Arctic Cloud Properties and Radiative Forcing From Observations and Their Role in Sea Ice Decline Predicted by the NCAR CCSM3 Model During the 21st Century - Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications - page 47

improve our understanding of the role of clouds in the Arctic
sea ice decline. The model consists of components simulating
the Earth's atmosphere, ocean, land surface, and sea ice con-
nected by a flux coupler [ Collins et al. , 2006]. This paper
shows results from the configuration used for climate change
simulations with a T85 grid for the atmosphere and land and
a 1° grid for the ocean and sea ice. Cloud amount is diag-
nosed by the relative humidity, atmospheric stability, and
convective mass fluxes [ Boville et al. , 2006]. Cloud ice and
liquid phase condensates are predicted separately [ Rasch and
Kristijansson , 1998; Zhang et al. , 2003]. This links the ra-
diative properties of the clouds with their formation and dis-
sipation. However, after advection, convective detrainment,
and sedimentation, the cloud phase is recalculated as a func-
tion of cloud temperature. Cloud liquid and ice are assumed
to coexist within a temperature range of -10°C and -40°C
with the ice fraction linearly increasing with decreasing tem-
perature [ Boville et al. , 2006]. Clouds are all liquid above
-10°C, and all ice below -40°C. Compared with observa-
tions, the model produces too much atmospheric moisture in
the polar regions and too little in the tropics and subtropics,
suggesting that the poleward moisture flux is excessive [ Col-
lins et al. , 2006]. As consequence of the excessive moisture
advection and allowing cloud liquid water to exist at low
temperatures, the model Arctic clouds contain large amount
of cloud liquid water [ Gorodetskaya et al. , 2008].
The sea ice in the CCSM3 is represented by a dynamic-
thermodynamic model that includes a subgrid-scale ice
thickness distribution, energy conserving thermodynamics,
and elastic-viscous-plastic dynamics [ Briegleb et al. , 2004].
The surface albedo for the visible and near infrared bands is a
function of ice and snow thickness and surface temperature.
In this paper, the model output during 1950-1999 is ob-
tained from the “Climate of the 20th century experiment”
20C3M scenario simulations conducted from 1870 to
present. The model output for 2000-2100 time period is from
the SRES A1B experiment, where simulations are initial-
ized with conditions from the end of the 20C3M simulations
and run to 2100 under imposed SRES A1B conditions. The
SRES A1B scenario assumes moderate population growth
and rapid economic growth according to the estimates in the
end of 1990s with atmospheric CO 2 concentration increase to
720 ppm by 2100 (doubling the 1990 amount) [ Houghton et
al. , 2001].

ground-based observations during 1 year from 1997 to 1998.
The CCSM3 model output is analyzed on monthly mean and
annual mean timescales on the 1° grid for the time period
from 1950 to 2100. Although the data are available on dif-
ferent time periods and spatial resolutions, they are used for
sensitivity estimates rather than for providing the absolute
magnitudes. Observations give a picture of the sea ice/cloud/
radiation relationships for modern climate. These relation-
ships are then applied to understand the role of cloud and
radiation changes on sea ice cover during the 21st century as
predicted by the CCSM3 model.

3. SEA ICE AND CLOUD EFFECTS ON RADIATION:
OBSERVATIONS

3.1. Effect of Sea Ice on the Top-of-Atmosphere Albedo

Top-of-atmosphere albedo represents a fraction of the in-
coming solar energy reflected from the surface and atmo-
sphere system. Clouds shield the surface from solar radiation
mitigating the effect of melting sea ice on the TOA albedo.
While the surface albedo decreases drastically when sea
ice cover declines (exposing dark ocean surface) the TOA
albedo changes are much smaller if the skies are cloudy.
The magnitude of the ice-albedo feedback is defined by the
changes in the TOA albedo, rather than in the surface albedo,
in response to the sea ice changes. Thus, cloud changes can
modify the magnitude of the ice-albedo feedback. To show
the role of clouds in mitigating the ice-albedo feedback, we
estimate sea ice effect on the SW radiation reflected at the
TOA for average cloud conditions.
In the Arctic, the ice-free ocean areas are almost always
associated with overcast skies, which reduces the differ-
ence between the TOA albedo over the open ocean and sea
ice. Figure 1 shows spatial distribution of the all-sky TOA
albedo over the Arctic and adjacent ocean area during March
and July, together with the climatological sea ice extent for
these months. During July the sea ice surface has the low-
est albedo because of ubiquitous melt ponds, while the sea
ice extent reaches the minimum area only in September. In
March, Arctic sea ice extent is at maximum, and insolation
starts to increase after the polar night (incoming SW flux at
the TOA averaged over the ocean north of 70°N reaches 100
W m -2 in March). The monthly mean TOA albedo over sea
ice in March lies between 60 and 70% (Figure 1a), reduc-
ing to 50-60% during the summer (Figure 1b). During the
summer, the ice-free ocean, including the area of the ocean
occupied by ice during the winter, has TOA albedo rang-
ing between 40 and 50% (Figure 1b). Thus, while the differ-
ence between the TOA albedo over the ice-free ocean and
over the sea ice is reduced by high cloud amount, seasonal

2.3. Methods

Satellite data analysis of sea ice effects on the TOA short-
wave fluxes is performed during the ERBE time period from
November 1984 to February 1990 on 2.5° grid. Cloud prop-
erties influencing radiative forcing are obtained from point