This figure is a presentation of the wavenumber-frequency power spectral peaks of the "convectively-coupled equatorial waves". The power spectra were calculated from a long (~18 years) twice-daily dataset of outgoing longwave radiation (OLR) between the latitudes of 15°S and 15°N. The power spectra have then been divided by a red-noise background spectrum to give the contours as shown above. (a) is for the antisymmetric component (w.r.t. the equator), and (b) is for the symmetric component. Contour interval of this power ratio is 0.1, and the shading begins at a value of 1.1, for which the spectral signatures are statistically significantly above the background at the 95% level (based on 500 d.o.f.). Superimposed are the dispersion curves of the respectively even and odd meridional-numbered equatorial waves as in Matsuno (1966) for the three equ ivalent depths of h=12, 25, and 50m. For further details, see Wheeler and Kiladis, 1999, J. Atmos. Sci. (Feb 1 issue) [Paper available] and Wheeler, Kiladis, and Webster, 2000, J. Atmos. Sci. (March 1 issue) [Paper available].
This figure shows the evolution of the "typical" 850mb circulation pattern (contours of streamfunction and wind vectors) and OLR anomalies (red and blue shading) of an observed (partly dry!) n = 1 equatorial Rossby (ER) wave. This result is obtained by first filtering the 850mb winds and satellite-derived Outgoing Longwave Radiation (OLR) to retain only fluctuations acting on the time scales between 6 and 30 days. Daily time series of the meridional wind difference between 7.5°N and 7.5°S at 160°W are then regressed against the filtered winds and OLR at all other points to produce the typical circulation and convection patterns associated with a peak in that wind difference. OLR is a good proxy for deep tropical convection on this time scale. Blue shading is thus an indication of enhanced convection, while red is an indication of suppressed convection. This result has been obtained using December through February 1985/86 through 1992/93 NMC analysis data. The circulations near the equator in these figures can be seen to be very similar to the structure of the n = 1 ER wave derived from the simple linear inviscid shallow water equations by Matsuno [J. Meteor. Soc. Jpn., 1966]. The individual circulation dipoles can be seen to propagate to the west while new circulations develop in their wake to their east. The period of the disturbance in this sequence is about 12 days and there also appears to be a slight eastward energy dispersion. This type of regression analysis has been performed using base points across the Pacific basin, and such planetary wavenumber ~6 circulations have been found to be quite prominent on this time scale. Also significant is the fact that they have a very deep structure (through the troposphere) with enhanced convection in the poleward flow, and suppressed convection in the equatorward flow. This is somewhat different to that found for other observed equatorial wave-like disturbances. For further details, see Kiladis and Wheeler, 1995, J. Geophys. Res. (Nov issue) [Abstract available].
This figure was created by filtering the observed 850mb winds and OLR between 4 and 6 days. Shown are the streamfunction anomalies (contours and purple/orange shading), wind vectors, and the OLR anomalies (blue is an indication of enhanced convection and red is suppressed convection). The westward phase speed of these waves and a period of about 5 days is readily apparent. Also of note is the assymetrical convection patterns associated with these waves. The latitudinal coverage of these plots is from 17.5°N to 17.5°S. Papua New Guniea and northern Australia appear on the left-hand-side of each frame.