• Bell, C., , L. Gray, , A. Charlton-Perez, , and M. Joshi, 2009: Stratospheric communication of El Niño teleconnections to European winter. J. Climate, 22, 40834096.

    • Search Google Scholar
    • Export Citation
  • Bourqui, M., , C. Taylor, , and K. Shine, 2005: A new fast stratospheric chemistry scheme in an intermediate general circulation model. II: Application to effects of future increases in greenhouse gases. Quart. J. Roy. Meteor. Soc., 131, 22432261.

    • Search Google Scholar
    • Export Citation
  • Braesicke, P., , and J. Pyle, 2004: Sensitivity of dynamics and ozone to different representations of SSTs in the unified model. Quart. J. Roy. Meteor. Soc., 130, 20332045.

    • Search Google Scholar
    • Export Citation
  • Brayshaw, D. J., , B. Hoskins, , and M. Blackburn, 2008: The storm-track response to idealized SST perturbations in an aquaplanet GCM. J. Atmos. Sci., 65, 28422860.

    • Search Google Scholar
    • Export Citation
  • Brönniman, S., , J. Luterbacher, , J. Staehelin, , T. Svendby, , G. Hansen, , and T. Svenøe, 2004: Extreme climate of the global troposphere and stratosphere in 1940–1942 related to El Niño. Nature, 431, 971974.

    • Search Google Scholar
    • Export Citation
  • Butchart, N., , S. Clough, , T. Palmer, , and P. Trevelyan, 1982: Simulations of an observed stratospheric warming with quasigeostrophic refractive index as a model diagnostic. Quart. J. Roy. Meteor. Soc., 108, 475502.

    • Search Google Scholar
    • Export Citation
  • Butchart, N., and Coauthors, 2006: Simulations of anthropogenic change in the strength of the Brewer-Dobson Circulation. Climate Dyn., 27, 727741.

    • Search Google Scholar
    • Export Citation
  • Charlton, A. J., , and L. M. Polvani, 2007: A new look at stratospheric sudden warmings. Part I: Climatology and modeling benchmarks. J. Climate, 20, 449469.

    • Search Google Scholar
    • Export Citation
  • Charney, J., , and P. Drazin, 1961: Propagation of planetary-scale disturbances from the lower to the upper atmosphere. J. Geophys. Res., 66, 83109.

    • Search Google Scholar
    • Export Citation
  • Chen, G., , and P. Zurita-Gotor, 2008: The tropospheric response to prescribed zonal forcing in an idealized atmospheric model. J. Atmos. Sci., 65, 22542271.

    • Search Google Scholar
    • Export Citation
  • Chen, G., , I. M. Held, , and W. A. Robinson, 2007: Sensitivity of the latitude of the surface westerlies to surface friction. J. Atmos. Sci., 64, 28992915.

    • Search Google Scholar
    • Export Citation
  • Chen, G., , R. A. Plumb, , and J. Lu, 2010: Sensitivities of zonal mean atmospheric circulation to SST warming in an aqua-planet model. Geophys. Res. Lett., 37, L12701, doi:10.1029/2010GL043473.

    • Search Google Scholar
    • Export Citation
  • Cohen, J., , M. Barlow, , P. K. Kushner, , and K. Saito, 2007: Stratosphere–troposphere coupling and links with Eurasian land surface variability. J. Climate, 20, 53355343.

    • Search Google Scholar
    • Export Citation
  • Fischer, A., , D. Shindell, , B. Winter, , M. Bourqui, , G. Faluvegi, , E. Rozanov, , M. Schraner, , and S. Brönniman, 2008: Stratospheric winter climate response to ENSO in three chemistry-climate models. Geophys. Res. Lett., 35, L13819, doi:10.1029/2008GL034289.

    • Search Google Scholar
    • Export Citation
  • Fletcher, C. G., , S. C. Hardiman, , P. J. Kushner, , and J. Cohen, 2009: The dynamical response to snow cover in a large ensemble of atmospheric GCM integrations. J. Climate, 22, 12081222.

    • Search Google Scholar
    • Export Citation
  • Forster, P. M. F., , and K. P. Shine, 1999: Stratospheric water vapour changes as a possible contributor to observed stratospheric cooling. Geophys. Res. Lett., 26, 33093312.

    • Search Google Scholar
    • Export Citation
  • Gerber, E. P., , and L. M. Polvani, 2009: Stratosphere–troposphere coupling in a relatively simple AGCM: The importance of stratospheric variability. J. Climate, 22, 19211933.

    • Search Google Scholar
    • Export Citation
  • Held, I. M., , and B. J. Soden, 2006: Robust responses of the hydrological cycle to global warming. J. Climate, 19, 56865699.

  • Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Elsevier, 535 pp.

  • Hoskins, B., , and A. Simmons, 1975: A multi-layer spectral model and the semi-implicit method. Quart. J. Roy. Meteor. Soc., 101, 637655.

    • Search Google Scholar
    • Export Citation
  • Lu, J., , G. Chen, , and D. M. Frierson, 2008: Response of the zonal-mean atmospheric circulation to El Niño versus global warming. J. Climate, 21, 58355851.

    • Search Google Scholar
    • Export Citation
  • McLandress, C., , and T. G. Shepherd, 2009: Simulated anthropogenic changes in the Brewer–Dobson circulation, including its extension to high latitudes. J. Climate, 22, 15161540.

    • Search Google Scholar
    • Export Citation
  • Morcrette, J.-J., 1991: Radiation and cloud radiative properties in the European Centre for Medium Range Weather Forecasts forecasting system. J. Geophys. Res., 96, 91219132.

    • Search Google Scholar
    • Export Citation
  • Newman, P. A., , E. R. Nash, , and J. E. Rosenfield, 2001: What controls the temperature of the Arctic stratosphere during the spring? J. Geophys. Res., 106, 19 99920 010.

    • Search Google Scholar
    • Export Citation
  • Olsen, M. A., , M. R. Schoeberl, , and J. E. Nielsen, 2007: Response of stratospheric circulation and stratosphere-troposphere exchange to changing sea surface temperatures. J. Geophys. Res., 112, D16104, doi:10.1029/2006JD008012.

    • Search Google Scholar
    • Export Citation
  • Peixoto, J. P., , and A. H. Oort, 1992: Physics of Climate. American Institute of Physics, 520 pp.

  • Philander, S. G., 1990: El Niño, La Niña, and the Southern Oscillation. Academic Press, 293 pp.

  • Rind, D., , D. Shindell, , P. Lonergan, , and N. Balachandran, 1998: Climate change and the middle atmosphere. Part III: The doubled CO2 climate revisited. J. Climate, 11, 876894.

    • Search Google Scholar
    • Export Citation
  • Ring, M. J., , and R. A. Plumb, 2007: Forced annular mode patterns in a simple atmospheric general circulation model. J. Atmos. Sci., 64, 36113626.

    • Search Google Scholar
    • Export Citation
  • Rosier, S. M., , and K. P. Shine, 2000: The effect of two decades of ozone change on stratospheric temperature as indicated by a general circulation model. Geophys. Res. Lett., 27, 26172620.

    • Search Google Scholar
    • Export Citation
  • Sassi, F., , D. Kinnison, , B. Boville, , R. R. Garcia, , and R. Roble, 2004: Effect of El Niño–Southern Oscillation on the dynamical, thermal, and chemical structure of the middle atmosphere. J. Geophys. Res., 109, D17108, doi:10.1029/2003JD004434.

    • Search Google Scholar
    • Export Citation
  • Scinocca, J., , and P. Haynes, 1998: Dynamical forcing of stratospheric planetary waves by tropospheric baroclinic eddies. J. Atmos. Sci., 55, 23612392.

    • Search Google Scholar
    • Export Citation
  • Shepherd, T., , K. Semeniuk, , and J. Koshyk, 1996: Sponge layer feedbacks in middle-atmosphere models. J. Geophys. Res., 101, 23 44723 464.

    • Search Google Scholar
    • Export Citation
  • Sigmond, M., , P. Siegmund, , E. Manzini, , and H. Kelder, 2004: A simulation of the separate effects of middle-atmospheric and tropospheric CO2 doubling. J. Climate, 17,23522367.

    • Search Google Scholar
    • Export Citation
  • Solomon, S., , D. Qin, , M. Manning, , M. Marquis, , K. Averyt, , M. M. B. Tignor, , H. L. Miller Jr., , and Z. Chen, Eds., 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

    • Search Google Scholar
    • Export Citation
  • Taguchi, M., , and D. L. Hartmann, 2006: Increased occurrence of stratospheric sudden warmings during El Niño as simulated by WACCM. J. Climate, 19, 324332.

    • Search Google Scholar
    • Export Citation
  • Taylor, C., , and M. Bourqui, 2005: A new fast stratospheric ozone chemistry scheme in an intermediate general circulation model. I: Description and evaluation. Quart. J. Roy. Meteor. Soc., 131, 22252242.

    • Search Google Scholar
    • Export Citation
  • Winter, B., , and M. S. Bourqui, 2010: Wave-forcing in the stratosphere under doubled-CO2 conditions in a 100-year coupled chemistry-climate model study. J. Geophys. Res., 115, D12126, doi:10.1029/2009JD012777.

    • Search Google Scholar
    • Export Citation
  • Winter, B., , and M. S. Bourqui, 2011: The impact of surface temperature variability on the climate change response in the Northern Hemisphere polar vortex. Geophys. Res. Lett., 38, L08808, doi:10.1029/2011GL047011.

    • Search Google Scholar
    • Export Citation
  • Wittman, M. A., , A. J. Charlton, , and L. M. Polvani, 2007: The effect of lower-stratospheric shear on baroclinic instability. J. Atmos. Sci., 64, 479496.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    January–February response in zonal-mean (left) T and (middle) u, and (right) December–January response in and F, to forcing in the NH equatorial and tropical bands EQ, N1, N2, and N3. Contours of control values are overlaid. In the right column, the zonal wind for each experiment is overlaid. See text for details.

  • View in gallery

    As in Fig. 1, but for the NH midlatitude band experiments.

  • View in gallery

    As in Fig. 1, but for the NH polar band experiments.

  • View in gallery

    January–February average response in the NH O3 number density. Contours of control values are overlaid.

  • View in gallery

    July–August average response in the SH zonal-mean wind for all SH band experiments. Note that the scale on the EQ panel is twice as big as on the others. Contours of control values are overlaid.

  • View in gallery

    Seasonal cycles of (left) at 100 hPa and (remaining columns) November–December average response in and its wavenumber-1 and -2 components for the (top to bottom) EQ, NM, and N8 experiments. Control contours are overlaid on all plots. Contour spacing is, left to right, 4, 4, 2, 1 K m s−1.

  • View in gallery

    January–February response in zonal-mean (left) T and (middle) u, and (right) December–January response in and F, to forcing in the combined-band experiments (top to bottom) N1N4, EQNM, and EQNMP. Contours of control values are overlaid on the left and middle columns; contours of the zonal wind for each experiment are overlaid on the right column. Note that the color scale on the first two columns is larger than on Figs. 1, 2, or 3. See Table 1 as well as explanations in the text.

  • View in gallery

    Seasonal cycles of (left) at 100 hPa and (remaining columns) November–December response in and its wavenumber-1 and -2 components for the combined-band experiments (top to bottom) N1N4, EQNM, and EQNMP. Contour spacing is as on Fig. 6.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 14 14 2
PDF Downloads 7 7 1

Sensitivity of the Stratospheric Circulation to the Latitude of Thermal Surface Forcing

View More View Less
  • 1 Department of Atmospheric and Oceanic Sciences, McGill University, Montréal, Quebec, Canada
  • | 2 Department of Atmospheric and Oceanic Sciences, and Department of Chemistry, McGill University, Montréal, Quebec, Canada
© Get Permissions
Full access

Abstract

Using the chemistry climate model Intermediate General Circulation Model–Fast Stratospheric Ozone Chemistry (IGCM-FASTOC), the authors analyze the response in the Northern Hemisphere winter stratosphere to idealized thermal forcing imposed at the surface. The forcing is a 2-K temperature anomaly added to the control surface temperature at all grid points within a latitudinal window of 10° or 30°. The bandwise forcing is applied systematically throughout all latitudes of the Northern Hemisphere. Thermal forcing applied anywhere equatorward of 20°N, or continuously from the equator to 30°N, increases planetary-wave generation in the troposphere and enhances the flux of wave activity propagating vertically into the stratosphere. Consequently, a greater flux of wave activity breaks in the polar vortex, increasing the Brewer–Dobson circulation and leading to a warm anomaly in the polar stratosphere. Ozone concentration increases at high latitudes and decreases at low latitudes. Thermal surface forcing imposed between 30° and 60°N has the reverse effect—decreased planetary-wave generation in the lower troposphere and reduced vertically propagating wave flux entering the stratosphere—and leads to a stronger and colder vortex. Thermal forcing applied poleward of 60°N has little effect on the tropospheric mean state but nonetheless decreases the planetary-scale eddy heat flux from the surface to the tropopause, resulting in a sufficient decrease of the vertical flux of wave activity for the vortex to be anomalously strong and cold. When surface forcing is imposed only poleward of 30°N, ozone concentration decreases at high latitudes but is not affected at low latitudes. Combining the forcing in an equatorial and an extratropical band leads to a response similar to that of the equatorial forcing, demonstrating that the subtropical surface temperature changes determine the sign of the surface-driven response in the vortex.

Supplemental information related to this paper is available at the Journals Online Web site: http://dx.doi.org/10.1175/2011JCLI4006.s1.

Corresponding author address: Barbara Winter, Department of Atmospheric and Oceanic Sciences, McGill University, Montréal QC H3A 2K6, Canada. E-mail: barbara.winter@mail.mcgill.ca

Abstract

Using the chemistry climate model Intermediate General Circulation Model–Fast Stratospheric Ozone Chemistry (IGCM-FASTOC), the authors analyze the response in the Northern Hemisphere winter stratosphere to idealized thermal forcing imposed at the surface. The forcing is a 2-K temperature anomaly added to the control surface temperature at all grid points within a latitudinal window of 10° or 30°. The bandwise forcing is applied systematically throughout all latitudes of the Northern Hemisphere. Thermal forcing applied anywhere equatorward of 20°N, or continuously from the equator to 30°N, increases planetary-wave generation in the troposphere and enhances the flux of wave activity propagating vertically into the stratosphere. Consequently, a greater flux of wave activity breaks in the polar vortex, increasing the Brewer–Dobson circulation and leading to a warm anomaly in the polar stratosphere. Ozone concentration increases at high latitudes and decreases at low latitudes. Thermal surface forcing imposed between 30° and 60°N has the reverse effect—decreased planetary-wave generation in the lower troposphere and reduced vertically propagating wave flux entering the stratosphere—and leads to a stronger and colder vortex. Thermal forcing applied poleward of 60°N has little effect on the tropospheric mean state but nonetheless decreases the planetary-scale eddy heat flux from the surface to the tropopause, resulting in a sufficient decrease of the vertical flux of wave activity for the vortex to be anomalously strong and cold. When surface forcing is imposed only poleward of 30°N, ozone concentration decreases at high latitudes but is not affected at low latitudes. Combining the forcing in an equatorial and an extratropical band leads to a response similar to that of the equatorial forcing, demonstrating that the subtropical surface temperature changes determine the sign of the surface-driven response in the vortex.

Supplemental information related to this paper is available at the Journals Online Web site: http://dx.doi.org/10.1175/2011JCLI4006.s1.

Corresponding author address: Barbara Winter, Department of Atmospheric and Oceanic Sciences, McGill University, Montréal QC H3A 2K6, Canada. E-mail: barbara.winter@mail.mcgill.ca

1. Introduction

Annually averaged surface temperatures, under climate change conditions, are expected to increase by 1°–2°C near the equator to 4°–6°C at mid- to high latitudes by the end of the twenty-first century, depending on the greenhouse gas emissions scenario (Solomon et al. 2007, Figs. 10–8 therein). The response in the winter lower stratosphere in high northern latitudes to similar emissions scenarios applied in stratosphere-resolving models is projected to include greater wave forcing in the polar vortex and thus an increased Brewer–Dobson circulation, with associated reduced vortex strength and higher temperatures in the polar lower stratosphere (Rind et al. 1998; Sigmond et al. 2004; Braesicke and Pyle 2004; Butchart et al. 2006; Olsen et al. 2007; Winter and Bourqui 2010). The question has then been asked whether the stratosphere in such experiments is responding to the radiative effect of the increased greenhouse gas loading throughout the atmosphere or to the changed surface temperatures, or both. Olsen et al. (2007) showed that, in the absence of any changes in greenhouse gas loading, increased SST alone results in greater wave forcing of the vortex through changes in the strength of the tropospheric jet and associated changes in the propagation characteristics of planetary waves. The study by Sigmond et al. (2004) compares the response in the stratosphere to a doubling of CO2 applied throughout the atmosphere to that of a CO2 doubling applied either in the troposphere alone or in the stratosphere alone and attributes two-thirds of the warming in the Arctic winter lower stratosphere to the tropospheric CO2 doubling. In these experiments, SSTs were adjusted to be consistent with the corresponding radiative effect of CO2 and were therefore higher in the tropospheric than in the stratospheric-doubling case. In the Northern Hemisphere, the stratospheric response to tropospheric CO2 doubling was found to be driven by the higher SSTs (Sigmond et al. 2004).

The potential importance of the surface temperature in modulating the strength of the northern polar vortex in winter has also been investigated with a focus on localized phenomena with a strong signal in the surface temperature, such as the El Niño–La Niña oscillation or Siberian snow cover changes. The signature of El Niño in the SST field is a warm anomaly in the central to eastern equatorial Pacific of about 2°C (though episodically up to 5°C) (Philander 1990). Its effect on the northern stratospheric winter vortex is found to be an increased forcing by planetary waves (Brönniman et al. 2004; Taguchi and Hartmann 2006; Bell et al. 2009). Localized surface temperature changes due to seasonal snow cover anomalies in Siberia were studied by Cohen et al. (2007) and Fletcher et al. (2009). In these studies, early winter increases in the snow cover extent over Siberia, by inducing a surface temperature anomaly of about −3°C via albedo changes, affect the local diabatic heating in such a way as to excite a Rossby wave response, which propagates to the stratosphere. Siberia was chosen for these experiments because it is a key area for generating vertically propagating Rossby waves (Cohen et al. 2007).

To isolate the processes involved in transmitting the surface signal to the stratosphere, other studies have focused on idealized surface forcings imposed on an idealized climate state. Brayshaw et al. (2008) imposed SST anomalies of finite length and width at different latitudes in an aquaplanet GCM and investigated the effects on the storm track (at 850 hPa) and on the tropospheric jet of varying both the gradient and the location of the anomaly. The changes in SST gradient patterns were found to be more relevant to the response in the wind field than the magnitude of the anomalies themselves, and the position of the anomaly gradient relative to the location of the subtropical jet was a key factor in determining whether the jet underwent a poleward or an equatorward shift. In Chen et al. (2007) and Chen and Zurita-Gotor (2008), the applied forcing was a surface friction and a westerly torque (not confined to the surface), respectively. The torque was moved incrementally from low to high latitudes, and the authors noted that the response of the jet was to shift equatorward for low-latitude forcing and poleward for high-latitude forcing, with a transition from one to the other when the forcing was at 35°N. The authors explained these jet displacements on the basis of shifts in the critical latitudes for tropospheric eddies (Chen and Zurita-Gotor 2008). More recently, Chen et al. (2010) investigated the different effects of imposed SST warming in high or low latitudes on an aquaplanet model having a moist atmosphere. The emphasis in these idealized studies is on the displacements of the tropospheric jets in general and particularly on the eddy processes leading to such shifts. Responses in the stratosphere to surface and tropospheric forcing are not explicitly considered [rather, changes in the lower-stratospheric winds are the starting point leading to the tropospheric jet shifts (Chen and Zurita-Gotor 2008; also Wittman et al. 2007)]. Note that it is pointed out in Gerber and Polvani (2009) that the absence of topography in aquaplanet models is a large factor in producing a shift of the jet. When topography is present, shifts in the jet are very small.

Between studies targeting specific processes in idealized climates and studies examining realistic surface perturbations in comprehensive GCMs there is a gap. Experiments with idealized surface forcings imposed on a realistic basic state atmosphere would help untangle the link between surface temperature changes and the stratospheric response to increased greenhouse gas forcing. Similarly to fully idealized studies focusing on the troposphere, such experiments can establish the surface region of largest influence on the stratosphere and the sign of the change. However, such experiments are not well represented in the literature. It is the aim of the present paper to help fill this gap, with particular attention to the Northern Hemisphere stratosphere in winter. Given that the surface temperature increase associated with climate change is global but has latitudinal structure, our underlying motivation is to relate the stratospheric response to a surface temperature change in a particular range of latitudes. We therefore investigate the effect on the stratospheric circulation of surface temperature changes in given latitude bands. This is the simplest way of controlling the location of the forcing and the subsequent change in the meridional thermal gradient, and allows the forcing to be applied systematically to a series of different latitudinal windows. Although the surface forcing is zonally uniform, the control surface field on which it is imposed is not; it has a realistic distribution of topography and specific humidity and thus of temperatures, whose zonal gradients are retained when the thermal anomaly is added.

The experimental setup and model are described in detail in section 2. Results for the Northern Hemisphere winter stratosphere are presented in section 3, and in section 4 we discuss the processes by which the surface forcings are communicated to the stratosphere. Section 5 investigates the results from the combination of forcing in two or more surface bands, and conclusions are given in section 6.

2. Model and numerical experiments

Our experiments are performed using a chemistry climate model (CCM), the Intermediate General Circulation Model–Fast Stratospheric Ozone Chemistry (IGCM-FASTOC) (Bourqui et al. 2005). Its atmospheric general circulation model component is the University of Reading IGCM, which is built on the spectral dynamical core of Hoskins and Simmons (1975). The model physics are described in Forster and Shine (1999) and Rosier and Shine (2000), and include orography, a land surface scheme, moist and dry convection, and a radiation scheme (Morcrette 1991), and they are designed to be computationally fast. Here, the IGCM is run at horizontal resolution T31 and has 26 vertical levels, with a lid at 0.1 hPa; 13 levels are within the stratosphere. Rayleigh friction is applied in the top three model levels (1 hPa or above).

The IGCM is coupled to the computationally fast chemistry scheme FASTOC introduced in Bourqui et al. (2005) and Taylor and Bourqui (2005). The chemically active region for the FASTOC scheme is between the tropopause and 4 hPa, and the advected chemically active species are O3, NOx, N2O5 and HNO3. The focus of our study is on idealized surface forcing in an otherwise undisturbed control climate, representing preozone hole conditions. Advected chemical species are relaxed to climatological values outside of the chemically active region. Chemistry and dynamics in the model interact through the feedbacks between ozone concentration and radiation. A description of the T31 version of the IGCM-FASTOC, including validation of zonal-mean fields, is given in Winter and Bourqui (2010). Note that the computational efficiency of this model is a critical factor in the present study, as it allows us to perform simulations of sufficient length for our results to meet statistical significance.

In all the simulations presented here, the IGCM-FASTOC is run for 50 years in timeslice mode, that is, the 50 years are not chronological in real time but each year can be regarded as a single member of an ensemble, taking its initial conditions from the last time step of another member. In the control case, global surface temperatures and specific humidities are prescribed using climatological monthly-mean output from a separate 100-yr IGCM-FASTOC simulation in which the atmosphere interacted with a 25-m mixed layer ocean and the land surface scheme. The prescribed surface temperatures and specific humidities are linearly interpolated between monthly means to give daily values. Note that the daily surface temperatures thereby become less variable than when the land surface scheme operates. The consequence on the model climate of limiting daily variability and removing interannual variability is to reduce the amount of wave driving in the polar vortex (Winter and Bourqui 2010; Olsen et al. 2007; Braesicke and Pyle 2004). However, this is unavoidable here because of the nature of our experiments. In the perturbed simulations, we add a temperature anomaly of 2 K to all gridpoints (land and sea) within latitudinal bands of 10° or 30° width, keeping surface temperatures outside the bands identical to the control values. The transform grid of the IGCM-FASTOC at T31 resolution has 24 Gaussian latitudes per hemisphere. To best represent the full width of the 10° bands, it was necessary to use some latitudes as endpoints in more than one band (e.g., the latitude 50°N is included in both the N5 and the N6 band; see Table 1). Six of the 10° bands correspond to four Gaussian latitudes in width and the others to three Gaussian latitudes. The wide 30° bands do not have endpoints in common. To prevent step-like gradients in the surface temperature fields, a running mean over five points was applied to the perturbed surface temperature field (the running mean was reduced to three points for 10° bands spanning only three Gaussian latitudes). This means that contiguous bands are not perfectly independent of one another, and the imposed smoothed temperature anomaly is not absolutely identical for all bands. It does not affect our results because the response above the boundary layer is insensitive to slight variations in width (this was tested in an additional run in which the N3 band was made wider by one gridpoint; the response was unchanged). However, the band pairs (N1, N2) and (N7, N8) have greater latitudinal overlap than the others, and this probably contributes to the similarities in their responses. The forcing value of 2 K was chosen to be consistent with real-world examples of surface heating (see section 1). Note that we also tested a 3-K anomaly and found the response to scale linearly in amplitude. The surface specific humidity field was adjusted in the bands to be consistent with the higher temperatures, with the assumption that surface relative humidity does not change under climate warming conditions (Held and Soden 2006).

Table 1.

Summary of numerical experiments. A surface temperature anomaly of 2 K is imposed at all grid points of each latitude band. The meridional extent of the bands is given below. All experiments were run for 50 years in timeslice mode, and climatological means are presented.

Table 1.

The perturbed experiments are as follows. In a first series, the thermal surface forcing is applied in the following three wide bands: an equatorial band (EQ) from the equator to 30°N; a midlatitude band (NM) from 30° to 60°N; and a polar band (NP) from 60° to 90°N. In a second series, the width of the bands is reduced to 10°, so that there are nine bands in the Northern Hemisphere, from N1 (equator to 10°N) to N9 (80° to 90°N). Finally, in a third series, we apply the surface forcing in a combination of bands: N1N4, forced from the equator to 10°N and from 30° to 40°N; EQNM, forced continuously from the equator to 60°N; and EQNMP, forced continuously from the equator to 90°N. The main focus of this paper is the results from latitudinal forcing in the Northern Hemisphere, although some results for the Southern Hemisphere are also presented in section 3c. Table 1 summarizes these experiments. The 50-yr duration of individual experiments provided good statistical significance on most results and therefore longer simulations were not made. Results are presented as differences between the 50-yr climatological fields of the perturbation experiments and the control. Bilateral Student’s t-test statistics are used to determine the robustness of results, and for clarity on the graphics, we apply gray shading to those areas of the plot not significant at the 5% level.

3. Response in the winter stratosphere

a. Seasonal-mean response in the Northern Hemisphere

We begin by considering the January–February climatology of the zonal-mean temperature and zonal wind in the Northern Hemisphere, as shown in Figs. 1, 2, and 3 by difference fields (perturbed band experiment – control). The temperature and wind anomalies of January–February are expected to be related to prior wave-forcing anomalies: Newman et al. (2001) showed that temperatures in the polar vortex are correlated most strongly with wave flux entering the stratosphere between 40° and 80°N a month earlier. We therefore show, in the right column of Figs. 1, 2, and 3, the December–January average difference in the Eliassen–Palm (EP) flux vectors and in its total divergence. To facilitate the graphical representation of these vectors, the EP flux components above 20 km are divided by density. Vector scaling is therefore different in the lower and upper portions of each panel but is consistent across all graphs. A convergence (negative divergence) of the EP flux acts to decelerate the zonal mean flow, and geostrophic balance is maintained by an induced poleward circulation with a descending branch near the pole. The adiabatic warming related to this descent, partly balanced by radiative relaxation, is in thermal wind balance with the deceleration in the polar vortex.

Fig. 1.
Fig. 1.

January–February response in zonal-mean (left) T and (middle) u, and (right) December–January response in and F, to forcing in the NH equatorial and tropical bands EQ, N1, N2, and N3. Contours of control values are overlaid. In the right column, the zonal wind for each experiment is overlaid. See text for details.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

Fig. 2.
Fig. 2.

As in Fig. 1, but for the NH midlatitude band experiments.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

Fig. 3.
Fig. 3.

As in Fig. 1, but for the NH polar band experiments.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

The top rows in Figs. 1, 2, and 3 give the results for the wide-band experiments [equatorial (EQ), Northern Midlatitude, and North Polar in Figs. 1, 2, and 3, respectively]. The lower rows give the results for the narrow 10° bands into which each wide band can be subdivided. A green or black line at the bottom of each panel indicates where the 2-K thermal forcing was applied. Refer to Table 1 for a summary of all the experiments and the extent of the bands. In Fig. 1, the thermal forcing is applied within 30° latitude of the equator. The thermal anomaly extends upward at the equator (EQ and N1 cases, left column, upper two rows) and follows isentropes upward and poleward in the tropics (EQ, N2, and N3 experiments). To maintain thermal wind balance, the changes in the meridional temperature gradient are accompanied by changes in the tropospheric jet. When the thermal forcing is confined within 10° latitude of the equator (N1, second row), the tropospheric jet is enhanced. When the surface forcing extends away from the equator but remains in the tropics (EQ, N2, and N3 experiments), the tropospheric jet undergoes a poleward shift (Fig. 1, center panels, rows 1, 3, and 4).

In the stratosphere, the effect of the thermal surface forcing in the equatorial and tropical latitudes in Fig. 1 is seen as a large warm anomaly at high latitudes and a significant weakening of the vortex. This response is found for experiments EQ, N1, and N2 but is absent when the thermal forcing is applied between 20° and 30°N (N3 case). While a weakening and slight poleward shift of the tropospheric jet is noted in the N3 case, there is no significant response in the stratosphere to the surface forcing.

The right column of Fig. 1 provides a key to the response in the stratospheric vortex winds. For experiments EQ, N1, and N2, there is increased EP flux into the stratosphere in December–January, especially poleward of 60°N, and stronger convergence of the flux into the upper stratosphere. Since EP flux convergence acts as a force on the flow, increased convergence is accompanied by the deceleration of the vortex seen in the center column of Fig. 1. The deceleration in turn allows greater propagation of waves into the core of the vortex, in accordance with the Charney–Drazin criterion. The zonal wind field for each experiment is overlaid (green contours) on the right panels of the figure. In the N3 experiment (bottom row of Fig. 1), the upward EP flux entering the stratosphere between 40° and 80°N is slightly decreased, with weaker convergence into the upper stratosphere. The accompanying acceleration of the vortex, however, is not above the statistical significance criterion in this experiment. The stratospheric response to thermal surface forcing thus changes in character between the N2 and N3 cases. The link between tropospheric and stratospheric responses will be analyzed in Section 4.

Figure 2 presents the same diagnostic fields as Fig. 1 but for the cases when thermal surface forcing was applied between 30° and 60°N, either over this entire span (experiment NM, top row) or over 10° sections at a time (experiments N4, N5, and N6, in rows two, three, and bottom, respectively). The effect throughout the mid- to high-latitude troposphere is to warm the extratropical lower troposphere and thereby decrease the meridional temperature gradient, with the attendant slowing of the tropospheric jet expected by thermal wind balance. When the surface thermal forcing is applied only north of 50°N, in the N6 case (Fig. 2, bottom row), the meridional temperature gradient is affected only in the lowest troposphere and the decrease in the jet is small and limited to its northern flank. The stratospheric response to the bandwise surface forcing is consistent in all panels of Fig. 2—there is a cold anomaly at high latitudes throughout the stratosphere, accompanied by an increase in vortex winds. The stratospheric response is of the same sign in all experiments, though weaker and not statistically significant in the N6 case. Applying the surface forcing to a narrow band of latitudes, such as in N4 and N5, yields a stronger response than applying the surface forcing to all latitudes from 30° to 60°N (i.e., the NM case). The changes in wave forcing are seen in the right-hand column of Fig. 2. In the NM, N4, and N6 cases, the December–January EP flux entering the stratosphere is decreased equatorward of 60°N but increased poleward of 60°N, while in the N5 case, the EP flux entering the stratosphere is decreased from 40° to 80°N. In all cases, the December–January upward EP flux in the high-latitude middle stratosphere is reduced (shown by the downward arrows), leading to its decreased convergence in the upper stratosphere and consequently to stronger vortex winds.

Finally, Fig. 3 presents the same diagnostics for the northern polar bands, covering the range 60° to 90°N. In the troposphere, the thermal response follows the pattern seen in the N6 experiment (Fig. 2, bottom row) and becomes limited to the surface in the N8 and N9 experiments. Therefore little or no weakening of the tropospheric jet is noted. The response in the stratosphere is not statistically significant in either the wide NP band experiment (thermal forcing is applied from 60° to 90°N) or in the N7 experiment (forcing applied from 60° to 70°N), but its sign in the latter is nevertheless consistent with that of the N8 and N9 responses (bottom two rows of Fig. 3). The EP flux entering the stratosphere shows a slight decrease equatorward, and increase poleward, of 60°N in the NP and N7 cases, and a decrease everywhere poleward of 40°N in the N8 and N9 cases (Fig. 3, right column). In the NP case, the midstratospheric vertical EP flux is stronger on the poleward flank of the vortex, but the associated increase in flux convergence is very small and the weakening of the January–February vortex (Fig. 3, top center) is not strong or statistically significant. In the N7, N8, and N9 cases, there is a strong reduction of EP flux into the vortex, particularly on the poleward side. The large cold anomaly evident in the high-latitude stratosphere in the N8 and N9 experiments, accompanied by a stronger vortex, is consistent with the reduced vertical component of the EP flux in these experiments and the associated reduced convergence of EP flux in the upper stratosphere.

On Fig. 2, the midlatitude 10° experiments gave results whose average would resemble the results of the 30° NM experiment. This is not the case in either of Figs. 1 or 3. In Fig. 1, the proximity of the surface forcing to the equator determines the character of the response. In Fig. 3, the three 10° experiments each present a response to the applied forcing that is consistent with the others, but the NP experiment (Fig. 3, top row) essentially shows no response to the surface forcing, aside from a slight weakening of the winds at the northern flank of the tropospheric jet. From these results, it is clear that the latitudinal width of thermal forcing at the surface, in addition to the location of the forcing itself, plays a role in determining the stratospheric response in high latitudes.

Wave forcing in the vortex drives the Brewer–Dobson circulation (BDC), in which ozone is transported from low to high latitudes in winter. We find the BDC to be significantly enhanced in the lower stratosphere in mid to high latitudes in the EQ, N1, and N2 experiments and weakened throughout the stratosphere from low to midlatitudes in the NM, N4, and N5 experiments (not shown). A slight but statistically significant weakening is also noted in the high-latitude upper stratosphere in the N8 and N9 cases. By affecting the strength of the BDC, changes in the wave forcing impact the concentration of ozone at the poles directly through changes in the transport of ozone, as well as indirectly through dynamically induced temperature changes that alter the reaction rates of ozone chemistry. Changes in ozone concentration feed back on local temperature and dynamics through changes in radiative heating or cooling. It is therefore of interest to consider the response in the ozone concentration to the thermal surface forcing in different bands. This is presented in Fig. 4. (Note that the flat bottom of the anomalies is due to the chemistry scheme being active only in the stratosphere.) We expect high-latitude ozone concentrations to be higher when the wave forcing response and BDC are stronger, and this is the case for experiments EQ, N1, N2, and N3 (Fig. 4, left column). The low-latitude decrease in these experiments suggests a more vigorous upwelling branch of the BDC there, which is also consistent with the lower temperatures seen in the tropical upper troposphere on Fig. 1, left column. In cases EQ, N1, and N2, the strongest increase in ozone concentration occurs poleward of the climatological maximum. In experiments NM, NP, and N4–N9, decreased wave forcing of the vortex is seen as a decrease in ozone concentration near the climatological maximum. In all cases, the response in the high-latitude ozone concentration is statistically robust.

Fig. 4.
Fig. 4.

January–February average response in the NH O3 number density. Contours of control values are overlaid.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

b. Seasonal cycle response in the Northern Hemisphere

In the preceding analysis (Figs. 1, 2, and 3), the changes in the mean flow were discussed for the period January–February because it corresponds to the annual maximum in the wave forcing near the top of the stratospheric polar vortex, and therefore to the period in which the mean flow is most influenced by wave drag. However, as mentioned earlier, the state of the winds in January and February is the result of prior wave forcing beginning in the fall. Early winter wave breaking, by affecting the vertical profile of zonal winds, influences the amount and level of breaking in subsequent months, a “preconditioning” of the vortex that is also invoked to explain why the polar vortex in some cases is more susceptible to destruction by sudden stratospheric warming events (for instance, Butchart et al. 1982; Charlton and Polvani 2007).

To provide more insight into the seasonal evolution of the changes in the mean flow and wave forcing, seasonal cycles of the same fields as in Figs. 1, 2, and 3 are provided as supplementary online material and will be briefly described here (available at the Journals Online Website: http://dx.doi.org/10.1175/2011JCLI4006.s1). On these height–time plots, we show the area-averaged response in the high latitudes, to capture the extent of the polar vortex. The two contrasting responses to the imposed forcing discussed for Figs. 1, 2, and 3 in fact persist through the seasonal cycle: when the forcing is between the equator and 20°N (cases N1 and N2), or continuously from the equator to 30°N (case EQ), the polar stratosphere warms. When the forcing is exclusively north of 30°N (cases NM, NP and N4 through N9), the polar stratosphere cools.

However, the seasonal cycles show interesting variations from case to case. The warming in cases EQ, N1, and N2 is seen throughout the winter from November through April, with a peak in November–December. Consistently, anomalously large wave forcing is also seen early in the winter (November and December). Subsequent wave forcing can be reduced or no longer statistically significant, yet the vortex is weakened for the remainder of the winter. When the forcing is applied in the midlatitude NM, N4, or N5 bands, the polar stratosphere is colder from November onward, with a minimum in February. The largest decrease in wave driving of the vortex occurs in January and February. These are normally the months of maximum wave drag, and if the drag is anomalously weak at this time, the vortex remains stronger throughout the winter and spring. The N3 case, with surface forcing imposed from 20° to 30°N, serves as a transition between these two opposite responses, with a warm anomaly only from October through December (associated with an early and weak pulse of anomalous wave activity) and cold anomalies in the winter months thereafter.

Finally, when the surface forcing is applied only poleward of 60°N, the same reduction of January–February wave forcing as for the midlatitude bands is seen but is of shorter duration. In a sense, there is another transition marked by the N6 and N7 cases, as the surface forcing is applied to either side of 60°N—the response goes from a cooling throughout the stratosphere continuously from the fall into the spring to a cooling localized in the midstratosphere mainly in January and February. Furthermore, once the forcing is applied poleward of 60°N, there is no longer any statistically significant response in the troposphere. The NP case does not exhibit any response resembling the N8 and N9 cases. Note that the response in the mean flow in January–February gives a good representation of the overall response in all experiments. Similarly, December–January is a good representation of the overall change in the wave forcing. This justifies the choices made in Figs. 1, 2, and 3.

c. Seasonal-mean response in the Southern Hemisphere

It is beyond the scope of this paper to investigate in detail the Southern Hemisphere’s winter stratospheric response to bandwise forcing imposed in the Southern Hemisphere. In the absence of strong land–sea contrasts and the stationary Rossby waves they engender, we do not expect an imposed zonally uniform warm surface anomaly to have as large an impact on the polar vortex as in the Northern Hemisphere. Nevertheless, we ran a complementary series of 10° latitude band experiments in the Southern Hemisphere, analogous to the cases N1 to N9 presented in this study. The response in the austral winter zonal-mean wind field (Fig. 5) differs from the Northern Hemisphere experiments in two significant ways: 1) all cases exhibit the same response in the stratosphere, that is, there is no transition zone; and 2) the zonal winds essentially weaken along the entire axis of maximum wind speed, from the troposphere to the stratosphere, with strong positive anomalies occurring on either side (with some variation by band). This suggests an annular mode–like response in the winds rather than a change in the strength of wave forcing of the vortex.

Fig. 5.
Fig. 5.

July–August average response in the SH zonal-mean wind for all SH band experiments. Note that the scale on the EQ panel is twice as big as on the others. Contours of control values are overlaid.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

4. From the troposphere to the stratosphere

Having established, in section 3, that the polar vortex is anomalously disrupted by wave activity when there is thermal surface forcing equatorward of 30°N, and anomalously strong when the surface forcing is poleward of 30°N, we now seek the origins of these responses in the troposphere. In this section, we limit our analysis to three of the experiments: EQ, whose response fields are similar to the low-latitude cases N1 and N2; NM, which is characteristic of the midlatitude cases N4, N5, and N6; and N8, whose response fields are similar to the high-latitude cases N7 and N9 and are sufficiently different, in the troposphere, from the NM responses to merit a separate category (the wide high-latitude band experiment NP did not give a statistically significant response).

In Cartesian coordinates under quasigeostrophic scaling, the horizontal and vertical components of the EP flux are defined, respectively, as (Holton 2004)
e1

Here, R is the dry air gas constant, ρo(z) is the reference density profile, N2 is the squared Brunt–Väisälä frequency, and H is the scale height. The left column on Fig. 6 presents the zonal-mean eddy heat flux at 100 hPa for the Northern Hemisphere. Note that this quantity is, through Eq. (1), proportional to the vertical component of the EP flux entering the stratosphere. Green horizontal lines on Fig. 6 demarcate the region of imposed thermal forcing at the surface. The control eddy heat flux is greatest between 40° and 80°N (see the black contours for the control simulation on Fig. 6) and its average within this sector was found to be strongly correlated with temperatures in the polar cap (60° to 90°N) at 50 hPa a month later by Newman et al. (2001). As discussed in section 3a (Figs. 1, 2, and 3), waves entering the stratosphere north of 40°N propagate upward and poleward and thereby affect the stratospheric polar temperature by breaking in the vortex. Comparing the temporal evolution of anomalies at 100 hPa in Fig. 6 (left column) with the seasonal cycle in EP flux divergence anomalies shown on the supplementary online figures confirms that at 100 hPa is a good proxy for stratospheric forcing. For the EQ experiment (Fig. 6, top left), the eddy heat flux entering the stratosphere is increased from October to February north of 50°N, with a maximum positive anomaly in December–January. The January–February temperature response field (Fig. 1, top left) accordingly exhibits a strong warm anomaly. On the middle left panel of Fig. 6, the NM results show the opposite, that is, a reduction north of 40°N in November and January, followed by a spring negative anomaly in March–April. The reduced flux in January leads to the negative temperature anomaly seen in the January–February average on Fig. 2 (top left). Similarly, the reduction in poleward of 40°N in the N8 case (Fig. 6, bottom left) from November into January leads to the cold anomaly in January–February seen in the area of the polar vortex in Fig. 3 (left column, second from bottom). In all three cases, the January–February average response in the vortex discussed in section 3 is set up at the beginning of winter. A change of opposite sign is generally seen in the eddy heat flux south of 40°N, but these waves propagate preferentially toward the equator, and do not affect the polar vortex in the stratosphere.

Fig. 6.
Fig. 6.

Seasonal cycles of (left) at 100 hPa and (remaining columns) November–December average response in and its wavenumber-1 and -2 components for the (top to bottom) EQ, NM, and N8 experiments. Control contours are overlaid on all plots. Contour spacing is, left to right, 4, 4, 2, 1 K m s−1.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

The variations in wave flux entering the stratosphere seen in our experiments can be due to changes in tropospheric wave sources, redirection of vertically propagating waves below 100 hPa, or a combination thereof. Low wavenumbers are favored for propagation into the winter stratosphere in general (e.g., Charney and Drazin 1961), and stationary Rossby waves in particular play an important role in wave forcing at high latitudes (McLandress and Shepherd 2009). The sources for such waves are primarily the land-sea thermal contrasts and mountain ranges of the Northern Hemisphere. To investigate changes in the stationary waves, we show, in columns two to four of Fig. 6, the eddy heat flux and its zonal-wavenumber 1 and 2 (denoted s1 and s2) components from the surface to 16 km (about 100 hPa). These are averaged over November and December to capture the early-winter response in the troposphere. (As stated in Section 2, all graphs in this paper show the climatological mean response over the 50 timeslice years of each experiment, therefore eddies can only be stationary.)

The control values of the November–December eddy heat flux (black contours on the second column of Fig. 6, spaced every 4 K m s−1) indicate maxima in the troposphere (centered at around 45°N and 4 km) and the stratosphere (centered at around 60°N), and a midlatitude minimum at around 12 km. In the EQ case, both maxima are strengthened and extended upward and northward in the troposphere and stratosphere, respectively. The minimum is weakened and extended upward. In the NM experiment (second column, middle row), the eddy heat flux maxima are weakened, particularly on the equatorward side of the tropospheric maximum, while the minimum at 12 km becomes slightly stronger. It indicates that the centers of greatest poleward transport of heat in stationary eddies are all attenuated as a result of surface forcing in the NM band, contrary to the response to EQ band forcing. In the N8 case (Fig. 6, second on bottom), both eddy heat flux maxima are again decreased, though only weakly in the troposphere. However, in contrast to case NM, the minimum at 12 km becomes more negative. The stationary thermal eddies shown on Fig. 6 are one mechanism by which heat is transported from low to high latitudes in a balanced climate. A thermal surface anomaly imposed in low latitudes (as in the EQ case) requires a greater poleward transport of heat in order for the climate to remain in equilibrium and, conversely, when a thermal forcing is imposed in higher latitudes, as in the NM and N8 cases, the poleward transport of heat must decrease. It is seen on Fig. 6 that the response to the surface forcing in the stationary eddy field is consistent with this energy balance requirement.

Turning to the s1 component of the eddy heat flux shown in the third column of Fig. 6, the control values (black contours, spaced every 2 K m s−1) show a maximum around 60°N, which gradually increases from the surface to the lower stratosphere. This is the latitude of the greatest zonal variability of mean winter temperature, stemming from the stationary eddies (Peixoto and Oort 1992). In the EQ experiment (third column, top row), an increase in the s1 component of the eddy heat flux is seen continuously from the surface to the lower stratosphere following two upward branches: one around 45°N, corresponding to an equatorward expansion throughout the troposphere of the region of the maximum, and the other branch around 65°N, corresponding to an upward expansion of the surface maximum there and a poleward strengthening of the upper-level maximum. Note that the 10° bands N1, N2, and N3, which occupy the same latitudes as the EQ band, also exhibit a bottom-to-top increase in the s1 component of the eddy heat flux (not shown). In the N1 case, however, the equatorward expansion of the 45°N branch seen in the EQ case is also present from the surface upward, while no accompanying increase in the surface maximum at 65°N occurs. The reverse is true for the N2 and N3 cases—each of these exhibits a stronger and higher surface maximum at 65°N as well as a strengthening at the poleward edge of the maximum near the tropopause, but no equatorward expansion at any level. In the NM case, the early winter decrease in total eddy heat flux at 100 hPa (middle row, left-most panel) can be connected to a decrease in the s1 component (second panel) throughout most of the troposphere. Here, we note a weakening along the equatorward edge of the control high-s1 zone from the near-surface upward and northward. A similar response is found for the 10° bands N4 through N9 (only N8 is shown, Fig. 6, bottom row) but not for the NP case (not shown), in which there is no response.

The stationary s2 control values (black contours on the third column of Fig. 6, spaced every 1 K m s−1) show two maxima around 50°–60°N, in the lower stratosphere and near the surface. At the surface the maximum corresponds roughly to the zonal-mean signature of the storm tracks. In all experiments the response in the s2 component of the eddy heat flux in November–December is considerably weaker than the s1 response and is often not of the same sign at the top and the bottom of the troposphere. For the bands having surface forcing near the equator (EQ, N1, and N2), the s2 response is positive at the top and bottom of the troposphere, and thus reinforces the s1 response. In most other cases the s2 response near the tropopause is of opposite sign to the s1 response.

Comparing the amplitude of the control values of on these last three columns of Fig. 6 (keeping in mind that the contour spacing changes) confirms that the vertically propagating wave flux entering the stratosphere at 60°N (about 12 K m s−1) is almost entirely composed of the s1 and s2 components (8 and 3 K m s−1, respectively), although these components account for only a fraction of the tropospheric maximum eddy heat flux. Yet an overall picture that emerges from considering the stationary s1 and s2 components in the troposphere is that the response found in the early part of the seasonal cycle in the total flux at 100 hPa (left-hand column) at least partly originates from the near-surface in the s1 component averaged over November and December. The change in the s2 component is secondary. This suggests that the application of thermal surface forcing in the sensitive latitude bands south (north) of 30°N leads to an increase (a decrease) in the tropospheric stationary eddies at higher latitudes, which then propagate from the near-surface upward to the stratosphere. Poleward transport of heat via the stationary synoptic-scale eddies can be represented by changes in the baroclinicity, as quantified by the growth rate of the maximum Eady mode. Plots of the climatological Eady growth rate calculated at about 5 km (not shown) are consistent with the second column of Fig. 6; in the EQ case, baroclinicity increases at around 40°N, where a strong positive anomaly occurs in the eddy heat flux, and the reverse takes place in the NM case (decreased baroclinicity where there is a negative heat flux anomaly). This is expected since the tropospheric thermal eddy maximum is dominated by contributions from higher wavenumbers, as indicated above. In the N8 case, there is no evident signal in the Eady growth rate (not shown), which may be due to limitations of this diagnostic at the high latitudes where the N8 surface forcing is imposed. However, it is clear from comparing the three left panels on the bottom of Fig. 6 that the s1 and s2 components of the eddy heat flux response make a larger contribution in the troposphere in this experiment than in the EQ and NM cases. The response in baroclinicity reflects the response in both stationary and transient synoptic-scale eddies. Both have the potential to affect the climatological storm-track response, and thereby to affect the stationary eddy heat flux on a planetary scale, for instance by modifying the strength and location of the storm-track exit regions either in the Pacific alone (leading to an s1 component in the response) or in both the Pacific and the Atlantic (s2). In this manner synoptic eddies can lead to planetary eddies, which are able to propagate to the stratosphere. Elucidating the processes that cause these changes in the lower troposphere is, however, beyond the scope of this paper. Another potential link between anomalies in synoptic and planetary eddies is the nonlinear interaction between shorter waves that can result in transfer of energy to lower wavenumbers. It was shown by Scinocca and Haynes (1998) that synoptic-scale waves can become reorganized into packets having envelopes of zonal wavenumber 2 (the absence of s1 waves in that paper was attributed to the absence of topography in the model used). These packets can propagate to the stratosphere (Scinocca and Haynes 1998).

It is clear from the results presented thus far that there is a transition between responses of opposite sign when the surface forcing is moved from south to north of 30°N. An argument can also be made for a transition zone near 60°N as follows: the NP, N6, and N7 experiments show little or no response in the averaged January–February stratospheric circulation (Fig. 3). The low sensitivity to surface forcing in these experiments may stem from a more variable seasonal cycle in the 100-hPa eddy heat flux response poleward of 40°N: in experiments NM and N4–N9, this seasonal cycle begins with a negative anomaly in the late fall (only cases NM and N8 are shown, on Fig. 6). In most experiments, the negative anomaly persists for the entire winter. In cases N6 and N7 alone (not shown), there is a brief period of positive 100-hPa eddy heat flux anomalies in high latitudes in January. This brief interruption is not statistically significant, but may sufficiently undermine the effect of the negative early winter wave flux to give the weak and statistically insignificant responses in the vortex seen in these two experiments.

5. Combined bands

Given the opposite responses to the surface forcing in the N1 and N4, and in the EQ and NM, experiments, we conducted the following three further experiments with a surface thermal forcing applied simultaneously in two or more latitude bands: case N1N4, with surface forcing from the equator to 10°N and from 30° to 40°N; case EQNM, with surface forcing continuously from the equator to 60°N; and EQNMP, with the forcing applied continuously from the equator to the north pole, covering the entire Northern Hemisphere. Responses to the forcing in these additional experiments are presented in Fig. 7 (zonal-mean temperature, zonal wind, EP flux and its divergence, analogously to Figs. 1, 2, and 3) and Fig. 8 (eddy heat flux), analogously to Fig. 6.

Fig. 7.
Fig. 7.

January–February response in zonal-mean (left) T and (middle) u, and (right) December–January response in and F, to forcing in the combined-band experiments (top to bottom) N1N4, EQNM, and EQNMP. Contours of control values are overlaid on the left and middle columns; contours of the zonal wind for each experiment are overlaid on the right column. Note that the color scale on the first two columns is larger than on Figs. 1, 2, or 3. See Table 1 as well as explanations in the text.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

Fig. 8.
Fig. 8.

Seasonal cycles of (left) at 100 hPa and (remaining columns) November–December response in and its wavenumber-1 and -2 components for the combined-band experiments (top to bottom) N1N4, EQNM, and EQNMP. Contour spacing is as on Fig. 6.

Citation: Journal of Climate 24, 20; 10.1175/2011JCLI4006.1

a. Experiment N1N4

The stratospheric response in the N1N4 experiment is similar to, but slightly weaker than, the N1 response: a warm anomaly is found in the vortex at the same location as—but about 1°C weaker than—in case N1, and correspondingly the vortex wind response is negative (cf. Fig. 7, bottom row and Fig. 1, second row, but note that the color scales are different). The tropospheric response again combines elements from both the N1 and the N4 cases. The troposphere is warmer throughout all latitudes and the jet is stronger, but not as much as in the N1 case. The December–January (average) increased EP flux convergence in the upper stratosphere in the N1N4 experiment occupies a smaller area than in the N1 case and is due almost entirely to the vertical component of the EP flux. The anomalous northward flux in the lower stratosphere seen in the N1 case is almost absent in the N1N4 response (cf. the second row right panels of Figs. 1 and 7).

The seasonal response in the zonal-mean eddy heat flux at 100 hPa is positive poleward of 50°N but significant only from late December to early February. There is no significant early winter response when surface forcing is in both the N1 and N4 bands. Note that, forced individually, the N1 and N4 early winter (i.e., November–December) eddy heat flux response (not shown) is, respectively, positive as for EQ and negative as for NM. It is possible that cancellation of these early winter effects leads to the lack of a response in the combined-band N1N4 experiment (Fig. 8, top row left). The response in November–December average of the eddy heat flux (Fig. 8, second column) resembles that of the EQ case (Fig. 6) in the troposphere, with a stronger and expanded maximum and a weakened minimum. However, the response in the lower-stratospheric maximum at 60°N is weak (and statistically insignificant, cf. top left) in these early winter months. A continuous response from the top of the troposphere to the near-surface in the s1 component of the eddy heat flux can nonetheless be observed in the November–December average (top row, third panel), though it is partly offset by the s2 response (top row right).

b. Experiment EQNM

The EQNM response in the stratosphere appears to be an enhanced version of the EQ response, with a stronger high-latitude, lower-stratospheric warm anomaly accompanied by a decrease in vortex strength (note that the scale for the zonal-mean temperature and wind responses on Fig. 7 are larger than their counterparts in Figs. 1, 2, and 3; the scale of the EP flux diagnostics is the same). The tropospheric response patterns retain characteristics from both the EQ and the NM experiments. The greatest temperature anomaly is located around 30°N as in the EQ case (Fig. 1), but with a stronger warming throughout the lower troposphere which extends to the high latitudes, as in the NM case (Fig. 2). Consequently, the tropospheric jet undergoes a weaker poleward shift than in the EQ experiment. The eddy heat flux response in the EQNM experiment resembles that of the EQ experiment but begins about a month later, with a maximum response in February. Consequently, the vertical profile of the eddy heat flux averaged over November–December in the EQNM experiment (middle row, second) is weaker than its EQ counterpart while largely exhibiting the same features. A notable difference is the weak positive response throughout the lower troposphere poleward of around 50°N; in the EQ case (Fig. 6), the response in this region was weakly negative. Decomposed by wavelength, the s1 component in the November–December average is somewhat weaker while the s2 response is stronger. Both responses are continuous from the surface to the stratosphere.

c. Experiment EQNMP

When surface forcing is applied continuously from the equator to the north pole (experiment EQNMP, Fig. 8, bottom row), the response in the 100-hPa eddy heat flux retains the character of the EQ and EQNM cases, with a strong December increase around 70°N preceded by a small November increase at 60°N. The November–December average in the eddy heat flux response (total as well as the s1 and s2 components) strongly resembles that of the EQ and EQNM cases. It was noted in section 4 that the surface heating imposed only in low latitudes, as in the EQ case, required poleward heat transport to maintain a balanced climate, and that the stationary eddies provide an important contribution to this transport. In the EQNMP experiment, the surface thermal forcing is applied at all gridpoints of the Northern Hemisphere, yet the eddy heat flux response is nearly identical to the EQ case. Because the IGCM-FASTOC model used in these experiments includes comprehensive physics and topography, a uniformly imposed surface thermal anomaly does not necessarily lead to a uniform near-surface temperature increase. As described in section 3, surface heating applied only in the high-latitude bands remains confined to the surface, while heating applied in low latitudes leads to convective heating throughout the tropical troposphere. The resulting changes in the meridional temperature gradient in the middle and upper troposphere therefore require additional poleward heat transport even when surface heating is also imposed at higher latitudes. The EQNMP results demonstrate that the response to heating imposed in the tropics dominates the response when surface heating is also imposed at higher latitudes.

6. Conclusions

Using a fully three-dimensional climate chemistry model, we investigated the stratospheric response in the Northern Hemisphere winter to latitudinal variations in thermal surface forcing by imposing 2-K temperature anomalies in a series of zonal bands of 10° and 30° latitude width. Our work provides a link between studies imposing idealized forcings on an idealized basic state (e.g., dry or moist aquaplanets, as in Chen et al. 2010; Chen and Zurita-Gotor 2008; Brayshaw et al. 2008; Ring and Plumb 2007), and those considering realistic, often localized, surface forcings in realistic GCMs (e.g., El Niño, as in Brönniman et al. 2004; Fischer et al. 2008; Lu et al. 2008; or snow anomalies, as in Cohen et al. 2007; Fletcher et al. 2009).

We note two fundamental responses to our surface forcing. First, when the forcing is applied continuously between 30° and 60°N, or within 10° latitude windows centered at 35°, 45°, 75°, or 85°N, the wave activity entering the winter stratosphere at 100 hPa poleward of 40°N is reduced, and so is the wave breaking in the vortex. The anomalously weak upward flux at 100 hPa begins in the early winter, allowing an anomalously strong vortex to form. The reduction in wave forcing higher in the stratosphere occurs after the vortex has reached its annual peak, so that it persists at full strength for a longer time. The associated response in zonal-mean temperature is a strong cooling in the high-latitude stratosphere. Responses of the same sign, but strongly reduced and not statistically significant, are found when the surface forcing is centered at 55° and 65°N or imposed continuously from 60° to 90°N. Second, when surface forcing is applied equatorward of 30°N, the response is of opposite sign. There is a greater upward flux of waves into the stratosphere poleward of 40°N from the fall into the winter, leading to anomalously strong vortex disruption in January and February. An anomalously warm high-latitude lower to midstratosphere accompanies the weakened vortex. The increased wave-forcing response occurs before the vortex is fully established, and has the potential to weaken the vortex for the entire winter, even when the wave forcing anomaly subsequently decreases. The upward flux of waves from the troposphere into the stratosphere is represented by the eddy heat flux at 100 hPa, whose response at that level can be traced to the near-surface response of its zonal wavenumber-1 component. This is an indication that the thermal surface forcings we apply lead to anomalous generation of stationary Rossby waves in the lower troposphere, which propagate into the stratosphere and lead to changes in vortex strength. Sources of these waves include changes in the land–sea thermal contrasts and changes in the strength and location of the climatological storm tracks and their exit regions, the latter mechanism linking reponses in the baroclinicity (and thus in synoptic-scale waves) and responses in planetary waves. In addition to this direct response of planetary waves to the applied surface forcing, a contribution involving planetary wave formation from the nonlinear interaction among synoptic-scale waves cannot be ruled out.

In the “fully idealized” experiments of Chen et al. (2010), Chen and Zurita-Gotor (2008), Brayshaw et al. (2008), and Ring and Plumb (2007) discussed in section 1, the position of the imposed surface or tropospheric forcing, or the location of the strongest gradient in the forcing relative to the position of the tropospheric jet, are of key importance in determining the strength and sign of the response. Chen and Zurita-Gotor (2008) identify 35°N as the latitude where the response changes sign. In our 10°-band experiments, a transition between responses of opposite sign takes place at the N3 band (20°–30°N): there is an increased wave forcing response in the fall and early winter, as in cases EQ, N1, and N2, followed by a decreased wave-forcing response through the winter, as in cases NM and N4 through N9. In the 30°-band experiments, a transition occurs between the EQ and NM bands, in which the surface forcing is applied south and north of 30°N, respectively. However, when the forcing extends continuously from the equator to a latitude north of 30°N, as in the EQNM and EQNMP experiments, the stratospheric response is similar to cases in which only tropical latitudes are forced. It was pointed out in Olsen et al. (2007) that the wave driving of the polar vortex is highly correlated with the zonal-mean difference between observed SSTs in latitudes 30°–40°N and those in latitudes 0°–10°N. Our N1 experiment increases the difference between the surface temperatures in these two regions, and, consistently with the results of Olsen et al. (2007), greater wave forcing in the vortex ensues. However, when surface heating is applied to both regions as in both our N1N4 and EQNM experiments, the wave driving of the vortex nonetheless remains similar to the N1 experiment. This highlights the nonlinear relationship between the thermal surface forcing in both of these bands and the stratospheric response.

The surface forcing we apply in this study—a 2-K warm anomaly over all longitudes in a range of latitudes—is different from natural surface forcings such as El Niño, in which the surface forcing is zonally varying. However, climatological distributions of wave sources in our model are also zonally varying and govern the longitudinal distribution of the response, so that imposed forcing in a complete latitude band can have a similar effect on the stratospheric circulation as a natural, localized surface forcing such as El Niño. It has been shown that El Niño leads to more wave forcing of the polar vortex (Sassi et al. 2004; Taguchi and Hartmann 2006, who focus on vortex variability; or Fischer et al. 2008, who compare ENSO-related increases in the Brewer–Dobson circulation in different models, including the IGCM-FASTOC used in the present study). At the surface, our EQ, N1, and N2 experiments, like El Niño, produce a classic positive Pacific–North America (PNA) pattern in the mean sea level pressure (not shown). The relationship between changes in the PNA pattern induced by both phases of the ENSO and circulation in the stratosphere is discussed in Sassi et al. (2004).

Unlike this similarity between imposed low-latitude heating and El Niño, surface forcing in mid or high latitudes alone does not have a counterpart in a realistic climate. In climate change experiments, these latitudes are shown to warm the most, but the surface heating is nonetheless global, and expected to range from 1°–2°C at the equator to to 4°–6°C in high latitudes [see for instance Solomon et al. (2007), their Fig. 10–8]. The 2-K surface thermal forcing applied at low latitudes in this paper is therefore too strong to invite direct comparison to climate change experiments, while our applied high-latitude forcing is too weak. Imposing a surface thermal forcing that respects the temperature gradient expected from greenhouse gas–induced climate change might attenuate the dominance of the equatorial response that we find in the combined-band experiments N1N4, EQNM, and EQNMP, although qualitatively the stratospheric response in these experiments resembles what is found in CO2-doubling experiments (Winter and Bourqui 2010; McLandress and Shepherd 2009). More important, the cause of the surface temperature increase in climate change simulations is the higher atmospheric concentration of greenhouse gases, which leads to air temperature increases throughout the troposphere and to a warming of the surface. The stratospheric circulation responds to the combination of in situ radiative cooling, tropospheric radiative heating, and the increased surface temperatures. These effects are not fully separable, although one may dominate the others in specific regions or seasons (Sigmond et al. 2004). The experiments presented here are concerned only with temperature forcing at the surface.

Finally, the results we present are of course subject to the model’s limitations. In particular, ocean points north of 70°N are sufficiently cold as to be ice covered during the winter months, with or without our imposed 2-K thermal anomaly. The IGCM-FASTOC does not allow for thinning of the ice or creation of open “leads,” therefore ice albedo is unaffected by the imposed surface forcing in the N8 and N9 experiments, and fluxes of surface heat and humidity respond only to the imposed temperature anomaly without causing ice albedo feedbacks. At lower latitudes, the simple soil scheme of the IGCM-FASTOC may not respond adequately to the imposed surface heating in regions that are very dry or very wet. In CO2-doubling experiments, for instance, we have found that surface warming in the IGCM-FASTOC matches the multimodel results from Solomon et al. (2007) very well over the ocean, but is weaker (by 1–2 K) over land in the Northern Hemisphere mid and high latitudes (see Winter and Bourqui 2010). This has the potential to reduce land–sea temperature contrasts, and thus to decrease an important source of planetary waves. The implication for the experiments presented here is that our results likely represent a lower limit on the response in wave activity. Wave activity is also inhibited because surface temperatures at all gridpoints are prescribed monthly means, interpolated linearly to give daily values, and thus have neither daily nor interannual variability. Consequences of suppressing this variability are discussed in Winter and Bourqui (2011). In the stratosphere, an important limitation of the model is its use of Rayleigh friction at the top three model levels (at or above 1 hPa). It is known that Rayleigh friction can induce an artificial countercirculation between the level of the deposition of resolved waves and the model lid, which may affect the modeled circulation below (Shepherd et al. 1996). However, this contamination of the circulation is strongest at the model top, and falls to about 10% of its maximum value by about two scale heights beneath the applied drag (Shepherd et al. 1996). The middle stratosphere where we see the strongest response in zonal-mean temperature is located about three scale heights beneath the lowest level in which Rayleigh friction is applied. The generation and vertical propagation of planetary waves in the troposphere, on which we focus in sections 4 and 5, are not affected by Rayleigh friction at the model top. Finally, the model’s spectral truncation at wavenumber 31 limits its capacity to capture all synoptic-scale baroclinic eddies.

Acknowledgments

We are grateful to Jacques Derome, Seok-Woo Son, Michael Sigmond, and Jean de Grandpré for helpful discussions and suggestions and to three anonymous reviewers for their careful reading of our manuscript and constructive comments. This study was funded by the Canadian Foundation for Climate and Atmospheric Sciences and the Fonds québécois de la recherche sur la nature et les technologies.

REFERENCES

  • Bell, C., , L. Gray, , A. Charlton-Perez, , and M. Joshi, 2009: Stratospheric communication of El Niño teleconnections to European winter. J. Climate, 22, 40834096.

    • Search Google Scholar
    • Export Citation
  • Bourqui, M., , C. Taylor, , and K. Shine, 2005: A new fast stratospheric chemistry scheme in an intermediate general circulation model. II: Application to effects of future increases in greenhouse gases. Quart. J. Roy. Meteor. Soc., 131, 22432261.

    • Search Google Scholar
    • Export Citation
  • Braesicke, P., , and J. Pyle, 2004: Sensitivity of dynamics and ozone to different representations of SSTs in the unified model. Quart. J. Roy. Meteor. Soc., 130, 20332045.

    • Search Google Scholar
    • Export Citation
  • Brayshaw, D. J., , B. Hoskins, , and M. Blackburn, 2008: The storm-track response to idealized SST perturbations in an aquaplanet GCM. J. Atmos. Sci., 65, 28422860.

    • Search Google Scholar
    • Export Citation
  • Brönniman, S., , J. Luterbacher, , J. Staehelin, , T. Svendby, , G. Hansen, , and T. Svenøe, 2004: Extreme climate of the global troposphere and stratosphere in 1940–1942 related to El Niño. Nature, 431, 971974.

    • Search Google Scholar
    • Export Citation
  • Butchart, N., , S. Clough, , T. Palmer, , and P. Trevelyan, 1982: Simulations of an observed stratospheric warming with quasigeostrophic refractive index as a model diagnostic. Quart. J. Roy. Meteor. Soc., 108, 475502.

    • Search Google Scholar
    • Export Citation
  • Butchart, N., and Coauthors, 2006: Simulations of anthropogenic change in the strength of the Brewer-Dobson Circulation. Climate Dyn., 27, 727741.

    • Search Google Scholar
    • Export Citation
  • Charlton, A. J., , and L. M. Polvani, 2007: A new look at stratospheric sudden warmings. Part I: Climatology and modeling benchmarks. J. Climate, 20, 449469.

    • Search Google Scholar
    • Export Citation
  • Charney, J., , and P. Drazin, 1961: Propagation of planetary-scale disturbances from the lower to the upper atmosphere. J. Geophys. Res., 66, 83109.

    • Search Google Scholar
    • Export Citation
  • Chen, G., , and P. Zurita-Gotor, 2008: The tropospheric response to prescribed zonal forcing in an idealized atmospheric model. J. Atmos. Sci., 65, 22542271.

    • Search Google Scholar
    • Export Citation
  • Chen, G., , I. M. Held, , and W. A. Robinson, 2007: Sensitivity of the latitude of the surface westerlies to surface friction. J. Atmos. Sci., 64, 28992915.

    • Search Google Scholar
    • Export Citation
  • Chen, G., , R. A. Plumb, , and J. Lu, 2010: Sensitivities of zonal mean atmospheric circulation to SST warming in an aqua-planet model. Geophys. Res. Lett., 37, L12701, doi:10.1029/2010GL043473.

    • Search Google Scholar
    • Export Citation
  • Cohen, J., , M. Barlow, , P. K. Kushner, , and K. Saito, 2007: Stratosphere–troposphere coupling and links with Eurasian land surface variability. J. Climate, 20, 53355343.

    • Search Google Scholar
    • Export Citation
  • Fischer, A., , D. Shindell, , B. Winter, , M. Bourqui, , G. Faluvegi, , E. Rozanov, , M. Schraner, , and S. Brönniman, 2008: Stratospheric winter climate response to ENSO in three chemistry-climate models. Geophys. Res. Lett., 35, L13819, doi:10.1029/2008GL034289.

    • Search Google Scholar
    • Export Citation
  • Fletcher, C. G., , S. C. Hardiman, , P. J. Kushner, , and J. Cohen, 2009: The dynamical response to snow cover in a large ensemble of atmospheric GCM integrations. J. Climate, 22, 12081222.

    • Search Google Scholar
    • Export Citation
  • Forster, P. M. F., , and K. P. Shine, 1999: Stratospheric water vapour changes as a possible contributor to observed stratospheric cooling. Geophys. Res. Lett., 26, 33093312.

    • Search Google Scholar
    • Export Citation
  • Gerber, E. P., , and L. M. Polvani, 2009: Stratosphere–troposphere coupling in a relatively simple AGCM: The importance of stratospheric variability. J. Climate, 22, 19211933.

    • Search Google Scholar
    • Export Citation
  • Held, I. M., , and B. J. Soden, 2006: Robust responses of the hydrological cycle to global warming. J. Climate, 19, 56865699.

  • Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Elsevier, 535 pp.

  • Hoskins, B., , and A. Simmons, 1975: A multi-layer spectral model and the semi-implicit method. Quart. J. Roy. Meteor. Soc., 101, 637655.

    • Search Google Scholar
    • Export Citation
  • Lu, J., , G. Chen, , and D. M. Frierson, 2008: Response of the zonal-mean atmospheric circulation to El Niño versus global warming. J. Climate, 21, 58355851.

    • Search Google Scholar
    • Export Citation
  • McLandress, C., , and T. G. Shepherd, 2009: Simulated anthropogenic changes in the Brewer–Dobson circulation, including its extension to high latitudes. J. Climate, 22, 15161540.

    • Search Google Scholar
    • Export Citation
  • Morcrette, J.-J., 1991: Radiation and cloud radiative properties in the European Centre for Medium Range Weather Forecasts forecasting system. J. Geophys. Res., 96, 91219132.

    • Search Google Scholar
    • Export Citation
  • Newman, P. A., , E. R. Nash, , and J. E. Rosenfield, 2001: What controls the temperature of the Arctic stratosphere during the spring? J. Geophys. Res., 106, 19 99920 010.

    • Search Google Scholar
    • Export Citation
  • Olsen, M. A., , M. R. Schoeberl, , and J. E. Nielsen, 2007: Response of stratospheric circulation and stratosphere-troposphere exchange to changing sea surface temperatures. J. Geophys. Res., 112, D16104, doi:10.1029/2006JD008012.

    • Search Google Scholar
    • Export Citation
  • Peixoto, J. P., , and A. H. Oort, 1992: Physics of Climate. American Institute of Physics, 520 pp.

  • Philander, S. G., 1990: El Niño, La Niña, and the Southern Oscillation. Academic Press, 293 pp.

  • Rind, D., , D. Shindell, , P. Lonergan, , and N. Balachandran, 1998: Climate change and the middle atmosphere. Part III: The doubled CO2 climate revisited. J. Climate, 11, 876894.

    • Search Google Scholar
    • Export Citation
  • Ring, M. J., , and R. A. Plumb, 2007: Forced annular mode patterns in a simple atmospheric general circulation model. J. Atmos. Sci., 64, 36113626.

    • Search Google Scholar
    • Export Citation
  • Rosier, S. M., , and K. P. Shine, 2000: The effect of two decades of ozone change on stratospheric temperature as indicated by a general circulation model. Geophys. Res. Lett., 27, 26172620.

    • Search Google Scholar
    • Export Citation
  • Sassi, F., , D. Kinnison, , B. Boville, , R. R. Garcia, , and R. Roble, 2004: Effect of El Niño–Southern Oscillation on the dynamical, thermal, and chemical structure of the middle atmosphere. J. Geophys. Res., 109, D17108, doi:10.1029/2003JD004434.

    • Search Google Scholar
    • Export Citation
  • Scinocca, J., , and P. Haynes, 1998: Dynamical forcing of stratospheric planetary waves by tropospheric baroclinic eddies. J. Atmos. Sci., 55, 23612392.

    • Search Google Scholar
    • Export Citation
  • Shepherd, T., , K. Semeniuk, , and J. Koshyk, 1996: Sponge layer feedbacks in middle-atmosphere models. J. Geophys. Res., 101, 23 44723 464.

    • Search Google Scholar
    • Export Citation
  • Sigmond, M., , P. Siegmund, , E. Manzini, , and H. Kelder, 2004: A simulation of the separate effects of middle-atmospheric and tropospheric CO2 doubling. J. Climate, 17,23522367.

    • Search Google Scholar
    • Export Citation
  • Solomon, S., , D. Qin, , M. Manning, , M. Marquis, , K. Averyt, , M. M. B. Tignor, , H. L. Miller Jr., , and Z. Chen, Eds., 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

    • Search Google Scholar
    • Export Citation
  • Taguchi, M., , and D. L. Hartmann, 2006: Increased occurrence of stratospheric sudden warmings during El Niño as simulated by WACCM. J. Climate, 19, 324332.

    • Search Google Scholar
    • Export Citation
  • Taylor, C., , and M. Bourqui, 2005: A new fast stratospheric ozone chemistry scheme in an intermediate general circulation model. I: Description and evaluation. Quart. J. Roy. Meteor. Soc., 131, 22252242.

    • Search Google Scholar
    • Export Citation
  • Winter, B., , and M. S. Bourqui, 2010: Wave-forcing in the stratosphere under doubled-CO2 conditions in a 100-year coupled chemistry-climate model study. J. Geophys. Res., 115, D12126, doi:10.1029/2009JD012777.

    • Search Google Scholar
    • Export Citation
  • Winter, B., , and M. S. Bourqui, 2011: The impact of surface temperature variability on the climate change response in the Northern Hemisphere polar vortex. Geophys. Res. Lett., 38, L08808, doi:10.1029/2011GL047011.

    • Search Google Scholar
    • Export Citation
  • Wittman, M. A., , A. J. Charlton, , and L. M. Polvani, 2007: The effect of lower-stratospheric shear on baroclinic instability. J. Atmos. Sci., 64, 479496.

    • Search Google Scholar
    • Export Citation
Save