AMIP I/AMIP II Differences: Model PNNL CCM2 (T42 L18)
1997
AMIP II Model
Designation
PNNL CCM2 (T42 L18) 1997
Most Similar AMIP
I Model
NCAR
CCM2 (T42 L18) 1992
AMIP I/AMIP II Model Differences
Equations of State
The AMIP II model uses the state variables of the AMIP
I model plus additional prognostic variables required for the prognostic
cloud
formation scheme. The latter includetotal moisture mixing ratio, ice
mixing ratio and number concentration, and condensation-conserved temperature.
Radiation
The cloud-radiative interactions of the AMIP II model are different from
those of the AMIP
I model. Convective clouds are treated as radiatively transparent.
The shortwave optical properties of large-scale clouds for the delta-Eddington
approximation (optical depth, single scattering albedo, and asymmetry factor)
are parameterized in terms of the cloud liquid water and ice water
paths, the droplet effective radius, and the crystal effective size. The
droplet effective radius is parameterized in terms of the cloud liquid
water content (diagnosed from the prognostic total moisture mixing ratio)
and a prescribed droplet number concentration. The ice crystal effective
size is parameterized in terms of the prognostic ice water content and
ice crystal number concentration. Longwave broad-band emissivity of large-scale
cloud is a negative exponential function of liquid water path, with the
cloud ice absorption coefficient specified as 1x10-4 m2Kg-1.
Large-scale cloud fills the entire grid square (i.e., cloud fraction is
1, with full vertical overlap). Cf. Ghan et al. (1997)[39]
for further details.
Convection
In order to better simulate cloud radiative forcing, the characteristic
convective adjustment time scale is increased to 5400 seconds in the AMIP
II model from its value of 3600 seconds in the AMIP
I model.
Cloud Formation
-
The prognostic scheme of Ghan et al. (1997)[39]
determines large-scale cloud formation in AMIP II, whereas a diagnostic
scheme is used in the AMIP
I model. The prognostic variables include total moisture mixing
ratio rw, condensation-conserved temperature Tc,
cloud ice mixing ratio ri and cloud ice number concentration
Ni. Cloud liquid droplet number is prescribed. Large-scale
cloud is assumed to fill the grid square (i.e., cloud fraction 1, with
full vertical overlap). The treated microphysical processes include: condensation
of water vapor and evaporation of cloud water and rain; nucleation of ice
crystals; vapor deposition and sublimation of cloud ice and snow; autoconversion
and accretion of cloud water; aggregation and collection of cloud ice;
melting of ice and snow; riming on ice and snow; and gravitational settling
of ice; rain, and snow. Saturation of water vapor with respect to
ice vs liquid water is clearly distinguished and the Bergeron-Findeisen
process is explicitly represented.
-
Unlike the AMIP
I model, a sub-grid scale convective cloud fraction is not computed,
since convective cloud is treated as radiatively transparent.
Precipitation
The AMIP II model treats grid-scale precipitation differently from the
AMIP
I model. Grid-scale precipitation is diagnostically determined
from conservation equations (neglecting tendency and advection terms) for
rain and snow that apply to the prognostic large-scale cloud formation
scheme. Subsequent evaporation of falling precipitation is not simulated.
Snow Cover
The AMIP II model treats snow prognostically, as opposed to its climatological
prescription in the AMIP
I model. Continental snow cover is treated as in the BATS1e land surface
scheme (cf. Dickinson at al. (1993)[40]).
See also Land Surface Processes.
Surface Characteristics
In the AMIP II model, continental surface characteristics (albedos and
roughnesses) are specified as in the BATS1e land surface scheme for 18
distinguished vegetation types (cf. Dickinson et al. (1993)[40])),
as opposed to the AMIP
I model's 10 surface types. (The surface types of both models
are derived from the Matthews (1983)[30]
1 x 1-degree, 32-type vegetation data set, however .) Dry and saturated
albedos in two wavelength regions have been associated with the eight soil
color classes. Otherwise, both models use the same formulations of
surface roughnesses and direct/diffuse-beam albedos on ocean and ice surfaces.
Land Surface Processes
The AMIP II model treats land surface processes as in the Biosphere-Atmosphere
Transfer Scheme Version 1e (BATS1e) of Dickinson et al. (1993)[40]),
a more complex approach than in the AMIP
I model:
-
Soil moisture in the absence of vegetation either infiltrates the ground
or is lost to surface runoff. The soil is represented by 3 layers
when treating soil moisture. Three parameters represent soil moisture:
water in the upper layer of soil, water in the soil rooting zone, and total
soil water. Soil properties are associated with a texture class ranging
from 1 = very coarse (equivalent to sand) to 12 = very fine (equivalent
to heavy clay) and an eight-color class. Soil surface runoff is based
on a parameterization that computes low surface runoff for soil moisture
at field capacity and complete surface runoff for saturated soil.
If the subsurface temperature is below freezing, the fraction of surface
runoff is increased.
Go to:
References
Return to:
Top of Page
AMIP II Model/Experiment Documentation
AMIP II Documentation Directory
AMIP Home Page
Last update December 6, 2000. For questions or comments, contact
Tom Phillips (phillips@pcmdi.llnl.gov)
or the AMIP Representative(s).
LLNL Disclaimers
UCRL-MI-135872