Evaluating the GECCO3 1948-2018 Ocean Synthesis - a configuration for initalizing the MPI-ESM climate model
Evaluating the GECCO3 1948-2018 Ocean Synthesis - a configuration for initalizing the MPI-ESM climate model
Abstract:
The paper describes an update of the GECCO ocean synthesis, now in its version 3, and provides an evaluation of the results with assimilated and independent data. GECCO3 covers the 71-year period 1948-2018 and different from its predecessor returns to a single assimilation window, instead of partitioning the period in 5-year long overlapping windows that was necessary before to yield convergence. Problems with the partitioning are documented and a solution to the convergence problem is presented. GECCO3 is intended to be used for the initialization of coupled climate models and is configured for the higher resolution version of the earth system model developed at the Max Planck Institute for Meteorology (MPI-ESM). It uses the bathymetry and grid of the MPI-ESM with quasi-uniform resolution of 0.4\dgr and provides the first global eddy-permitting synthesis based on the adjoint method. The synthesis features additionally the estimation of various mixing parameters and can regionally choose between explicit or parameterized eddy-fluxes.
Except for the altimeter data in tropical regions, GECCO3 is in better agreement with the assimilated data than GECCO2. The improvements relative to the in-situ data partly result from the much larger amount of Argo data, which show lower model-data differences. Global heat content changes are in good agreements with recent estimates but show uptake almost exclusively in the top 700 m. An alternative version of GECCO3, created by starting from different first-guess control parameters, was used to evaluate the uncertainty of the estimated parameters and state due to lack of convergence. This estimate suggests a large uncertainty related to the uptake of heat into the lower layers, while estimates of mean meridional transports of heat and freshwater are not affected.
Except for the altimeter data in tropical regions, GECCO3 is in better agreement with the assimilated data than GECCO2. The improvements relative to the in-situ data partly result from the much larger amount of Argo data, which show lower model-data differences. Global heat content changes are in good agreements with recent estimates but show uptake almost exclusively in the top 700 m. An alternative version of GECCO3, created by starting from different first-guess control parameters, was used to evaluate the uncertainty of the estimated parameters and state due to lack of convergence. This estimate suggests a large uncertainty related to the uptake of heat into the lower layers, while estimates of mean meridional transports of heat and freshwater are not affected.