Coupled
Colorado State
Model
(CCoSM)
Model Description & Documentation
The Coupled Colorado State Model (CCoSM) is both a collection of climate component models and a framework for their coupling. All component models are discretized on a geodesic grid with hexagons and pentagons. A component may be dynamic, with full prediction of prognostic variables, or dummy, with key variables prescribed from outside data sets. These components are coupled in a consistent fashio
n with interfacial fluxes computed in the coupler and passed to the components i
n a conservative and efficient manner. All components share common utility softw
are for geodesic grid generation and operators.
Atmosphere
- Dummy (prescribed) atmosphere:
This atmosphere component reads high frequency surface meteorology and fluxes from either ERA-40 reanalysis interpolated to the geodesic grid, or previous sigma-coordinate (BUGS5) atmosphere runs.
- Sigma Coordinate Atmosphere (BUGS5):
The atmosphere has prognostic equations for potential temperature, vorticity and divergence, surface pressure, specific humidity, cumulus kinetic energy (CKE), mixing ratios of cloud water, cloud ice, rain and snow, and the planetary boundary layer (PBL) depth. The discretization of the advection is highly conservative. The vertical coordinate is a generalized sigma coordinate, where the PBL top is a coordinate surface. Deep convection is parameterized with a modified Arakawa-Schubert scheme with ice, prognostic CKE, cumulus friction, and multiple cloud bases. The large-scale cloud microphysical scheme is Fowler-Randall-Rutledge, with cumulus detrainment as a source of cloud water and/or ice. The radiation is the Stephens parameterization. Gravity-wave drag is parameterized with a simple Palmer-like scheme. This atmosphere can also be run uncoupled to the surface, e.g. Held-Suarez simulations.
- Quasi-Lagrangian Coordinate Atmosphere (BUGS6):
The atmosphere has prognostic equations for potential temperature, vorticity and divergence, surface pressure, specific humidity, cumulus kinetic energy (CKE), mixing ratios of cloud water, cloud ice, rain and snow, and the planetary boundary layer (PBL) depth. The discretization of the advection is highly conservative. The vertical coordinate is a generalized hybrid coordinate with sigma-lika characteristics near the surface transitioning to isentropic in the upper atmosphere. Additionally, the grid staggering of the variables is Charney-Phillips, with vorticity and divergence at the layer centers, and potential temperature and moisture at the layer edges. The parameterized physics are adapted from the sigma-coordinate model: deep convection is parameterized with a modified Arakawa-Schubert scheme with ice, prognostic CKE, cumulus friction, and multiple cloud bases; the large-scale cloud microphysical scheme is Fowler-Randall-Rutledge, with cumulus detrainment as a source of cloud water and/or ice; the radiation is the Stephens parameterization; gravity-wave drag is parameterized with a simple Palmer-like scheme. This atmosphere can also be run uncoupled to the surface, e.g. Held-Suarez simulations.
Ocean
- Dummy (prescribed) ocean/ice:
This ocean/ice component reads in prescribed monthly sea surface temperature and sea ice cover. The data read has been interpolated to the geodesic grid. Data options are AMIP or climatology. The sea ice thermodynamics are one-layer Semtner. The ocean also features prognostic slab ocean option.
- Dynamic ocean (DAFFY):
The ocean has prognostic equations for momentum, temperature, salinity and the free surface height. Vertical coordinate options are a z-level coordinate and an isopycnic coordinate. Horizontal transport is done by monotone flux-corrected transport, and vertical transport by monotone remapping. KPP is used to parameterize the ocean boundary-layer and convection. This ocean can also be run uncoupled to any atmosphere.
Sea Ice
- Dynamic sea ice:
The sea ice predicts ice concentration, volume and energy content. There are multiple thickness categories possible, multiple layers in each category, and snow is accumulated on the ice. The dynamics are based on the Hunke and Dukowicz elastic-viscous-plastic rheology. The thermodynamics use either a Semtner scheme or CICE thermodynamics. Options for advecting the ice include llux-corrected transport or incremental remapping.The sea ice can also be run uncoupled.
Coupler
- Coupler:
The coupler coordinates the integration of the other climate components. It computes interfacial fluxes using state variables received from the components, and passes those fluxes on to the components. If the atmosphere and the surface are at different resolutions the coupler conservatively interpolates between the grids using SCRIP. The coupler operations are performed on the surface resolution, which is equal to or finer than the atmospheric resolution. Atmospheric PBL physics are also computed here, as they are assumed to be tightly coupled to surface conditions. Bulk-aerodynamic formula are used to compute the energy water and momentum fluxes over the ocean and sea ice. The land-surface parameterization, SiB2, is also computed here.