Ocean DA experimets with 6-hour cycles, targeting a future weakly-coupled atmosphere-ocean DA system

     KIAPS (Korea Institute of Atmospheric Prediction Systems) has recently been adding Earth-system components to the atmospheric version of KIM (Korean Integrated Model), to produce a prediction model that considers various geophysical factors such as atmosphere, ocean, sea-ice, and land, to support skillful forecasting out from the very short range (~6 hours) to the extended medium range (up to ~30 days). The ocean component is a version of the NEMO (Nucleus for European Modelling of the Ocean) ocean model. For data assimilation (DA), our first goal is to develop a weakly-coupled DA system that combines KIM’s existing atmosphere/land DA system with a separate DA system for NEMO. This ocean DA system is based on NEMOVAR, and uses a 3-Dimension Variational – First Guess at Appropriate Time (3DVar-FGAT) DA method. The assimilated observations include SST observations from satellites and moored buoys, temperature profiles collected by Argo floats, and satellite observations of sea-level anomaly and sea ice concentration.
     The ocean DA system is being developed from an existing ocean-only NEMOVAR system that is based on 24-hour DA windows. In order to make this system compatible with the existing atmospheric DA system, we have changed the ocean DA cycling strategy to match the 6-hourly strategy used by the atmospheric system, including reductions to the ocean observation cutoff times. After introducing the details of this system, we will present results from experiments designed to test (a) the impact of changing the DA window lengths, without changing observation usage, and (b) the further impact of using earlier observation cutoff times.