J. Atmos. Sci., 45, 3521-3525
Department of Meteorology, University of Hawaii at Manoa, Honolulu, Hawaii
(2 February 1988 and 2 May 1988)
In the air-sea interaction model for intraseasonal oscillation (Emmanuel 1987), the amplification and phase propagation of unstable waves are controlled by the east-west asymmetries in circulation-dependent surface heat fluxes. The interpretation of the intraseasonal oscillation is based on the assumption that the mean surface winds are easterlies with significant strength. The observed mean surface winds, however, exhibit remarkable longitudinal variations. Significant surface easterlies prevail only over the central-eastern Pacific and Atlantic oceans. The annual mean of the surface westerlies averaged between 0° and 160°E is near zero (-0.25 m s-1). The absence of meaningful surface easterlies in a near-hemispheric scale is expected to severely limit the model's applicability to observed tropical atmosphere. Furthermore, observed 30-60 day waves are found to amplify over the Indian and western Pacific oceans but the absence of mean surface easterlies implies that the evaporation-wind feedback is not favored in these regions. Over the western hemisphere equatorial oceans strong easterlies could support east-west asymmetries in surface heat fluxes; but the sea surface temperature is low due to wind-induced upwelling and the troposphere is stable, thus it might be inadequate to assume that convections immediately redistribute anomalous heat fluxes aloft and cause the 30-60 day wave to develop. In fact the wave circulation exhibits significant weakening as these waves travel across the eastern Pacific and Atlantic oceans, regardless of the evaporation-wind feedback being favored there.
Although Neelin et al. (1987) found that evaporation-wind feedback is important to intraseasonal oscillation in a GCM with zonally symmetric climate (the heating representation is applicable to this idealized model tropics), they also noticed that the existence of the spectral peak on intraseasonal time scale does not depend on this mechanism because in a "swamp" model which rolls out evaporation-wind feedback, the intraseasonal oscillation still exists and the wave structure is essentially the same. Further analysis of the model used by Neelin et al. (1987) show that the eddy available potential energy is mainly generated and then converted to mechanical energy in the surface convergence phase. The air-sea interaction model, however, predicts that no available potential energy is generated in the surface convergence phase and no kinematic energy is converted from eddy available potential energy. Future observational diagnosis is suggested to determine the relative importance of the precipitation--convergence feedback and the evaporation-wind feedback in supporting the tropical intraseasonal oscillation.
Back to top | Back to menu