FESOM2 simulations with increasing sea-ice model complexity under different atmospheric forcings

• Main Authors: Zampieri, Lorenzo, Kauker, Frank, Fröhle, Jörg, Sumata, Hiroshi, Hunke, Elizabeth C., Goessling, Helge F.
• Format: Dataset
• Language: English
• Published: Zenodo 2020
• Subjects: Sea-ice simulation; Parameter optimization; JAMES; Arctic; Antarctic; Green's function; Model development; Antarc*; Sea ice
• Online Access: https://dx.doi.org/10.5281/zenodo.4038252; https://zenodo.org/record/4038252

Introduction This dataset has been compiled in support of the paper "Impact of sea-ice model complexity on the performance of an unstructured sea-ice/ocean model under different atmospheric forcings" by Zampieri et al., submitted to the Journal of Advances in Modeling Earth Systems (JAMES) published by the American Geophysical Union (AGU). Scientific description of the dataset The dataset contains the results of sea-ice simulations performed with the Finite-volumE Sea ice-Ocean Model version 2 (FESOM2), based on six model configurations: C1-E, C1-N, C2-E, C2-N, C3-E, and C3-N. As described in the paper, the complexity of the sea-ice model increases from the setup C1 to C3. The suffix -E and -N indicate respectively the ERA5 and NCEP atmospheric forcings used as boundary conditions for the FESOM2 model. As two iterations of the Green's function approach for the optimization of the parameter space have been performed, each configuration features three separate simulations: a control run (cnt), a first-round of optimization (opt_1), and a second and final round of optimization (opt_2). The parameter optimization is based on various sea-ice observations retrieved over the period 2002–2015. In total, 18 simulations compose the dataset (6 configurations x 3 realizations). The following 2D monthly-averaged variables are provided: the sea-ice concentration, the sea-ice thickness, the meridional and zonal components of the sea-ice velocity, and the snow thickness on top of the sea ice. The fields are defined on a global unstructured mesh denominated "CORE2", which is also included in the database. Technical description of the dataset As an unstructured model output is not widely diffused in the sea-ice community, we include here some suggestions for handling and analyzing the simulation results. The files can be interpolated to a regular grid using the following CDO commands: Add grid description to model file: cdo setgrid,CORE2_mesh.nc var.fesom.yyyy.nc temp.nc Interpolate to regular grid: cdo remapycon,r360x180 temp.nc var.fesom.interpolated.yyyy.nc Furthermore, the python package pyfesom2 can be used for plotting the unstructured model data and for interpolating it to a regular grid. The R package spheRlab can be used for plotting the model data directly on its unstructured grid and for performing further analysis. More information can be found on the FESOM website . The following naming convention is adopted for the model variables: a_ice → sea-ice concentration m_ice → sea-ice volume per unit area of ice m_snow → snow-volume per unit area of ice vice → meridional component of the sea-ice velocity uice → zonal component of the sea-ice velocity Three types of simulation are included: cnt → control run before the parameters optimization (2000–2019) opt_1 → after the first iteration of the parameter optimization method (2000–2015) opt_2 → after the second iteration of the parameter optimization method (2000–2019) Do not hesitate to contact the corresponding author ([e-mail]) for additional information about the data processing and for any other issue with this dataset. : We acknowledge the European Union's Horizon 2020 Research and Innovation program project APPLICATE (grant 727862) and the Federal Ministry of Education and Research of Germany (BMBF) in the framework of SSIP (grant 01LN1701A) for funding this research. ECH acknowledges support from the Energy Exascale Earth System Modeling (E3SM) project of the US Department of Energy's Office of Science, Biological and Environmental Research division. Furthermore, we are grateful to the German Climate Computing Centre (DKRZ) for granting computational resources through the BMBF computing project "Impact of sea ice parameterizations on polar predictions".