Investigation of the Rossby Wave Source components: Implications for the Predictability of the MJO and a New Pathway of the Influence of Mid-Latitude Transients

The Rossby Wave Source (RWS) diagnostic was designed to show the effective source of the barotropic component of the extra-tropical response to tropical heating. By combining the upper-level stretching term in the vorticity equation with the advection of the absolute vorticity by the divergent component of the flow (the “advection” term), the RWS becomes largest in the upper-level Westerly background flow in the sub-tropics, outside the near-equatorial Easterlies, thus allowing for stationary Rossby wave propagation. While the RWS has been widely used in conjunction with the extra-tropical teleconnections associated with the MJO, (which are routinely interpreted using stationary wave concepts), uncertainties in its application remain. It is not clear for which latitudes the RWS isolates the effective forcing from the extra-tropical response, and if this is even possible. Additionally, the role of the stretching term vis-à-vis the advection remains under-explored. In the context of MJO teleconnections, the predictability of the RWS (and not that of the tropical heating) is the limiting factor for the predictability of the extra-tropical response. We report on a suite of 51-member ensembles of the coupled model of the European Centre for Medium-Range Weather Forecasts with atmospheric resolution of 36 km. Ensembles of two-month reforecasts were carried out for 13 initial dates during MJO phase 3, with 8 of these having 01 November start dates for different years and 5 having 01 January start dates. In these reforecasts, no initial condition perturbations were applied, and the stochastic perturbations to the parameterized tendencies were turned off except in the tropical Indo-Pacific region. We find that the initial signal of the tropical MJO propagation is carried by the advection term in the RWS, but that subsequently the stretching term dominates as the signal disperses. Further analysis of the stretching term shows dominant eastward propagation in the sub-tropics throughout the forecasts, which we hypothesize is due to the sub-tropical edge of mid-latitude baroclinic disturbances. This hypothesis is supported by previous research showing that the leading empirical orthogonal functions of the RWS are associated with propagating wave trains, and suggests that a new mechanism by which transient eddies affect the slowly varying flow – namely by modulating the RWS that carries the tropical signal. The growth of the ensemble spread of the RWS is dominated by the stretching terms for zonal wavenumbers greater than 5 (again consistent with baroclinic systems playing a role), but a new finding is that the spread of the stretching and advection components cancel to a large extent for planetary waves (zonal waves 1 – 3). We plan to further investigate the mechanism of this cancellation, by which the total RWS is more predictable than either of its components, in the ensemble suite described above and other suites of reforecasts. This analysis will enhance our understanding or the predictability of the MJO.