Search for: All records

Previous research has shown that the equatorial upper tropospheric circulation signal associated with the Madden Julian Oscillation (MJO) over the Indian Ocean behaves like a Kelvin wave, with the eastward-propagation of the associated zonal wind anomaly caused by acceleration by the geopotential gradient force in quadrature with the wind anomaly, with the resultant signal amplified or decayed as the MJO wind advects background zonal wind in regions of background confluent or diffluent flow, with its phase speed adjusted by Doppler shifting by background zonal wind. This paper assesses these previously diagnosed mechanisms in the GEFS V12 forecast model, showing that similar mechanisms occur with the model MJO, but weaker and less organized. Results suggest that, relative to the validation data, the model stalls the MJO upper tropospheric zonal wind anomaly during diffluent background conditions near the Maritime Continent and weakens its amplitude more rapidly than validation data. The stalled propagation leads the model MJO to persist the signals of advection of and by the background wind, but at a reduced rate as the scale of the model MJO wind anomaly diminishes. Results show that beyond the stronger Doppler effect in the model, stalling the model MJO results from filling of the relative geopotential trough collocated with the MJO easterly wind anomaly, leading to breakdown of the geopotential gradient force term that is responsible for propagation of the wind anomaly. Thus, when the Maritime Continent region experiences zonally diffluent flow, the reason for the stronger Maritime Continent barrier effect in the model is that the model does not persist the Kelvin wave propagation mechanism that continues the eastward movement of the easterly wind anomaly in observations.