Geoscience Reference
In-Depth Information
prising that daily model turbulent heat fluxes differ by an or-
der of magnitude between summer and winter. Ignoring wind
speed and drag coefficients, sensible heat flux is proportional
to the temperature difference between the ocean and near-
surface air, which is on the order of 1°K in summer and 10°K
in winter. The winds are generally stronger in winter, which
increases the fluxes, but the polar atmosphere is also rela-
tively stable which damps turbulent fluxes. The sensible heat
flux differences between winter and summer can be explained
by the seasonal difference in vertical temperature gradients.
Parallel arguments can be made for latent heat fluxes. Plots
of model net surface solar flux, albedo, surface temperature,
cloud cover, and specific humidity are described in the text
but will not be shown. Anomalies of net surface solar heat
flux are directed into the ocean and are on the order of 15-30
W m
-2
where high-albedo ice is replaced by a lower-albedo
ocean. However, the shortwave anomalies do not impact our
simulation since the ocean temperature and ice are fixed and
are not included in the net heat flux calculation. In nature, the
enhanced solar flux into the surface would melt more ice or
act to warm the ocean in the shallow ice-free seas. However,
we specify the observed evolution of sea ice and argue that
any ice melt from increased net solar radiation into the ocean
is represented by the observed sea ice conditions.
The surface temperature anomalies associated with the
reduced ice area are between 0.5° and 1.5°C where clima-
tological ocean sea surface temperatures replace sea ice.
The surface air temperature (Plate 2b) warms throughout the
Arctic with strongest warming present over the Kara-Bar-
ents and East Siberian seas and over eastern Siberia with
anomalies between 0.5° and 1.5°C. This low-level warm-
ing is associated with small decreases in sea level pressure
and geopotential heights that are not statistically significant
(Plate 3). The air temperature warming is relatively shallow
over the Arctic with no significant anomalies at or above
925 hPa. There is a significant increase in convective pre-
cipitation (1 mm d
-1
), convective clouds (1-2%), and mid-
dle-level clouds (2-4%) in the Laptev Sea where sea ice is
reduced. In the Kara Sea over the reduced ice extent, there
is a significant decrease in total cloud cover (2-4%), which
results from less low, medium, and high clouds. There are no
significant changes in large-scale precipitation over the Arc-
tic. The CCM3 positive SLP bias in the control simulation
discussed earlier could play a role in the small convective
response over the Arctic.
ern Siberia (65°N, 165°E) and over the ocean storm track
region (55°-60°N). In far eastern Siberia, negative sensible
and longwave heat flux anomalies total 5 W m
-2
(Plate 2a)
while downward solar heat flux is reduced by 5 W m
-2
, result-
ing in a net heat flux change of near zero. Surface tempera-
ture and surface air temperatures are warmer by up to 1.0°C
and 0.5°C (Plate 2b), respectively. Increased convective and
large-scale precipitation is collocated with increases in spe-
cific humidity (up to 2 g kg
-1
). Total cloudiness in eastern
Siberia increases by up to 4% with more clouds at all levels.
The SLP and geopotential height anomalies are weakly neg-
ative and not significant over eastern Siberia. Since the net
surface heat flux anomalies are weak, the warmer moister
atmosphere results from southerly advection associated with
the circulation changes over the North Pacific.
The SLP response is characterized by a significant anoma-
lous high over the North Pacific with a central maximum of
2 hPa (Plate 3a). At 500 and 200 hPa the anomalous high in
the North Pacific reaches 20 and 30 m (Plates 3b and 3c), re-
spectively, displaying an equivalent barotropic structure. This
pattern is characteristic of the equilibrium response to a mid-
latitude heating source attributed to transient eddy feedbacks
that results from the interaction of the forced anomalous flow
and the storm tracks [
Kushnir and Lau
, 1992;
Ting and Peng
,
1995;
Peng
and Whittaker
, 1999]. The response to reduced
sea ice does not project on the dominant modes of model vari-
ability, and this is discussed further in section 4. Additionally,
these SLP and 500-hPa patterns compare favorably with ob-
served circulation anomalies associated with reduced Eurasian
sea ice, and this is briefly addressed at the end of section 4.
The model displays a response in the North Pacific storm
track region with significant total precipitation anomalies
(Plate 4). Anomalies of large-scale precipitation are about
twice as large as those of convective precipitation (not
shown). The magnitude of the total precipitation response
reaches values of 25% of the mean climatological precipita-
tion (Plate 4, contours). The mean and anomalous precipita-
tion patterns suggest a weakening of the main North Pacific
storm track and a slight enhancement on the poleward side.
In other words, the storm track shifted northward and weak-
ened, which is consistent with a weakened meridional tem-
perature gradient [see
Hartmann
, 1994, section 9.5]. The
mechanisms associated with the strengthening of the sub-
tropical high and the storm track changes are not well under-
stood, and a further discussion is included in section 4. Note
that the precipitation maximum on the south coast of Alaska
is associated with orography, and the reduced onshore winds
associated with the SLP response is likely responsible for
the coastal precipitation anomalies.
Storm track variability as indicated by 2- to 8-day band-
pass-filtered variance statistics for 500-hPa height variance,
3.2. Midlatitude Response
The model response to reduced western Arctic sea ice in
the North Pacific is characterized by changes in the large-
scale circulation and displays significant anomalies in east-
Search WWH ::
Custom Search