• Andrews, D. G., , J. R. Holton, , and C. B. Leovy, 1987: Middle Atmosphere Dynamics. Academic Press, 489 pp.

  • Boehm, M. T., , and S. Lee, 2003: The implications of tropical Rossby waves for tropical tropopause cirrus formation and for the equatorial upwelling of the Brewer–Dobson circulation. J. Atmos. Sci., 60 , 247261.

    • Search Google Scholar
    • Export Citation
  • Brewer, A. W., 1949: Evidence for a world circulation provided by the measurements of helium and water vapor distribution in the stratosphere. Quart. J. Roy. Meteor. Soc., 75 , 351363.

    • Search Google Scholar
    • Export Citation
  • Butchart, N., , and A. A. Scaife, 2001: Removal of chlorofluorocarbons by increased mass exchange between the stratosphere and troposphere in a changing climate. Nature, 410 , 799802.

    • 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
  • Chae, H. C., , and S. C. Sherwood, 2007: Annual temperature cycle of the tropical tropopause: A simple model study. J. Geophys. Res., 112 , D19111. doi:10.1029/2006JD007956.

    • Search Google Scholar
    • Export Citation
  • Christy, J. R., , and S. J. Drouilhet Jr., 1994: Variability in daily, zonal mean lower-stratospheric temperatures. J. Climate, 7 , 106120.

    • Search Google Scholar
    • Export Citation
  • Deckert, R., , and M. Dameris, 2008: Higher tropical SSTs strengthen the tropical upwelling via deep convection. Geophys. Res. Lett., 35 , L10813. doi:10.1029/2008GL033719.

    • Search Google Scholar
    • Export Citation
  • Dhomse, S., , M. Weber, , and J. Burrows, 2008: The relationship between tropospheric wave forcing and tropical lower stratospheric water vapor. Atmos. Chem. Phys., 8 , 471480.

    • Search Google Scholar
    • Export Citation
  • Dobson, G. M. B., 1956: Origin and distribution of polyatomic molecules in the atmosphere. Proc. Roy. Soc. London, 236A , 187193.

  • Eichelberger, S. J., , and D. L. Hartmann, 2005: Changes in the strength of the Brewer–Dobson circulation in a simple AGCM. Geophys. Res. Lett., 32 , L15807. doi:10.1029/2005GL022924.

    • Search Google Scholar
    • Export Citation
  • Folkins, I., , P. Bernath, , C. Boone, , G. Lesins, , N. Livesey, , A. M. Thompson, , K. Walker, , and J. C. Witte, 2006: Seasonal cycles of O3, CO, and convective outflow at the tropical tropopause. Geophys. Res. Lett., 33 , L16802. doi:10.1029/2006GL026602.

    • Search Google Scholar
    • Export Citation
  • Fu, Q., , C. M. Johanson, , J. M. Wallace, , and T. Reichler, 2006: Enhanced mid-latitude tropospheric warming in satellite measurements. Science, 312 , 1179.

    • Search Google Scholar
    • Export Citation
  • Fu, Q., , S. Solomon, , and P. Lin, 2009: On the seasonal dependence of tropical lower-stratospheric temperature trends. Atmos. Chem. Phys. Discuss., 9 , 2181921846.

    • Search Google Scholar
    • Export Citation
  • Garcia, R. R., , and W. J. Randel, 2008: Acceleration of the Brewer–Dobson circulation due to increases in greenhouse gases. J. Atmos. Sci., 65 , 27312739.

    • Search Google Scholar
    • Export Citation
  • Geller, M. A., , T. Zhou, , and K. Hamilton, 2008: Morphology of tropical upwelling in the lower stratosphere. J. Atmos. Sci., 65 , 23602374.

    • Search Google Scholar
    • Export Citation
  • Haynes, P. H., , and M. E. McIntyre, 1987: On the evolution of vorticity and potential vorticity in the presence of diabatic heating and frictional and other forces. J. Atmos. Sci., 44 , 828841.

    • Search Google Scholar
    • Export Citation
  • Haynes, P. H., , C. J. Marks, , M. E. McIntyre, , T. G. Shepherd, , and K. P. Shine, 1991: On the “downward control” of extratropical diabatic circulations by eddy-induced mean zonal forces. J. Atmos. Sci., 48 , 651678.

    • Search Google Scholar
    • Export Citation
  • Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Academic Press, 535 pp.

  • Holton, J. R., , P. H. Haynes, , M. E. McIntyre, , A. R. Douglass, , R. B. Rood, , and L. Pfister, 1995: Stratosphere–troposphere exchange. Rev. Geophys., 33 , 403439.

    • Search Google Scholar
    • Export Citation
  • Iwasaki, T., 1992: General circulation diagnosis in the pressure-isentrope hybrid vertical coordinate. J. Meteor. Soc. Japan, 70 , 673687.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kerr-Munslow, A. M., , and W. A. Norton, 2006: Tropical wave driving of the annual cycle in tropical tropopause temperatures. Part I: ECMWF analyses. J. Atmos. Sci., 63 , 14101419.

    • Search Google Scholar
    • Export Citation
  • Lanzante, J., 2009: Comment on “Trends in the temperature and water vapor content of the tropical lower stratosphere: Sea surface connection” by Karen H. Rosenlof and George C. Reid. J. Geophys. Res., 114 , D12104. doi:10.1029/2008JD010542.

    • Search Google Scholar
    • Export Citation
  • Li, F., , J. Austin, , and J. Wilson, 2008: The strength of the Brewer–Dobson circulation in a changing climate: Coupled chemistry–climate model simulations. J. Climate, 21 , 4057.

    • Search Google Scholar
    • Export Citation
  • Mote, P. W., , K. H. Rosenlof, , J. R. Holton, , R. S. Harwood, , and J. W. Waters, 1995: Seasonal variations of water vapor in the tropical lower stratosphere. Geophys. Res. Lett., 22 , 10931096.

    • Search Google Scholar
    • Export Citation
  • Mote, P. W., and Coauthors, 1996: An atmospheric tape recorder: The imprint of tropical tropopause temperatures on stratospheric water vapor. J. Geophys. Res., 101 , 39894006.

    • Search Google Scholar
    • Export Citation
  • Niwano, M., , K. Yamazaki, , and M. Shiotani, 2003: Seasonal and QBO variations of ascent rate in the tropical lower stratosphere as inferred from UARS HALOE trace gas data. J. Geophys. Res., 108 , 4794. doi:10.1029/2003JD003871.

    • Search Google Scholar
    • Export Citation
  • Norton, W. A., 2006: Tropical wave driving of the annual cycle in tropical tropopause temperatures. Part II: Model results. J. Atmos. Sci., 63 , 14201431.

    • Search Google Scholar
    • Export Citation
  • Plumb, R. A., , and J. Eluszkiewicz, 1999: The Brewer–Dobson circulation: Dynamics of the tropical upwelling. J. Atmos. Sci., 56 , 868890.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , R. Garcia, , and F. Wu, 2002a: Time-dependent upwelling in the tropical lower stratosphere estimated from the zonal-mean momentum budget. J. Atmos. Sci., 59 , 21412152.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , F. Wu, , and R. Stolarski, 2002b: Changes in column ozone correlated with the stratospheric EP flux. J. Meteor. Soc. Japan, 80 , 849862.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , M. Park, , F. Wu, , and N. Livesey, 2007: A large annual cycle in ozone above the tropical tropopause linked to the Brewer–Dobson circulation. J. Atmos. Sci., 64 , 44794488.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , R. Garcia, , and F. Wu, 2008: Dynamical balances and tropical stratospheric upwelling. J. Atmos. Sci., 65 , 35843595.

  • Reed, R. J., , and C. L. Vlcek, 1969: The annual temperature variation in the lower tropical stratosphere. J. Atmos. Sci., 26 , 163167.

  • Rosenlof, K. H., 1995: Seasonal cycle of the residual mean meridional circulation in the stratosphere. J. Geophys. Res., 100 , 51735191.

    • Search Google Scholar
    • Export Citation
  • Rosenlof, K. H., , and G. C. Reid, 2008: Trends in the temperature and water vapor content of the tropical lower stratosphere: Sea surface connection. J. Geophys. Res., 113 , D06107. doi:10.1029/2007JD009109.

    • Search Google Scholar
    • Export Citation
  • Rosenlof, K. H., , and G. C. Reid, 2009: Reply to comment by John R. Lanzante on “Trends in the temperature and water vapor content of the tropical lower stratosphere: Sea surface connection”. J. Geophys. Res., 114 , D12105. doi:10.1029/2008JD011265.

    • Search Google Scholar
    • Export Citation
  • Russell III, J. M., and Coauthors, 1993: The Halogen Occultation Experiment. J. Geophys. Res., 98 , 1077710797.

  • Salby, M. L., , and P. F. Callaghan, 2002: Interannual changes of the stratospheric circulation: Relationship to ozone and tropospheric structure. J. Climate, 15 , 36733685.

    • Search Google Scholar
    • Export Citation
  • Schoeberl, M. R., , B. N. Duncan, , A. R. Douglass, , J. Waters, , N. Livesey, , W. Read, , and M. Filipiak, 2006: The carbon monoxide tape recorder. Geophys. Res. Lett., 33 , L12811. doi:10.1029/2006GL026178.

    • Search Google Scholar
    • Export Citation
  • Schoeberl, M. R., and Coauthors, 2008: QBO and annual cycle variations in tropical lower stratosphere trace gases from HALOE and Aura MLS observations. J. Geophys. Res., 113 , D05301. doi:10.1029/2007JD008678.

    • Search Google Scholar
    • Export Citation
  • Scott, R. K., 2002: Wave-driven mean tropical upwelling in the lower stratosphere. J. Atmos. Sci., 59 , 27452759.

  • Semeniuk, K., , and T. G. Shepherd, 2001: Mechanisms for tropical upwelling in the stratosphere. J. Atmos. Sci., 58 , 30973115.

  • Shine, K. P., and Coauthors, 2003: A comparison of model-simulated trends in stratospheric temperatures. Quart. J. Roy. Meteor. Soc., 129A , 15651588.

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

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880–2006). J. Climate, 21 , 22832296.

    • Search Google Scholar
    • Export Citation
  • Spencer, R. W., , J. R. Christy, , and N. C. Grody, 1990: Global atmospheric temperature monitoring with satellite microwave measurements: Method and results 1979–84. J. Climate, 3 , 11111128.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and S. Solomon, 2005: Recent stratospheric climate trends as evidenced in radiosonde data: Global structure and tropospheric linkages. J. Climate, 18 , 47854795.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., and Coauthors, 2007: Observations: Surface and atmospheric climate change. Climate Change 2007: The Physical Science Basis. S. Solomon et al., Eds., Cambridge University Press.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131 , 29613012.

  • Yulaeva, E., , J. R. Holton, , and J. M. Wallace, 1994: On the cause of the annual cycle in tropical lower-stratospheric temperatures. J. Atmos. Sci., 51 , 169174.

    • Search Google Scholar
    • Export Citation
  • Zhou, T., , M. A. Geller, , and K. Hamilton, 2006: The roles of the Hadley circulation and downward control in tropical upwelling. J. Atmos. Sci., 63 , 27402757.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Monthly (a) mean and (b) anomaly (the departure from climatological-mean annual cycle) time series of global-mean T.

  • View in gallery

    Meridional profile of climatological (1979–2007) annual-mean T. The gray line represents the climatological annual mean of global-mean T (208.6 K).

  • View in gallery

    Meridional profiles of (a) total, (b) odd, and (c) even components of monthly mean T* (* denotes departure from global mean) for the individual months of Jan and Feb (black) and Jul and Aug (gray) during 1979–2007.

  • View in gallery

    The two leading EOFs and their corresponding standardized PC time series of the monthly mean, even component of T. EOFs 1 and 2 explain 89% and 5% of the total variance, respectively.

  • View in gallery

    Monthly mean time series of the even component of polar cap (60°N/S to pole) T, zonal wind (inverted) in the midlatitude stratosphere (40°–70°N/S, 10–30 hPa) based on monthly ERA-40 data, and tropical (equator to 30°N/S) T* (inverted), with standardized PC 1 of the monthly mean, even component of T (see Fig. 4). Peak-to-peak amplitudes are indicated on the left of each time series.

  • View in gallery

    Monthly mean time series plotted as a function of calendar month (with Jan–Jun repeated): polar cap (60°N/S to pole) T; zonal wind (inverted) in the midlatitude stratosphere (40–70°N/S, 10–30 hPa) based on monthly ERA-40 data; standardized PC 1 of monthly mean, even component of T (see Fig. 4); tropical (equator to 30°N/S) T* (inverted); EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on pentad NCEP–NCAR reanalysis data; and EPW index (vertical momentum flux averaged over 10°S–10°N, 100–300 hPa) based on monthly ERA-40 data. Peak-to-peak amplitudes are indicated on the left of each time series.

  • View in gallery

    Meridional profiles of (top) total and (bottom) even components of monthly anomaly T* for the individual months of (left) Jan/Feb and (right) Jul/Aug during 1979–2007. Profiles with equatorial T* below (above) 0°C are in black (gray).

  • View in gallery

    The two leading EOFs and their corresponding standardized PC time series of monthly anomaly, even component of T. EOFs 1 and 2 explain 55% and 25% of the total variance, respectively.

  • View in gallery

    Monthly anomaly time series of the even component of polar cap (60°N/S to pole) T, zonal wind (inverted) in the midlatitude stratosphere (40°–70°N/S, 10–30 hPa) based on monthly ERA-40 data, and tropical (equator to 30°N/S) T* (inverted), with standardized PC 1 of monthly anomaly, even component of T (see Fig. 8). Peak-to-peak amplitudes are indicated on the left of each time series.

  • View in gallery

    The leading EOFs and their corresponding standardized PC time series of monthly anomaly, even component of T* based on (top) high-pass filtered and (bottom) low-pass filtered data. EOF 1 of high-pass and low-pass filtered data explains 81% and 70% of their respective total variance.

  • View in gallery

    Monthly (a) mean and (b) anomaly time series of standardized PC 1 of even component of T (see Figs. 4 and 8) and EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on pentad NCEP–NCAR reanalysis data. Peak-to-peak amplitudes are indicated on the left of each time series.

  • View in gallery

    Monthly anomaly time series of standardized PC 1 of even component of T* and EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on (a) high-pass and (b) low-pass filtered pentad NCEP–NCAR reanalysis data. Peak-to-peak amplitudes are indicated on the left of each time series.

  • View in gallery

    Meridional profiles of (top) regression and (bottom) correlation coefficients between monthly anomaly, even component of T* and monthly anomaly time series of EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on high-pass (solid) and low-pass (dashed) filtered pentad NCEP–NCAR reanalysis data.

  • View in gallery

    A regression of the zonally averaged meridional heat flux of the previous month upon the monthly anomaly time series of tropical (30°S–30°N) T*. The contour interval is 1 m s−1 K. Positive (negative) regression coefficients are indicated by solid (dashed) lines. Correlations >0.4 are shaded in gray.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 20 20 2
PDF Downloads 13 13 3

To What Extent Does High-Latitude Wave Forcing Drive Tropical Upwelling in the Brewer–Dobson Circulation?

View More View Less
  • 1 Department of Atmospheric Sciences, University of Washington, Seattle, Washington
© Get Permissions
Full access

Abstract

The causes of the annual cycle and nonseasonal variability in the globally averaged, equator-to-pole Brewer–Dobson circulation (BDC; defined here as the equatorially symmetric component of the Lagrangian-mean meridional circulation) are investigated based on zonally averaged, lower-stratospheric temperature data from satellite-borne Microwave Sounding Unit (MSU) and Advanced Microwave Sounding Unit (AMSU). Time-varying vertical velocities in the BDC are inferred from departures of the meridional temperature profiles from the respective radiative equilibrium temperature profiles. Equatorward of ∼45°N/S, the annual-mean profile of lower-stratospheric temperature and the seasonal and nonseasonal variations about it project almost exclusively onto the equatorially symmetric component. The climatological-mean annual cycle accounts for nearly 90% of the month-to-month variance of the equatorially symmetric component of the temperature field; January/February is colder than July/August equatorward of ∼45°N/S and warmer than July/August poleward of that latitude. The equator-to-subpolar temperature contrast roughly doubles from July/August to January/February, implying an approximate doubling of the strength of the BDC. The nonseasonal variability is dominated by a similar pattern. Tropical upwelling in the BDC, as inferred from of the temperature field, varies in response to variations in eddy heat fluxes at high latitudes with comparable strength on the intraseasonal and interannual time scales; it does not appear to be correlated with equatorial tropospheric planetary wave activity or with variations in wave forcing in subtropical lower stratosphere. It is concluded that high-latitude wave forcing plays an important role in modulating tropical upwelling in the BDC across a wide range of frequencies.

Corresponding author address: Rei Ueyama, Department of Atmospheric Sciences, University of Washington, Box 351640, Seattle, WA 98195-1640. Email: rei@atmos.washington.edu

Abstract

The causes of the annual cycle and nonseasonal variability in the globally averaged, equator-to-pole Brewer–Dobson circulation (BDC; defined here as the equatorially symmetric component of the Lagrangian-mean meridional circulation) are investigated based on zonally averaged, lower-stratospheric temperature data from satellite-borne Microwave Sounding Unit (MSU) and Advanced Microwave Sounding Unit (AMSU). Time-varying vertical velocities in the BDC are inferred from departures of the meridional temperature profiles from the respective radiative equilibrium temperature profiles. Equatorward of ∼45°N/S, the annual-mean profile of lower-stratospheric temperature and the seasonal and nonseasonal variations about it project almost exclusively onto the equatorially symmetric component. The climatological-mean annual cycle accounts for nearly 90% of the month-to-month variance of the equatorially symmetric component of the temperature field; January/February is colder than July/August equatorward of ∼45°N/S and warmer than July/August poleward of that latitude. The equator-to-subpolar temperature contrast roughly doubles from July/August to January/February, implying an approximate doubling of the strength of the BDC. The nonseasonal variability is dominated by a similar pattern. Tropical upwelling in the BDC, as inferred from of the temperature field, varies in response to variations in eddy heat fluxes at high latitudes with comparable strength on the intraseasonal and interannual time scales; it does not appear to be correlated with equatorial tropospheric planetary wave activity or with variations in wave forcing in subtropical lower stratosphere. It is concluded that high-latitude wave forcing plays an important role in modulating tropical upwelling in the BDC across a wide range of frequencies.

Corresponding author address: Rei Ueyama, Department of Atmospheric Sciences, University of Washington, Box 351640, Seattle, WA 98195-1640. Email: rei@atmos.washington.edu

1. Introduction

The stratosphere is ventilated by the Brewer–Dobson circulation (BDC), which is marked by low-latitude ascent and high-latitude descent (Brewer 1949; Dobson 1956; Andrews et al. 1987). Adiabatic warming of descending air in the high-latitude branch of the BDC keeps polar temperatures in the winter hemisphere above radiative equilibrium (e.g., Andrews et al. 1987). In a similar manner, adiabatic cooling of ascending air entering the tropical lower stratosphere gives rise to a distinctive “tropical cold point” just above the 100-hPa level with temperatures on the order of −80°C. The rate of tropical upwelling, or the strength of the BDC, determines the tropical cold point temperature and consequently the water vapor mixing ratio of the “freeze-dried” air entering the stratosphere. It also controls the concentrations of long-lived chemical species of tropospheric origin in the lower stratosphere.

The existence of pronounced annual cycles in tropical cold point temperature (Reed and Vlcek 1969) and in the mixing ratios of water vapor, ozone, methane, and other chemical species in the tropical lower stratosphere (Russell et al. 1993; Mote et al. 1995, 1996; Niwano et al. 2003; Folkins et al. 2006; Schoeberl et al. 2006; Randel et al. 2007; Schoeberl et al. 2008) suggests that tropical upwelling in the BDC is substantially stronger in January than in July. The BDC also exhibits more subtle nonseasonal variability on intraseasonal and interannual time scales as discussed below. Long-term trends in the strength of the BDC and the associated rate of ventilation of the stratosphere have been the focus of studies by Butchart and Scaife (2001), Sigmond et al. (2004), Eichelberger and Hartmann (2005), Butchart et al. (2006), Li et al. (2008), Garcia and Randel (2008), and Fu et al. (2009).

The annual-mean BDC is thermally indirect with upwelling of cold air in the tropics and descent of warmer air at high latitudes and hence must be mechanically driven. Haynes and McIntyre (1987) and Haynes et al. (1991) argued that it is driven by the breaking of vertically propagating Rossby waves and gravity waves in the extratropics. In these studies, the tropical upwelling in the BDC is interpreted in terms of an extratropical suction pump that operates in accordance with the downward control principle, which states that the steady-state vertical mass flux across any given level is controlled by the vertically integrated (along angular momentum contours) zonal force above that level.

It is widely believed that wave breaking in the extratropical stratosphere plays an important role in driving the BDC, but it is not clear how much of that forcing derives from the wintertime polar night jet region and how much of the forcing is due to wave breaking at lower latitudes. The downward control principle predicts that the time-mean response to wave forcing is local in the latitude domain but that a wave forcing can exert a nonlocal control under transient conditions whereby the induced mean meridional circulation extends horizontally away from the forcing region (Haynes et al. 1991; Holton et al. 1995). Thus, it is conceivable that transient variations in the strength of the BDC and the associated variations in tropical upwelling could largely be driven by high-latitude wave forcing.

Randel et al. (2002a,b) showed observational evidence that tropical upwelling, as inferred from temperature and ozone perturbations, varies in response to week-to-week fluctuations in the high-latitude Eliassen–Palm flux, in accordance with diagnostics based on the downward control principle. Results of a concurrent observational study by Salby and Callaghan (2002) suggest that tropical upwelling varies in synchrony with the high-latitude Eliassen–Palm flux on the interannual time scale as well. Dhomse et al. (2008) showed evidence of a year-to-year modulation of tropical lower-stratospheric water vapor by mid- to high-latitude planetary wave driving as inferred from eddy heat fluxes in the extratropical stratosphere. Iwasaki (1992), Yulaeva et al. (1994), and Chae and Sherwood (2007) argued, largely on the basis of circumstantial evidence, that the annual cycle in tropical upwelling is associated with enhanced high-latitude planetary wave forcing during the northern winter. The statistically significant anticorrelations between high-latitude Eliassen–Palm flux and tropical temperature tendencies documented in Fig. 5 of Salby and Callaghan (2002) and those between extratropical eddy heat fluxes and tropical lower-stratospheric water vapor mixing ratios documented in Fig. 2 of Dhomse et al. (2008) lend credence to this hypothesis.

Based on the conventional interpretation of the downward control principle, one might expect that a forcing on time scales much longer than the radiative relaxation time should induce a localized response that resembles the steady-state response (e.g., Holton et al. 1995). However, the episodic character of the eddy forcing in the polar night jet region, as discussed by Randel et al. (2002a), could conceivably give rise to a nonlinear low-frequency response. A series of wave “bursts” that decelerate the wintertime polar night jet, interspersed with quiescent periods during which the jet strengthens in response to radiative cooling over the polar cap region, might produce a cumulative transient BDC response that resembles the adiabatic response illustrated in the top panel of Fig. 4 of Holton et al. (1995) and reproduced as Fig. 12.7 of Holton (2004). Radiative relaxation toward the equilibrium temperature profile during the quiescent intervals would also tend to force tropical upwelling, yielding a BDC response that is stronger in the time mean and more persistent than might be expected on the basis of the eddy forcing alone. Hence, it is conceivable that variations in high-latitude wave forcing could contribute to and perhaps even dominate the forcing of the time mean and/or variability in the BDC.

Diagnostics of the forcing of tropical upwelling based on the downward control principle by Rosenlof (1995), Boehm and Lee (2003), and Kerr-Munslow and Norton (2006), however, give the general impression that high-latitude wave forcing does not influence tropical upwelling on a time scale of seasons or longer. To maintain and/or perturb tropical upwelling with a realistic magnitude and structure in numerical simulations of the BDC, it appears to be necessary to prescribe some form of wave drag that extends into subtropical latitudes (Plumb and Eluszkiewicz 1999; Semeniuk and Shepherd 2001; Scott 2002; Zhou et al. 2006; Geller et al. 2008). As an alternative to high-latitude wave driving, Norton (2006) argued that seasonally varying equatorial planetary waves and their associated deep convection could force an annual cycle in tropical upwelling. Extending this line of argument, Deckert and Dameris (2008) and Rosenlof and Reid (2008) have suggested that a secular trend toward higher tropical sea surface temperatures (SSTs) and more vigorous tropical convection could act to strengthen the BDC.

In this paper, we will show additional evidence that variations in high-latitude wave forcing affect the strength of tropical upwelling in the BDC on intraseasonal, annual, and interannual time scales. To simplify the analysis of the extratropical forcing of tropical upwelling, irrespective of the hemisphere in which the forcing occurs, we decompose the BDC-related fields into equatorially symmetric and asymmetric components. We confirm the results of Salby and Callaghan (2002) and Dhomse et al. (2008), which indicate that an appreciable fraction of the interannual variability of tropical upwelling in the two-sided but globally averaged, equator-to-pole BDC (i.e., the equatorially symmetric component of the BDC) can be attributed to high-latitude eddy forcing. We will also show empirical evidence casting doubt on the notion that tropical SSTs or variations in the intensity of tropospheric baroclinic waves in the subtropics exert a strong influence on the tropical upwelling in the BDC.

Section 2 describes the data used in this study. Section 3 presents results of an analysis of the time-varying meridional profile of zonally averaged, lower-stratospheric temperature. We show that equatorward of ∼45°N/S, both the annual-mean temperature profile and the seasonal and nonseasonal variability about the mean project almost exclusively onto the equatorially symmetric component, for which the distribution of upwelling and downwelling in the BDC can be inferred quite simply from the thermodynamic energy balance (i.e., from variations in the departures of the meridional temperature profiles from the respective radiative equilibrium temperature profiles). The inference is straightforward because temporal variations in radiative forcing (apart from volcanic eruptions) are strictly seasonal and project almost exclusively onto the equatorially asymmetric component. For example, the equatorially symmetric component of the meridional profiles of January and July radiative heating are identical, apart from the small effect of the eccentricity of the earth’s orbit. Hence, the rate of upwelling and downwelling in the equatorially symmetric component of the BDC can be determined from the observed temperatures without reference to the time-varying distribution of radiative equilibrium temperature. In this study, we consider (i) the annual-mean profile, (ii) temporal variations about the mean, which are shown to be dominated by the climatological-mean annual cycle, and (iii) nonseasonal variations filtered to separate the intraseasonal and interannual variability; empirical orthogonal function (EOF) analysis of the equatorially symmetric component of the zonally averaged temperature field is used to isolate the dominant modes of variability in (ii) and (iii). It is shown that the dominant mode of variability at all frequencies is global in scale with vertical motions of opposing sign in low and high latitudes. In section 4, we document the statistically significant linear correlations between the strength of the tropical upwelling in the BDC and the high-latitude eddy forcing on intraseasonal as well as interannual time scales. In section 5, we show evidence that tropical SSTs and eddy forcing at low latitudes are not the dominant causes of variability in the strength of the tropical upwelling in the BDC. Our findings are summarized and discussed in the final section.

2. Data

The analysis is based on monthly, zonally averaged brightness temperature (T) fields derived from the lower-stratospheric channel of version 3.2 of the Microwave Sounding Unit (MSU)/Advanced Microwave Sounding Unit (AMSU) carried aboard National Oceanic and Atmospheric Administration (NOAA) satellites. The data are gridded at 2.5° × 2.5° resolution and extend to 87.5°N/S. The period of record 1979–2007 is used in this study. The weighting function of the lower-stratospheric channel is concentrated mainly in the 15–19 km (30–150 hPa) layer, as detailed in the Remote Sensing Systems Web site (see www.ssmi.com/msu/).

The monthly mean time series of global-mean T (weighted by cosine of latitude) shown in Fig. 1a exhibits a well-defined annual cycle that is highly reproducible from year to year, with the exception of those following the eruptions of El Chichón (1982) and Mt. Pinatubo (1991). As documented by Spencer et al. (1990), Christy and Drouilhet (1994), and Yulaeva et al. (1994), the climatological-mean annual cycle of global-mean T exhibits a maximum between August and September and a minimum between January and February, with peak-to-peak amplitude of ∼0.5 K. The posteruption intervals stand out even more clearly in the anomaly time series (i.e., the departures from the climatological-mean annual cycle) shown in Fig. 1b. The anomaly time series also exhibits a cooling trend due to stratospheric ozone depletion, changes in water vapor content, and the buildup of greenhouse gases (Shine et al. 2003; Thompson and Solomon 2005). A similar plot appears as Fig. 3.17 of Trenberth et al. (2007). Apart from these features, the nonseasonal variability of global-mean T is very small.

In the analysis that follows, we will make use of the variable T* defined as the local T minus the global-mean T (i.e., the departures from each month’s global-mean T). Using T* in place of T effectively eliminates the signatures of the volcanic eruptions and the cooling trend of the global stratosphere in the interannual variability while having only a very small influence on the local T.

We also make use of global European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40; Uppala et al. 2005) and National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research (NCAR) Reanalysis (Kalnay et al. 1996) data for the 23-yr period of record, 1979–2001, to investigate the relationship between lower-stratospheric temperature and the forcing of the mean flow by wave activity. The variables analyzed were zonally averaged zonal wind u, meridional eddy heat flux vT ′, and the vertical eddy flux of zonal momentum uw′. The zonal wind and vertical eddy fluxes of zonal momentum are based on monthly mean ERA-40 data, the highest temporal resolution of the ECMWF Reanalysis data that was available to us at the time the study was conducted. To include the contribution of the variability with periods shorter than a month, the meridional eddy heat fluxes were computed based on nonoverlapping, 5-day (pentad)-mean NCEP–NCAR Reanalysis data rather than the monthly mean ERA-40 data. The meridional eddy heat fluxes in the two datasets were compared and found to be very similar, so the reliability of the dataset used to represent them is not an issue. To compute the eddy fluxes, each variable is first decomposed into zonal means and departures from the zonal mean, and the departures of the two variables are multiplied to yield a flux time series. The climatological (23-yr) monthly and pentad means are then subtracted from their respective time series to obtain time series of the nonseasonal variability. Finally, the eddy fluxes are zonally averaged; in the case of vT ′, successive pentads are averaged to create monthly mean data.

3. Results

a. Annual mean

The meridional profile of the annual-mean T is shown in Fig. 2, superimposed on a reference line of the annual-mean, global-mean T (208.6 K). Consistent with prior results of Christy and Drouilhet (1994), T increases with latitude up to ∼55°N/S, beyond which it decreases toward the pole in both hemispheres. Annual-mean T within the deep tropics (equatorward of ∼15°N/S) is uniform at ∼203.8 K, approximately 5 K below the global mean. The coolness of the tropical belt relative to subpolar latitudesmust be a consequence of the eddy-driven, thermally indirect BDC that drives temperatures away from the radiative equilibrium temperature profile.

In the annual mean, the radiative equilibrium temperature profile in the lower stratosphere is almost perfectly equatorially symmetric. The annual-mean T mirrors this symmetry, except at latitudes poleward of ∼55°N/S. The weaker subpolar-to-polar temperature gradient in the Northern Hemisphere compared to the Southern Hemisphere reflects the larger land–sea temperature contrasts and rougher orography of the Northern Hemisphere, which give rise to more vigorous wave breaking in the stratosphere, resulting in a stronger poleward mass flux in the BDC with stronger descent and consequently stronger adiabatic warming over the northern polar cap region (Iwasaki 1992; Chae and Sherwood 2007). The more pronounced polar minimum in the Southern Hemisphere may also be related to the radiative impact of the loss of stratospheric ozone in recent decades.

b. The variability about the mean

The top panel of Fig. 3 shows meridional profiles of T* (departures from each month’s global-mean T) for the individual months of January and February (hereafter Jan/Feb) and July and August (hereafter Jul/Aug) during the 29-yr period of record. The Jan/Feb and Jul/Aug profiles cross at ∼45° latitude in both hemispheres. Equatorward of this crossing latitude, temperatures are consistently lower in Jan/Feb than Jul/Aug and their profiles all exhibit the same nearly parabolic shape that resembles the BDC-induced structure in the annual-mean T (Fig. 2). The compactness of the Jan/Feb and Jul/Aug bundles of profiles indicates that the seasonal variations in T* are much larger than the year-to-year variations, except at high northern latitudes during winter. Removing the equatorial stratospheric quasi-biennial oscillation signal by linear regression (not shown) further reduces the year-to-year variability in T* within the equatorial belt, but only slightly. Poleward of ∼45°N/S, winters are colder than summers, with smaller seasonal contrast in the Northern Hemisphere than in the Southern Hemisphere. The anomalously high temperatures over the Arctic in Jan/Feb of some years (e.g., 1987) are due to the occurrence of sudden warming events.

We now decompose the meridional temperature profiles into even and odd components with respect to symmetry about the equator. The even (equatorially symmetric) component is defined as the mean of zonally averaged temperatures at corresponding latitudes in the Northern and Southern Hemispheres, (TNH + TSH)/2. The odd (equatorially asymmetric) component is defined as half the difference between zonally averaged temperatures at corresponding latitudes in the Northern and Southern Hemispheres, which are (TNHTSH)/2 and (TSHTNH)/2, respectively. By construction, the odd component of the meridional profiles is identically equal to zero on the equator at all times. In fact, the odd component of the Jan/Feb and Jul/Aug T* profiles shown in Fig. 3b are remarkably flat out to beyond 20°N/S; the low-latitude variability in T* is almost entirely captured by the even component shown in Fig. 3c.

Given that the annual cycle in the radiative equilibrium temperature projects almost exclusively onto the odd component, as explained near the end of section 1, it follows that the seasonal variability in the even component of temperature must be almost exclusively dynamically driven;1 the difference between the Jan/Feb and Jul/Aug meridional profiles of the even component of T* in Fig. 3c is a response to differences between the eddy-driven, equator-to-pole BDC in those months. The equator-to-subpolar temperature contrast in the even component of T and T* is almost twice as large in Jan/Feb as in Jul/Aug. The two sets of profiles intersect at ∼45°N/S; Jan/Feb is colder (warmer) than Jul/Aug equatorward (poleward) of this crossing latitude. Hence, the annual cycle in T* can be characterized as an enhancement of the equator-to-subpolar temperature contrast during the northern winter, which is presumably caused by the strengthening of the BDC. Tropical upwelling in the BDC, as inferred from the concave meridional profiles of the even component of T in Fig. 3c, is stronger during Jan/Feb than in Jul/Aug at latitudes extending from the equator to as far poleward as ∼45°N/S.

To place these results concerning the annual cycle in the context of the total variability, EOF analysis is performed on the even component of T (including the global mean), weighting the data in proportion to their contributions to the area-weighted variance (i.e., by the square root of the cosine of latitude). The EOF patterns displayed in the paper are obtained by regressing the even component of the temperature field upon the respective standardized principal component (PC) time series. Thus, the amplitudes of the patterns provide an indication of the typical amplitudes of the anomalies observed in association with fluctuations in the respective EOF modes.

The leading mode of variability, shown in the top panels of Fig. 4, accounts for 89% of the total month-to-month variance of the even-component T field. The corresponding standardized PC 1 time series is dominated by the annual cycle, with extrema in Jan/Feb and Jul/Aug, consistent with the profiles in Fig. 3c. During positive excursions of PC 1 observed in Jan/Feb of each year, temperatures equatorward of the crossing latitude at 45°N/S are below the annual mean by ∼2°C over most of this latitudinal range. The fluctuations in PC 1 are almost perfectly sinusoidal in shape and quite reproducible from one cycle to the next.

Most of the remaining variance of the even component of T is accounted for by the second EOF shown in the lower panel of Fig. 4, which has positive loadings at all latitudes. The corresponding PC time series closely mirrors the behavior of the global-mean T anomaly (Fig. 1b). The bulge at ∼30°N/S resembles the meridional profile of the observed temperature trend from 1979 to 2005 documented by Fu et al. (2006). When EOF analysis is performed on the T* field from which the variance associated with fluctuations in global-mean T is removed, a virtually identical leading EOF/PC is obtained that accounts for 93% of the variance (not shown).

In view of the dominance of the leading EOF of the even component of T, its PC time series, repeated in Fig. 5, is a useful index of the fluctuations in the strength of the equator-to-pole BDC. Also shown in Fig. 5 are time series of the even component of tropical (equator to 30°N/S) T*, polar cap (60°N/S to the pole) T, and zonal wind in the midlatitude stratosphere (40°–70°N/S, 10–30 hPa). The tropical T* and zonal wind time series are inverted to make them easier to compare with the other series. Correlations between each of these time series are shown in Table 1. The variability in all four of the time series is dominated by the annual cycle, which is quite smooth and regular from year to year, even in the presence of the quasi-biennial oscillation and El Niño–Southern Oscillation (ENSO). The similarity between the polar cap and inverted tropical temperature time series confirms the high degree of cancellation between low- and high-latitude temperature variations noted by Yulaeva et al. (1994). The midlatitude stratospheric zonal winds are strongly correlated with temperature indices of the BDC, as required by thermal wind balance; months of strong equator-to-pole temperature gradient are marked by strong (westerly) vertical shear and vice versa. Upon careful inspection of Fig. 5, it is evident that the similarity between the time series extends beyond the annual cycle. Subtle differences in the shapes of the seasonal evolution from one year to the next are also similar in the four time series. These relationships are explored in further detail in the next subsection.

Figure 6 shows the same time series plotted as a function of calendar month, with individual years superimposed on the same time axis. This mode of presentation emphasizes the shape of the annual march, and it reveals more clearly the seasonal dependence of the year-to-year variability of the time series. The four time series are remarkably similar, both with respect to the mean and the relative amplitude and seasonality of the interannual variability. Nonseasonal variability, as manifested in the diversity of the annual marches, tends to be much larger during Jan/Feb than Jul/Aug.

c. Nonseasonal variability

The top panels of Fig. 7 show meridional profiles of monthly T* anomalies (departures from the climatological-mean annual cycle) for Jan/Feb and Jul/Aug of each calendar year in the 29-yr period of record. Compared with the monthly mean profiles of T* (Fig. 3a), the amplitudes of the anomaly profiles are smaller by about a factor of 2 within the tropics. It is evident from an inspection of the even component of the profiles (shown in the bottom panels of Fig. 7) that anomalously low tropical temperatures are not always accompanied by anomalous high-latitude warmth (and vice versa) as in Fig. 3c. However, there is still a tendency for anticorrelation between low- and high-latitude temperatures.

Modes that reflect the structure of the anomalies about the seasonally varying climatological mean are obtained by performing EOF analysis on the even component of the monthly T anomaly field. The leading EOF of the even component of the T anomaly, shown in the top panel of Fig. 8, accounts for 55% of the total variance. This mode is dominated by the annual cycle and resembles the profile of EOF 1 of the monthly mean variability (Fig. 4), except that the crossing latitude is 47.5°N/S, a few degrees poleward of the crossing latitude in Fig. 4. Its counterpart for T* (not shown) is virtually identical and accounts for 72% of the variance. The second mode, shown in the bottom panels of Fig. 8, is dominated by variations in global-mean temperature, as reflected in the exclusively positive loadings of EOF 2, as well as in the resemblance between its PC 2 time series and global-mean T anomaly time series (Fig. 1b). Monthly anomaly time series of the various indices of the BDC are shown in Fig. 9. The correlations between these anomaly time series (Table 2a) are not as strong as those for the time series that include the annual cycle (Table 1), but they are statistically significant at above the 99% confidence level, based on the one-tailed Student’s t test with the number of degrees of freedom computed from the autocorrelations of the time series.

The meridional structure of the intraseasonal and interannual variability in the BDC is examined separately by applying successive centered five- and three-month running mean smoothing operators to the even component of the monthly anomaly fields to form the low-pass filtered version, which represents the interannual variability. To exclude the contribution of volcanic eruptions and the long-term cooling trend in lower-stratospheric temperatures to the interannual variability, the analysis is performed on the T* field. The corresponding high-pass filtered series, obtained by subtracting the low-pass filtered time series from the unfiltered series, represents the intraseasonal variability.

The leading EOFs of the even component of high-pass and low-pass filtered T*, shown in Fig. 10, both explain large fractions of the total variance (81% and 70%, respectively). The high-pass filtered pattern is characterized by remarkably uniform amplitudes equatorward of the crossing latitude at 50°N/S and much larger temperature anomalies of opposing sign poleward of that latitude. In the corresponding low-pass filtered pattern, the crossing latitude is 37.5°N/S and the amplitude equatorward of this latitude is not as uniform as that in the high-pass filtered pattern. These results indicate that the dominant structures of the seasonal and nonseasonal variability in the equator-to-pole BDC are similar, both consisting of “see-saws” with crossing latitudes of 37.5°N/S for the intraseasonal variability, 45°N/S for the annual cycle, and 50°N/S for the interannual variability.

4. The role of high-latitude wave forcing

In this section, we examine the direct connection between high-latitude wave forcing and tropical (equator to 30°N/S) lower-stratospheric temperature. As a measure of high-latitude wave forcing, we use an Eliassen–Palm (EP) flux index based on the zonally averaged eddy heat flux υT ′ averaged over 50°–80°N/S (weighted by cosine of latitude) and the 50-, 30-, 20-, and 10-hPa levels, which roughly correspond to the region occupied by the zonally averaged polar night jet. Since eddy heat flux is directly proportional to the vertical component of the EP flux, the high-latitude mean υT ′ is commonly used to indicate the amount of wave activity entering the extratropical stratosphere. Our wave-forcing index is a linear combination of the current and the previous month’s values, where the relative weights are determined by a least squares best fit designed to maximize the correlation with the PC 1 time series shown in Figs. 4, 5, 8, and 9; the weights for the anomaly time series are 0.31 for the current month and 0.60 for the previous month. The resulting wave-forcing index has a one-month autocorrelation of 0.35.

Monthly mean and anomaly time series of the EP flux index are shown in Figs. 11a,b together with the corresponding PC 1 time series, repeated from Figs. 4, 5, 8, and 9. The annual cycle in the EP flux index is nearly in phase with the annual cycle in PC 1, with the strongest forcing during the northern winter. Shown in Tables 1 and 2a are the correlation coefficients between the EP flux index and various indices of the strength of the BDC. The correlations between the EP flux index and PC 1 are quite strong for both the mean and anomaly time series (0.82 and 0.65, respectively), as are the corresponding correlations between the EP flux index and tropical T* (−0.83 and −0.65, respectively). Hence, it is evident that variations in high-latitude wave forcing account for an appreciable fraction of the temporal variance of the strength of the tropical upwelling in the BDC.

Figure 12 shows monthly anomaly time series of the EP flux index and PC 1 time series based on high-pass and low-pass filtered data, as defined in section 3c. The resemblance is striking at both frequencies; correlations are both in the range 0.6 to 0.7, as shown in Tables 2b,c. High-pass and low-pass filtered EP flux indices are both well correlated with tropical temperature as well as with the other filtered BDC indices. To show that these high correlations are not artifacts of the filtering process, filtered versions of the various indices were recomputed based on annualized time series consisting of seasonal November through March means, which represent the time of year when the EP flux indices are highest (Fig. 6). The correlation between the tropical T* and the EP flux index time series based on these data is −0.72. To further substantiate this result, Fig. 13 shows meridional profiles of regression coefficients (top panel) and correlation coefficients (bottom panel) between the even component of T* and the EP flux index based on high-pass and low-pass filtered data. The profiles of the regression coefficients resemble the meridional structure of the leading EOFs (Fig. 10). The correlation coefficients in the tropical belt are comparable for the high-pass and low-pass filtered data, with values on the order of −0.6. The slight dip in the correlation in the equatorial belt relative to the subtropics for the low-pass filtered data may be due to the presence of the quasi-biennial oscillation.

The EP flux index plotted as a function of calendar month in Fig. 6 clearly shows the extended “active season” of the planetary waves, which extends from the month of the breakdown of the Antarctic polar night jet in November through the northern winter. During this season, the fluxes are large and highly variable from month to month and from year to year.

5. The role of other forcings

a. The role of forcing from below

Modeling studies of Boehm and Lee (2003) and Norton (2006) showed that the upward propagation of equatorial planetary waves (EPWs) from the troposphere is capable of influencing the rate of upwelling across the tropical tropopause. An enhancement of the EPWs leads to enhanced wave breaking in the lower stratosphere, which induces an easterly acceleration of the mean flow at that level. Enhanced equatorial upwelling in the layer below is required to maintain thermal wind balance. Norton (2006) proposed that the annual cycle in the strength of the BDC is induced by a weakening of the EPWs during the northern summer, when the core of deep tropical convection shifts away from the equator because of enhanced heating over the Tibetan plateau. Indeed, the vertical momentum flux in the equatorial belt (10°S–10°N, 100–300 hPa), shown in Fig. 6, exhibits a pronounced minimum during Jul/Aug that is no less reproducible from one year to the next than the EP flux indices. On the other hand, it seems unlikely that the seasonality of EPWs could be the primary cause of the pronounced annual cycle or the nonseasonal variability in the BDC; nonseasonal variability in these fluxes is virtually uncorrelated with nonseasonal BDC variability (r = −0.03) as represented by PC 1 of the even component of monthly T anomaly field (Figs. 8 and 9). It is conceivable that vertical momentum flux as resolved by daily data could yield a correlation that is higher than that based on monthly data. Even so, the role of the EPWs in modulating the strength of the BDC is questionable since the EPWs are restricted to much lower latitudes than the BDC-related upwelling, as can be inferred from the crossing latitudes of the leading EOFs (Figs. 4, 8, and 10). In numerical experiments of Norton (2006), the response of the tropical upwelling to fluctuations in the EPW forcing was found to be comparable to the width of the equatorial waveguide itself (∼15°N/S), whereas the crossing latitude of the T profile in the annual cycle is ∼45°N/S.

Rosenlof and Reid (2008) suggested that the downward trends in tropical lower-stratospheric temperature and water vapor content over the past few decades could be occurring in response to rising SSTs in the western Pacific warm pool. By way of evidence, they noted a maximum anticorrelation (r = −0.44) between monthly tropical cold point temperature anomalies averaged over a region slightly to the west of the western Pacific warm pool and SST anomalies averaged over the warm pool region. To determine whether this correlation is indicative of a coherent pattern of SST variations that is linked to variations in the strength of the BDC, we regressed gridded data of the extended reconstruction SST version 3 (ERSST.v3; Smith et al. 2008) upon PC 1 of the nonseasonal variability. While positive regression coefficients prevail over most of the western Pacific warm pool region, the values range only up to 0.15 and the regression pattern (not shown) lacks spatial coherence. Variability of the BDC (PC 1 time series) and SST anomalies averaged over the warm pool region (10°S–10°N, 120°E–180°) are weakly positively correlated (r = 0.17 and 0.16 for unfiltered and low-pass filtered time series, respectively). The stronger local anticorrelations between SST and cold point temperature reported by Rosenlof and Reid (2008) may have more to do with the influence of SST anomalies on the EPW than on the zonally symmetric, equator-to-pole BDC; deep convection, large-scale ascent, elevated geopotential heights in the upper troposphere, and depressed cold point temperatures all tend to overlie regions of positive SST anomalies. Lanzante (2009) has argued that the cooling trend in the tropical lower stratosphere documented by Rosenlof and Reid (2008) and Rosenlof and Reid (2009) is an artifact related to a change in the radiosonde instrument type during the last decade or so.

b. The role of eddy forcing in subtropical lower stratosphere

Rosenlof (1995) concluded that much of the forcing of the climatological-mean annual cycle in tropical lower-stratospheric temperatures is due to wave breaking at ∼30° latitude. Randel et al. (2008) showed evidence supporting this view based on climatological-mean eddy flux statistics. They noted that substantial EP flux divergences at northern subtropical latitudes extend as high as 70-hPa level during the northern winter, inducing strong tropical upwelling, while wave activity at corresponding southern latitudes during the southern winter does not extend as far. If the subtropical momentum flux convergence by these waves does indeed play an important role in the annual cycle in the BDC, it is reasonable to expect that nonseasonal variability in the EP fluxes associated with these waves might force some of the nonseasonal variability in the BDC as well. To test this hypothesis, we regressed the total (even and odd components combined) field of vT ′ for the previous month upon the monthly anomaly time series of tropical T* (Fig. 9). Consistent with the results of the previous section, the regression pattern shown in Fig. 14 exhibits prominent maxima in the vicinity of the polar night jet, with the Northern and Southern Hemispheres contributing in roughly equal measure to the forcing of the tropical upwelling. However, no feature of comparable strength is evident at subtropical latitudes, as would be expected if tropospheric waves penetrating into the lower stratosphere in this region were playing a prominent role in forcing the nonseasonal variability in the BDC. Nor are statistically significant subtropical features evident in patterns obtained by regressing uv′ and EP flux divergence fields upon the tropical T* time series (not shown). Further work is needed to reconcile these results with conclusions of earlier works.

6. Summary and discussion

Variations in MSU/AMSU zonally averaged, lower-stratospheric temperatures (T) have been decomposed into even and odd components with respect to their symmetry about the equator. Seasonal variations in radiative equilibrium temperature are dominated by the odd component. In this study, we have focused on variations in the even component, whose variability is mainly in response to variations in the eddy-driven Brewer–Dobson circulation (BDC). The distribution of upwelling and downwelling in the BDC is qualitatively inferred from the observed seasonal and nonseasonal variations in the meridional profiles of the even component of temperature. Whereas most previous studies have considered the eddy driving in the Northern and Southern Hemispheres separately, here they are combined in the “even component.”

We find that nearly 90% of the month-to-month variance of the temporal variations about the annual-mean reference profile, with ascent at low latitudes and descent at high latitudes, is accounted for by a single mode of variability dominated by the annual cycle. The equator-to-subpolar temperature contrast roughly doubles from Jul/Aug to Jan/Feb, implying an approximate factor of 2 difference in the strength of the BDC between those months. The crossing latitude in the annual cycle (i.e., the latitude poleward of which T is warmer and the descent in the BDC is presumably stronger in Jan/Feb than in Jul/Aug) is located at 45°N/S. The leading mode of the nonseasonal variability in the BDC, which accounts for over half of the total variance about a seasonally varying reference profile, exhibits a meridional structure similar to that of the climatological-mean annual cycle. The dominance of a planetary-scale structure of the BDC variability, with a single crossing latitude and relatively uniform amplitudes equatorward of that latitude, prevails on intraseasonal, annual, and interannual time scales. The only notable frequency-dependent feature is the crossing latitude, which is 12.5° farther poleward in the intraseasonal variability than in the interannual variability.

Three different indices of the strength of the BDC have been shown to vary in synchrony with one another in the annual cycle and in the nonseasonal variability across a wide range of frequencies: (i) inverted tropical temperature, (ii) polar cap temperature, and (iii) inverted zonal wind in the midlatitude stratosphere. These time series all exhibit a strong lagged response to month-to-month variations in the intensity of high-latitude wave forcing represented by the EP flux index constructed from a weighted sum of current and 1-month previous eddy heat fluxes in the region 50°–80°N/S and 10–50 hPa. Equally strong relationships are observed on intraseasonal and interannual time scales, as evidenced by visual inspection of the time series and supporting calculations of the temporal correlation coefficients. The meridional structures of the leading EOFs of the even component of T* based on the filtered nonseasonal time series closely mirror the corresponding profiles of T* regressed on an index of high-latitude eddy forcing; in particular, crossing latitudes for high-pass and low-pass filtered profiles are located at 50°N/S and 37.5°N/S, respectively.

We have shown that year-to-year variations in tropical upwelling in the BDC are significantly correlated with variations in the high-latitude EP flux index, thereby confirming and extending previous findings of Randel et al. (2002a,b), Salby and Callaghan (2002), and Dhomse et al. (2008). Fu et al. (2009) offer additional evidence, showing that the 30-yr trends in tropical MSU/AMSU lower-stratospheric temperatures are highly correlated with dynamically induced trends in high-latitude temperatures on a calendar month–by–calendar month basis. The strong similarity between the meridional structure of the leading EOFs of the even component of T* (Fig. 10) and the pattern derived by regressing the even component of T* upon the time series of the EP flux index (Fig. 13) suggests that high-latitude eddy forcing strongly influences the equator-to-pole BDC on both intraseasonal and interannual time scales. These results support the notion that the pronounced annual cycle in the strength of the BDC is a consequence of the larger land–sea temperature contrasts and rougher orography in the Northern Hemisphere compared to the Southern Hemisphere, which give rise to stronger eddy forcing at high latitudes during the northern winter and thereby induce stronger tropical upwelling.

We have also considered the possible roles of equatorial planetary waves and wave forcing in the subtropical lower stratosphere in influencing the variability of the BDC, but found no significant temporal correlations between the eddy forcing and tropical upwelling on either interannual or intraseasonal time scales. The locations of the crossing latitudes identified in this study would be difficult to explain on the basis of tropical or subtropical eddy forcing alone.

Acknowledgments

We thank Drs. Qiang Fu, William Randel, Dargan Frierson, Karen Rosenlof, Conway Leovy, and Mark Weber, as well as an anonymous reviewer, for many discussions and helpful comments over the course of this work. MSU/AMSU data are produced by Remote Sensing Systems and sponsored by the NOAA Climate and Global Change Program. This work was supported by the Climate Dynamics Program Office of the National Science Foundation under Grant 0812802.

REFERENCES

  • Andrews, D. G., , J. R. Holton, , and C. B. Leovy, 1987: Middle Atmosphere Dynamics. Academic Press, 489 pp.

  • Boehm, M. T., , and S. Lee, 2003: The implications of tropical Rossby waves for tropical tropopause cirrus formation and for the equatorial upwelling of the Brewer–Dobson circulation. J. Atmos. Sci., 60 , 247261.

    • Search Google Scholar
    • Export Citation
  • Brewer, A. W., 1949: Evidence for a world circulation provided by the measurements of helium and water vapor distribution in the stratosphere. Quart. J. Roy. Meteor. Soc., 75 , 351363.

    • Search Google Scholar
    • Export Citation
  • Butchart, N., , and A. A. Scaife, 2001: Removal of chlorofluorocarbons by increased mass exchange between the stratosphere and troposphere in a changing climate. Nature, 410 , 799802.

    • 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
  • Chae, H. C., , and S. C. Sherwood, 2007: Annual temperature cycle of the tropical tropopause: A simple model study. J. Geophys. Res., 112 , D19111. doi:10.1029/2006JD007956.

    • Search Google Scholar
    • Export Citation
  • Christy, J. R., , and S. J. Drouilhet Jr., 1994: Variability in daily, zonal mean lower-stratospheric temperatures. J. Climate, 7 , 106120.

    • Search Google Scholar
    • Export Citation
  • Deckert, R., , and M. Dameris, 2008: Higher tropical SSTs strengthen the tropical upwelling via deep convection. Geophys. Res. Lett., 35 , L10813. doi:10.1029/2008GL033719.

    • Search Google Scholar
    • Export Citation
  • Dhomse, S., , M. Weber, , and J. Burrows, 2008: The relationship between tropospheric wave forcing and tropical lower stratospheric water vapor. Atmos. Chem. Phys., 8 , 471480.

    • Search Google Scholar
    • Export Citation
  • Dobson, G. M. B., 1956: Origin and distribution of polyatomic molecules in the atmosphere. Proc. Roy. Soc. London, 236A , 187193.

  • Eichelberger, S. J., , and D. L. Hartmann, 2005: Changes in the strength of the Brewer–Dobson circulation in a simple AGCM. Geophys. Res. Lett., 32 , L15807. doi:10.1029/2005GL022924.

    • Search Google Scholar
    • Export Citation
  • Folkins, I., , P. Bernath, , C. Boone, , G. Lesins, , N. Livesey, , A. M. Thompson, , K. Walker, , and J. C. Witte, 2006: Seasonal cycles of O3, CO, and convective outflow at the tropical tropopause. Geophys. Res. Lett., 33 , L16802. doi:10.1029/2006GL026602.

    • Search Google Scholar
    • Export Citation
  • Fu, Q., , C. M. Johanson, , J. M. Wallace, , and T. Reichler, 2006: Enhanced mid-latitude tropospheric warming in satellite measurements. Science, 312 , 1179.

    • Search Google Scholar
    • Export Citation
  • Fu, Q., , S. Solomon, , and P. Lin, 2009: On the seasonal dependence of tropical lower-stratospheric temperature trends. Atmos. Chem. Phys. Discuss., 9 , 2181921846.

    • Search Google Scholar
    • Export Citation
  • Garcia, R. R., , and W. J. Randel, 2008: Acceleration of the Brewer–Dobson circulation due to increases in greenhouse gases. J. Atmos. Sci., 65 , 27312739.

    • Search Google Scholar
    • Export Citation
  • Geller, M. A., , T. Zhou, , and K. Hamilton, 2008: Morphology of tropical upwelling in the lower stratosphere. J. Atmos. Sci., 65 , 23602374.

    • Search Google Scholar
    • Export Citation
  • Haynes, P. H., , and M. E. McIntyre, 1987: On the evolution of vorticity and potential vorticity in the presence of diabatic heating and frictional and other forces. J. Atmos. Sci., 44 , 828841.

    • Search Google Scholar
    • Export Citation
  • Haynes, P. H., , C. J. Marks, , M. E. McIntyre, , T. G. Shepherd, , and K. P. Shine, 1991: On the “downward control” of extratropical diabatic circulations by eddy-induced mean zonal forces. J. Atmos. Sci., 48 , 651678.

    • Search Google Scholar
    • Export Citation
  • Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Academic Press, 535 pp.

  • Holton, J. R., , P. H. Haynes, , M. E. McIntyre, , A. R. Douglass, , R. B. Rood, , and L. Pfister, 1995: Stratosphere–troposphere exchange. Rev. Geophys., 33 , 403439.

    • Search Google Scholar
    • Export Citation
  • Iwasaki, T., 1992: General circulation diagnosis in the pressure-isentrope hybrid vertical coordinate. J. Meteor. Soc. Japan, 70 , 673687.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kerr-Munslow, A. M., , and W. A. Norton, 2006: Tropical wave driving of the annual cycle in tropical tropopause temperatures. Part I: ECMWF analyses. J. Atmos. Sci., 63 , 14101419.

    • Search Google Scholar
    • Export Citation
  • Lanzante, J., 2009: Comment on “Trends in the temperature and water vapor content of the tropical lower stratosphere: Sea surface connection” by Karen H. Rosenlof and George C. Reid. J. Geophys. Res., 114 , D12104. doi:10.1029/2008JD010542.

    • Search Google Scholar
    • Export Citation
  • Li, F., , J. Austin, , and J. Wilson, 2008: The strength of the Brewer–Dobson circulation in a changing climate: Coupled chemistry–climate model simulations. J. Climate, 21 , 4057.

    • Search Google Scholar
    • Export Citation
  • Mote, P. W., , K. H. Rosenlof, , J. R. Holton, , R. S. Harwood, , and J. W. Waters, 1995: Seasonal variations of water vapor in the tropical lower stratosphere. Geophys. Res. Lett., 22 , 10931096.

    • Search Google Scholar
    • Export Citation
  • Mote, P. W., and Coauthors, 1996: An atmospheric tape recorder: The imprint of tropical tropopause temperatures on stratospheric water vapor. J. Geophys. Res., 101 , 39894006.

    • Search Google Scholar
    • Export Citation
  • Niwano, M., , K. Yamazaki, , and M. Shiotani, 2003: Seasonal and QBO variations of ascent rate in the tropical lower stratosphere as inferred from UARS HALOE trace gas data. J. Geophys. Res., 108 , 4794. doi:10.1029/2003JD003871.

    • Search Google Scholar
    • Export Citation
  • Norton, W. A., 2006: Tropical wave driving of the annual cycle in tropical tropopause temperatures. Part II: Model results. J. Atmos. Sci., 63 , 14201431.

    • Search Google Scholar
    • Export Citation
  • Plumb, R. A., , and J. Eluszkiewicz, 1999: The Brewer–Dobson circulation: Dynamics of the tropical upwelling. J. Atmos. Sci., 56 , 868890.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , R. Garcia, , and F. Wu, 2002a: Time-dependent upwelling in the tropical lower stratosphere estimated from the zonal-mean momentum budget. J. Atmos. Sci., 59 , 21412152.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , F. Wu, , and R. Stolarski, 2002b: Changes in column ozone correlated with the stratospheric EP flux. J. Meteor. Soc. Japan, 80 , 849862.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , M. Park, , F. Wu, , and N. Livesey, 2007: A large annual cycle in ozone above the tropical tropopause linked to the Brewer–Dobson circulation. J. Atmos. Sci., 64 , 44794488.

    • Search Google Scholar
    • Export Citation
  • Randel, W. J., , R. Garcia, , and F. Wu, 2008: Dynamical balances and tropical stratospheric upwelling. J. Atmos. Sci., 65 , 35843595.

  • Reed, R. J., , and C. L. Vlcek, 1969: The annual temperature variation in the lower tropical stratosphere. J. Atmos. Sci., 26 , 163167.

  • Rosenlof, K. H., 1995: Seasonal cycle of the residual mean meridional circulation in the stratosphere. J. Geophys. Res., 100 , 51735191.

    • Search Google Scholar
    • Export Citation
  • Rosenlof, K. H., , and G. C. Reid, 2008: Trends in the temperature and water vapor content of the tropical lower stratosphere: Sea surface connection. J. Geophys. Res., 113 , D06107. doi:10.1029/2007JD009109.

    • Search Google Scholar
    • Export Citation
  • Rosenlof, K. H., , and G. C. Reid, 2009: Reply to comment by John R. Lanzante on “Trends in the temperature and water vapor content of the tropical lower stratosphere: Sea surface connection”. J. Geophys. Res., 114 , D12105. doi:10.1029/2008JD011265.

    • Search Google Scholar
    • Export Citation
  • Russell III, J. M., and Coauthors, 1993: The Halogen Occultation Experiment. J. Geophys. Res., 98 , 1077710797.

  • Salby, M. L., , and P. F. Callaghan, 2002: Interannual changes of the stratospheric circulation: Relationship to ozone and tropospheric structure. J. Climate, 15 , 36733685.

    • Search Google Scholar
    • Export Citation
  • Schoeberl, M. R., , B. N. Duncan, , A. R. Douglass, , J. Waters, , N. Livesey, , W. Read, , and M. Filipiak, 2006: The carbon monoxide tape recorder. Geophys. Res. Lett., 33 , L12811. doi:10.1029/2006GL026178.

    • Search Google Scholar
    • Export Citation
  • Schoeberl, M. R., and Coauthors, 2008: QBO and annual cycle variations in tropical lower stratosphere trace gases from HALOE and Aura MLS observations. J. Geophys. Res., 113 , D05301. doi:10.1029/2007JD008678.

    • Search Google Scholar
    • Export Citation
  • Scott, R. K., 2002: Wave-driven mean tropical upwelling in the lower stratosphere. J. Atmos. Sci., 59 , 27452759.

  • Semeniuk, K., , and T. G. Shepherd, 2001: Mechanisms for tropical upwelling in the stratosphere. J. Atmos. Sci., 58 , 30973115.

  • Shine, K. P., and Coauthors, 2003: A comparison of model-simulated trends in stratospheric temperatures. Quart. J. Roy. Meteor. Soc., 129A , 15651588.

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

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880–2006). J. Climate, 21 , 22832296.

    • Search Google Scholar
    • Export Citation
  • Spencer, R. W., , J. R. Christy, , and N. C. Grody, 1990: Global atmospheric temperature monitoring with satellite microwave measurements: Method and results 1979–84. J. Climate, 3 , 11111128.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and S. Solomon, 2005: Recent stratospheric climate trends as evidenced in radiosonde data: Global structure and tropospheric linkages. J. Climate, 18 , 47854795.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., and Coauthors, 2007: Observations: Surface and atmospheric climate change. Climate Change 2007: The Physical Science Basis. S. Solomon et al., Eds., Cambridge University Press.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131 , 29613012.

  • Yulaeva, E., , J. R. Holton, , and J. M. Wallace, 1994: On the cause of the annual cycle in tropical lower-stratospheric temperatures. J. Atmos. Sci., 51 , 169174.

    • Search Google Scholar
    • Export Citation
  • Zhou, T., , M. A. Geller, , and K. Hamilton, 2006: The roles of the Hadley circulation and downward control in tropical upwelling. J. Atmos. Sci., 63 , 27402757.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Monthly (a) mean and (b) anomaly (the departure from climatological-mean annual cycle) time series of global-mean T.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 2.
Fig. 2.

Meridional profile of climatological (1979–2007) annual-mean T. The gray line represents the climatological annual mean of global-mean T (208.6 K).

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 3.
Fig. 3.

Meridional profiles of (a) total, (b) odd, and (c) even components of monthly mean T* (* denotes departure from global mean) for the individual months of Jan and Feb (black) and Jul and Aug (gray) during 1979–2007.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 4.
Fig. 4.

The two leading EOFs and their corresponding standardized PC time series of the monthly mean, even component of T. EOFs 1 and 2 explain 89% and 5% of the total variance, respectively.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 5.
Fig. 5.

Monthly mean time series of the even component of polar cap (60°N/S to pole) T, zonal wind (inverted) in the midlatitude stratosphere (40°–70°N/S, 10–30 hPa) based on monthly ERA-40 data, and tropical (equator to 30°N/S) T* (inverted), with standardized PC 1 of the monthly mean, even component of T (see Fig. 4). Peak-to-peak amplitudes are indicated on the left of each time series.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 6.
Fig. 6.

Monthly mean time series plotted as a function of calendar month (with Jan–Jun repeated): polar cap (60°N/S to pole) T; zonal wind (inverted) in the midlatitude stratosphere (40–70°N/S, 10–30 hPa) based on monthly ERA-40 data; standardized PC 1 of monthly mean, even component of T (see Fig. 4); tropical (equator to 30°N/S) T* (inverted); EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on pentad NCEP–NCAR reanalysis data; and EPW index (vertical momentum flux averaged over 10°S–10°N, 100–300 hPa) based on monthly ERA-40 data. Peak-to-peak amplitudes are indicated on the left of each time series.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 7.
Fig. 7.

Meridional profiles of (top) total and (bottom) even components of monthly anomaly T* for the individual months of (left) Jan/Feb and (right) Jul/Aug during 1979–2007. Profiles with equatorial T* below (above) 0°C are in black (gray).

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 8.
Fig. 8.

The two leading EOFs and their corresponding standardized PC time series of monthly anomaly, even component of T. EOFs 1 and 2 explain 55% and 25% of the total variance, respectively.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 9.
Fig. 9.

Monthly anomaly time series of the even component of polar cap (60°N/S to pole) T, zonal wind (inverted) in the midlatitude stratosphere (40°–70°N/S, 10–30 hPa) based on monthly ERA-40 data, and tropical (equator to 30°N/S) T* (inverted), with standardized PC 1 of monthly anomaly, even component of T (see Fig. 8). Peak-to-peak amplitudes are indicated on the left of each time series.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 10.
Fig. 10.

The leading EOFs and their corresponding standardized PC time series of monthly anomaly, even component of T* based on (top) high-pass filtered and (bottom) low-pass filtered data. EOF 1 of high-pass and low-pass filtered data explains 81% and 70% of their respective total variance.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 11.
Fig. 11.

Monthly (a) mean and (b) anomaly time series of standardized PC 1 of even component of T (see Figs. 4 and 8) and EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on pentad NCEP–NCAR reanalysis data. Peak-to-peak amplitudes are indicated on the left of each time series.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 12.
Fig. 12.

Monthly anomaly time series of standardized PC 1 of even component of T* and EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on (a) high-pass and (b) low-pass filtered pentad NCEP–NCAR reanalysis data. Peak-to-peak amplitudes are indicated on the left of each time series.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 13.
Fig. 13.

Meridional profiles of (top) regression and (bottom) correlation coefficients between monthly anomaly, even component of T* and monthly anomaly time series of EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on high-pass (solid) and low-pass (dashed) filtered pentad NCEP–NCAR reanalysis data.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Fig. 14.
Fig. 14.

A regression of the zonally averaged meridional heat flux of the previous month upon the monthly anomaly time series of tropical (30°S–30°N) T*. The contour interval is 1 m s−1 K. Positive (negative) regression coefficients are indicated by solid (dashed) lines. Correlations >0.4 are shaded in gray.

Citation: Journal of the Atmospheric Sciences 67, 4; 10.1175/2009JAS3216.1

Table 1.

Correlation coefficients (>99% significance level) between monthly mean time series of the even components of tropical T* (equator to 30°N/S); polar cap T (60°N/S to pole); midlatitude stratospheric zonal wind based on monthly ERA-40 data (40°–70°N/S, 10–30 hPa); standardized PC 1 of monthly mean, even component of T (see Fig. 4); and EP flux index constructed from lag 0 and −1 month eddy heat fluxes (50°–80°N/S, 10–50 hPa) based on pentad NCEP–NCAR reanalysis data.

Table 1.

Table 2a. As in Table 1, but for monthly anomaly time series of the even components of tropical T*, polar cap T, zonal wind U, EP flux index, and standardized PC 1 of monthly anomaly (see Fig. 8).

i1520-0469-67-4-1232-t02a

Table 2b. As in Table 2a, but for high-pass filtered data.

i1520-0469-67-4-1232-t02b

Table 2c. As in Table 2a, but for low-pass filtered data.

i1520-0469-67-4-1232-t02c

1

Nonseasonal variations in T* at high latitudes of the Northern and Southern Hemispheres are virtually uncorrelated and hence project equally on the even and odd components of the variability. The even component is emphasized in this study because it captures virtually all of the anticorrelation between low- and high-latitude temperatures that exists by virtue of the equator-to-pole BDC. The nonseasonal variability in the odd component of the high-latitude temperature is largely unrelated to the nonseasonal variability in tropical temperature.

Save