The role of anomalous zonal winds on the convectively coupled equatorial waves

The role of anomalous zonal winds on the convectively coupled equatorial waves

Guiying Yang (swsyangg@met.rdg.ac.uk), Brian Hoskins and Julia Slingo, CGAM, University of Reading.

In Yang et al. (2002), a new methodology which isolates individual equatorial modes has been developed and applied to ECMWF Reanalysis (ERA15) and satellite observed window brightness temperature (T_b). The robust spatial and temporal consistency of the results between the two independent data sources has given confidence that the methodology has been successful in detecting various convectively coupled equatorial waves. Most of those waves appear to have convection associated with their low level convergence, as suggested by simple equatorial wave theory. As an example Figure 78a shows the theoretical horizontal structures of the n=1 Rossby (R1) and Kelvin waves. If the low-level convergence and upper level divergence in the first internal mode, with associated mid-tropospheric ascent, provide the organization for convection, then we would expect this convection to occur in the blue-green shaded convergence regions in Figure 78a. Such convection does not radically change the dry wave structure.

However, for convectively coupled waves, wind-dependent surface fluxes of moist entropy may play an important organizational role for convection (e.g., Emanuel (1987). Observational studies have shown that in the Western Pacific region MJO-related deep convection tends to occur in regimes of enhanced surface westerly flow, rather than in-phase with the low level convergence (e.g., Zhang 1996). It is suggested, therefore, that convection associated with equatorial waves may, in some cases, occur where there are anomalous zonal winds. Since warmest SSTs occur in the near equatorial region, wind-induced surface energy fluxes and hence equatorial convection are likely to occur in the Kelvin and R1 waves, and possibly also the n=2 Rossby (R2) wave. These regions are shown as the blue and red ovals in Figure 78. Whether the convection coincides with the blue or the red oval depends on the sign and magnitude of the background zonal wind. It is possible that the heating associated with the equatorial convection, induced by the wind-induced fluxes, may alter the theoretical structures of the Kelvin and Rossby waves. For the Kelvin wave, this could be just a modification of the dry structure. However, for the Rossby waves, the associated equatorial convection could lead to a radical change in structure and behaviour. Evidence has been found in Yang et al. 2002 that equatorial or near equatorial convection is indeed sometimes associated with the zonal wind extrema in the lower level Rossby and Kelvin waves, consistent with it being associated with a maximum in surface energy fluxes.

Figure 78. (a) Dimensional horizontal winds (vector) and divergence (colour) solutions of theoretical equatorial n=1 Rossby (R1) and Kelvin waves.Units are ms^-1 for winds and s^-1 for divergence. Divergence field of the theoretical structure indicate off-equatorial (equatorial) convection for the R1 (Kelvin) wave. Colour circles indicate the possible convective region induced by wind-dependent surface energy fluxes. (b) 850 hPa horizontal winds analysed from the ERA daily data projected onto the westward moving R1 and eastward moving Kelvin wave structure. Superimposed colours show the brightness temperature a T_b for the same days. Units are ms^-1 for winds and K for T_b. 'A' and 'B' indicate regions of equatorial convection which appear to be connected with zonal wind maxima of the R1 and Kelvin waves.Their propagation is highlighted by the two green arrow lines. (c) Seasonal histograms (colour) and linear regression lines of T_b and 850 hPa zonal wind u extrema associated with the R1 and Kelvin waves in the Eastern Hemisphere. Solid lines are regression lines for all T_b and dashed lines are for |T_b| greater or equal than 2K. All lines are significant at the 99% level or greater. In (b) and (c) the westward R1 and eastward Kelvin waves, and associated Tb, are for zonal wavenumber k=-2 ~ -10 (k=2~10) and period=3 ~30 days

As a further investigation, Figure 78b shows another example of an observed lower level westward moving R1 wave and an eastward moving Kelvin wave during July and September 1992. It is seen that a westward moving equatorial convection 'A' and eastward moving convection 'B' are closely related to the maximum westerly wind anomaly of the lower level waves, as indicated in the red circles of Figure 78a, rather than in the regions of low-level convergence. It is interesting to see also that, suppressed equatorial convection tends to be accompanied by equatorial easterly flow. Noting that the ambient flow at this time in these regions is westerly, the different behaviours further support the relationship of surface energy fluxes and convection. It is also found that the strong off-equatorial convection centred to the north of equatorial convection 'A' is closely related to anomalous westerly winds at 2-12° N associated with a lower level R2 wave and also coupled with a low level WMRG wave there (not shown). Since zonal winds for R2 waves are antisymmetric about the equator, the counterpart flow in the Southern Hemisphere is strong easterly which suppresses convection there and leads to an asymmetric off equatorial convection around 160° E region.

Figure 78c gives a seasonal statistical analysis of the relationship between equatorial convection and zonal wind (u) for the lower level R1 and Kelvin waves for the Eastern Hemisphere. Both histograms and linear regression lines indicate strong relationships between Tb and u for the two waves. Enhanced (suppressed) convection is closely related to westerly (easterly) winds for these waves. For |Tb| > 2K, 1 ms^-1 westerly wind is associated with a -3.85K and -6.16K in T_b for the two waves, respectively. The corresponding statistical analysis for the R2 wave (with T_b and u near 10° N) shows a similar result (not shown). All these regression relationships are very significant, at the 99% level or greater. It appears that wind-dependent surface energy fluxes are indeed playing an important role in determining the location of the convection associated with equatorial waves and that the convective response to the equatorial zonal winds associated with Rossby waves modifies the structure of the wave so that it no longer closely resembles its theoretical counterpart.

Neelin, J. D., I. M. Held, and K. H. Cook., 1987: Evaporation-wind feedback and low-frequency variability in the tropical atmosphere. J. Atmos. Sci., 44, 2341-2348.

Yang, G. Y, B. Hoskins, and J. Slingo, 2003: Convectively coupled equatorial waves: A new methodology for identifying wave structures in observational data. J. Atmos. Sci.,in press.

Zhang, G. J., 1996: Atmospheric intraseasonal variability at the surface in the tropical Western Pacific ocean. J. Atmos. Sci., 53, 739-758.