Introduction

In recent years, the compilation of measurements from the stratosphere has shown a strong decrease of temperatures during the last few decades. This decrease is to be seen in contrast to a tropospheric temperature increase, and usually attributed to anthropogenic changes of greenhouse gases as CO2, water vapour, and ozone. Model simulations on CO2 increase, that take into account the circulation in the mesosphere / lower thermosphere region also indicate a negative stratospheric temperature trend (RIND et al., 1990; BERGER and DAMERIS, 1993; AKMAEV and FORMICHEV, 1998) but they usually assume a double CO2 scenario. Only recently, Akmaev et al. (2000) used realistic change rates of CO2 – an increase of 15% - to estimate the response of the middle atmosphere (MA) to anthropogenic greenhouse gas changes for the last 3 – 4 decades. The effect of CO2-increase on the middle atmosphere basically consists of a decrease in temperature.

The effect of atmospheric constituent changes on the middle atmosphere dynamics, however, has only rarely been estimated. RIND et al. (1990) discussed these effects, and found a weakening of the stratospheric/mesospheric jets due to decrease of the meridional temperature gradients. However, they used a model that did not include the mesopause region. Therefore, the upper branch of the middle atmosphere residual circulation, which consist of upward winds in the summer polar MA, a southward jet in the mesopause region, and subsidence in the winter polar MA is not realistically represented in their study. This circulation is primarily driven by momentum deposition of gravity waves (GW) in the MA, which are filtered through the stratospheric and mesospheric summer easterly jets, so that only eastward travelling GW are remaining in the mesopause region and, breaking there, cause an eastward mean flow acceleration and westerly winds in the following. This acceleration is transformed into a southward residual circulation through the Coriolis force. Thus, if the MA filtering of GW is less effective due to reduced zonal easterly winds, this should result in a decreased southward flow. Using northern hemisphere midlatitude radar data, it has been shown, that this meridional winds have decreased during the last 30 years (BREMER at al., 1997; JACOBI et al., 1997, 2001).

In this paper we used the COMMA-LIM mechanistic circulation model to simulate the effect of the CO2 and O3 change during the last 30 years on the mesopause region circulation, with special emphasis on the meridional winds there. The results are compared with the results from long-term measurements at midlatitudes.

Model description and experimental setup

The COMMA model is a global mechanistic model from 0 - 135 km with logarithmic pressure coordinates. It contains a full radiation scheme calculating solar heating and infrared cooling rates of the most prominent absorbers and emitters in the middle atmosphere. A simple troposphere is provided that represents the large scale circulation features. Several dissipation processes from turbulent mixing and ionospheric processes as molecular heat conduction and ion drag are also included. Model equations are the spherical non-linear primitive equations in flux formulation. The GW are parameterised after HOLTON and ZHU (1984). Hydrostatic balance is assumed. The model is described in detail by LANGE (2001), see also BERGER and DAMERIS (1993).

Two model runs are considered here, both considering northern hemisphere summer solstice conditions. The first one is a control run, assuming current mixing ratios of CO2 (360 ppmV), while ozone is represented after the climatology of FORTUIN and LANGEMATZ (1994). For the simulation of anthropogenically induced changes, a 30 ppmV increase of CO2 is assumed, while ozone was decreased up to 10% in high latitudes using observations presented by BOJKOV (1995). The envelope of the zonal mean ozone depletion during the summer months is shown in Figure 1. The trend is estimated from observations of total ozone in the two periods 1964-1980 and 1980-1993. From this an inverted symmetric Gaussian fit with respect to the equator is used in the model scenario. The model approximation is optimised to get the best fit in the northern hemisphere. The given values were extrapolated to a thirty-year period. This run is called the “10% run” in the following.

Model and observational results

Mean modelled zonal prevailing winds are shown in Figure 2. It can be seen that for the 10% run the stratosphere/mesosphere easterly jet is reduced. The maximum zonal wind reduction amounts to about 20% at 70 km near the wind maximum. Following the thermal wind relation, the zonal easterly jet decreases due to the decreased ozone heating and increased infrared CO2 cooling that leads to a decrease of the latitudinal temperature gradients. In the lower thermosphere, the westerly zonal wind jet above 90 km is weakened in the 10% run. This is due to the less effective blocking of westward travelling GW by the reduced easterly jet below, so that waves with high phase speed now can propagate through the stratosphere and deposit their momentum in the mesosphere and lower thermosphere. Thus they are partly unmaking the effect of the eastward travelling GW that are not affected by the stratospheric and mesospheric mean flow. In the right panel of Figure 2 the acceleration of the zonal mean flow by GW is shown for the two runs; a decrease of momentum deposition of about 15% can be seen near 80 km.

In the northern hemisphere owing to the Coriolis force in a quasi-stationary case any zonal acceleration is transformed into a meridional residual circulation, so that eastward acceleration leads to a southward meridional flow. If the wave momentum transfer and thus the acceleration is reduced, this should result in a decrease of the mean meridional wind. Such a decrease is actually observed, as is shown in Figure 3, where summer mean mesopause region meridional winds measured over Collm, Germany are presented. The winds are measured applying the LF D1 radio wave total reflection closely spaced receiver method (see SCHMINDER et al., 1997; JACOBI et al., 1997, and references therein). They refer to a mean height of 90 – 100 km, which is also the height of the summer meridional wind maximum over Collm (see, e.g., SCHMINDER at al., 1997). The data are updated from JACOBI et al. (1997). The results confirm earlier findings of BREMER at al. (1997), who compared data from different stations and found positive trends for most datasets available. During the last 20 years, a decrease of the meridional wind of about 3 m/s is observed, and the long-term trend is statistically significant at the 95% level (see JACOBI et al., 2001). As was also pointed out by JACOBI et al., such a decrease is also visible in Canadian data, but over Canada the meridional wind maximum is found at lower altitudes, probably owing to different GW activity influenced by the Rocky Mountains.

The height-latitude cross sections of the control run meridional winds and the changes between control run and 10% run are shown in Figure 4. In the model runs, the southward wind jet is found at lower altitudes than measured over Collm, but as was shown by JACOBI et al. (2001) there is some longitudinal height variation of the relevant wind structures, so that the model results can still be considered as realistic. In the 10% run the southward jet is reduced by about 25%. This is somewhat less than seen in the measurements, but still represents the same order of magnitude .

Discussion and Conclusions

Using a simple mechanistic circulation model with full radiation parameterisation scheme we could show that the mesospheric circulation is clearly affected by atmospheric CO2 and O3 changes, if these changes are in the order of magnitude that is already observed in the atmosphere. In particular, a reduction of the summer stratospheric/mesospheric jet by about 20% is found, which results in a change of the filtering of GW and in the following to a reduction of the residual circulation by about 25%. Taking into account that this residual circulation is driving the meridional transport of trace gases, the consequences for the composition of the middle atmosphere could be significant. Furthermore, we could show that owing to the strong response of the MA system to greenhouse gas and absorber density monitoring of the upper middle atmosphere could provide a tool for indicating for climate trends more easily and earlier than using lower atmospheric data.

We have to point out, however, that a similar agreement of measurements and model results is not found concerning the zonal prevailing wind. Long-term trend estimations of this component still reveal inconclusive results, which are partly contradicting when data from different longitudes, or from different time intervals are analysed (see BREMER et al., 1997, JACOBI et al., 1997, 2001). It has to be taken into account that the MA zonal wind obviously varies on time scales of years to decades, partly owing to an 11-year solar cycle effect on the MLT circulation, and also being steered by planetary wave activity that can be modified by interdecadal tropospheric circulation variations. Further investigation of zonal wind variability is therefore strongly required to obtain a comprehensive picture of tropospheric-lower thermospheric coupling processes.

In the simulations presented here, we did not take into account changes of stratospheric water vapour, which has also increased in recent years and which is also considered as a climate gas (SMITH et al., 1999, 2001). Furthermore, for the sake of simplicity the model did not include the forcing of planetary waves at the lower boundary. Change of tropospheric circulation patterns, however, and consequently changes in planetary wave activity also affects the mean flow and may give rise to some modifications of the results shown here. Last but not least, GW in circulation models have to be parameterised, and in particular the originally launched GW spectrum is, although in principle basing upon observations, virtually arbitrary. First global climatologies of GW activity are available only recently (TSUDA et al., 2000; MCLANDRESS et al. 2000). In particular, space-based GPS-measurements of tropospheric and stratospheric temperature variances should be able to deliver reliable climatologies of GW parameters in the lower stratosphere which could significantly improve the modelling of climate change impact on the upper middle atmosphere.

Acknowledgements

This study was supported by the Deutsche Forschungsgemeinschaft under grant JA-836/6-1, and by INTAS under grant 991-1186. We thank the COMMA modelling group for support with the basic version of the model. The ozone climatology was kindly supplied by the FU Berlin stratospheric research group.

References

AKMAEV, R.A., V.I. FORMICHEV, 1998: Cooling of the mesosphere and lower thermosphere due to doubling of CO2.  Ann. Geophysicae 16, 1501-1512.

AKMAEV, R.A., V.I. FORMICHEV, 2000: A model estimate of cooling in the mesosphere and lower thermosphere due to CO2 increase over the last 3-4 decades. Geophys. Res. Lett. 27, 2113-2116.

BERGER, U., M. DAMERIS, 1993: Cooling of the upper atmosphere due to CO2 increases: a model study. Ann. Geophysicae 11, 809-819.

BOJKOV, R.D., 1995: Estimating the global ozone characteristics during the last 30 years. J. Geophys. Res. 100, 16,537-16,551.

BREMER, J., R. SCHMINDER, K.M. GREISIGER, P. HOFFMANN, D. KÜRSCHNER, W. SINGER, 1997:  Solar cycle dependence and long-term trends in the wind field of the mesosphere/lower thermosphere.  J. Atmos. Solar-Terr. Phys. 59, 497-509.

FORTUIN, J.P.F., U. LANGEMATZ, 1994: An update on the global ozone climatology and on concurrent ozone and temperature trends. SPIE, Atmospheric Sensing and Modeling 2311, 207-216.

HOLTON, J.R., X. ZHU, 1984: A further study of gravity wave drag induced drag and diffusion in the mesosphere. J. Atmos. Sci. 41, 2653-2662.

JACOBI, CH., R. SCHMINDER, D. KÜRSCHNER, J. BREMER, K.M. GREISIGER, P. HOFFMANN, W. SINGER, 1997: Long-term trends in the mesopause wind field obtained from D1 LF wind measurements at Collm, Germany. Adv. Space Res. 20, 2085-2088.

JACOBI, CH., M. LANGE, D. KÜRSCHNER, A.H. MANSON, C.E. MEEK, 2001: A long-term comparison of Saskatoon MF radar and Collm LF D1 mesosphere-lower thermosphere wind measurements. Phys. Chem. Earth 26, 419–424.

LANGE, M., 2001: Modellstudien zum CO2-Anstieg und O3-Abbau in der mittleren Atmosphäre und Einfluß des Polarwirbels auf die zonale Symmetrie des Windfeldes in der Mesopausenregion. Reports of the Institute for Meteorology 25, University of Leipzig, 121 pp.

MCLANDRESS, C., M.J. ALEXANDER, D.L. WU, 2000: Micowave Limb Sounder observations of gravity waves in the stratosphere: A climatology and interpretation. J. Geophys. Res. 105, 11947-11967.

SCHMINDER, R., KÜRSCHNER, D., SINGER, W., HOFFMANN, P., KEUER, D., J. BREMER, 1997 Representative height-time cross-sections of the upper atmosphere wind field over Central Europe 1990-1996. J. Atmos. Solar-Terr. Phys. 59, 2177–2184.

SMITH, C.A., R. TOUMI, J.D. HAIGH, 1999: Radiative forcing due to trends in stratospheric water vapour. Geophys. Res. Lett. 27, 1687-1690.

SMITH, C.A., J.D. HAIGH, R. TOUMI, 2001: Seasonal trends in stratospheric water vapour. Geophys. Res. Lett. 27, 1687-1690.

TSUDA, T., M. NISHIDA, C. ROCKEN, R.H. WARE, 2000: A global morphology of gravity waves by GPS/MET data. J. Geophys. Res. 105, 7252-7273.
 
 

Back to Collm results page

Last modification: 13.12. 2001