Scalable Coupled Ensemble Data Assimilation with AWI-CM and PDAF

• Main Authors: Nerger, Lars, Tang, Qi, Sidorenko, Dmitry
• Format: Conference Object
• Language: unknown
• Published: 2020
• Subjects: Sea ice
• Online Access: https://epic.awi.de/id/eprint/51253/; https://epic.awi.de/id/eprint/51253/1/poster_AWICM-PDAF_weiss2.pdf; https://hdl.handle.net/10013/epic.02903a63-6a9d-40de-9dac-bf5b44447da0; https://hdl.handle.net/

We discuss a strategy to build a highly scalable and flexible data assimilation system on the basis of the Parallel Data Assimilation Framework (PDAF, http://pdaf.awi.de) using the example of the coupled climate model AWI-CM (Sidorenko et al., Climate Dynamics, 44 (2015) 757-780). AWI-CM consists of the finite-element sea ice-ocean model FESOM, which uses an unstructured model grid, and the model ECHAM6 for the atmosphere. The model compartments are coupled using OASIS3-MCT. The model system consists of two separate executable programs for the ocean and atmosphere. The assimilation system is generated by online-coupling of AWI-CM and PDAF. This modifies AWI-CM to perform ensemble forecasting and data assimilation and allows to fully keep the ensemble information in memory avoiding costly file operations and model restarts. The resulting assimilation system supports to apply the assimilation both in-compartment (i.e. weakly-coupled) as well as cross-compartment (i.e. strongly-coupled). Discussed are the structure and computational performance of the assimilation system as well as results from the assimilation of sea surface temperature and ocean profile data sets into a realistic configuration of AWI-CM.