• Allen, D. R., , R. M. Bevilacqua, , G. E. Nedoluha, , C. E. Randall, , and G. L. Manney, 2003: Unusual stratospheric transport and mixing during the 2002 Antarctic winter. Geophys. Res. Lett., 30 .1599, doi:10.1029/2003GL017117.

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

  • Baldwin, M., , T. Hirooka, , A. O’Neill, , and S. Yoden, 2003: Major stratospheric warming in the Southern Hemisphere in 2002: Dynamical aspects of the ozone hole split. SPARC Newsletter, No. 20, SPARC Office, Toronto, ON, Canada, 24–26.

    • Search Google Scholar
    • Export Citation
  • Charney, J. G., , and M. E. Stern, 1962: On the stability of internal baroclinic jets in a rotating atmosphere. J. Atmos. Sci., 19 , 159172.

    • Search Google Scholar
    • Export Citation
  • Fairlie, T. D. A., , and A. O’Neill, 1988: The stratospheric major warming of winter 1984/85: Observations and dynamical inferences. Quart. J. Roy. Meteor. Soc., 114 , 557577.

    • Search Google Scholar
    • Export Citation
  • Fairlie, T. D. A., , M. Fisher, , and A. O’Neill, 1990: The development of narrow baroclinic zones and other small-scale structure in the stratosphere during simulated major warmings. Quart. J. Roy. Meteor. Soc., 116 , 287315.

    • Search Google Scholar
    • Export Citation
  • Farrara, J., , M. Fisher, , C. R. Mechoso, , and A. O’Neill, 1992: Planetary-scale disturbances in the southern stratosphere during early winter. J. Atmos. Sci., 49 , 17571775.

    • Search Google Scholar
    • Export Citation
  • Garcia, R., , T. J. Dunkerton, , R. S. Liebermann, , and R. A. Vincent, 1997: Climatology of the semiannual oscillation of the tropical middle atmosphere. J. Geophys. Res., 102 , 2601926032.

    • Search Google Scholar
    • Export Citation
  • Gray, L. J., , S. Sparrow, , M. Juckes, , A. O’Neill, , and D. G. Andrews, 2003: Flow regimes in the winter stratosphere of the northern hemisphere. Quart. J. Roy. Meteor. Soc., 129 , 925945.

    • Search Google Scholar
    • Export Citation
  • Harwood, R. S., 1975: The temperature structure of the southern hemisphere stratosphere, August–October 1971. Quart. J. Roy. Meteor. Soc., 101 , 7592.

    • Search Google Scholar
    • Export Citation
  • Holton, J. R., , and H-C. Tan, 1982: The influence of the equatorial quasi biennial oscillation on the global circulation at 50 mb. J. Meteor. Soc. Japan, 60 , 140148.

    • Search Google Scholar
    • Export Citation
  • Hoppel, K., , R. Bevilacqua, , D. Allen, , G. Nedoluha, , and C. Randall, 2003: POAM III observations of the anomalous 2002 Antarctic ozone hole. Geophys. Res. Lett., 30 .1394, doi:10.1029/2003GL016899.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., 1991: Towards a PV-Theta view of the general circulation. Tellus, 43 , 2735.

  • Hoskins, B. J., , M. E. McIntyre, , and A. W. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 111 , 877946.

    • Search Google Scholar
    • Export Citation
  • Kushner, P. J., , and L. M. Polvani, 2005: A very large, spontaneous stratospheric sudden warming in a simple AGCM: A prototype for the Southern Hemisphere warming of 2002? J. Atmos. Sci., 62 , 890897.

    • Search Google Scholar
    • Export Citation
  • Labitzke, K. G., , and H. van Loon, 1999: The Stratosphere: Phenomena, History and Relevance. Springer-Verlag, 179 pp.

  • Lahoz, W. A., and Coauthors, 1996: Vortex dynamics and the evolution of water vapour in the stratosphere of the southern hemisphere. Quart. J. Roy. Meteor. Soc., 122 , 423450.

    • Search Google Scholar
    • Export Citation
  • Leovy, C. B., , and P. J. Webster, 1976: Stratospheric long waves: Comparison of thermal structure in the Northern and Southern Hemispheres. J. Atmos. Sci., 33 , 16241638.

    • Search Google Scholar
    • Export Citation
  • Lorenc, A. C., and Coauthors, 2000: The Met. Office global three-dimensional variational data assimilation scheme. Quart. J. Roy. Meteor. Soc., 126 , 29913012.

    • Search Google Scholar
    • Export Citation
  • Mechoso, C. R., , A. O’Neill, , V. D. Pope, , and J. D. Farrara, 1988: A study of the stratospheric final warming of 1982 in the southern hemisphere. Quart. J. Roy. Meteor. Soc., 114 , 13651384.

    • Search Google Scholar
    • Export Citation
  • O’Neill, A., 2003: Stratospheric sudden warmings. Encyclopedia of Atmospheric Sciences, J. R. Holton, J. A. Curry, and J. A. Pyle, Eds., Academic Press, 1342–1353.

    • Search Google Scholar
    • Export Citation
  • O’Neill, A., , and V. D. Pope, 1988: Simulations of linear and nonlinear disturbances in the stratosphere. Quart. J. Roy. Meteor. Soc., 114 , 10631110.

    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., 1981: Diagnostic study of a wavenumber-2 stratospheric sudden warming in the transformed Eulerian-mean formalism. J. Atmos. Sci., 38 , 844855.

    • Search Google Scholar
    • Export Citation
  • Plumb, R. A., 1981: Instability of the distorted polar night jet: A theory of stratospheric warmings. J. Atmos. Sci., 38 , 25142531.

  • Simmons, A., , M. Hortal, , G. Kelly, , A. McNally, , A. Untch, , and S. Uppala, 2005: ECMWF analyses and forecasts of stratospheric winter vortex breakup: September 2002 in the Southern Hemisphere and related events. J. Atmos. Sci., 62 , 668689.

    • Search Google Scholar
    • Export Citation
  • Swinbank, R., , and A. O’Neill, 1994: A stratosphere–troposphere data assimilation system. Mon. Wea. Rev., 122 , 686702.

  • Taguchi, M., , and S. Yoden, 2002: Internal interannual variations of the troposphere–stratosphere coupled system in a simple global model. Part II: Millennium integrations. J. Atmos. Sci., 59 , 30373050.

    • Search Google Scholar
    • Export Citation
  • Tung, K. K., , and R. S. Lindzen, 1979: A theory of stationary long waves. Part II: Resonant Rossby waves in the presence of realistic vertical shears. Mon. Wea. Rev., 107 , 735750.

    • Search Google Scholar
    • Export Citation
  • Waugh, D. W., , and D. G. Dritschel, 1991: The stability of filamentary vorticity in two-dimensional geophysical vortex-dynamics models. J. Fluid Mech., 231 , 575598.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Schematic of the typical circulation in the middle stratosphere of the SH during (a) Sep and (b) Oct. Anticyclones are shaded; cyclones are not shaded

  • View in gallery

    Fields of geopotential height in the middle stratosphere of the SH during 2002 on (a) 5 Aug, (b) 22 Aug, (c) 30 Aug, and (d) 11 Sep. Units are km, and data are from UKMO

  • View in gallery

    Fields of geopotential height (contours) and temperature (colors) for the 10-hPa surface of the SH during 2002 on (a) 21 Sep, (b) 24 Sep, (c) 26 Sep, and (d) 8 Oct. Units are km for geopotential height; K for temperature. Data are from UKMO

  • View in gallery

    Same as in Fig. 3, but for the 100-hPa surface

  • View in gallery

    Fields of geopotential height (contours) and temperature (colors) for the 215-hPa surface of the SH during 2002 on (a) 24 and (b) 26 Sep. Units are km for geopotential height; K for temperature. Data are from UKMO

  • View in gallery

    Zonal-mean wind (m s−1) at 10 hPa near 60°S during 2002. Shading indicates the range of values in the period 1992–2002. Data are from UKMO

  • View in gallery

    Fields of Ertel’s PV on the 850-K isentropic surface (near 10 hPa) of the SH during 2002 at (a) 1200 UTC on 22 Sep, (b) 1200 UTC on 25 Sep, (c) 1200 UTC on 27 Sep, (d) 0000 UTC, 30 Sep, (e) 1200 UTC on 2 Oct, (f) 0000 UTC on 4 Oct, (g) 1800 UTC on 6 Oct, and (h) 1200 UTC on 9 Oct. Data are from ECMWF

  • View in gallery

    (Continued)

  • View in gallery

    Fields of Ertel’s PV on isentropic surfaces in the upper troposphere/lower stratosphere of the SH: (left) the 395-K surface at (top) 1200 UTC on 25 Sep and (bottom) 1200 UTC on 27 Sep; (right) 1200 UTC on 25 Sep for the (top) 350- and (bottom) 315-K surfaces. Data are from ECMWF

  • View in gallery

    Evolution of the vertical component of the EP flux at 60°S on the 100-hPa surface during 2002. The shaded region denotes the range of values during the period 1992–2001. Units are kg s−2. Data are from UKMO

  • View in gallery

    Evolution of rms variance of geopotential height around the latitude circle at 60°S during 2002. Units are m. Data are from UKMO

  • View in gallery

    Evolution of zonal-mean winds at the equator during 2002. Units are m s−1. Data are from UKMO

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 144 143 15
PDF Downloads 111 111 9

The Splitting of the Stratospheric Polar Vortex in the Southern Hemisphere, September 2002: Dynamical Evolution

View More View Less
  • 1 Department of Meteorology, University of Reading, Reading, United Kingdom
© Get Permissions
Full access

Abstract

The polar vortex of the Southern Hemisphere (SH) split dramatically during September 2002. The large-scale dynamical effects were manifest throughout the stratosphere and upper troposphere, corresponding to two distinct cyclonic centers in the upper troposphere–stratosphere system. High-resolution (T511) ECMWF analyses, supplemented by analyses from the Met Office, are used to present a detailed dynamical analysis of the event. First, the anomalous evolution of the SH polar vortex is placed in the context of the evolution that is usually witnessed during spring. Then high-resolution fields of potential vorticity (PV) from ECMWF are used to reveal several dynamical features of the split. Vortex fragments are rapidly sheared out into sheets of high (modulus) PV, which subsequently roll up into distinct synoptic-scale vortices. It is proposed that the stratospheric circulation becomes hydrodynamically unstable through a significant depth of the troposphere–stratosphere system as the polar vortex elongates.

Corresponding author address: Prof. Alan O’Neill, Department of Meteorology, University of Reading, Whiteknights, Reading, RG6 6BB, United Kingdom. Email: alan@met.reading.ac.uk

Abstract

The polar vortex of the Southern Hemisphere (SH) split dramatically during September 2002. The large-scale dynamical effects were manifest throughout the stratosphere and upper troposphere, corresponding to two distinct cyclonic centers in the upper troposphere–stratosphere system. High-resolution (T511) ECMWF analyses, supplemented by analyses from the Met Office, are used to present a detailed dynamical analysis of the event. First, the anomalous evolution of the SH polar vortex is placed in the context of the evolution that is usually witnessed during spring. Then high-resolution fields of potential vorticity (PV) from ECMWF are used to reveal several dynamical features of the split. Vortex fragments are rapidly sheared out into sheets of high (modulus) PV, which subsequently roll up into distinct synoptic-scale vortices. It is proposed that the stratospheric circulation becomes hydrodynamically unstable through a significant depth of the troposphere–stratosphere system as the polar vortex elongates.

Corresponding author address: Prof. Alan O’Neill, Department of Meteorology, University of Reading, Whiteknights, Reading, RG6 6BB, United Kingdom. Email: alan@met.reading.ac.uk

1. Introduction

On 24 September 2002, the stratospheric polar vortex in the Southern Hemisphere (SH) split in two, as did the ozone hole that it harbored at its center. This event, which has all the hallmarks of a so-called major warming, was a complete surprise to atmospheric scientists, much as the discovery of an Antarctic ozone hole had been almost 2 decades earlier. It was well known that the stratosphere of the Southern Hemisphere was often dynamically active on a planetary scale, especially in spring, exhibiting a sequence of so-called minor warmings in the run up to the seasonal breakdown of the polar vortex, or final warming, in late spring. But major warmings, characterized by a breakdown of the polar vortex in winter or early spring, were thought to be confined to the stratosphere of the Northern Hemisphere, where planetary waves are generally stronger, owing to stronger tropospheric wave amplitudes.

That the first observed major warming in the Southern Hemisphere was of the “wave-2” type (with two cyclonic vortices diametrically opposed across the South Pole) is also remarkable, since wave-2 major warmings are much rarer than “wave-1” major warmings (single vortex shifted off the pole) in the Northern Hemisphere (see, e.g., the review by O’Neill 2003).

This paper analyzes the dynamics of the event using meteorological datasets that include high-resolution fields of potential vorticity (PV) that give a higher-resolution view of an observed stratospheric warming than has been seen before. Although brief studies of the SH 2002 event have been published by Allen et al. (2003) and Hoppel et al. (2003) we seek to provide a more complete account of the three-dimensional phenomenology of the event for this special issue.

2. Meteorological analyses

The global meteorological analyses used in this paper derive from two sources: the U.K. Met Office (UKMO) and the European Centre for Medium-Range Weather Forecasts (ECMWF). UKMO analyses are used to portray the basic meteorological development of the SH 2002 event and to construct time series of variables that place the event in the context of other years. The ECMWF analyses shown here are only recently available and are used to portray the event with isentropic maps of Ertel’s PV. They give a higher-resolution “action movie” of an actual stratospheric warming than has been seen before.

a. UKMO analyses

The analyses and their construction have been described in detail by Swinbank and O’Neill (1994). They comprise global, gridded fields of temperature, geopotential height, and wind components on a 2.5° latitude by 3.75° longitude grid. These fields are held on 22 pressure levels from 1000 to 0.316 hPa, at standard values of pressure adopted by the National Aeronautics and Space Administration (NASA) Upper Atmosphere Research Satellite (UARS) project. The analyses are produced by means of data assimilation, incorporating observations used in operational weather forecasting (no UARS observations are included in the analyses used here). On 15 November 2000, the assimilation was improved from the analysis correction method, described by Swinbank and O’Neill, to a three-dimensional variational data assimilation (3DVAR) method, described by Lorenc et al. (2000). The analyses are daily (1200 UTC), spanning the period 17 October 1991 to the present.

b. ECMWF analyses

High-resolution PV fields are derived from analyses produced by the ECMWF operational numerical weather prediction system, described by Simmons et al. (2005). Horizontal resolution is equivalent to T511, corresponding to a smallest resolved horizontal wavelength of 40 km. There are 60 levels between the earth’s surface and 70 km. The analyses are produced using a multivariate, four-dimensional variational data assimilation system (4DVAR), with the variational algorithm applied to observational data over 12-h periods. In the stratosphere, ozone is assimilated (but not water vapor). The high-resolution system was introduced in 2001, and the analyses are available for the four main synoptic hours: 0000, 0600, 1200, and 1800 UTC.

3. The split of the SH polar vortex

a. Typical flow regime in the SH stratosphere during spring

The dynamics of the split of the SH polar vortex needs to be considered in the context of its usual springtime evolution. This evolution has been described by Mechoso et al. (1988). While there is year-to-year variability in the SH stratosphere, Mechoso et al. noted several characteristic features of the springtime flow regime.

Figure 1 is a schematic diagram that attempts to depict these features. In September, the stratospheric flow is dominated at all levels by a large westerly vortex. A key point to note for later is that the vortex is often elongated, in the presence of a pair of eastward-traveling anticyclones about 180° apart in longitude. These anticyclones are at different stages of their life cycle (so that at times only one may be present). As described by Mechoso et al. (1988), they tend to form over the Indian Ocean (near 90°E), travel eastward around the polar vortex (possibly as a result of an eastward component to the velocity field induced by the PV associated with the polar vortex), and start to decay over the Pacific (near 90°W). A life cycle of “anticyclogenesis”, eastward translation, and decay repeats through September and early October on a time scale of about 10 days. The accompanying steady eastward rotation of the axis of elongation of the polar vortex results in an eastward-traveling wave 2 in the geopotential height field, as noted by Harwood (1975) and Leovy and Webster (1976). Though the polar vortex has occasionally been seen to be highly elongated when straddled by a pair of unusually strong anticyclones, its splitting is evidently a rare, if not unprecedented, event.

This flow regime generally changes in October, when an eastward-traveling anticyclone, instead of decaying over the Pacific, becomes quasi-stationary in a preferred region in or near to the quadrant 90°E–180°, where it persists (Fig. 1b). The polar vortex weakens from the top downward as cyclonic PV is drawn away by the now intensifying quasi-stationary anticyclone (Farrara et al. 1992; Lahoz et al. 1996). This transition in flow regime leads to the replacement of westerly winds by easterly winds in the stratosphere (the “final warming”) around about the same time each year, with a variability of a few weeks (Farrara et al. 1992).

b. Change in the stratospheric flow regime a month before the vortex split

Toward the end of August 2002, a month before the vortex split, the flow regime of the stratosphere seemed already to be in an anomalous state compared with that for typical years. Figure 2 shows fields of geopotential height at 10 hPa during August and early September. During the first week of August (Fig. 2a), the middle stratosphere exhibited the characteristic flow regime of a typical September, sketched in Fig. 1a. The polar vortex was elongated in the presence of a pair of eastward-moving transient anticyclones.

By the last week of August, however, this mobile flow regime changed: an eastward-moving anticyclone became quasi-stationary in the quadrant 90°E–180° (Fig. 2b). This quadrant was noted above to be the preferred region where a strong quasi-stationary anticyclone develops during October in a typical year (cf. Fig. 2b with the schematic Fig. 1b). Thereafter, anticyclones continued to develop upstream, travel eastward, and eventually merge with the quasi-stationary anticyclone, thereby reinforcing it (through the combination of two areas of anticyclonic PV). Two instances are shown in late August and early September in Figs. 2c and 2d. This strengthening of the quasi-stationary anticyclone by vortex merger, which in 2002 took place during September, was observed by Mechoso et al. (1988) to take place during a typical October. It seems, therefore, that in the run up to the split of the polar vortex, the flow regime in the stratosphere around springtime exhibited the characteristic flow regimes and change in regime of other years, but about 1 month earlier than normal.

c. Meteorological overview of the splitting of the vortex

An overview of the splitting of the vortex, as depicted by fields of geopotential height and temperature at 10 hPa, is shown in Fig. 3. On 21 September 2002, the cold stratospheric vortex was weakening and was displaced from the pole in the presence of an intensifying, quasi-stationary anticyclone, lying mostly in the quadrant 90°E–180°, the preferred location for a quasi-stationary anticyclone at the time of the final warming (Fig. 3a). As the anticyclone further intensified, the polar vortex elongated, and by 24 September it was splitting (Fig. 3b). Accompanying rapid local temperature increases are attributable to strong, localized adiabatic descent associated with ageostrophic motions in intensifying westerly jet streams (e.g., Fairlie et al. 1990). By 26 September, the polar vortex had split, at 10 hPa (Fig. 3c), into two vortices of similar intensity, with a “tongue” of anticyclonic circulation stretching across the pole, and with a region of locally high temperature between each vortex and the anticyclone. By 8 October, a single, much-weakened vortex was reestablished over the pole, accompanied by a quasi-stationary anticyclone near its preferred location (Fig. 3d).

A clear split in the vortex extended up to at least 1 hPa in the upper stratosphere. In the lower stratosphere, the polar vortex did not actually split—in the sense that a continuous westerly jet continued to encircle the pole—but two distinct cyclonic centers were present in the lower stratosphere at the time of the split aloft. Figure 4 shows a sequence of maps of geopotential height and temperature at 100 hPa, matching in time those of Fig. 3. On 21 September, there was a single cyclone over the pole (Fig. 4a). The intense anticyclone at 10 hPa noted above is (by hydrostatic balance) an upper-level manifestation of the large temperature perturbation in the quadrant 90°E–180° at 100 hPa. By 24 September, as the vortex aloft was beginning to split, cyclogenesis near 90°W over the Drake Passage (between South America and Antarctica) led to a second cyclonic center (Fig. 4b), this two-center pattern persisting while the polar vortex split aloft. By 8 October, a single circumpolar vortex was reestablished, and the zonal asymmetry in the temperature pattern was weaker than in the run up to the vortex split.

The formation of two distinct cyclonic centers can be traced into the upper troposphere, as shown in Fig. 5 for the 215-hPa surface on 24 and 26 September. In particular, the cyclonic center resulting from cyclogenesis over the Drake Passage can be seen. At this level, the zonal asymmetry in the geopotential and temperature patterns are, however, weaker than they are aloft. That the stratospheric warming was a so-called major warming according to the World Meteorological Organization (WMO) definition—which requires a reversal of zonal-mean winds from westerlies to easterlies at 10 hPa at 60° latitude (e.g., Labitzke and van Loon 1999)—is confirmed by the time series of zonal-mean wind at 10 hPa shown in Fig. 6. Also shown (by shading) is the range of values of this quantity during the previous decade. Zonal-mean winds decelerated below this range shortly after mid-August 2002 as a quasi-stationary anticyclone developed in the stratosphere (described in section 3b). Zonal-mean winds remained below this range for most of the next 4 weeks, decelerating sharply at the end of September to reach an easterly wind speed in excess of 20 m s−1. Much weaker zonal-mean westerly winds were reestablished during October as a cyclonic remnant of the polar vortex moved toward the pole. The seasonal reversal of zonal-mean westerlies to easterlies, or final warming, occurred at the end of October at about the usual time.

d. Dynamical features of the split revealed by high-resolution fields of PV

The arguments for using isentropic fields of PV to analyze meteorological flows are now well known (see Hoskins et al. 1985). The sequence of high-resolution (T511) PV fields from ECMWF on the 850-K isentropic surface, displayed in Fig. 7, shows the split of the vortex in dramatic fashion, revealing dynamical aspects that cannot be discerned from the figures shown earlier. On 22 September (Fig. 7a), just before the split, the PV field shows a highly elongated polar vortex (blue) surrounded by a “membrane” of strong PV gradients (yellow). (The analogy with a cell membrane seems appropriate since strong PV gradients resist lateral displacements.) The adjacent quasi-stationary anticyclone, centered between Australia and Antarctica, was a region of weak PV gradients where mixing of air drawn off the vortex (yellow streamers) and air drawn from the extratropics (red streamers) was taking place. Similar mixing was also taking place in a much weaker anticyclonic circulation over South America. The polar vortex was, in effect, squeezed between the asymmetrical pincers formed by the strong and weak anticyclones.

On 25 September (Fig. 7b), the polar vortex had split in two, save for a narrow streamer of air (yellow) from the edge of the vortex (a remnant of the membrane). As the anticyclone from the Australian side moved between them, coalescing with the anticyclone over South America to form one large anticyclone, the cyclonic vortices separated and underwent strong quasi-horizontal shearing (Fig. 7c for 27 September), during which high (modulus) PV was drawn around the anticyclone. This stretching out process rapidly weakened the two cyclonic vortices: the vortex in the Western Hemisphere was almost completely destroyed (apart from small remnants), leaving the vortex in the Eastern Hemisphere at less than half the size of the original polar vortex.

By 30 September (Fig. 7d), thin streamers of vortex air stretched over planetary distances. Since the gradient of PV was of opposite sign on either side of the streamers (Charney and Stern 1962) and, additionally, the streamers were located a large distance away from the shear induced by the main part of the vortex (Waugh and Dritschel 1991), the necessary condition for hydrodynamic instability was met. The streamers broke up into individual, synoptic-scale cyclonic vortices (Figs. 7e and 7f). A few of these small-scale remnants drifted into the cross-polar jet stream between the remaining vortex and the anticyclone (Figs. 7g and 7h), and their associated PV was wrapped around the outside of the vortex, so that what was once inner-vortex air became outer-vortex air.

At the time of the split in the middle stratosphere, Fig. 8 shows PV distributions for isentropic surfaces in the upper troposphere and lower stratosphere—the region dubbed the “middle world” by Hoskins (1991). (The isentropic surfaces cross the sloping dynamical tropopause at midlatitudes, but nowhere do they reach the ground.) On the 395-K surface, the vortex does not quite split (the outer membrane of strong PV gradients, colored yellow, remained intact), but there has been a separation of inner-vortex air (colored dark blue) into two distinct air masses. On isentropic surfaces a few kilometers lower in altitude, the PV distribution was much more convoluted, and the double-lobe structure of PV at higher altitudes was no longer evident.

Figure 8 (bottom left) also shows the PV distribution on the 395-K isentropic surface shortly after the split. The separated air masses have started to recombine, with the air mass from the Western Hemisphere wrapping cyclonically around the air mass in the Eastern Hemisphere, resulting in strong mixing within the outer membrane of strong PV gradients.

The central issue of the dynamical connection between the fields depicted in Figs. 7 and 8 will be taken up in section 4.

4. Discussion

The splitting of the SH polar vortex in September 2002 bears a strong resemblance to examples of the wave-2 type of major warming that have been recorded in the Northern Hemisphere, for instance, during February 1979 (Palmer 1981), January 1985 (Fairlie and O’Neill 1988), and February 1999 (our own analysis; not shown). If one regards the warming as taking place on a time scale of about 10 days, then in all of these cases, the split develops from a state in which the polar vortex is already elongated and therefore much more liable to being split than an axially symmetric vortex would be. Also, in all cases, this vortex elongation was associated with the presence of a strong, quasi-stationary anticyclone in the stratosphere, giving a strong prior wave-1 pattern (recall Figs. 3a and 7a for September 2002). The tendency noted by Mechoso et al. (1988) for the SH polar vortex to be elongated during spring, during the run up to the final warming, renders it more susceptible to splitting, even though splitting is evidently a very rare event in the Southern Hemisphere.

O’Neill and Pope (1988) have emphasized that, on the 10-day time scale, the flow preceding a major warming is strongly asymmetric and that strong warmings develop through the nonlinear interactions of large, deep vortices associated with large-scale asymmetry in the tropospheric circulation. For the event in September 2002, a dynamical association with the underlying topography is indicated by the apparent “locking on” of a strong anticyclone to a preferred geographical location for such features, which took place at the end of August and continued through September.

But what actually triggered the split? It was noted in section 3c that a double-lobed structure developed in the PV distribution in the lowermost stratosphere, associated with localized cyclogenesis over the Drake Passage. There was a local PV anomaly associated with this cyclogenesis (Fig. 8), which would have induced a PV anomaly aloft. Because the cyclogenesis occurred under the tip of an elongated stratospheric vortex, the induced anomaly would have enhanced the cyclonic circulation at the tip and may have sufficed to cut off a distinct cyclone from the remnant of the polar vortex. According to this provisional hypothesis, the split was driven from below when an event in the lower atmosphere met with a favorable asymmetric stratospheric circulation.

To examine further the notion of forcing from below, consider the time series of the vertical component of the Eliassen–Palm (EP) flux at 100 hPa, 60°S, shown in Fig. 9. This diagnostic is commonly interpreted as a measure of the upward propagation of wave activity from below (e.g., Andrews et al. 1987). At the time of the split in late September, there was indeed an exceptionally strong positive (upward) EP flux at 100 hPa. (Notice, incidentally, that according to this measure of dynamical activity, the preceding southern winter was disturbed in 2002 but not exceptionally so.)

It is not clear, however, that a physical interpretation of this time series indicating “forcing from below” is entirely warranted. The transient perturbations affecting the troposphere–stratosphere system span a considerable depth of the atmosphere. This is shown in Fig. 10, which displays the height–time variability of the variance of geopotential height at 60°S during spring 2002 (a simple measure of eddy activity). The variability appears to be synchronous through the upper troposphere and stratosphere, throughout August and September, up to and including the time of the split. The corresponding time series of the index of the Arctic Oscillation shown by Baldwin et al. (2003) gives a similar impression (though these two diagnostics occasionally differ on the apparent height range of the tropospheric connectivity). It is therefore possible that at certain critical times—specifically during the splitting of the vortex—that the troposphere and stratosphere act dynamically as a single system; the tropospheric “master”/stratospheric “slave” model of their interactions breaks down. In these circumstances, the exceptional peak in the EP flux at 100 hPa in Fig. 9 is, at least in part, a symptom of the exceptional stratospheric warming above rather than indicative of the cause.

This notion does not contradict the importance of the role of asymmetries in the underlying topography as an essential prerequisite for the observed, large-scale variability of the troposphere–stratosphere system and therefore for the occurrence of stratospheric warmings in the system. The idea of a troposphere–stratosphere system acting in concert recalls the idea, advanced by Tung and Lindzen (1979) and Plumb (1981), that the stratospheric state could create a resonant amplification of tropospheric waves. It is not clear, however, that the physical concept of resonance applies in flows such as that of September 2002, which are better described in terms of vortex interactions than of amplifying waves.

For the event of September 2002, the issues mentioned above could be addressed with GCM simulations to determine whether the evolution of the flow in the upper troposphere/lower stratosphere depicted in Fig. 8 was sensitive to stratospheric conditions over a 10-day period, for example, spanning the split. This could be achieved with an ensemble of experiments initialized with a variety of stratospheric states.

Some work along these lines has already been done at ECMWF (T. N. Palmer 2004, personal communication) for the 2002 event. In an ensemble of forecasts initialized 1 week before the split, some with small differences in initial state in the stratosphere only and some with small differences in the troposphere only, about half of the ensuing forecasts predicted a split. These experiments imply that as the polar vortex elongated (through internal nonlinear interactions in the troposphere–stratosphere system), it may have become hydrodynamically unstable, and that this instability might have affected the whole troposphere–stratosphere system. If a large depth of the troposphere–stratosphere system did become hydrodynamically unstable as the warming matured, then it is possible that, at such times, the stratospheric circulation would have to be accurately represented in weather forecasting models to achieve good skill in extended-range (e.g., 10 day) tropospheric weather forecasts.

Were there circumstances favoring a split in the SH polar vortex during spring 2002? One possibility relates to the winds in the tropical stratosphere (Fig. 11). In the lower stratosphere, the phase of the quasi-biennial oscillation (QBO) meant that winds were westerly in the latter half of the year. In the paradigm of a tropospheric planetary wave maker producing upward-propagating, quasi-linear waves, westerlies in the lower stratosphere would, according to Holton and Tan (1982), lead to refraction of planetary waves toward the Tropics away from the midlatitude stratosphere, rendering warmings less likely. If, however, the dynamics is inherently nonlinear, as appears to be the case, then this argument may be incomplete. Whether or not the vortex splits may, in fact, be more sensitive to equatorial winds in the upper stratosphere than to those in the lower stratosphere. Numerical experiments by Gray et al. (2003) demonstrated that easterly winds in the tropical upper stratosphere favor the development of strong stratospheric warmings.

Tropical winds in the upper stratosphere were unusual in 2002 (for reasons that are not clear). Figure 11 shows that the westerly phase of the semiannual oscillation did not descend into the upper stratosphere during the year as it usually does (cf. with the time series shown in Garcia et al. 1997). Moreover, easterly winds were somewhat stronger than normal during September when the vortex split (compared with the data shown by Garcia et al.). Ensembles of simulations with a GCM could be used to determine whether these conditions enhanced the probability of a split vortex, in line with the results of Gray et al., and played a role in the earlier-than-usual change of stratospheric flow regime noted in section 3a.

How rare is a split vortex in the Southern Hemisphere? The work of Taguchi and Yoden (2002) sheds light on this question. They used 1000-yr integrations with a simple GCM to investigate interannual variations in the troposphere–stratosphere coupled system. In a run corresponding to conditions in the Southern Hemisphere, they found a far-from-Gaussian distribution of polar temperature in the upper stratosphere from April to September [Kushner and Polvani (2005) found that the tails of the Southern Hemisphere stratospheric temperature distribution have an exponential character]. About five years in a thousand had monthly temperature anomalies over six standard deviations warmer than normal. Such a rare event is likely to be missed in our current record of meteorological analyses of the stratosphere, which spans less than 50 years.

5. Conclusions

The principal conclusions are as follows.

  • The first observed example of a split in the stratospheric polar vortex of the Southern Hemisphere occurred during the last week of September 2002. The vortex split from the middle to the upper stratosphere. The disturbance was manifest as far down as the region of the upper troposphere/lower stratosphere, where the vortex did not split, but developed a double-lobed structure. The preferred alignment of the quasi-stationary arrangement of vortices during the split indicates a dynamical connection with the underlying topography.
  • The event bears the hallmarks of a major warming of the “wave-2” type, witnessed occasionally in the stratosphere of the Northern Hemisphere. As with such warmings, the elongation of the polar vortex in the Southern Hemisphere prior to the split made it more susceptible to splitting. It was proposed that the dynamics of the split involved nonlinear dynamics in the combined troposphere–stratosphere system and that the concept of forcing from the lower atmosphere might be inappropriate. (Numerical experiments were suggested to test this hypothesis.) It was also proposed that, as the vortex elongated, it became hydrodynamically unstable; that this instability could have affected the whole of the upper troposphere and stratosphere; and that at such times the state of the stratosphere would need to be accurately represented in weather prediction models.
  • Unusually strong easterly winds in the tropical upper stratosphere might have favored the splitting of the vortex during 2002, as well as an earlier-than-normal flow transition in the stratosphere (both hypotheses can be tested, in principle, by numerical experiments). Such events could be rare but not unprecedented in the Southern Hemisphere. In particular, it would be premature to invoke climate change to account for them when their statistical properties are so poorly known.

Acknowledgments

AO and WL are part of the Date Assimilation Research Centre, which is funded by NERC. AJC was funded by an NERC studentship and a CASE award from the UK Met Office. PB is funded by the NCAS Centre for Global Atmospheric Modelling.

REFERENCES

  • Allen, D. R., , R. M. Bevilacqua, , G. E. Nedoluha, , C. E. Randall, , and G. L. Manney, 2003: Unusual stratospheric transport and mixing during the 2002 Antarctic winter. Geophys. Res. Lett., 30 .1599, doi:10.1029/2003GL017117.

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

  • Baldwin, M., , T. Hirooka, , A. O’Neill, , and S. Yoden, 2003: Major stratospheric warming in the Southern Hemisphere in 2002: Dynamical aspects of the ozone hole split. SPARC Newsletter, No. 20, SPARC Office, Toronto, ON, Canada, 24–26.

    • Search Google Scholar
    • Export Citation
  • Charney, J. G., , and M. E. Stern, 1962: On the stability of internal baroclinic jets in a rotating atmosphere. J. Atmos. Sci., 19 , 159172.

    • Search Google Scholar
    • Export Citation
  • Fairlie, T. D. A., , and A. O’Neill, 1988: The stratospheric major warming of winter 1984/85: Observations and dynamical inferences. Quart. J. Roy. Meteor. Soc., 114 , 557577.

    • Search Google Scholar
    • Export Citation
  • Fairlie, T. D. A., , M. Fisher, , and A. O’Neill, 1990: The development of narrow baroclinic zones and other small-scale structure in the stratosphere during simulated major warmings. Quart. J. Roy. Meteor. Soc., 116 , 287315.

    • Search Google Scholar
    • Export Citation
  • Farrara, J., , M. Fisher, , C. R. Mechoso, , and A. O’Neill, 1992: Planetary-scale disturbances in the southern stratosphere during early winter. J. Atmos. Sci., 49 , 17571775.

    • Search Google Scholar
    • Export Citation
  • Garcia, R., , T. J. Dunkerton, , R. S. Liebermann, , and R. A. Vincent, 1997: Climatology of the semiannual oscillation of the tropical middle atmosphere. J. Geophys. Res., 102 , 2601926032.

    • Search Google Scholar
    • Export Citation
  • Gray, L. J., , S. Sparrow, , M. Juckes, , A. O’Neill, , and D. G. Andrews, 2003: Flow regimes in the winter stratosphere of the northern hemisphere. Quart. J. Roy. Meteor. Soc., 129 , 925945.

    • Search Google Scholar
    • Export Citation
  • Harwood, R. S., 1975: The temperature structure of the southern hemisphere stratosphere, August–October 1971. Quart. J. Roy. Meteor. Soc., 101 , 7592.

    • Search Google Scholar
    • Export Citation
  • Holton, J. R., , and H-C. Tan, 1982: The influence of the equatorial quasi biennial oscillation on the global circulation at 50 mb. J. Meteor. Soc. Japan, 60 , 140148.

    • Search Google Scholar
    • Export Citation
  • Hoppel, K., , R. Bevilacqua, , D. Allen, , G. Nedoluha, , and C. Randall, 2003: POAM III observations of the anomalous 2002 Antarctic ozone hole. Geophys. Res. Lett., 30 .1394, doi:10.1029/2003GL016899.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., 1991: Towards a PV-Theta view of the general circulation. Tellus, 43 , 2735.

  • Hoskins, B. J., , M. E. McIntyre, , and A. W. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 111 , 877946.

    • Search Google Scholar
    • Export Citation
  • Kushner, P. J., , and L. M. Polvani, 2005: A very large, spontaneous stratospheric sudden warming in a simple AGCM: A prototype for the Southern Hemisphere warming of 2002? J. Atmos. Sci., 62 , 890897.

    • Search Google Scholar
    • Export Citation
  • Labitzke, K. G., , and H. van Loon, 1999: The Stratosphere: Phenomena, History and Relevance. Springer-Verlag, 179 pp.

  • Lahoz, W. A., and Coauthors, 1996: Vortex dynamics and the evolution of water vapour in the stratosphere of the southern hemisphere. Quart. J. Roy. Meteor. Soc., 122 , 423450.

    • Search Google Scholar
    • Export Citation
  • Leovy, C. B., , and P. J. Webster, 1976: Stratospheric long waves: Comparison of thermal structure in the Northern and Southern Hemispheres. J. Atmos. Sci., 33 , 16241638.

    • Search Google Scholar
    • Export Citation
  • Lorenc, A. C., and Coauthors, 2000: The Met. Office global three-dimensional variational data assimilation scheme. Quart. J. Roy. Meteor. Soc., 126 , 29913012.

    • Search Google Scholar
    • Export Citation
  • Mechoso, C. R., , A. O’Neill, , V. D. Pope, , and J. D. Farrara, 1988: A study of the stratospheric final warming of 1982 in the southern hemisphere. Quart. J. Roy. Meteor. Soc., 114 , 13651384.

    • Search Google Scholar
    • Export Citation
  • O’Neill, A., 2003: Stratospheric sudden warmings. Encyclopedia of Atmospheric Sciences, J. R. Holton, J. A. Curry, and J. A. Pyle, Eds., Academic Press, 1342–1353.

    • Search Google Scholar
    • Export Citation
  • O’Neill, A., , and V. D. Pope, 1988: Simulations of linear and nonlinear disturbances in the stratosphere. Quart. J. Roy. Meteor. Soc., 114 , 10631110.

    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., 1981: Diagnostic study of a wavenumber-2 stratospheric sudden warming in the transformed Eulerian-mean formalism. J. Atmos. Sci., 38 , 844855.

    • Search Google Scholar
    • Export Citation
  • Plumb, R. A., 1981: Instability of the distorted polar night jet: A theory of stratospheric warmings. J. Atmos. Sci., 38 , 25142531.

  • Simmons, A., , M. Hortal, , G. Kelly, , A. McNally, , A. Untch, , and S. Uppala, 2005: ECMWF analyses and forecasts of stratospheric winter vortex breakup: September 2002 in the Southern Hemisphere and related events. J. Atmos. Sci., 62 , 668689.

    • Search Google Scholar
    • Export Citation
  • Swinbank, R., , and A. O’Neill, 1994: A stratosphere–troposphere data assimilation system. Mon. Wea. Rev., 122 , 686702.

  • Taguchi, M., , and S. Yoden, 2002: Internal interannual variations of the troposphere–stratosphere coupled system in a simple global model. Part II: Millennium integrations. J. Atmos. Sci., 59 , 30373050.

    • Search Google Scholar
    • Export Citation
  • Tung, K. K., , and R. S. Lindzen, 1979: A theory of stationary long waves. Part II: Resonant Rossby waves in the presence of realistic vertical shears. Mon. Wea. Rev., 107 , 735750.

    • Search Google Scholar
    • Export Citation
  • Waugh, D. W., , and D. G. Dritschel, 1991: The stability of filamentary vorticity in two-dimensional geophysical vortex-dynamics models. J. Fluid Mech., 231 , 575598.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Schematic of the typical circulation in the middle stratosphere of the SH during (a) Sep and (b) Oct. Anticyclones are shaded; cyclones are not shaded

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 2.
Fig. 2.

Fields of geopotential height in the middle stratosphere of the SH during 2002 on (a) 5 Aug, (b) 22 Aug, (c) 30 Aug, and (d) 11 Sep. Units are km, and data are from UKMO

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 3.
Fig. 3.

Fields of geopotential height (contours) and temperature (colors) for the 10-hPa surface of the SH during 2002 on (a) 21 Sep, (b) 24 Sep, (c) 26 Sep, and (d) 8 Oct. Units are km for geopotential height; K for temperature. Data are from UKMO

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 4.
Fig. 4.

Same as in Fig. 3, but for the 100-hPa surface

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 5.
Fig. 5.

Fields of geopotential height (contours) and temperature (colors) for the 215-hPa surface of the SH during 2002 on (a) 24 and (b) 26 Sep. Units are km for geopotential height; K for temperature. Data are from UKMO

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 6.
Fig. 6.

Zonal-mean wind (m s−1) at 10 hPa near 60°S during 2002. Shading indicates the range of values in the period 1992–2002. Data are from UKMO

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 7.
Fig. 7.

Fields of Ertel’s PV on the 850-K isentropic surface (near 10 hPa) of the SH during 2002 at (a) 1200 UTC on 22 Sep, (b) 1200 UTC on 25 Sep, (c) 1200 UTC on 27 Sep, (d) 0000 UTC, 30 Sep, (e) 1200 UTC on 2 Oct, (f) 0000 UTC on 4 Oct, (g) 1800 UTC on 6 Oct, and (h) 1200 UTC on 9 Oct. Data are from ECMWF

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 7.
Fig. 7.

(Continued)

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 8.
Fig. 8.

Fields of Ertel’s PV on isentropic surfaces in the upper troposphere/lower stratosphere of the SH: (left) the 395-K surface at (top) 1200 UTC on 25 Sep and (bottom) 1200 UTC on 27 Sep; (right) 1200 UTC on 25 Sep for the (top) 350- and (bottom) 315-K surfaces. Data are from ECMWF

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 9.
Fig. 9.

Evolution of the vertical component of the EP flux at 60°S on the 100-hPa surface during 2002. The shaded region denotes the range of values during the period 1992–2001. Units are kg s−2. Data are from UKMO

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 10.
Fig. 10.

Evolution of rms variance of geopotential height around the latitude circle at 60°S during 2002. Units are m. Data are from UKMO

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Fig. 11.
Fig. 11.

Evolution of zonal-mean winds at the equator during 2002. Units are m s−1. Data are from UKMO

Citation: Journal of the Atmospheric Sciences 62, 3; 10.1175/JAS-3318.1

Save