Tropical–Extratropical Interactions on Intraseasonal Time Scales in a Global Spectral Model

Abstract

Observational studies have revealed some coherent extratropical patterns associated with the tropical Madden?Julian (MJ) wave. This study is an attempt to clarify and constrain the interpretation of these patterns by investigating tropical?extratropical interactions on intraseasonal time scales in a global spectral model (GSM). Forcing representative of northern winter is used. A simple heating-only cumulus parameterization scheme is included to generate the MJ wave. The wave period in the model falls within the 30?60 day range observed and has a structure consistent with observations. Various statistical techniques including compositing, empirical orthogonal function (EOF) analysis, and singular value decomposition (SVD) have been used to identify those extratropical patterns associated with the tropical MJ wave. Under zonally symmetric external conditions (no topography) the MJ wave maintains a highly regular amplitude and phase speed. Nevertheless, there is no statistically significant coherent variability between the tropics and extratropics?no matter how much one field lags the other, and despite the frequent appearance of upper-level equatorial waveguides. Significance is determined by a Monte Carlo data scrambling method. When topography is included, the MJ wave has a more variable amplitude in both space and time. All the statistical analyses reveal consistent planetary-scale extratropical patterns associated with different phases of the tropical wave. EOF and SVD analyses indicate that the MJ wave can explain about 10% of the variance in the extratropical 250-mb height field on intraseasonal time scales. Monte Carlo significance testing indicates that about 5% can be attributed to physical processes and 5% to chance. The signal of tropical?extratropical interaction in the mountain case can be exposed more clearly by reducing the radiative driving by half and by using a specified propagating heat source as a proxy MJ wave. In this case up to 40% of the extratropical variance can be explained by the tropical wave, while the principal patterns of interaction remain similar to those obtained with strong driving. The authors conclude that topography is essential to the propagation of a coherent MJ signal into the extratropics, while topography tends to disrupt the MJ wave in the tropics. The MJ wave (and its associated extratropical patterns) must be maintained in the presence of topography by wave activity penetrating the tropics from higher latitudes. A sufficiently high level of eddy activity in the extratropics is necessary for this to occur.