Geography Reference
In-Depth Information
constituents such as carbon dioxide, ozone, and even cloudiness are often employed,
but the more complete models do utilize model predicted zonally and time varying
cloudiness in their radiation codes.
Boundary layer fluxes of momentum, heat, and moisture are parameterized in
most AGCMs by the use of bulk aerodynamic formulas (see Section 5.3.1). Typi-
cally the fluxes are specified to be proportional to the magnitude of the horizontal
velocity at the lowest atmospheric level times the difference between the field
variable at the boundary and its value at the lowest atmospheric level. In some
models the boundary layer is explicitly resolved by locating several prediction
levels within the lowest 2 km and utilizing the model predicted boundary layer
static stability in the parameterization of turbulent fluxes.
The hydrological cycle is usually represented by a combination of parameter-
ization and explicit prediction. The water vapor mixing ratio is generally one of
the explicitly predicted fields. The distributions of layer clouds and large-scale
precipitation are then determined from the predicted distribution of humidity by
requiring that when the predicted humidity exceeds 100% enough vapor is con-
densed to reduce the mixing ratio to saturation or less. Parameterizations in terms
of the mean state thermal and humidity structure must be used to represent the
distributions of convective clouds and precipitation.
10.8.4
The NCAR Climate System Model
The Climate System Model (CSM) developed at the National Center for Atmo-
spheric Research (NCAR) is a coupled global climate model consisting of an
AGCM, an ocean GCM, a land surface biophysics model, and a sea-ice model. The
atmospheric component is a spectral model (see Section 13.5) with about 2.8
◦
hor-
izontal resolution and 18 layers in the vertical with the top layer centered at 4.8 hPa.
The ocean component of the CSM has 2.4
◦
resolution in longitude and variable
resolution in latitude, with smallest grid spacing of 1.2
◦
at the equator and in the
Arctic, and largest spacing of 2.3
◦
at midlatitudes. In the vertical, the ocean model
has 45 levels, with more than half these concentrated in the upper kilometer of the
ocean. The land surface model runs on the same grid as the AGCM, whereas the
sea-ice model runs on the ocean model grid.
Because characteristic timescales for the various components of the CSM are
very different, the individual components are integrated to approximate equilib-
rium states with fixed external forcing, and then the components are coupled
together and stepped forward until the coupled climate system equilibrates. Solstice
season zonal mean winds simulated by the CSM for late 20th century conditions
are shown in Fig. 10.23. Overall there is excellent agreement between the model
and observations.
Meridional cross sections of mean zonal wind (m s
−
1
) for the two solstice seasons.
(Left) Results from CSM simulation; (right) observed climatology. (After Dai et al., 2001.)
Fig. 10.23
