13JCSDA Exploring 4D variational and hybrid ocean DA methods with JEDI based Sea-ice, Ocean, and Coupled Assimilation (SOCA)

The Sea-Ice, Ocean, and Coupled Assimilation (SOCA) project is a data assimilation (DA) system that interfaces between the generic components of the JEDI system and ocean models, the Modular Ocean Model (MOM6). SOCA provides a suite of DA methods ranging from variational based (3DVAR, 3D-FGAT), ensemble based (Local Ensemble Transform Kalman Filter: LETKF, LETKF-EnVAR), to 3D and 4D hybrid (LETKF-EnVAR and hybrid EDA-EnVar) DA methods.

This study focuses on exploring the capabilities of 4D variational (4dVar) methods in SOCA. Implementing 4dVar methods requires a tangent linear model (TLM) of MOM6 which isn;t available and would be very involved and expensive to compute and maintain. To address this issue, JEDI system's newly integrated hybrid tangent linear model (HTLM) is utilized. The HTLM combines a simple tangent linear model,which is generally less expensive and straightforward to compute, with the information available from ensembles of TLMs. The feasibility of the 4dVAR is evaluated by comparing the computational and scientific performance of the HTLM-4dVar with existing DA methods in SOCA. The 1⁄4-degree MOM6 configuration for the regional North Atlantic domain serves as the test-bed for the implementation and development of the 4dVar DA system. These methods assimilate satellite Altimeter Absolute Dynamic Topography (ADT) products from CRYOSAT-2, JASON-2 & 3, Saral, SENTINEL-3A, 3B, & 6A; and satellite sea surface temperatures (SST) product from VIIRS- NPP.