Simulation of the mesospheric ozone response to natural and anthropogenic
H. Schmidt and G. P. Brasseur
Max Planck Institute for Meteorology, Hamburg, Germany
two simulations of 20 years each with permanent solar
Introduction minimum and solar maximum conditions, respectively,
based on solar spectra from Lean et al. (1997) were
The newly developed Hamburg Model of the Neutral
performed. Additionally a 10-year run with doubled CO2
and Ionized Atmosphere (HAMMONIA) is a general
mixing ratio (720 ppm) is compared to the solar minimum
circulation and chemistry model covering the atmospheric
altitude range from the Earth's surface up to about 250 km.
It is designed to study interactions between chemistry,
dynamics and radiation in the whole atmosphere but in
particular the mesosphere, lower thermosphere (MLT)
Due to the space limitations we concentrate here only
region, and the coupling between the atmospheric spheres.
on zonal mean results above the stratopause for some
This paper concentrates on the response of trace gases
selected parameters and the month of January. For solar
(in particular ozone) in the MLT region to natural and
maximum conditions the ozone pattern seems to be
anthropogenic climate variability. Results of different
amplified with respect to solar minimum. The mixing ratio
simulations with HAMMONIA for low and high solar
increases for the secondary (in the mesopause region) and
activity on the one hand and for present day and doubled
the tertiary (winter high latitudes between 0.1 and 0.01
CO2 concentration on the other hand are presented. Similar
hPa) maxima, and decreases for the minimum below the
studies were until now performed only with 2D models (e.g.
mesopause (Figs. 1 a and c). The doubling of CO2 leads to
Khosravi et al., 2002) or with 3D GCMs that use
an increase of ozone almost everywhere (Fig. 1e) which is
prescribed chemical fields for the computation of radiative
very likely a consequence of reaction rates changing due to
heating (e.g. Akmaev and Fomichev, 1998).
the temperature decrease (Fig. 1f). The small areas of
Studying the response of the MLT region to different
ozone decrease at polar latitudes are due to changes in
types of forcing is interesting with respect to at least two
dynamics. The total solar heating (direct + chemical
issues of the current climate change discussion: 1) Recently
heating, not shown) increases for solar maximum by about
there have been numerous and sometimes contradictory
0.5 to 3 K/day around the mesopause. At this height region
reports on observed temperature trends in this altitude
the increase is more due to the change in chemical heating
region (for a review see Beig et al., 2003) which still wait
than a direct radiative effect. For CO2 doubling, heating
for a concise explanation. Some people also argue that
rate changes are less significant. Concerning temperature,
MLT trends might serve as an early indicator for climate
maximum solar irradiance leads to an increase in the
change (Thomas, 1996). 2) It is still unclear how strongly
mesopause region by about 2 to 7 K (Fig. 1d). This value
and through which mechanisms solar variability influences
agrees quite well with some numbers listed by Beig et al.
the Earth’s atmosphere. For the numerical simulation of
(2003) for different analyses of observational data. CO 2
these phenomena it should be helpful to include the
doubling would lead to a simulated temperature decrease
atmospheric altitude range upward of, say, 50 km where the
everywhere above the tropopause with smallest values in
part of the solar spectrum is absorbed that shows the largest
the mesopause region. Please note, that all results are
presented with pressure as the vertical reference system.
Using height coordinates would lead to significantly
different results in particular for the comparison of the
Model description and setup of numerical “360 ppm” and “720 ppm” CO2 simulations. This is due to
experiments the shrinking of the atmosphere for doubled CO2 and the
HAMMONIA combines the 3D dynamics and physics strong vertical gradients in e.g. ozone and temperature.
from the ECHAM5 (Roeckner et al., 2003) and
MAECHAM4 (Manzini et al., 1979) models with the
MOZART3 chemistry scheme (an offspring of the Acknowledgments The authors would like to thank M.
Charron (now at Meteorological Service of Canada), E. Manzini
MOZART2 scheme, Horowitz et al., 2003, that should be (now at INGV, Bologna, Italy), M. Giorgetta (MPI, Hamburg), T.
valid for a very large altitude range) and some extensions Diehl (now at NASA, MD, USA), V. Fomichev (York University,
to account for important processes in the upper atmosphere. Toronto, Canada), D. Kinnison, S. Walters, D. Marsh, and R.
These new parameterizations include the treatment of Garcia (all NCAR, Boulder, USA) who contributed to the model
development. J. Lean (Naval Res. Lab., Washington, USA) is
molecular diffusion and heat conduction, chemical heating, acknowledged for providing the solar irradiance data. The work
the ion drag (Hong and Lindzen, 1976), solar heating was funded by the German ministry for education and science
shortward of 250 nm (from Richards et al., 1994, for the (BMBF). Computations were performed at the (German climate
EUV, 5-105 nm, and from the MOZART 3 computing center (DKRZ).
photo-dissociation rate computation for 120 to 250 nm),
and non-LTE effects in infrared cooling (Fomichev et al., References
1998). Akmaev, R. A. and V. I. Fomichev, Cooling of the mesosphere
In order to study the effect of the 11-year solar cycle, and lower thermosphere due to doubling of CO2. Ann.
Geophysicae, 16, 1501-1512, 1998. in solar mid- and near-ultraviolet radiation (200-400nm), J.
Beig, G. et al., Review of mesospheric temperature trends, Geophys. Res., 102, 29939-29956, 1997.
Reviews of Geophysics, doi:10.1029/2002RG000121, 2003. Manzini, E., N. A. McFarlane, and C. McLandress, Impact of the
Fomichev, V. I., J. P. Blanchet, and D. S. Turner, Matrix Doppler spread parameterization on the simulation of the
parameterization of the 15 mum CO2 band cooling in the middle atmosphere circulation using the MA/ECHAM4
middle and upper atmosphere for variable CO2 concentration, general circulation model, J. Geophys. Res., 102,
J. Geophys. Res., 103, 11505-11528, 1998. 25751-25762, 1997.
Hong, S. S., and R. S. Lindzen, Solar semidiurnal tide in the Richards, P. G., J. A. Fennelly, and D. G. Torr, A solar EUV flux
thermosphere, J. Atm. Sc., 33, 135-153, 1976. model for aeronomic calculations, J. Geophys Res., 99,
Horowitz, L. W., et al., A global simulation of tropospheric ozone 8981-8992, 1994.
and related tracers: Description and evaluation of MOZART, Roeckner, E. et al., The atmospheric general circulation model
version 2, J. Geophys Res , doi:10.1029/2002JD002853, 2003. ECHAM 5. PART I: Model description, Rep. 349, MPI for
Khosravi, R., G. Brasseur, A. Smith, D. Rusch, S. Walters, S. Meteorology, Hamburg, Germany, 2003.
Chabrillat, and G. Kockarts, Response of the mesosphere to Thomas, G. E., Is the polar mesosphere the miner’s canary of
human-induced perturbations and solar variability calculated global change?, Adv. Space Res., 18, 149, 1996.
by a 2-D model, J. Geophys. Res., doi:10.1029/2001JD001235,
Lean, J. L., G. J. Rottmann, H. L. Kyle, T. N. Woods, J. R. Hickey,
and L. C. Puga, Detection and parameterization of variations
Figure 1. Ozone mixing ratio (a) and temperature (b, K) for the reference simulations (“solar minimum” , 360 ppm CO2). c) and d)
show differences “solar maximum” – “solar minimum”, e) and f) show differences “720 ppm CO2” – “360 ppm CO2”. c), e): ozone (%).
d), f): temperature (K). Statistical confidence levels of 90% and 99% are indicated by light and dark gray shading, respectively. All
figures show zonal mean values for January as simulated by HAMMONIA.