Data Assimilative Ocean Circulation Modeling Systems of KIOST

Abstract

Data Assimilation System of the KIOST (Korea Institute of Ocean Science and Technology) (DASK) has been developed based on the Ensemble Optimal Interpolation(EnOI). In the case where the computational resource is limited, the Ensemble Optimal Interpolation may provide an operational and cost-effective alternative to the Ensemble Kalman Filter (EnKF Oke et al., 2007). In fact, the EnOI estimates the background error covariance by using a stationary ensemble instead of ensemble model runs of the EnKF (Evensen, 2003).The DASK has been applied to a fully coupled climate model, GFDL CM2.1. While the ocean observation data are assimilated into the ocean component model, the other component models are freely integrated. By applying the DASK to the GFDL CM2.1, we have produced KIOST climate reanalysis from 1947 to 2012. To assess oceanic processes and ocean climate variability of the climate reanalysis, we examined the North Pacific Intermediate Water, El Nino Southern Oscillation, Pacific Decadal Oscillation, Indian Ocean Dipole, upper 300 m heat content as well as mean global temperature and salinity. It is noteworthy that the DASK performs better particularly in the North Pacific and North Atlantic even though the DASK applies the Ensemble Optimal Interpolation, which is numerically simpler than the Ensemble Kalman Filter applied by the ECDA (GFDL Ensemble Coupled Data Assimilation Systems Chang et al., 2013), and a lal Interpolation may provide an operational and cost-effective alternative to the Ensemble Kalman Filter (EnKF Oke et al., 2007). In fact, the EnOI estimates the background error covariance by using a stationary ensemble instead of ensemble model runs of the EnKF (Evensen, 2003).The DASK has been applied to a fully coupled climate model, GFDL CM2.1. While the ocean observation data are assimilated into the ocean component model, the other component models are freely integrated. By applying the DASK to the GFDL CM2.1, we have produced KIOST climate reanalysis from 1947 to 2012. To assess oceanic processes and ocean climate variability of the climate reanalysis, we examined the North Pacific Intermediate Water, El Nino Southern Oscillation, Pacific Decadal Oscillation, Indian Ocean Dipole, upper 300 m heat content as well as mean global temperature and salinity. It is noteworthy that the DASK performs better particularly in the North Pacific and North Atlantic even though the DASK applies the Ensemble Optimal Interpolation, which is numerically simpler than the Ensemble Kalman Filter applied by the ECDA (GFDL Ensemble Coupled Data Assimilation Systems Chang et al., 2013), and a l