Process studies of the Equatorial Pacific Ocean and its adjustment to intraseasonal wind forcing

The role of atmospheric intraseasonal variability on the evolution of the Equatorial Pacific Ocean is addressed in a series of process study experiments. The adjustment of the upper ocean to a Westerly Win Event (WWE) is modelled with the Océan Parallélisé (OPA) GCM to examine the sensitivity of the oceanic response to the variability of the background state of the Equatorial ocean. Different ocean conditions are configured in an idealised Equatorial ocean basin, depending on the strength of the background Trade winds and the horizontal viscosity. An analytical WWE is applied and we study the Kelvin wave adjustment and its sensitivity to the background ocean conditions. The model TIW field exhibits considerable temporal and spatial variability that is sensitive to the model configuration. A viscosity regime of order 103m2s-1 is necessary to simulate a realistic TIW field. The Kelvin wave generates velocity and SST anomalies when it encounters the TIW field and leads to a phase shift of the TIW field. The Kelvin wave is decomposed into its normal modes and we find that the surface signature of the Kelvin wave is predominantly due to the second baroclinic mode. The Kelvin wave generally has a bimodal structure consisting of the first and second baroclinic modes. The amplitude of the second baroclinic mode is sensitive to the background mean state and is significantly diminished by a strengthened Equatorial circulation and steepened thermocline in the increased background wind field experiments and the low viscosity regime, leading to a weaker surface ocean response. The results show that the model response to a WWE and its impact on the upper ocean depends on the oceanic conditions at the time of the wind perturbation, with the projection of the WWE onto the ocean modes determining the Kelvin wave structure. The analysis is extended to consider two observed WWEs, one in March 1997 that led to a strong ocean adjustment during the growth of the 1997 El Niño, and one in December 2002 that did not lead to strong surface anomalies.