• Introduction
• Goals of this study
• Observed Circulation
• The UCLA-GCM
• The performed simulations
• Diagnostic Tools
• Results and Remarks
• References

FIGURES

• Figure 1
• Figure 2
• Figure 3
• Figure 4
• Figure 5
• Figure 6
• Figure 7
• Figure 8
• Figure 9
• Figure 10
• Figure 11
• Figure 12
• Figure 13
• Figure 14
• Figure 15

Introduction

Interannual (IA) variability in the Southern Hemisphere (SH) extratropical lower stratosphere exhibits a maximum during late winter and spring (August-November) polewards of 50†S (Shiotani et al. 1993, Randel and Newman 1997). There are many aspects of the IA variability of the SH stratosphere that are not well understood at present (see section 5 in Randel and Newman 1997).

It has also been established that it is during the austral spring when large-scale teleconnection patterns linked to the El NiÒo ñ Southern Oscillation phenomenon (and surface climate anomalies, particularly in South America) are present in the upper levels of the SH tropospheric circulation (Karoly 1989, Arkin 1982, Aceituno 1989). These SH upper-troposphere anomalies have been described as Rossby wave-like teleconnection patterns (Kousky and Bell 1992, Farrara et al. 1989, Mo and Ghil 1987), and referred as the Pacific South America (PSA) pattern, a wave 3 pattern, with nodes from New Zealand to Tierra del Fuego. Qualitative explanation of these type of global teleconnection patterns including their wave train nature is usually based on the theory of stationary Rossby wave dispersion on the sphere (Held 1983, Hoskins and Karoly 1981).

Lower stratosphere anomalous circulations are directly related to upper troposphere anomalies through changes in the tropopause height. Wave number 3 have been observed as a component of lower stratosphere anomalous features (ozone, wind, etc.) in the SH, and also strong wave number 1 features have been documented in observational studies, particularly for total ozone perturbations (Ambrizzi et al 1998).

Goals of this study

Based in the strong impact on the SH tropospheric circulation associated to El NiÒo (EN) warm events, during the austral spring, and the particular characteristic of the springtime SH circulations with regards to planetary wave dispersion and vertically influencing planetary wave driving into the SH stratosphere, we will, in this study:

1) Describe the observed statistical relationships between El NiÒo (EN), Pacific warm events, and the SH lower stratosphere anomalous circulation;

2) Use an atmospheric General Circulation Model (aGCM) to simulate the climatological and anomalous (EN, warm events) October features and compare them.

3) Diagnose the simulated-circulations by using EP-fluxes, vorticity analysis and Hovmuller diagrams, in order to characterize and better understand the types of circulation which are established in the upper troposphere and lower stratosphere simulated-atmosphere.

Observed Circulation [1]

Changes in the circulation patterns of the stratospheric SH polar vortex are observed during years of extreme sea surface temperature (SST) anomalies at the equatorial Pacific Ocean during the austral spring.

Differences between 50 hPa ("observed") climatological zonal wind for October (Fig. 1) and the composite of ("observed") El NiÒo-Southern Oscillation (ENSO) years 1982-86-91 and 97 (warm events) for zonal wind at the same level (Fig. 2), show a vortex anomaly resulting on a "shift" towards higher latitudes during these years to the southwest of South America. The composite anomaly of the 50 hPa zonal wind (Fig. 3) shows the strongest values (+/- 4 to 7 m/s on a background field of aprox. 40 m/s) between 160†W and 80†W (with peaks of the anomalies around 130†W-120†W). In this area the vortex strengthens at higher latitudes, around 75†S, and weakens around 55†S. That anomalous structure of the 50 hPa circulation is followed by similar anomalies of the zonal wind to the east, on the Atlantic flank of South America on the Antartic Peninsula, but with the sign reversed and smaller zonal scale.

Ý

Those features result on a wavy structure of the stratospheric 50 hPa circulation during October of an EN (warm) event, or, in other words, a wave-like structure of the anomalous 50 hPa zonal wind field, perhaps forced from the central tropical Pacific region as in other EN-forced phenomena.

Correlation field between NiÒo 3-4 index and zonal wind at 50hPa (Fig. 4) show values higher than 0.4 at middle and high latitudes, with very similar pattern to that of circulation.

Note that for the Fig. 4 data which span 21 individual data for each year from 1979 to 1999, any correlation coefficient above 0.43 is statistically significant to the 95% level.

Therefore, it seems that the Central Pacific equatorial SST anomalies have a real effect on the Southern Hemisphere low stratosphere circulation, resulting on a less zonally-symmetric circulation during ENSO years than in the climatology (at that level).

[1] NCEP-Reanalysis- Data, Figs. and correlation maps were taken from and built based on the NCEP-Reanalysis, provided by the NOAA-CIRES CDC: www.cdc.noaa.gov/ ; (Kalnay, E. and Co-authors (1996).

The UCLA-GCM

The UCLA atmospheric GCM (Version 6.95) is a state-of-the art grid point model of the global atmosphere extending from the Earth's surface to a height of 50 km. The model predicts the horizontal wind, potential temperature, water vapor mixing ratio, planetary boundary layer (PBL) depth and the surface pressure, as well as the surface temperature and snow depth over land. The horizontal finite differencing of the primitive equations is done on a staggered Arakawa "C" grid and is based on a fourth order version of the scheme of Arakawa and Lamb (1977 and 1981). The vertical coordinate used is the modified sigma-coordinate of Arakawa and Suarez (1983).

Parameterization of physical processes are: cumulus convection: Arakawa-Schubert (1974), radiative heating: Harshvardan et al. (1987, 1989), PBL processes: Suarez et al. (1983) and gravity waves drag: Kim and Arakawa (1995). The ozone mixing ratios used in the radiation calculations are prescribed as a function of latitude, height and time based on values from a monthly UGAMP climatology (Li and Shine 1995) as used by Kim et al. (1998).

The performed simulations

The spatial resolution is of 4† latitude x 5† longitude, with 16 layers in vertical direction, from surface to the 1 hPa level. (one in the PBL, eight in troposphere and seven in stratosphere) for any of the simulations performed here. Time step was of 6 minutes in the calculation of predicted quantities. The physical processes are calculated each hour of simulated time.

We simulate this process by using the UCLA-GCM with two ensembles of 5 simulations each. The two ensembles are driven with different prescribed SST temperature. The first ensemble has the climatological SST for October ("Control" ensemble) as boundary condition, while the other ("Anomaly" ensemble) is driven (or forced) by imposing a monthly average of the anomalous SST for October 1997¹ (EN97, Fig. 5). The ensemble mean of the "Control" simulation fields are considered as the GCM representation of the October-climatological atmospheric circulation, whether the ensemble mean of the "Anomaly" simulations fields are considered as the GCM representation of the atmospheric circulation corresponding to October 1997. Any of the five individual simulation (for each of the two ensembles) is run after specifying the initial condition (with small random modifications among them) for September 15 (simulated date) and last 46 days (but we focus on the last 31 days, corresponding to October).

Ý

Diagnostic Tools.

In order to better understand the mechanism involved in the establishment and maintenance of those circulation we use two diagnostic tools which are presented here: E-P flux and balance of vorticity analysis.

Eliassen-Palm (E-P) flux diagrams

A concise diagnostic for the propagation characteristic and mean flow forcing of the large scale planetary waves is provided by E-P flux cross sections (Edmon et al. 1980; Andrews et al. 1987). We follow here Section 3.5, from Andrews et al. 1987 and partially Section 3 (particularly c) from Randel and Newman, 1997.

Estimates of the planetary wave EP flux and its divergence are calculated here for the "Control" simulation (Fig. 10) as indication of the October "climatological" (simulated by the model) mean flow capability of the large scales wave to drive meridional residual circulations and/or changes in the zonal mean wind.

We show in Fig. 10 EP flux diagram and its associated wave forcing (), (F is EP flux vector, as defined in Randel and Newman 1997), for the "Control" simulation, based on the sigma levels used by the model. We can conclude that most of the more relevant features of the SH springtime phenomena are present in the October background circulation simulated by our experiments. Particularly, important values of the springtime EP flux are obtained for the SH, specially in low stratospheric levels, which is a very important feature if we want to be able to adequately simulate driving of the stratospheric circulation through tropospheric anomalous circulations (as those imposed when a strong EN events develops in the tropical Pacific, as in our experiment).

Vorticity Analysis

We study the mechanisms involved in the polar vortex wave forcing by making a vorticity analysis.

A vorticity equation for the monthly means (noted with overbars), can be derived from the basic momentum equation solved by the model, and we obtain:

(1)

( in this case primes indicate deviation from the overbared magnitude; a complete deduction of this equation can be read in Cazes and Pisciottano 1998).

Equation 1 is written in vertical sigma coordinate, sigma dot is the respective vertical velocity and p is adequately defined.

The first three terms in the right side of (Eq. 1) are the usual Rossby wave mechanisms: the advection of planetary vorticity (beta-effect), the advection of the relative vorticity and the induction of vorticity by divergence, respectively. The fourth term represents the forcing (to the monthly variables) due to (intramonthly) transients. Figures 13 and 14 show different terms of the equation of balance of vorticity for the "Anomaly" simulation, representing the features associated to this strong EN event.

Results and Remarks

Comparison of the simulated climatological October ("Control") and EN97 October ("Anomaly") 50 hPa zonal wind (Figs. 6 and 7) to the corresponding "observed" fields (Figs. 1 and 2) show that the model is capable of simulate the most important features of the October climatology as well as those associated to an EN (warm) event. This fact can be seen much better by looking at the field of differences ("An"-"Cr") between the two ensembles of the 50 hPa zonal wind (Fig. 8) which shows a clean agreement with observed features (Fig. 3), with increased westerlies in the southern part of the vortex and decreased westerlies in the northern part between 140†W and 80†W. This anomaly has an area with t-Student statistical significance higher than 95% (Canavos 1988) in the region of zonal wind positive anomalies (shadow areas show statistical significance, Fig. 8). This anomaly can also be observed at high latitudes at 200hPa in the zonal wind difference between ensembles (Fig. 9), and it is a manifestation of the barotropic characteristics of the atmosphere there, typical of large scale patterns. The reversed sign anomalies of the 50 hPa zonal wind present to the southeast of South America in the "observed" field (Fig. 3) are also present in the simulated one (Fig. 8) although statistical significance is not so clear than for the former feature present to the southwest of South America

These features are shown also by using Hovmuller (longitude-time, October simulated) diagrams of the anomalous meridional wind at two different levels and for the difference between ensembles. The analysis are made at 62†S, and at 200 and 50 hPa. Hovmuller diagram at 200 hPa (Fig 11) shows a quasi-stationary disturbance corresponding to a wave number 3 between 0-180W, which lasts more than ten days installed in the general eastward zonally propagating wave field. There are also shorter propagating waves (number 6) in the last days of the month, between 0-180E.

In the diagram at 50 hPa (Fig 12) it can be noted, also, the wave number 3, very similar to the one noted at the 200 hPa diagram, which reflects the main observed circulation anomaly at this height. The simulation shows a quasi-equivalent barotropic vertical structure of the atmospheric circulation, at less for the wave number 3 type of features, at that region, associated to the (large scale and) strongly anomalous boundary conditions imposed. This could be interpreted as a direct influence of the upper troposphere in the lower stratosphere through changes in the tropopause height. In the 50 hPa diagram, there are also some waves associated with the shorter waves seen at 200 hPa diagram (but of larger zonal scale and of no barotropic equivalence), which are coherent with the vertical propagation characteristics of the tropospheric shorter waves (mainly 4-7) which are not able to propagate into the stratosphere (Randel and Held 1991).

Ý

Ý

The simulation shows a quasi-equivalent barotropic vertical structure, which is consistent at 200mb with the PSA pattern, a Rossby-wave type structure, and at 50hPa with the observed pattern for EN composite.

Additional analysis of the wave activity propagation is obtained from the study of the simulated averaged EP vectors (Fig 10) for the "Control" ensemble which shows the upward propagation of waves and the generation of smaller scale waves by nonlinear and mixing processes.

To characterize the observed pattern at s =.245 (almost coincident to ~=200hPa) as a Rossby wave structure, we also have used a vorticity analysis. By doing so it can be seen that there is a cancellation between the first three terms of the vorticity equation (Eq 1): the term representing the advection of relative vorticity (Fig. 13), which cancels the addition of the terms of advection of planetary vorticity and vorticity induction through divergence (Fig. 14). The addition of these terms is one order of magnitude smaller than them in all those areas that show important values in Figs 13 and 14, which is typical of a Rossby wave.

Ý

Ý

We can conclude that the most relevant physical mechanisms associated to the circulation anomaly reproduced by the model are the advection of planetary vorticity, the advection of the relative vorticity and the divergence-induced vorticity, other than the transient effects and the non-linear interactions. (However, we should notice that the model operating at this resolution tends to underestimate the transient activity). As the model reproduce the most relevant features of the 50 hPa anomalies associated to the EN97 warm event (i.e. the "shift" of the vortex to higher latitudes in the Pacific sector), it may be said that the intramonthly transients are not very important in the establishment of the lower stratospheric anomalous circulation, and a Rossby wave type mechanism is the most relevant way through which the anomalous tropospheric tropical heating associated to the warm event is influencing both the tropospheric SH extratropical springtime circulation and also the lower stratospheric SH circulation, in an apparently (quasi)barotropically equivalent structure (wave 3 type).

From observation of the precipitation differences between ensembles (Fig. 15), it can be seen the characteristic precipitation pattern associated with an ENSO warm event, with an increase (above climatology) in precipitation at tropical Pacific areas to the east of the most important positive SST anomalies imposed, and an adjacent horseshoe shaped U-pattern of negative anomalies associated to low level anomalous divergence. The simulated precipitation anomalies and the associated anomalous tropical heating must be good enough in amount as well as in location and vertical distribution in order to allow a simulation of the associated anomalous circulation with that level of agreement to the observed ones. We conclude that the model used can be seen as an useful tool to better describe and understand the observed anomalies associated to tropical heating anomalies, at less for some part of the large scale upper troposphere and lower stratosphere features, as the polar vortex alteration studied here.

The model capture very important features of the extratropical tropospheric and lower stratospheric anomalous circulation in the SH (for example the vortex modifications, in place, pattern and intensity) as well as the most relevant characteristics of the wave driven and forcing phenomena associated to the climatological October circulation.

The results obtained are promising in both senses: 1^st) we are able to better understand an interesting phenomenon which links tropospheric and stratospheric anomalous circulation and, 2^nd) we have tested and, now, better-know the UCLA-GCM as a relevant tool, capable of represent some of the most important effects involved in those phenomena. In any case we must be very careful in the interpretation of these results in view of the limitations intrinsic to the simulation performed. For example, as the ozone mixing ratios used in the radiation calculations are prescribed as a function of latitude, high and time but based in climatological values, no influence of the modified and transient circulation is expected via ozone transport on the stratospheric heating in the simulations, etc. Even with all the limitation that can be pointed out, new numerical experiments similar to this one are being planned at IMFIA-FI-UR in order to improve the understanding of the links between large scale phenomena in the SH troposphere and stratosphere.

Acknowledgments. This study was supported by IAI through Project IAI-ISP3-0076. A complete version of it is at the IMFIA web-site.

References

Aceituno, 1989: On the functioning of the SO in the South American sector. Part II: Upper-air circulation. J. Climate, 2, 341-355.

Ambrizzi,, Kayano and Stephenson, 1998: A comparison of global tropospheric teleconnections using observed satellite and general circulation model total ozone column data for 1979-91. Climate Dynamics, 14, 133-150.

Andrews, Holton and Leovy 1987: Middle atmosphere dynamics. Academic Press. San Diego, CA, 489 pp.

Arakawa and Lamb, 1977: Computational design of the basic dynamic process of the UCLA-GCM. Meth. in Comp. Phys., v17, Academic Press, 173-265.

Arakawa and Lamb, 1981: A potential enstrophy and energy conserving scheme for the shallow water equations. Mon. Wea. Rev., 100, 18-36.

Arakawa and Shubert, 1974: Interaction of a cumulus cloud ensemble with the large-scale environment, Part I. J. Atmos. Sci., 31, 674-701.

Arakawa and. Suarez, 1983: Vertical differencing of the primitive equations in sigma coordinates. Mon. Wea. Rev., 111, 34-45.

Arkin, 1982: The Relationship Between Interannual Variability in the 200mb Tropical Wind Field and the Southern Oscillation. Mon. Wea. Rev., 110, 1393-1404.

Canavos, 1988. Probabilidad y EstadÌstica. Aplicaciones y MÈtodos. Mc. Graw-Hill 1988, Mexico, 651 pp.

Cazes and Pisciottano, 1998: Variability of precipitation related to ENSO in southeastern South America. A numerical study and diagnostic of simulated circulation. Extended Abstracs, VIII Congreso Iberoamericano de Meteorologia, 26/30 Octubre 1998, Brasilia-DF, Brasil.

Edmon, Hoskins and McIntyre, 1980: Eliassen-Palm cross sections for the troposphere. J. Atmos. Sci., 37, 2600-2616.

Farrara,, Ghil, Mechoso and Mo 1989: Empirical Orthogonal Functions and Multiple Flow Regimes in the Southern Hemisphere Winter. J. Atmos. Sci., 46, 3219-3223.

Harshvardan, Randall and Corsetti, 1987: A fast radiation parameterization for General Circulation Models. J. Geophys. Res., 92, 1009-1016.

Harshvardan, Randall, Corsett and Dazlich, 1989: Earth radiation budget and cloudiness simulations with a General Circulation Model. J. Atmos. Sci., 46, 1922-1942.

Held 1983: Stationary and quasi-stationary eddies in the extratropical troposphere. In: Hoskins, B.J. R.P. Pearce (eds) Theory. Large-scale dynamical processes in the atmosphere. Academic Press. pp 127-168.

Hoskins and Karoly 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38, 1179-1196.

Kalnay and Co-authors, 1996: The NCEP/NCAR Reanalysis 40-year Project. Bull. Amer. Meteor. Soc. 77, 437-471.

Karoly, 1989: Southern Hemisphere circulation features associated with ENSO events. J. Climate, 2, 1239-1252.

Kim and Arakawa, 1995: Improvement of orographic gravity-wave parameterization using a mesoscale gravity-wave model. J. Atmos. Sci., 52, 1875-1902.

Kim, Farrara and Mechoso, 1998: Sensitivity of AGCM simulations to modifications in the ozone distribution and refinements in selected physical parameterizations. J. Meteor. Soc. Japan,, 76, 695-709.

Kousky. and Bell 1992: Atlas for Southern Hemisphere 500-mb teleconnection patterns derived from National Meteorological Center Analyses. NOAA Atlas 9, US Department of Commerce. 99pp.

Li and Shine, 1995: A 4-dimensional ozone climatology for UGAMP models. UGAMP Internal Report No. 35.

Mo and Ghil 1987: Statistics and dynamics of persistent anomalies. J. Atmos. Sci., 44, 877-901.

Randel and Held, 1991: Phase speed spectra of transient eddy fluxes and critical layer absoroption. J. Atmos. Sci., 48, 2496-2508.

Randel and Newman 1997: The stratosphere in the SH. Chapter 6 of the AMS Monograph Meteorology of the Southern Hemisphere. Cooperative Research Centre for Southern Hemisphere Meteorology. Technical Report N† 2, 63 pp.

Shiotani, Shimoda and Hirota, 1993: Interannual variability of the stratospheric circulation in the Southern Hemisphere. Q.J.R. Meteorol. Soc., 119, 531-546.

Suarez Arakawa and Randall, 1983: The parameterization of the PBL in the UCLA-GCM: formulation and results. Mon. Wea. Rev., 111, 2224-2243.