The Probability Density Distribution of the Planetary-Scale Atmospheric Wave Amplitude Revisited

Abstract

The investigation for evidence of multiple statistical flow regimes in the amplitude of the planetary-scale waves presented earlier by Hansen and Sutera is revisited. A methodology for generating Monte Carlo tests of the statistical significance of univariate probability density estimates of temporally correlated time series is proposed and applied to a 42-winter time series of the amplitude of the planetary-scale waves. The statistical significance of bimodality in the planetary wave amplitude can be established from a 28-winter dataset as well as the 42-winter dataset, but not from the 16-winter dataset originally used by Hansen and Sutera. Corroboration of this result is found from comparable analysis of an extended general circulation model integration and from examination of the results of other independent observational studies. Several analysis schemes find flow regimes with a similar pattern that features large amplitudes in the planetary-scale waves.