A flexible bandpass filter design procedure applied to midlatitude intraseasonal variability

A general digital bandpass filtering procedure is presented whose advantage s over other simple filtering methods are 1) versatility in its design, bec ause only a set of three parameters is needed to calculate the filter weigh ts via a simple analytic expression, 2) good performance at the transition band depending on the number of weights considered, and 3) the reduction of the Gibbs oscillations in the pass band of a given raw filter by convolvin g it with a convergence window. In order to illustrate the method, the filt er has been first used to assess the ability of the Meteo-France general ci rculation model ARPEGE to simulate the midtropospheric low-frequency intras easonal variability in the Northern Hemisphere. The filter examined here al lows one to assess the model drawbacks in different frequency bands. As a s econd example, the synoptic-scale baroclinic fluctuations in midlatitudes h ave also been studied. It is shown that the horizontal and vertical structu re of these fluctuations does not depend very much on the frequency band up to a period of 10 days, but shows an increase in zonal wavelength as lower -frequency fluctuations are considered.