.. _icon: ICON ==== The ICOsahedral Nonhydrostatic (ICON) model is developed by the ICON partnership: initially developed in a partnership of the Max-Planck Institute of Meteorology (MPI-M) and the German Weather Service (DWD), and is now further developed and maintained in a larger consortium, including the German Climate Computing Centre (DKRZ), the Swiss Federal Institute of Technology (ETH), MeteoSwiss, and the Karlsruhe Institute of Technology (KIT). :cite:`Hohenegger2023` present a detailed first description of the km-scale version of ICON. ICON employs an Arakawa C-type stenciling (icosahedral-triangular C grid) of mass and momentum for each one of its coupled Earth system components: atmosphere, land, and ocean. For more information about the dynamical core and grid, the reader is referred to :cite:`Zaengl2015`, the chapter 3 of the ICON tutorial :cite:`icontutorial`, and the `ICON model documentation `_. The supported and public ICON grids for atmosphere and ocean can be found on the `ICON grids page `_. Below you can find links to the different simulations as well as an introduction to the ICON model. Simulations ########### .. toctree:: :maxdepth: 1 control historical_projection Atmosphere component #################### The atmospheric component is modelled with non-hydrostatic equations. It retains only parameterization for the radiant energy transfer, which is based on RRTMGP (Rapid Radiative Transfer model for General circulation model applications, :cite:`Pincus2019`), a single moment bulk microphysical parameterization consisting of five condensate habits :cite:`Baldauf2011`, and a 3D Smagorinsky turbulence mixing :cite:`Dipankar2015` :cite:`Lee2022`. Additionally, it includes a simple plume implementation for modelling the anthropogenic aerosol optical properties and an associated Twomey effect :cite:`Stevens2017`. The vertical structure of the atmosphere is discretized as a terrain-following stretched vertical grid with 90 levels. The vertical discretization is finely spaced close to the surface than at the model top at 75km, with a damping layer with increasing strength from 44km upwards. The atmospheric physical step, e.g., transport and microphysics, is called every 40s with five dynamical substeps, while the radiation step is called every 10 minutes for the horizontal resolution of 5 km. Land component ############## The land component, land surface processes, including biochemistry, are simulated by the Jena Scheme for Biosphere-Atmosphere Coupling in Hamburg (JSBACH, :cite:`Reick2021`). The land component is coupled to the atmosphere through the atmospheric diffusion equation for vertical turbulent transport following the Richtmyer and Morton numerical scheme :cite:`Richtmyer1994`. The atmosphere receives information about albedo, roughness length, and the required parameters to compute latent and sensible heat fluxes. At the same time, the land component interacts with the ocean component via the hydrological discharge model of :cite:`Hagemann1997`, with river directions determined by the steepest descent :cite:`Riddick2021`. The land component is called at the same time as the main atmospheric processes and integrated on the same grid. Ocean component ############### The ocean component (ICON-O) solves the hydrostatic Boussinesq equations :cite:`Korn2018` :cite:`Korn2022`. In contrast to the atmosphere component, ICON-O uses a novel concept of Hilbert space compatible reconstructions to calculate volume and tracer fluxes on the staggered grid :cite:`Korn2018`. ICON-O poses two vertical coordinates, Z and Z*, to track changes in the movement of the ocean surface and avoid numerical discretization errors due to negative layer thickness. Z vertical coordinate describe the fixed depth of the ocean, while Z* stretches vertically with the free surface of the ocean (change with sea surface height). :cite:`Korn2022` describe in detail the remaining parameterization of ICON-O, which includes parameterization for the vertical turbulent mixing and velocity dissipation. Additionally, ICON-O contains the sea ice model with a thermodynamic component, describing freezing and melting by a single-category (zero-layer formulation, :cite:`Semtner1976`), and a dynamical component based on the sea ice dynamics component of the Finite-Element Sea Ice Model (FESIM, :cite:`Danilov2015`). The ocean component is called every 5 minutes for the horizontal resolution of 5 km and exchanges information with the atmosphere through Yet Another Coupler (YAC, :cite:`Hanke2016`) every 10 minutes, alongside hydrological discharge from the land component. The atmosphere components provide different physical processes to the ocean component over ocean and ice grid cells, i.e., the zonal and meridional wind stresses, the surface freshwater flux (rain and snow), evaporation rate, shortwave and longwave radiation, and latent and sensible heat fluxes. While the ocean component provides information regarding the sea surface temperature, the zonal and meridional sea surface velocity components, the ice and snow thickness, and the ice concentration. .. _icon_initial_conditions_and_forcing: Initial conditions and forcing ############################## ICON initial conditions are interpolated onto the horizontal grid at a pre-processing step. The orography is derived from the Global Land One-km Base Elevation Project (GLOBE). The atmosphere, solid moisture, snow cover, and soil and surface temperature are derived from the European Centre for Medium-Range Weather Forecast (ECMWF) without being spun up for the chosen start date. The ocean initial state is taken from a spun-up ocean model run with climatology forcing as described in :cite:`Hohenegger2023`. Global mean concentrations of greenhouse, ozone, and aerosol gases are set to their respective values of the simulated year. Between 1990-2014, a climatology is used and after 2015, SSP3-7.0 is used. .. Bibliography .. ############ .. .. bibliography:: ref_icon.bib .. :filter: docname in docnames .. .. _autosubmit_docs: https://autosubmit.readthedocs.io/en/master/