Model UKMO (HadCM2): Elaborations
Participation
Model UKMO (HadCM2) is an entry in both the CMIP1 and CMIP2
intercomparisons,
and is providing an optional extended set of output data in compliance
with the
CMIP2+ initiative.
Spinup/Initialization
Starting from Levitus (1982) climatological
September values of oceanic potential temperature and salinity, the
510-year
spinup was implemented mostly in a fully coupled mode, alternating
between
relaxation boundary conditions on SST and SSS . Reference SST and SSS
climatologies
for the relaxation phases were based on Levitus
(1982) monthly-mean surface conditions (but with some adjustments
made
in SSS near Antarctica during Austral winter). Flux adjustments for
heat
and fresh water, which were calibrated from 5-day mean differences
between
the modeled and reference SST and SSS averaged over a 25-year run in
the
preceding relaxation phase, also were applied as prescribed,
noninteractive
terms. Cf. Johns et al. 1997
for further
details.
Land Surface Processes
- A vegetation canopy model (cf. Warrilow
et al. 1986 and Shuttleworth 1988)
includes effects of moisture condensation, precipitation interception,
direct evaporation from wet leaves and from surface ponding, and
evapotranspiration
via root uptake of soil moisture. The fractional coverage and
water-storage
capacity of the canopy vary spatially by vegetation type, with a small
storage added to represent surface ponding. The canopy intercepts a
portion
of the precipitation, which is exponentially distributed over each grid
box. Throughfall of canopy condensate and intercepted precipitation
occurs
in proportion to the degree of fullness of the canopy; evaporation from
the wet canopy occurs in the same proportion, as a fraction of the
local
potential evaporation. Cf. Gregory
and
Smith (1990) for further details.
- Soil moisture is predicted from a single-layer model with
spatially
nonuniform
water-holding capacity; it is augmented by snowmelt, precipitation, and
the throughfall of canopy condensate. This moisture infiltrates the
soil
at a rate depending on saturated soil hydraulic conductivity enhanced
by
effects of root systems that vary spatially by vegetation type. The
noninfiltrated
moisture is treated as surface runoff. Subsurface runoff from
gravitational
drainage is also parameterized as a function of spatially varying
saturated
soil hydraulic conductivity and of the ratio of soil moisture to its
saturated
value (cf. Eagleson 1978). Surface and
subsurface
runoff are converted into river outflow using predefined geographical
catchments
over land and associated coastal outflow points defined relative to the
model grid. Because river transport is not modeled explicitly, the
runoff
is distributed instantaneously to these outflow points.
- Soil moisture is depleted by evaporation at a fraction beta
of
the
local potential rate. The value of beta depends on the ratio of
soil moisture to a spatially varying critical value, and on the ratio
of
the stomatal resistance to aerodynamic resistance (cf. Monteith
1965). The critical soil moisture and stomatal resistance vary
spatially
by
soil/vegetation type. Cf. Gregory
and
Smith (1990) and Smith (1990b) for
further
details.
- Soil temperature is predicted after Warrilow
et al. (1986) from heat conduction in four layers. The depth of the
topmost soil layer is given by the penetration of the diurnal wave,
which
depends on the spatially varying soil heat conductivity/capacity. The
lower
soil layers are, respectively, about 3.9, 14.1, and 44.7 times the
depth
of this top layer. The top boundary condition for heat conduction is
the
net downward surface energy balance, including the latent heat of
fusion
for snowmelt; the bottom boundary condition is zero heat flux. Heat
insulation
by snow is modeled by reducing the thermal conductivity between the top
two soil layers; however, subsurface moisture does not affect the
thermodynamic
properties of the soil. Cf. Smith (1990b)
for
further details.
Sea Ice
- Sea ice is assumed to have a constant salinity of 0.6
percent.
Ocean
salinity is affected by sea ice formation and melt, sublimation,
precipitation
and snow melt, where the latter freshwater fluxes are assumed to pass
directly
through ice leads to the ocean below. The weight of snow accumulating
on
sea ice may force the ice-snow interface below the water line, thereby
increasing the ice depth by the formation of "white ice" (cf. Ledley
1985). Snow cover does not affect the surface albedo of the ice,
however.
- Sea ice thermodynamics are based on the zero-layer model of Semtner
(1976), with the inclusion of an ice concentration parameter A
after Hibler (1979). (A has
a maximum
of 0.995 in the Arctic and 0.980 in the Antarctic, since completely
unbroken
ice cover is rarely observed.) Internal brine pockets are not
represented.
Atmospheric surface fluxes and temperatures are calculated separately
over
the ice and the leads, with a linear temperature profile assumed in the
ice interior. Shortwave radiation does not penetrate the ice
interior,
but does warm the ocean where there are leads, just as it would at open
ocean grid squares. All other surface heat fluxes into the leads are
split,
so that a fraction equal to the ice concentration is used to form/melt
ice, with the remainder applied to the ocean model. The heat flux
from the ocean below the ice is proportional to the difference between
the temperature of the topmost ocean layer and that of the base of the
ice (assumed to be -1.8 deg C).
- Sea ice dynamics follow the simple free-drift parameterization of
Bryan
(1969b), for which the ice internal pressure gradient is
neglected.
The current in the topmost ocean layer advects (by an upstream scheme)
the ice thickness, concentration and snow depth. A rough
representation
of rheology prevents convergence of ice once its depth exceeds 4
m.
Mechanical energy due to wind mixing, applied under the ice, is
calculated
using the drag coefficient appropriate for leads; it is weighted by the
leads fraction of each grid square. See Johns
et al. (1996) for further details.
Chief Differences from Closest
AMIP
Model
The atmospheric component of coupled model HadCM2 differs from features
of the AMIP model UKMO
HadAM1 (2.5x3.75 L19) 1993 in the following respects:
Chemistry
Errors in horizontal interpolation of the ozone reference climatology
that
were present in the AMIP
model now are corrected. (The effect of these corrections is
thought
to be minor.)
Convection
The calculation of forced detrainment in the AMIP
model is modified to ensure internal consistency with the cloud top
and with the choice of reference level for the initial mass flux when a
parcel becomes unsaturated due to the addition of latent heat.
Cloud Formation
The parameterization of stratiform cloud formation is modified as
follows.
In the AMIP
model, the cloud condensate depends on the spatial fluctuations of
temperature within the grid box, which are assumed to take the form of
a "top-hat" probability distribution. In coupled model HadCM2, the
cloud
condensate depends instead on the spatial fluctuations of liquid
water
temperature, and these are assumed to follow a triangular
probability
distribution.
Surface
Characteristics
The range and temperature-dependence of the albedo of sea ice differ
from
those of the AMIP
model. In coupled model HadCM2, the sea ice albedo varies from 0.80
for ice surface temperatures at or below -10 deg C to 0.50 at 0 deg C
(cf. Johns
et al. 1997).
References
Bryan, K., (1969b) Climate and ocean
circulation.
III The ocean model. Mon. Wea. Rev., 97: 806-827.
Eagleson, P.S., 1978: Climate, soil and
vegetation. Water Resources Res., 14, 705-776.
Gregory, D., and R.N.B. Smith,
1990: Canopy, surface, and soil hydrology. Unified Model Documentation
Paper No. 25, United Kingdom Meteorological Office, Bracknell,
Berkshire
RG12 2SZ, UK.
Hibler, 1979: A dynamic-thermodynamic sea
ice model. J. Phys. Oceanogr., 9: 817-846.
Johns. T.C., 1996: A description of
the Second Hadley Centre Coupled Model (HadCM2). Climate Research
Technical
Note 71, Hadley Centre, United Kingdom Meteorological
Office, Bracknell Berkshire RG12 2SY, United Kingdom, 19 pp.
Johns, T.C., R.E. Carnell, J.F.
Crossley,
J.M. Gregory, J.F.B. Mitchell, C.A. Senior, S.F.B. Tett, and R.A. Wood,
1997: The second Hadley Centre coupled ocean-atmosphere GCM: Model
description,
spinup and validation. Climate Dyn., 13, 103-134.
Ledley, T.S., (1985) Sea ice:
multiyear
cycles and white ice. J. Geophy. Res., 90, 5676-5686.
Levitus, S., 1982: Climatological atlas
of
the world's oceans. NOAA Professional Paper 13, 173 pp.
Monteith, J.L., 1965: Evaporation and
environment.
Symp.
Soc. Exptl. Biol., 19, 205-234.
Semtner, A.J., 1976: A model for the
thermodynamic
growth of sea ice in numerical investigations of climate. J.
Phys.
Oceanogr., 6, 379-389.
Shuttleworth, W.J., 1988:
Macrohydrology:
The new challenge for process hydrology. J. Hydrol., 100,
31-56.
Smith, R.N.B., 1990b: Subsurface, surface,
and boundary layer processes. Unified Model Documentation Paper No. 24,
United Kingdom Meteorological Office, Bracknell, Berkshire RG12 2SZ,
UK.
Warrilow, D.A., A.B. Sangster,
and
A. Slingo, 1986: Modelling of land surface processes and their
influence
on European climate. DCTN 38, Dynamical Climatology Branch, United
Kingdom
Meteorological Office, Bracknell, Berkshire RG12 2SZ, UK.
Top
CMIP Documentation Directory
Last update 15 May, 2002. This page is maintained by Tom
Phillips
(phillips@pcmdi.llnl.gov).
LLNL Disclaimers
UCRL-MI-126431