SOLA, 2005, Vol. 1, 125 128, doi: 10.2151/sola. 2005 033
125
Influence of ENSO on the Stratosphere-Troposphere Coupling during Stratospheric Sudden Warming Events Hideo Shiogama and Hitoshi Mukougawa Disaster Prevention Research Institute, Kyoto University, Kyoto, Japan
Influence of El Niño/Southern Oscillation (ENSO) on the stratosphere-troposphere dynamical linkage during stratospheric sudden warming (SSW) events is investigated based upon the composite analysis using 44 winter record of NCEP-NCAR reanalysis data. Stratospheric waves with zonal wavenumber 1 (2), which are linked with tropospheric teleconnections over the Pacific-North America-Atlantic sector (the Eurasian sector), are important for the onset of SSWs during the warm (cold) phase of ENSO. During the cold phase, easterly anomalies associated with SSWs penetrate into the troposphere, and annular-mode-like height anomalies appear subsequently in the troposphere. On the other hand, during the warm phase, easterly anomalies are restricted in the stratosphere, and zonally asymmetric height anomalies are observed in the troposphere about 10 days after the warming peak in the stratosphere. Modulation of storm track activities due to lower stratospheric anomalies is found to be important for the downward penetration of zonal wind anomalies into the troposphere.
1. Introduction Stratospheric sudden warmings (SSWs) are the most impressive phenomena among various intraseasonal variations of the stratospheric circulation in the Northern Hemisphere winter. The interaction between stratospheric zonal flows and planetary waves propagating from the troposphere is the basic mechanism for SSWs (e.g., Matsuno 1971). Recently, observational studies have found downward migration of zonal wind anomalies from the stratosphere to the troposphere (Baldwin and Dunkerton 1999, 2001; Kodera et al. 2000). The framework of wave-mean flow interaction can explain the basic mechanism of the descending signal within the stratosphere (Kuroda and Kodera 1999). When the zonal wind anomalies descend to near the tropopause, they tend to be linked to the occurrence of the Northern Hemisphere annular mode (NAM) (Thompson and Wallace 1998, 2000) that is the tropospheric leading mode (Baldwin and Dunkerton 2001). However, the detailed mechanism for the downward penetration of zonal wind anomalies into the troposphere has not been resolved. The present study is aimed at examining whether this type of intraseasonal stratosphere-troposphere dynamical coupling is influenced by the familiar external forcing, El Niño/Southern Oscillation (ENSO). It is well known that ENSO affects on the interannual and intraseasonal variabilities of extratropical atmospheric motions in the troposphere and the stratosphere (Horel and Wallace 1981; Wallace and Chang 1982; Kimoto 1989). For example, Quadrelli and Wallace (2002) showed that ENSO modulates the structure of the tropospheric leading mode. Van Loon and Labitzke Corresponding author and present affiliation: Hideo Shiogama, National Institute for Environmental Studies, Tsukuba, 3058506, Japan. E-mail:
[email protected]. ©2005, the Meteorological Society of Japan.
(1987) identified a tendency for the stratospheric polar vortex to be anomalously weak during the warm phase of ENSO. In this paper, we investigate ENSO effects upon the intraseasonal descending motion of zonal wind anomalies. In particular, we try to clarify influences of ENSO on the stratosphere-troposphere dynamical linkage during SSW events with large zonal wind anomalies in the stratosphere based upon the composite analysis for the two extreme phases of ENSO.
2. Data and methods We used the National Centers for Environmental Prediction - National Center for Atmospheric Research (NCEP-NCAR) daily reanalysis data (Kalnay et al. 1996) for the 45-year period 1957 2001. The dataset is presented on a 2.5° × 2.5° latitude-longitude grid at 17 vertical levels from 1000 hPa to 10 hPa. The analysis is done for the 44 extended winter seasons (from November to April). We also used monthly mean sea surface temperature (SST) data of Met Office HadISST (Rayner et al. 2003) on a 1° × 1° grid for the same period analyzed. The ENSO cycle is determined based on the wintertime seasonal mean SST averaged over the area 5.5°N 5.5°S, 179.5°W 90.5°W, according to Quadrelli and Wallace (2002). Warm and cold phases of ENSO are defined as the 15 warmest SST winters and the 15 coldest SST ones, respectively. For the warm (cold) composite, the daily anomaly is computed as the departure from a daily climatology that is defined as the 15-year mean of the 10-day moving averaged value for each calendar day based upon 15 warmest (coldest) years. To focus on intraseasonal variability, the time mean value for each year is further removed from the daily anomaly. In the present study, 10-day low-passed (high-passed) anomalies computed by the Lanczos filter (Duchon 1979) are referred to as low- (high-) frequency anomalies. All composites are made on the low-frequency anomalies. To describe variations of the stratospheric polar vortex, we use a NAM index at 10 hPa. The NAM at 10 hPa is determined as the principal mode of empirical orthogonal function for the monthly mean Z10 (10 hPa height) anomalies north of 20°N. The NAM index is computed by projecting the low-frequency Z10 anomaly onto the NAM mode. A period when the NAM index drops below 1 standard deviation (std) and rises again above 1 std is defined as a weak polar vortex event. The date when the NAM index attains its minimum value during an event is defined as a key-day. Key-days occur in November and April are removed from the composite. These allow 18 and 14 events to be employed in the warm and cold composites, respectively. All of these events correspond to major or minor SSWs according to the criterion of the World Meteorological Organization. The statistical significance of those composite maps shown below is assessed by a nonparametric resampling procedure (Bootstrap test; see Efron 1979; also see Wilks 1995) with 1000 resampled data sets of 18 and 14 events for the warm and cold winters, respectively. The assumed null hypothesis is that the ensemble average of observed anomalies is zero. A nonparametric estimate of
126
Shiogama and Mukougawa, ENSO and Stratospheric Sudden Warmings
Fig. 1. Composites of zonal mean zonal wind anomalies on pressure-time sections at 60°N for (a) the warm, (b) cold winters, and (c) cold winters minus warm winters. Solid (dashed) contours denote positive (negative) anomalies. The contour interval is 1 ms 1 , with the zero contour omitted. Light and heavy red shading indicate positive anomalies that are significant at least at the 90% and 95% confidence levels, respectively. Blue shading shows significant negative anomalies.
the probability density function around zero can be made by randomly generating artificial data batches from the existing samples with replacement for a large number of trials in the following manner: (1) {ui; i = 1,..,n} denotes the n samples of observation (18 and 14 samples for the warm and cold winters, respectively), and u– is the average; (2) zero centered samples, {uic = ui u– ; i = 1,..,n}, are made to estimate zero centered probability density function; (3) a random resampling of {uic ; i = 1,..,n} with replacement generates an artificial data batch {b ujc ; j = 1,..,n}, and we compute the average of them, bu– c; (4) 1000 resamplings according to (3) are made to produce {bu– c; b = 1,..,1000}; (5) the obtained histogram of {b u– c ; b = 1,..,1000} is used to estimate zero centered probability density function, and the nonzero test of u– is done with this estimated probability density function, i.e., it is tested, for 95% two-tailed test, whether u– is outside the 2.5% 97.5% range; (6) in practice, u– and b u– c are transformed to t = n1/2u– /(std of u) and t = n1/2 bu– c/(std of buc) (Student’s t form), since it is well known that this transformation makes the nonzero test more robust (Bootstrap t test; see Hall 1988). Additionally, when we test significance of the composite for an m-day averaged value, we compute the m day average for each event before the aforementioned procedures (1) (6). In this case ui in (1) represents the m-day mean for the i-th event.
3. Results Figures 1a and 1b show pressure-time cross sections of zonal mean zonal wind anomalies. Downward propagation of negative zonal wind anomalies in the stratosphere is evident irrespective of the ENSO phase. On the other hand, the time evolution of zonal wind anomalies in the troposphere after day 0 is quite different between the warm and cold winters. Significant tropospheric easterly anomalies are evident for the cold winters, while the anomalies in the warm composite have no statistical significance. Figure 1c shows the difference of zonal-mean zonal wind anomalies (cold winters minus warm winters). Significant negative differences are
Fig. 2. Top two panels are Z50 anomaly composites for the periods between day 15 and day 5 of (a) the warm and (b) cold winters, respectively. Bottom two panels are as in top panels except for Z250 anomalies. The composites are computed after truncation of height anomalies at zonal wavenumber 4, and are weighted by sin 45°N/ sin (lat.). Solid (dashed) contours indicate positive (negative) anomalies. The contour interval is 20 m (12 m) in the top (bottom) panels, with the zero contour omitted. Shading is as in Fig. 1.
found in the troposphere after the SSW around day +10. There are also significant positive differences for the period between day 27 and day 8. The characteristic of the planetary wave during the onset of SSWs is also different between the two extremes of ENSO. In the stratosphere, zonal wavenumber (WN) 1 component is dominant in the anomaly field for warm winters, whereas WN 2 component also becomes conspicuous in the anomaly field for the cold winters (Figs. 2a and 2b). The difference of the dominant zonal wavenumber is also seen in the composites of total Z50 field (not shown). For the warm winters, a positive anomaly over North America with the opposite one over the Eurasian sector is evident. For the cold winters, there are significantly large anticyclonic and cyclonic anomalies over Alaska and Siberia, respectively. Anomalies over Scandinavia and Greenland are rather weak. The bottom two panels of Fig. 2 show the corresponding Z250 anomalies for both winters. In the warm winters (Fig. 2c), a PNA-like wave train (Wallace and Gutzler 1981) with a negative anomaly over Aleutian Islands coexists with an anticyclonic anomaly south of Greenland. The analysis on the wave activity flux formulated by Takaya and Nakamura (1997, 2001) shows that these waves induce upward wave activity flux anomalies, resulting in amplification of the stratospheric WN 1 wave (not shown). In the cold winters (Fig. 2d), a hemispheric wave pattern with WN 2 and WN 3 is dominant. Baroclinic structure of the negative anomaly over the Eurasian sector (Figs. 2b and 2d) accounts for the anomalous upward wave activity flux. During the mature stage of SSWs, positive Z50 anomalies over the North Pole are dominant for both winters (Fig. 3). However, it should be also noted that WN 1 (Fig. 3a) and WN 2 (Fig. 3b) components are
SOLA, 2005, Vol. 1, 125 128, doi: 10.2151/sola. 2005 033
127
Fig. 3. Same as Figs. 2a and 2b except for the periods between day 5 and day +5 of (a) the warm and (b) cold winters, respectively.
perceivable in the Z50 anomalies for the warm and cold winters, respectively. About 10 days after the key-days, the difference of the dominant WN in anomalies is still seen in the lower stratosphere (Figs. 4a and 4b). Striking differences in height anomaly patterns are also observed in the troposphere in accordance with zonal wind anomalies. In the cold winters, the tropospheric height anomalies bear a resemblance to the NAM with positive anomalies over the North Pole (Figs. 4d and 4f). On the other hand, for the warm winters, upper tropospheric anomalies (Fig. 4c) have an almost similar pattern to Fig. 2c with the opposite polarity, especially over the Pacific. In the Z1000 anomaly field (Fig 4e), a dipole structure extending up to the lower stratosphere is evident over the Pacific, whereas a monopole negative anomaly is seen over the Atlantic. These facts suggest that the stratospheric height anomalies persisting during the course of SSW events are intimately related to the difference in the response of the tropospheric circulations after the SSW events. Downward influence of the stratospheric height anomalies on the tropospheric circulation is directly assessed by the inversion of the observed stratospheric potential vorticity (PV) anomalies (Hoskins et al. 1985). In fact, tropospheric circulation anomalies computed by the inversion of stratospheric (from 200 hPa to 10 hPa) PV anomalies are similar to the NAM for the cold winters, whereas a negative height anomaly over the Pacific with the opposite one over the North Pole is induced in the troposphere for the warm winters (not shown). However, the magnitude of the tropospheric height anomalies induced by the observed stratospheric PV anomalies is about 10% of the observed one. Thus, some positive feedback mechanism is necessary to account for the amplification of the induced anomalies. One of the most promising agents for this feedback would be high-frequency eddies through their vorticity flux divergence, since the induced anomalies have large signals in the storm track region. Figure 5 shows tropospheric height tendency anomalies due to the dynamical feedback of high-frequency eddies. For the cold winters, negative height tendency anomalies in the mid-latitudes, which are associated with the negative zonal wind anomalies in the troposphere in Fig. 1b, extend from the Pacific to the Atlantic (Fig. 5b). On the other hand, for the warm winters, anomalous highfrequency eddies exert stronger forcing over the Pacific to reinforce the dipole height anomaly pattern (Fig. 5a). However, they also induce a weak dipole anomaly with the opposite polarity over the Atlantic, which tends to reduce the tropospheric easterly anomalies. The magnitude of the feedback forcing is sufficiently large to explain the tropospheric height anomalies after the SSWs. Therefore, we think that the modulation of highfrequency eddy activities by the lower stratospheric
Fig. 4. As in Fig. 2 except for (top) Z100 anomalies, (middle) Z250 anomalies, and (bottom) Z1000 anomalies for the periods between day +5 and day +15. The contour intervals are (top) 15 m, (middle) 12 m, and (bottom) 7 m, with the zero contour omitted.
circulation anomalies plays an important role for the penetration of zonal wind anomalies into the troposphere.
4. Discussion and conclusions The present study demonstrated that ENSO has a significant effect on the lifecycle of SSWs. The stratospheric WN 1 wave is responsible for the onset of SSWs during the warm winters of ENSO. On the other hand, WN 2 component also becomes important for the cold winters. During the cold phase, easterly anomalies associated with SSWs tend to penetrate across the tropopause, and induce the NAM-like height anomalies in the troposphere after SSWs. On the other hand, during the warm phase, the easterly anomaly tends to be restricted in the stratosphere. Tropospheric height anomaly after SSWs exhibits a non-annular mode structure with zonal asymmetry. During the onset period of SSWs in the warm (cold) phase, tropospheric teleconnection patterns over the Pacific-North America-Atlantic sector (the Eurasian sector) are connected to the stratospheric WN 1 (WN 2) wave. It is well known that the tropospheric teleconnections are linked to the phase of ENSO (e.g., Horel and
128
Shiogama and Mukougawa, ENSO and Stratospheric Sudden Warmings
Fig. 5. Composites of height tendency anomalies due to highfrequency eddy vorticity flux divergence (Holopainen et al. 1982) that are integrated between day 0 and day +10 of (a) the warm and (b) cold winters, respectively. They are vertically averaged from 1000 hPa to 100 hPa, and are weighted by sin 45°N/sin(lat.). Solid (dashed) contours indicate positive (negative) anomalies. The contour interval is 20 m, with the zero contour omitted. Shading is as in Fig. 1.
Wallace 1981). Kimoto (1989) showed that the weather regime with positive PNA index characterized by a negative height anomaly over Aleutian Islands occurs more frequently in the warm winters than in the cold winters. For the warm phase, enhanced wave activity associated with the PNA tends to force the planetary waves with WN 1 in the stratosphere as indicated by Itoh and Harada (2004). On the other hand, for the cold phase, the dominance of WN 2 component in the stratosphere could be due to the suppression of WN 1 component associated with the reduced activity of the PNA in the troposphere. After the peak period of SSWs, WN 1 and WN 2 components are still evident in the lower stratosphere for the warm and cold winters, respectively. This difference in the stratospheric height anomalies could induce the different tropospheric anomalies. The dynamical feedback associated with the modulated storm track activities due to the lower stratospheric circulation anomalies may play an important role to induce a characteristic tropospheric anomaly pattern for each winter. In this study, we have shown that WN 2 component of stratospheric planetary waves triggering SSWs is important for the downward propagation of zonal wind anomalies into the troposphere. The numerical study of Yoden et al. (1999), which indicated the tendency of the downward penetration of zonal wind anomalies associated with WN 2 type SSWs in a general circulation model, supports our results. For future studies, we will examine the detailed mechanism of the downward penetration of zonal wind anomalies into the troposphere due to high frequency eddies by using numerical models.
Acknowledgments Drs. K. Kodera and M. Taguchi provided helpful comments and their unpublished manuscripts. The authors also thank the two anonymous reviewers and the editor, Dr. T. Hirooka for their comments to improve the manuscript. This work was supported by a Grant-inAid for Scientific Research (B) 15340155 from the Ministry of Education, Science, Culture, Sports and Technology, Japan.
Baldwin, M. P., and T. J. Dunkerton, 1999: Propagation of the
Arctic Oscillation from the stratosphere to the troposphere, J. Geophys. Res., 104, 30937 30946. Baldwin, M. P., and T. J. Dunkerton, 2001: Stratospheric harbingers of anomalous weather regimes, Science, 294, 581 584. Duchon, C. E., 1979: Lanczos filtering in one and two dimensions, J. Appl. Meteor., 18, 1016 1022. Efron, B., 1979: Bootstrap methods: another look at the jackknife, Ann. Statist., 7, 1 26. Hall, P., 1988: Theoretical comparison of bootstrap confidence intervals (with discussion), Ann. Statist., 16, 927 985. Holopainen, E. O., L. Rontu, and N. C. Lau, 1982: The effect of largescale transient eddies on the time-mean flow in the atmosphere, J. Atmos. Sci., 39, 1972 1984. Horel, J. D., and J. M. Wallace, 1981: Planetary-scale atmospheric phenomena associated with the Southern Oscillation, Mon. Wea. Rev., 109, 813 829. 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, 877 946. Itoh, H., and K. Harada, 2004: Coupling between tropospheric and stratospheric leading modes, J. Climate, 17, 320 336. Kalnay, E., and coauthors, 1996: The NCEP/NCAR 40-Year reanalysis project, Bull. Amer. Meteor. Soc., 77, 437 471. Kimoto, M., 1989: Multiple flow regimes in the Northern Hemisphere winter. Ph. D. Thesis, University of California, Los Angeles, 210 pp. Kodera, K., Y. Kuroda, and S. Pawson, 2000: Stratospheric sudden warmings and slowly propagating zonal-mean zonal wind anomalies, J. Geophys. Res., 105, 12351 12359. Kuroda, Y., and K. Kodera, 1999: Role of planetary waves in the stratosphere-troposphere coupled variability in the northern hemisphere winter, Geophys. Res. Lett., 26, 2375 2378. Matsuno, T., 1971: A dynamical model of stratospheric sudden warming, J. Atoms. Sci., 27, 871 883. Quadrelli, R., and J. M. Wallace, 2002: Dependence of the structure of the Northern Hemisphere annular mode on the polarity of ENSO, Geophys. Res. Lett., 29, doi:10.1029/ 2002G-L015807. Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century, J. Geophys. Res., 108, No. D14, 4407, doi:10.1029/ 2002JD002670, 2003. Takaya, K., and H. Nakamura, 1997: A formulation of a waveactivity flux for stationary Rossby waves on a zonally varying basic flow, Geophys. Res. Lett., 24, 2985 2988. Takaya, K., and H. Nakamura, 2001: A formulation of a phaseindependent wave-activity flux for stationary and migratory quasi-geostrophic eddies on a zonally-varying basic flow, J. Atoms. Sci., 58, 608 627. Thompson, D. W., and J. M. Wallace, 1998: The arctic oscillation signature in the wintertime geopotential height, Geophys. Res. Lett., 25, 1297 1300. Thompson, D. W., and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability, J. Climate, 13, 1000 1016. van Loon, H., and K. Labitzke, 1987: The Southern Oscillation. Part V: The anomalies in the lower stratosphere of the Northern Hemisphere in winter and a comparison with the quasi-biennial oscillation, Mon. Wea. Rev., 115, 357 369. Yoden, S., T. Yamaga, S. Pawson, and U. Langematz, 1999: A composite analysis of the stratospheric sudden warmings simulated in a perpetual January integration of the Berlin TSM GCM, J. Meteor. Soc. Japan, 77, 481 445. Wallace, J. M., and F.-C. Chang, 1982: Interannual variability of the wintertime polar vortex in the Northern Hemisphere middle stratosphere, J. Meteor. Soc. Japan, 60, 149 155. Wallace, J. M., and D. S. Gutzler, 1981: Teleconnections in the geopotential height field during the Northern Hemisphere winter, Mon. Wea. Rev., 109, 784 812. Wilks, D. S., 1995: Statistical Methods in the Atmospheric Sciences, Academic Press, New York, 468 pp. Manuscript received 16 May 2005, accepted 12 July 2005 SOLA: http://www.jstage.jst.go.jp/browse/sola/
