DATA ASSIMILATION OF THE GLOBAL OCEAN USING THE 4D LOCAL ENSEMBLE TRANSFORM KALMAN FILTER (4D-LETKF) AND THE MODULAR OCEAN MODEL (MOM2)

Abstract

The 4D Local Ensemble Transform Kalman Filter (4D-LETKF), originally designed for atmospheric applications, has been adapted and applied to the Geophysical Fluid Dynamics Laboratory's (GFDL) Modular Ocean Model (MOM2). This new ocean assimilation system provides an estimation of the evolving errors in the global oceanic domain for all state variables. Multiple configurations of LETKF have been designed to manage observation coverage that is sparse relative to the model resolution. An Optimal Interpolation (OI) method, implemented through the Simple Ocean Data Assimilation (SODA) system, has also been applied to MOM2 for use as a benchmark. Retrospective 7-year analyses using the two systems are compared for validation. The oceanic 4D-LETKF assimilation system is demonstrated to be an effective method for data assimilation of the global ocean as determined by comparisons of global and regional `observation minus forecast' RMS, as well as comparisons with temperature/salinity relationships and independent observations of altimetry and velocity.