Roles of ocean initialization and wind bias correction on ENSO predictability

Abstract

El Ni&ntilde o and Southern Oscillation is one of the most well-known and important climate phenomena.Although the ENSO appears in the tropical Pacific, it interacts with climate variability over the world,which impacts the human life by various ways. KIOST has been developed an ENSO predictionsystem by applying the ocean data assimilation and wind bias correction to a fully coupled climatemodel, GFDL CM2.1. The ocean observation data are assimilated into its ocean component modelthrough the data assimilation system of the KIOST (DASK) while other component models are freelyintegrated. Even though atmospheric observation variables are not assimilated, the wind bias of theDASK has been corrected through applying a simple wind bias correction when calculating the airseafluxes. We evaluated the variability of the ocean climate in the climate reanalysis by the DASKfrom 1947 to 2012. The DASK represents global temperature and salinity well, not only at the surfacebut also at intermediate depths in the ocean. The DASK’s ocean climate variability also matches wellwith observations of the ENSO, Pacific Decadal Oscillation and Indian Ocean Dipole. The heat contentof the DASK shows a good correlation with real-world observations. In this study, we use thereanalysis data from the DASK as an initial condition of our ENSO prediction system. To evaluate theENSO prediction system, hindcast experiments hy various ways. KIOST has been developed an ENSO predictionsystem by applying the ocean data assimilation and wind bias correction to a fully coupled climatemodel, GFDL CM2.1. The ocean observation data are assimilated into its ocean component modelthrough the data assimilation system of the KIOST (DASK) while other component models are freelyintegrated. Even though atmospheric observation variables are not assimilated, the wind bias of theDASK has been corrected through applying a simple wind bias correction when calculating the airseafluxes. We evaluated the variability of the ocean climate in the climate reanalysis by the DASKfrom 1947 to 2012. The DASK represents global temperature and salinity well, not only at the surfacebut also at intermediate depths in the ocean. The DASK’s ocean climate variability also matches wellwith observations of the ENSO, Pacific Decadal Oscillation and Indian Ocean Dipole. The heat contentof the DASK shows a good correlation with real-world observations. In this study, we use thereanalysis data from the DASK as an initial condition of our ENSO prediction system. To evaluate theENSO prediction system, hindcast experiments h