Abstract
The sensitivity of various zonal mean general circulation statistics to the choice of the averaging period used to define them is tested with upper-air data for the Northern Hemisphere taken from the NMC global analysis for the winter of 1976–77. We find that averaging periods of less than about 10 days do not permit a clear separation of total eddy momentum and heat fluxes into their transient and standing eddy components. Between 10 and 30 days, the definition of these components is less sensitive to the specific averaging period chosen during the winter. We illustrate one use of monitoring general circulation statistics on short time scales by studying the evolution of 10-day mean eddy fluxes of sensible heat and their relation to changes in the meridional temperature gradient during this winter. It appears, for this one season at least, that the standing waves regulated the temperature structure of midlatitudes, whereas the transient waves merely responded to the temperature gradient that was imposed.
