5-10% of
typical cold-winter chemical ozone loss. We demonstrate a technique
for excluding effects of cross vortex boundary mixing on the tracer
correlation and separately calculate the overestimation of ozone loss
introduced by internal anomalous mixing as a result of strong
differential descent within the polar vortex. Internal anomalous
mixing alone would impose an error on the order of +100 ppm on the
ozone loss estimate; even for a well-isolated (90% of air mass
identity preserving) vortex, cross vortex boundary mixing introduces
an error on the order of -200 ppm into the tracer-tracer correlation
based ozone loss analysis. The methodology developed here allows a
validation of the tracer-tracer correlation technique on chemistry
climate model data. More importantly, we are now in a position to
quantify chemical ozone loss for model simulations where passive ozone
is not available. Specifically, for the purpose of process-oriented
validation of CCMs, we propose the inclusion of tracer-tracer
correlation derived ozone loss as a quantifier of polar winter
chemical ozone l
We compare NOGAPS-ALPHA simulations of stratospheric
ozone with in-situ and satellite-based ozone profile measurements to
evaluate the performance of four different linearized ozone
photochemistry parameterizations commonly used in both research
coupled-chemistry climate models and operational weather forecast
systems. NOGAPS-ALPHA is a prototype high-altitude version of the
general circulation model within the Navy's Operational Global
Atmospheric Prediction System. This Eulerian spectral model currently
extends from the surface up to 0.005 hPa (~85 km). Ozone has recently
been added to NOGAPS-ALPHA as a prognostic variable with the ultimate
goal of improving both assimilation of satellite radiances and
computation of stratospheric heating rates in the operational forecast
system. NOGAPS-ALPHA currently parameterizes gas-phase stratospheric
ozone photochemistry in terms of local deviations in the model ozone
mixing ratio, temperature, and overlying ozone column amounts from a
prescribed climatological mean state using a linearized Taylor series
expansion of odd-oxygen production and loss rates based on coefficients
derived from offline models containing full stratospheric ozone
photochemistry. NOGAPS-ALPHA ozone simulations using four different
sets of these derived photochemical coefficients (currently used by,
e.g., the ECMWF IFS, NCEP GFS, and NASA GISS climate model) are
compared with in-situ lidar observations and satellite-based ozone
profile measurements obtained during the second SAGE III Ozone Loss and
Validation Experiment (SOLVE2) of January-February 2003 and the recent
Polar Aura Validation Experiment (PAVE) of January-February 2005.
Overall, the best results are obtained using photochemical coefficients
derived from the NRL CHEM2D middle atmosphere photochemical transport
model. We also examine two separate methods to parameterize the
additional effects of wintertime ozone loss via heterogeneous chemistry
on polar stratospheric clouds.
The U.K. Chemistry and Aerosols project aims to couple
state-of-the-art whole-atmosphere chemistry and aerosol modules to the
"New Dynamics" Unified Model, creating a research tool for
climate-chemistry studies for the U.K. and international community.
Chemistry is specified in a flexible manner, allowing for different
schemes to be easily implemented in the model. Here we present first
results obtained with the middle-atmosphere version of the model with a
model to at 85 km, covered with 60 model levels. A comparison versus
various observations reveals strengths and remaining problems.
The Whole Atmosphere Community Climate Model (WACCM)
extends from the surface to 140 km and is based on the physics package
of the NCAR Community Climate Model, version 3 (CCM3). Several
multi-decadal climate simulations with observed SSTs and GHGs were
recently run to determine the extent to which observed trends are
reproduced in the model. This work will compare output from these runs
to satellite data and gridded meteorological analyses in various ways.
Monthly means in temperature and chemical species will be compared to
climatological means derived from satellite occultation measurements.
Seasonal and interannual variability in the model will also be compared
to variability observed in the occultation data (for chemical species)
and variability in Met Office analyzed wind and temperature fields (for
model dynamics). In addition to these general intercomparisons, we will
examine the structure and variability of stratospheric cyclones and
anticyclones in WACCM, look for “low-ozone pockets” in the
anticyclones, and look at variability in polar processes.
Bromine plays an important role in stratospheric ozone
depletion. In spite of its importance, in the past
bromine chemistry has not received much attention in the
CCM community. With the availability of global
observations of stratospheric bromine monoxide (BrO)
profiles from the SCIAMACHY instrument onboad ENVISAT
there is now the opportunity to test and validate
modeled bromine chemistry. Moreover, the BrO
observations provide indications for the total
stratospheric bromine loading. We will present a first
2-year "climatology" of BrO profiles from SCIAMACHY. A
comparison with constrained model calculations indicate
that the BrO observations are generally consistent with
our current understanding of the stratospheric bromine
chemistry and a total bromine loading of 18+-3ppt for the
period 2002 to 2004.
Vertical profiles of BrO are retrieved from SCIAMACHY limb
scatter data over much of the globe. Comparisons with
balloon-borne SAOZ-BrO and LPMA/DOAS measurements of BrO
indicate biases in the SCIAMACHY retrievals of less than 20%.
We find best agreement with the observed vertical and
latitudinal distribution of BrO for model results that include
a 5 to 7 pptv contribution to Bry primarily from the breakdown of
very short lived bromocarbons. This presentation will emphasize
the retrieval method and data validation efforts.
Models attempt to represent as much of our knowledge of how the atmosphere works as is possible within computer constraints. Both chemistry-transport models (CTMs) and chemistry-climate models (CCMs) are evaluated by comparison to a suite of data selected to test their representations of important processes. The next step in model evaluation is its ability to reproduce the multi-decadal response of ozone and other constituents to the natural and man-made forcings that have occurred during the time of the measurement record. We compare our CTM simulation of the last 30 years with the long-term data records for ozone column and profile measured by both satellite and
ground-based instruments. We further compare to measurements of other key species from the Network for Detection of Stratospheric Change. The CTM simulation uses the same chemistry package that is integrated into our CCM. Thus its evaluation is an important first step in understanding the coupled CCM.
The CTM results for ozone indicate the need for some caution in interpreting evidence for early indications of ozone trend slow-down. Interannual variability can easily mask or enhance a perturbation like that from Pinatubo. The combination of El Chichon and Pinatubo can be confused with solar cycle in time-series analyses of data. Together these effects add uncertainty to detecting the difference between a continuing trend and a slow-down of trend.
The simulation indicates that Pinatubo should have had a larger effect on ozone in the southern mid-latitudes than in the northern hemisphere. Ozone data from the Lauder station and from TOMS show no apparent effect of Pinatubo despite the clear effect on nitrogen dioxide. We have run our CTM with and without the volcanic aerosols to isolate the model's response from interannual variability of the dynamics. These studies indicate that there is some possibility that interannual variability in dynamics could mask the Pinatubo effect.
The French MOCAGE CTM was mainly developped for air pollution forecasting and now, is operationally used at the Meteo-France.It describes both stratospheric and tropospheric chemistry with associated physical processes such as surface emissions and dry deposition, convection with associated scavenging. The top of this model is located around 35 hPa, therefore, a climatic version has been developped in the aim to make the coupling with the ARPEGE-climat GCM. This climatic version of MOCAGE will be presented and assumptions used to reduce time-consumption will be described.
First, the MOCAGE-Climat CTM has been run in an off-line mode, using the ECMWF operational meteorological analyses during the period 2001-2004.
The resulting chemical species distributions will be evaluated, especially against long-term observations (TOMS, MOZAIC, O3-sondes, etc). We will look at both climatological means but also will focused on some periods of interest such as the dipole ozone hole event occuring in September 2002.A description of the ARPEGE GCM will be also given and the coupling technique be exposed.
After a future ozone prediction calculation using the CCSR/NIES CCM with T21 horizontal resolution by Nagashima et al. [2002], we developed a CCM with T42 horizontal resolution, which includes the gravity wave parameterization by Hines [1997] and bromine chemistry. The model also includes the heterogeneous reaction scheme on STS, NAT, and ICE by Sessler et al. [1996]. The chemical kinetics and photochemical data were updated with JPL-2003. In the model, CFC-11, -12, -113, CCl4, CH3CCl3, CH3Cl, HCFC-22, Halon-1211, -1301, and CH3Br are considered as Cly and Bry source gases. Calculations are conducted following the forcing requirements of CCMVal for the past and future atmosphere. In the new version of the CCM, the cold bias in the winter polar lower stratosphere is improved with a zonal mean minimum temperature of 180 K. The model outputs of dynamics, radiation, and chemistry are shown. The results and analyses of the calculations with the REF1 and REF2 scenarios also will be shown.
Three ensemble simulations of the GFDL coupled chemistry-climate model AMTRAC (Atmospheric Model with TRansport And Chemistry) will be run for the period 1960-2100. At the time of writing the simulations have reached 2017, 2026 and 2032 and are scheduled to complete during September. During the past, water vapour in the stratosphere increased episodically. Changes in tropical tropopause temperature are shown to have a controlling influence on the concentration of H2O entering the stratosphere, in agreement with other published work. Transport from the region then occurs over the timescale of the age of air, which in the model approaches 5 years in the upper stratosphere. Superimposed on these impacts is the steady effect of methane oxidation from the model photochemistry. During volcanic eruptions, the model radiative processes led to enhanced tropical tropopause temperatures and enhanced water vapour. On the timescale of the age of air this leads to a steady increase in water in the upper stratosphere which exceeds that of methane oxidation. The water vapour concentration therefore varies according to the time since the last major eruption. For the period 1981 to 1997 water vapour increased by about 15%, due in part to the two eruptions El Chichon and Mt. Pinatubo, thereby explaining in part the Boulder data. After 1997, water vapour decreased (as observed) back to the long term trend from methane oxidation.
For the future, ozone is simulated to recover and this leads to an increase in tropical tropopause temperature and a further increase in water vapour. The net water increase is about twice that of the methane oxidation term and moreover is not episodic as the volcanic aerosol was constant. These changes themselves have impacts on another tracer, which has generally been supposed to be constant: the stratospheric mean age of air. Model simulations of the age of air show a clear decrease of about 20% from 1960 to 2000 and a slight increase thereafter (until 2032). This shows the need to understand the behaviour of tracers in a coupled system in the context of other processes such as ozone depletion and volcanic eruptions.
SCOUT-O3 is a European Integrated Project bringing together over 100 scientists in 59 partner institutions. The project’s central aim is to provide forecasts of the evolution of the stratospheric coupled chemistry/climate system and impact on atmospheric ozone and surface UV. Ten climate-chemistry modelling groups are participating and model validation will be coordinated with the worldwide SPARC CCMVal activity. Model results from the European groups, and one model group from outside Europe (CCSR/NIES), have been obtained to assess the abilities and deficiencies of current models.
Of the 11 CCMs, 9 have already been used in multi-year transient simulations which include the well observed 20-year period 1980-1999. In this study output from the models for this period is compared and evaluated against various observational datasets. The aim in the first part of the ongoing study was to update the Austin et al. (2003) and Pawson et al. (2000) assessments (e.g., global mean of temperature, monthly mean of total ozone, decadal mean of total ozone). In addition, several diagnostics according to the CCMVal table (Eyring et al., 2005) have been applied, with a focus to evaluate transport and dynamics in CCMs.
Austin, J., et al., 2003: Uncertainties and assessments of chemistry-climate models of the stratosphere. Atmos. Chem. Phys., 3, 1-27.
Eyring V., et al., 2005: A strategy for process-oriented validation of coupled chemistry-climate models. Bull. Am. Meteorol. Soc., in press.
Pawson, S., et al., 2000: The GCM-Reality Intercomparison Project for SPARC: Scientific Issues and Initial Results. Bull. Am. Meteorol. Soc., 81, 781-796.
Ozone distributions in the UT/LS region are sensitive to
atmospheric transport processes and can therefore be used to test model
representations of near-tropopause transport. We compare UT/LS monthly
mean, tropopause-referenced ozone climatologies constructed from
multiple years of ozonesonde and SAGE II satellite data to ozone mixing
ratio and vertical gradient distributions generated by the NASA Global
Modeling Initiative (GMI) CTM. The model includes a full description of
stratospheric and tropospheric physicochemical processes extending from
the surface to above the stratopause, with a vertical resolution near
the tropopause of approximately1 km. The CTM was integrated for 5 years
post-spinup, and was driven by meteorological data from a run of the
NASA GEOS-4 AGCM which was constrained by 1994-1998 sea surface
temperatures. The five years of model output provide an initial
indication of model long-term average O3 vertical profiles and
gradients as well as interannual variability. Initial results suggest
improved O3 STE compared to previous GMI simulations driven by NCAR
MACCM3 meteorological data. Though significantly improved, model
extratropical STE still appears to be higher than suggested by the O3
observations. These comparisons demonstrate the utility of the
climatologies for evaluation of near tropopause transport of O3 in
global models of the stratosphere and troposphere.
One result of the November 2003 “Process-oriented
validation of coupled chemistry-climate models” workshop is
identification of a series of diagnostic tests that are used to
evaluate model performance. We are using those diagnostics to evaluate
a 20-year “time-slice” simulation for year 2000. This evaluation will
provide insight into the applicability of the diagnostics in the
context of the interannual variability of the simulation, and will
provide a basis for understanding the difference between simulations
with boundary conditions for 1980 and 2000. The simulation uses a
version of the GEOS-4 general circulation model coupled with the
stratospheric chemistry previously used in the GSFC off-line chemistry
and transport model (CTM). We also intend to include diagnostics that
have been developed since the 2003 workshop for evaluation of our CTM.
For example, in addition to evaluating the vertical propagation of the
tropical tape recorder we will compare the simulated horizontal
structure of the water vapor tape recorder with that observed by the
Aura Microwave Limb Sounder (MLS). Addition of the horizontal
diagnostic of the tape recorder signal provides information needed to
evaluate overall mass transport in the simulation. We will also apply a
novel diagnostic of the seasonal mass transport in the lowermost
stratosphere as derived from the difference between ground-based column
measurements of HCl and the stratospheric column as determined from HCl
profiles observed by the Halogen Occultati
New satellite data sets from the AIRS instrument are examined to develop a more detailed climatology of ozone in the upper troposphere and lower stratosphere, and the results for ozone and water vapor compared to chemical transport model simulations and dynamic fields. Examples of validation will be presented. Specific processes such as stratospheric intrusions, the Asian monsoon complex and stratosphere troposphere exchange are explored to try to understand how a more complete picture of the basic chemical distribution (Ozone and Water Vapor) of the UT/LS can supplement current stratospheric and in-situ observations. The prospects for using this data to evaluate key model transport processes in the UT/LS region are discussed.
We present measurements of NO, NOy, O3, N2O and other long-lived tracers within the lowermost stratosphere (LMS) over Europe obtained during the SPURT project. The measurements cover each of the four seasons during two years between November 2001 and July 2003, and probe the entire latitude/altitude of the LMS: from 5°N to 85°N equivalent latitude, and from 290 to 375 K potential temperature. An overview of typical distributions of these trace gases will be presented in the equivalent latitude - potential temperature framework. The results are discussed in the light of atmospheric
transport processes determining the tracer distributions
in the lowermost stratosphere. A special focus is given on the observed
seasonality in tracer-tracer correlations and we further show that the
influence of the troposphere is most prominent in a layer extending up
to 25 K potential temperature above the local tropopause.
Modelling of trace gas distributions in the UTLS-region provides a
challenging task since the underlying transport processes span the range from subgrid to large scale dynamics.
Here we present the results of simulations with the new chemistry circulation model ECHAM5/MESSy with 90 levels from the surface up to 80km providing a vertical resolution of about 600m in the UTLS-region.
A detailed comparison with in-situ data obtained during SPURT shows a good agreement between the model and the observations.
Although transport from the upper tropical tropopause into the extratropical lowermost stratosphere seems to be overestimated by the model, the distribution of CO is captured well.
In particular the model is able to reproduce the structure of the vertical CO-profiles relative to the tropopause as well as slope changes in the correlation between ozone and CO. Both findings indicate that the underlying transport processes across the
extratropical tropopause qualitatively are captured right by the model, which is crucial for a better quantification of stratosphere-troposphere-exchange.
It is well known that there are large interannual
variations in the timing of the Arctic vortex breakup. This makes an
interannual variation in N2O distribution in the lower stratosphere.
Holton [1986] and Mahlman et al. [1986] showed a clear relationship
among the annual mean meridional tracer slope, advection and horizontal
diffusion. In this paper, the year-to-year lower stratospheric N2O
distributions are analyzed in the early and late vortex breakup years
with the probability distribution function (PDF) technique. The
isentropic diffusion coefficient (Kyy) and the vertical advection are
calculated to quantify the effects of horizontal diffusion and vertical
advection on the N2O concentration. The data are from a Center for
Climate System Research/National Institute for Environmental Studies
(CCSR/NIES) nudging chemical transport model (CTM), which used
NCAR/NCEP reanalysis and CIRA data for the nudging processes, for 45
years from 1958 to 2002. Results showed that there are clear
differences in the N2O distribution in spring and early summer between
the early vortex breakup years and the late years. In the early breakup
years, the N2O concentration in the north of 45oN at the 600 K
isentropic surface is lower than that in the late years. The N2O shows
inhomogeneous distribution after the vortex breakup, which is related
to the large Kyy and strong vertical advection in the mid- and high-
latitude lower stratosphere. In the late breakup years, the N2O
distribution is more uniform than that in the early years, with small
Kyy and weak vertical advection. Studies also show that the vortex
breakup date defined dynamically is not the only factor for the above
N2O difference.
A transport equation based on mass-weighted isentropic zonal means applies to diagnosis for the meridional constituent transports of the MRI and NIES nudging chemical transport model (CTM). The transport equation offers some conceptual and computational advantages over conventional methods such as TEM.
This diagnostic tool can simply and exactly represent the eddy transport terms. Adiabatic diffusion is shown to be parallel to isentropic surfaces. Another advantage lies in the mean-meridional transport. Although it is almost similar to the TEM, significant differences can be found near the Antarctic polar vortex due to nongeostrophic effects. Furthermore, the isentropic diagnosis expresses a strong equatorward flux near the lower boundary, while the TEM hardly does this because of inadequate treatment of the lower boundary conditions. The isentropic diagnosis enables us to perform accurate budget analysis
The present diagnosis offers accurate validation of the mean and eddy meridional constituent transport of the CTM or CCM. We study the CTM driven by nudged GCM system for past ozone reanalysis. Changing meteorological variables to be assimilated into GCM affect both the mean and eddy transport, and reanalyzed ozone distribution. When temperature is nudged into a GCM in addition to wind fields, the mean-meridional ozone transport becomes stronger in the lower stratosphere and weaker in the troposphere, since the temperature nudging brings an external forcing to the equation of motion and break down the dynamical consistency in relation to the systematic cold bias in the GCM. Temperature nudging also makes wave breaking and eddy mixing stronger in the mid-latitude lower stratosphere.
We also compare diagnostic results of different CTMs, MRI and NIES. The eddy transport is very different between MRI and NIES model. Difference in chemical processes is considered to influence the strength of the eddy transport. The NIES model has larger spatial and temporal variations of chemical ozone tendency, compared to the MRI model, which provides more active equatorward eddy transport particularly in the lower stratosphere. As a result, the eddy transport is more dominant contributor to ozone distribution in the NIES model than in the MRI model. In addition, the eddy transport, rather than the mean transport, is very sensitive to model resolutions. Increasing horizontal resolution reduces eddy transport and results in a more realistic constituent simulation.
Validating the representation of tropical deep convection in climate
models is an important issue of general importance. From a
stratospheric chemistry point of view, tropical deep convection is
of importance as it controls the transport of short-lived source
gases into the stratosphere (e.g. WMO 2002, Chapter 2). Here we
present calculations of tropical mean profiles of idealized
short-lived source gases with a range of different lifetimes, which
can be compared with observations of source-gases such as CH3I and
CHBr3. The calculations are performed with two different
1-dimensional models: (a) a prognostic radiative-convective model
that uses a sophisticated convection parameterization and (b) the
diagnostic model of Folkins and Martin (2005) that is constrained by
observed tropical mean temperature and humidity profiles. The
results of the two models are qualitatively in mutual agreement, but
are rather different from previously published model results such as
those in WMO 2002, Chapter 2. This highlights the uncertainties in
the representation of convective transport in current models and the
need for further validation.
We report results from multiple linear regression analysis of total ozone and temperature long-term observations (TOMS/SBUV, NCEP-reanalyses, sondes and lidars), as well as 1960 to 1999 simulations by two versions of the ECHAM4/CHEM chemistry-climate model (ECHAM4.L39(DLR)/CHEM or MAECHAM/CHEM). The focus is on the comparison of observed and modelled interannual variations, due to trends, QBO, 11-year solar cycle, intensity of the polar vortices, tropospheric meteorological parameters and other factors. The model runs are transient experiments, where observed sea surface temperatures, increasing source gases (CO2, CFCs, CH4, N2O...), 11-year solar cycle, volcanic aerosols and the QBO are all accounted for.
Total ozone and lower stratospheric temperature show very similar variation patterns. Both models reproduce the observed patterns of the various modes of variability surprisingly well. Main contributors to interannual variations are trend (up to –30 DU/ decade or
-1.5K/decade), QBO (up to 25 DU or 2.5 K peak to peak), intensity of the polar vortices (up to 50 DU or 5 K peak to peak) and tropospheric weather (up to 30 DU or 3 K peak to peak).
At low latitudes most variation patterns are zonally symmetric.
At higher latitudes, however, strong, zonally non-symmetric signals for QBO, solar cycle and other influences, are found close to the Aleutian Islands or south of Australia. Similar features appear in the model runs, but often at different longitudes than in the observed data sets.
Chemistry and transport models are essential tools for chemistry climate research because they can perform sensitivity experiments that test climate model components, for example, the chemical mechanism or meteorological inputs. CTMs can test aspects of model implementation, such as horizontal or vertical resolution, so that we can understand how choices in implementation affect the results calculated. In this study, we have examined the performance of several stratospheric processes under a variety of implementations in order to identify the consequences of various model ‘shortcuts’. Observationally based diagnostics are used to identify how model processes are affected by the shortcuts and to assess whether those shortcuts are acceptable.
A series of sensitivity experiments were performed with the Goddard chemistry and transport model using meteorological input from the Finite Volume General Circulation Model (FVGCM). Each experiment changed a single aspect of implementation in order to isolate the effects of that change. Implementation tests included CTM lid height, reduction of the vertical and horizontal resolution of the original wind fields, and removal of small scale wave forcing. The simulations were compared against each other to evaluate sensitivity to implementation, and compared to observationally based diagnostics to assess how well the model represented stratospheric transport processes.
The results showed that a faithful representation of lower stratospheric transport processes requires 2x2.5 resolution. Simulations at 4x5 failed to produce adequate barriers to horizontal mixing across the subtropics or the vortex edge. The 4x5 simulation was unable to sufficiently contain the perturbed chemistry of the Antarctic vortex, which has important consequences for a realistic simulation of the ozone hole and its recovery. Upper stratospheric transport was much better represented when the CTM lid was in the upper rather than the middle mesosphere. In general, both the lower CTM lid and the reduced horizontal resolution sped up the residual circulation below 10 hPa and had the opposite effect above. Counter to expectation, simulations with the fastest tropical ascent also produced the oldest mean stratospheric ages. Rapid ascent was accompanied by increased horizontal mixing with midlatitude air, increasing the age of the ascending tropical air.
Previous analysis has shown that the dynamical properties of the Antarctic vortex provide strong constraints on the area within which chemical ozone depletion can occur (Bodeker et al., 2002). In particular it was shown that the size of the dynamical vortex had not changed at all over the period 1979 to 1998 while the size of the Antarctic ozone hole (O3<220 DU) had grown, encroaching towards the edge of the larger dynamical vortex. More recent complimentary studies have shown that temperatures close to the vortex edge exhibit strong control over the size of the Antarctic ozone hole (Newman et al., 2004). These temperatures, which affect both the size of the Antarctic ozone hole and the strength of the polar night jet, will be affected by future greenhouse gas concentrations and the severity of Antarctic ozone depletion itself. Therefore, interactions between stratospheric chemistry and radiation in this collar region of the vortex, and their feedbacks, are likely to influence the rate at which the size of the Antarctic ozone hole will change in the future, and perhaps the containment properties of the vortex.
We will investigate how well the CCMs participating in CCMVal reproduce this behaviour. We will first calculate daily meridional profiles (by equivalent latitude) of the meridional impermiability, (Bodeker et al., 2002), based on NCEP/NCAR reanalyses, and of total column ozone, based on the NIWA assimilated total column ozone data base (Bodeker et al., 2005), over the period 1979 to 2004. We will then compare these with similar meridional profiles calculated using the output from the CCMVal REF1 runs for all of the CCMs for which the relevant data are available. Comparison of measurement and model based results over the period 1980 to 2004 will be used to verify the extent to which the models can capture the historical encroachment of the area of Antarctic ozone depletion over the area of the dynamical vortex. Similar analyses based on output from all available CCMVal REF2 runs will be used to assess how the interplay of chemistry and changes in the temperature of structure of the stratosphere (driven by changes in ozone and greenhouse gases) will affect the future evolution of the size of the Antarctic ozone hole.
Solar effects on the stratospheric climate are
investigated using a CCSR/NIES GCM. The model is almost the same as
that in Nagashima et al. (2002) that is based on an old version
CCSR/NIES climate model. The model has ozone and PSC chemistry in the
stratosphere. Parameterized photo-chemistry is included in the
wavelength less than 200nm in the model. 20-year model simulations are
done in each solar maximum and minimum case corresponding to 11-year
solar cycle. After that, differences are calculated. Qualitative
structures are similar to observed, though the magnitude is smaller
than that observed in the upper stratosphere. We further investigate
winter ozone response in the atmosphere. There is large anomaly with
significance in Canadian region on January and around Antarctic on
August in the solar max period. The height anomaly on January also
extends to lower atmosphere around Pacific region.
We will describe results from simulations of a CCM with detailed
stratospheric chemistry. The CCM is based on the U.K. Met Office Unified Model and the stratospheric chemistry modules are the same as that used in the SLIMCAT/TOMCAT off-line CTM.
We will compare the simulations of the CCM and CTM for the past few decades and investigate the effect of the different transport/temperatures on the quality of the simulations. In particular we will look at the trends/variability in mid-latitude and polar ozone.
The UM generates its own QBO via its gravity wave
parameterisation. We will also show the modelled QBO in short-lived and
long-lived chemical tracers and discuss the effect of O3 coupling on
the QBO signals.