Aller au contenu. | Aller à la navigation

Laboratoire d’Etudes en Géophysique et Océanographie Spatiales

Outils personnels

This is SunRain Plone Theme


Vous êtes ici : Accueil / Members / Boris Dewitte / bxd_data / resources / d98


Par Webmaster Legos Dernière modification 31/07/2012 12:11

Sensitivity of an intermediate ocean-atmosphere coupled model of the Tropical Pacific to its oceanic vertical structure


Boris Dewitte

An intermediate coupled model (ICM) consisting of a 3 baroclinic mode tropical Pacific ocean and a Gill(1980)'s tropical atmosphere is presented. The mixed layer is similar to the Zebiak&Cane coupled atmosphere-ocean model (Hereafter ZC model) but uses a new mean state and a different parameterization for subsurface temperature. The impact of the differences in the parametrizations and mean states between the ICM and the ZC model is discussed. A 100-yr simulation is made and the long term variability of the ICM is compared to the one of the ZC model. The ICM reproduces regular ENSO events with a spectrum with peaks at the periods 3.3 and 3.6 years. Furthermore due to the presence of higher-order vertical modes in the oceanic component of the model the zonal current anomalies variability is much more realistic than the one obtained in the ZC model. The nature of the processes sustaining the oscillations in the ICM is then investigated based on the coupled instabilities and 'delayed oscillator' theories. It is shown that the oceanic vertical structure determines the ENSO period in the model. High order modes tend to generate warm anomalies in the central Pacific and set an equilibrium between the atmosphere and vertical advection leading to unstable Kelvin mode propagating slowly eastward during El Niño event. Conversely gravest modes favour unstable Rossby modes. The cold phase is a fast westward propagation as in the ZC model which is not very sensitive to the vertical structure of the ocean. Rossby wave reflection at the western meridional boundary is important in the model for the shifts from a warm (cold) phase to a cold (warm) phase consistently with the 'delayed oscillator' theory. The efficiency of the shift is however dependent on the vertical structure of the reflected Kelvin waves. For instance a second baroclinic Kelvin wave is more efficient in initiating the warming than a single first baroclinic Kelvin wave. Results of sensitivity tests to the parameterization of thermocline displacement anomalies based on realistic density profiles are consistent with recent studies suggesting the modulation of occurence of the El Niño events by tropical-extratropical teleconnections.

Acknowledgements. I am grateful to Gilles Reverdin for thoroughly reading the manuscript and for many invaluable discussions. This work has also benefitted from fruitful discussions with Claire Perigaud and Marc Pontaud. I would like to thank Ying-Quei Chen, David Battisti and Steve Zebiak for their help validating the multimode ocean model. Discussions with Yves DuPenhoat, Darya Gushchina and Rodrigo Abarca del Rio were also appreciated. Thanks are due to David Legler for providing the updated FSU pseudostress data. This work was supported by CNRS (Centre National de la Recherche Scientifique).

Actions sur le document