Abstract
With a 9-level general circulation model, all attempt was made to simulate numerically the breakdown of the circumpolar vortex in the winter stratosphere for the case of March 1965. The marching computations were started 2 and 5 days prior to the breakdown. The simulation of the vortex elongation and destruction was, to a certain extent, successful, but the split vortex in the prediction erroneously merged again after 8 days. The sudden warming was not simulated at all. The development of the Aleutian high associated with the vortex breakdown was not well computed. Studies are made on zonally averaged quantities pertaining to the tropospheric and stratospheric circulations and their coupling. The increase of eddy kinetic energy at the time of the amplification of zonal wavenumber 2 is discussed both for the numerical simulation and for the observed fields. It is reconfirmed that the eddy kinetic energy in the stratosphere is primarily supplied from below in the form of vertical flux of geopotential. The propagation of wave energy takes place through a rather narrow zonal belt at high latitude. A possible relation between the stratospheric vortex destruction and the tropospheric process of meandering westerlies is discussed in terms of vertical transmission of wave energy.
