| description abstract | The winter response of the atmosphere to El Niño events in the Pacific is studied both from a 14-year integration of the Center for Ocean?Land?Atmosphere GCM using observed SSTs from January 1979 to February 1993 and from the corresponding analyses of ECMWF. Emphasis is put on the shift in the high-frequency transients that define the Pacific storm track during warm events. Warm and normal ensembles are defined on the basis of the GCM?s diabatic heating field in the tropical Pacific, which falls in one of two states. During the winters of 1982/83, 1986/87, and 1991/92, the heating averaged between 6°S and 6°N lies in the range of 100?200 W?m?2 all the way across the basin. The remaining 10 ?normal? years all show no large tropical diabatic heating anomalies in the mid or eastern Pacific. The difference between warm and normal ensembles for the mean fields of zonal wind u and height z indicates an eastward and equatorward extension of the midlatitude Pacific jet, associated with a similar extension of the transient feedback on the mean flow as measured by the convergence of vorticity flux. The increase in high-frequency (periods of ?2 to 10 days) transient kinetic energy in the eastern portion of the Pacific during El Niño extends across Mexico. The high-frequency transient vertical and meridional sensible heat fluxes, and the low-level diabatic heating and baroclinicity of the mean state (measured by Ri?1/2) also indicate an eastward and equatorward shift of the entire storm track complex in the GCM and the analyses. The shift is consistent with the increased midlatitude shear during El Niño that accompanies the overall tropical warming. While the GCM storm track shift has strong similarities to that in the analyses, the GCM?s large systematic errors in the mid Pacific (eastward extension of the Pacific jet and the storm track) lead to an underestimation of the response to El Niño, which has a very similar form. However, the GCM also shows a spurious tendency to move the storm tracks equatorward in the far western Pacific, as seen in the large positive anomalies in all dynamical indicators of the storm tracks at latitudes 20°?30°N. | |
