The midlatitude response to tropical Pacific SST anomalies involves changes in transient eddy propagation, but the processes leading to the transient eddy changes are still not clear. In a recent study, we used a series of controlled general circulation model (GCM) experiments in which an imposed tropical Pacific sea-surface temperature (SST) anomaly is turned on abruptly and the response is analyzed in terms of its high- and low-frequency parts, to show that the changes in transient eddies induced by El Niño Southern Oscillation (ENSO) arise from changes in wave refraction on the altered mean flow. In this work, we use a quasi-geostrophic linear model and a linear stationary wave model, to interpret the GCM experiments and obtain the sequence of events that lead from a tropical SST anomaly to the quasi-equilibrium change in the mean and transient atmospheric circulation. The initial direct response of the mean flow is confined to the tropical and subtropical Pacific, similar to what is obtained from a stationary wave model. This tropical-subtropical mean flow change initiates a transient eddy response, which induces a midlatitude mean flow anomaly. The wave-mean flow system evolves towards a state in which the eddy anomalies maintain the mean flow anomalies, allowing them to persist. It is further shown that, while eddy momentum fluxes persistently accelerate and decelerate the subtropical and midlatitude mean flow, the eddy heat flux effect on the zonal mean flow is much more variable, and only marginally significant. The linear quasi-geostrophic model calculations capture the evolution of eddy momentum flux anomalies equatorwards of 60°N quite well, suggesting linear wave refraction can explain the midlatitude ENSO anomalies. However, other processes, like stationary waves or changes in the nonlinear stage of eddy life cycles, are needed to explain the ENSO-related anomalies at high latitudes, polewards of around 60°N.
|Number of pages||15|
|Journal||Quarterly Journal of the Royal Meteorological Society|
|State||Published - Oct 2010|
- Quasi-geostrophic linear model