Role of the initial ocean state for the 2006 El Nino

[1] In this study, we present results which are designed to isolate the role of the ocean initial condition prior to the 2006 El Nino. Analysis of the differences between model experiments with the same forcing but different initial conditions shows several key dynamical elements of the 2006 El Nino that were evident in the initial state of the ocean. Between November 2005 and March 2006, the eastern Pacific was cold due to an upwelling Kelvin wave. By May 2006, the combined impact of downwelling Kelvin and upwelling Rossby waves had carried warm SST eastward over the entire equatorial Pacific. From May until September, westward advection of cold waters by a downwelling Rossby wave temporarily reduced temperatures in the tropical Pacific. The El Nino was fully manifest when the second downwelling Kelvin wave traversed the Pacific in October to December 2006. A statistical atmospheric model forced by the sea surface temperature anomalies from these experiments isolates the atmospheric response to oceanic initial condition and can help guide coupled model forecasts. East of the dateline, easterlies limit the early growth of this event prior to May 2006 through destructive interference of any downwelling signal. During May 2006, warm SST is associated with westerly wind anomalies which stretch across the entire basin. However, cold advection from an upwelling Rossby wave limits the duration of these westerlies. After July, westerlies prevail over the central Pacific as the 2006 El Nino develops into a fully coupled event. Although high quality global ocean observations allow ever improving estimates of the initial ocean state for coupled ocean-atmosphere prediction models, the amplitude of 2006 El Nino was under-predicted prior to September 2006 as measured by composite forecast indices. Our methodology of isolating the role of the initial state of the ocean reproduces all the major features of the 2006 event and has merit for use in conjunction with coupled-model forecasts to improve long-range predictions of El Nino.