40
A.D. Del Genio
-Midlatitude ocean storm tracks.
Off the east coasts of North America and Asia, and throughout the "roaring 40's" latitude band
of the Southern Hemisphere, baroclinic wave activity is more persistent than over the midlatitude continents. The cloud regime in these locations thus also consists of midlevel optically
thick nimbostratus, but with a higher proportion of moderate optical thickness low stratus
than is found over the continents. Because of their almost ubiquitous cloudiness and their
lower cloud tops relative to the tropical convection zones, these are the most negative net cloud
forcing regions on the planet. It is this cloud regime that causes most of the error in implied
ocean heat transport in GCMs. Polar amplification of climate warming, combined with the dependence of baroclinic instability on meridional temperature gradient, suggests that cloudiness
here should decrease in a greenhouse climate change, but changes in the optical properties of
these clouds associated with a more vigorous hydrologic cycle are harder to anticipate. Despite
their possible climatic importance, little attention has been paid to this cloud regime by either
the climate modeling or observation community.
-Polar regions.
Little is known about the prevailing cloud types in the polar regions. Current wisdom suggests
that the cloud distribution is bimodal, with boundary layer cloudiness controlled by surface
heat and moisture fluxes and midlevel cloudiness due to poleward moisture transport by midlatitude storms (Curry and Ebert, 1992). Multilayer cloudiness may be common, as well as
"cloudless" ice crystal precipitation that complicates attempts to compare data sets. Because
of the difficulty in distinguishing snow lice-covered surfaces from clouds, traditional satellite
detection techniques have large uncertainties. The ISCCP seasonal cycle of polar cloudiness
disagrees completely with that inferred by surface observers, and the ERBE algorithm gets the
incorrect sign of cloud forcing at these latitudes. Consequently, most GCM modelers pay no attention to their polar cloudiness, absent an accepted validation standard. But snow lice-albedo
feedback and polar amplification of climate warming, central to the overall question of climate
sensitivity, may well be determined by climate changes in polar clouds.
b) Regional dynamic variability effects.
Since clouds are to a great extent tracers of the general circulation, it is logical to approach
the question of regional climate feedbacks from the standpoint of dynamics. Systematic trends
in specific components of the seasonal or annual mean dynamics, such as the baroclinic wave
issue discussed above, are one example. But the current climate also varies considerably on
interannual time scales, as a result of ocean-atmosphere interactions. Might the regional pattern
of climate changes in a warmer world bear any resemblance to the preferred modes of interannual
variability of the current climate?
The most direct example of observed interannual variability is the eastward shift in tropical
Pacific SST that occurs during ENSO. Under non-ENSO conditions, rising motion, deep convection, and heavy precipitation characterize the west Pacific warm pool region and the maritime
continent, while subsidence, low stratus, and drier conditions prevail in the east Pacific and
along the west coast of South America. ENSO warming of the central and east Pacific shifts
the Walker circulation eastward and causes regional climate anomalies on either side of the
Pacific, e.g., drought in Australia and Indonesia and flood conditions over South America.
It is not clear how the tropical ocean circulation will change in response to greenhouse gas
forcing. But there are purely energetic reasons for anticipating that longitudinal Pacific SST
gradients might weaken as climate warms, causing the regional pattern of climate change in
the tropics to resemble that for the warm phase of ENSO. First, because of the prevailing trade
winds near the equator and attendant ocean upwelling off the west coast of South America, the
ocean mixed layer and thermocline are shallower in the east Pacific than in the west. Thus,
the effective ocean heat capacity is lower in the east, and the eastern ocean surface should
A.D. Del Genio
-Midlatitude ocean storm tracks.
Off the east coasts of North America and Asia, and throughout the "roaring 40's" latitude band
of the Southern Hemisphere, baroclinic wave activity is more persistent than over the midlatitude continents. The cloud regime in these locations thus also consists of midlevel optically
thick nimbostratus, but with a higher proportion of moderate optical thickness low stratus
than is found over the continents. Because of their almost ubiquitous cloudiness and their
lower cloud tops relative to the tropical convection zones, these are the most negative net cloud
forcing regions on the planet. It is this cloud regime that causes most of the error in implied
ocean heat transport in GCMs. Polar amplification of climate warming, combined with the dependence of baroclinic instability on meridional temperature gradient, suggests that cloudiness
here should decrease in a greenhouse climate change, but changes in the optical properties of
these clouds associated with a more vigorous hydrologic cycle are harder to anticipate. Despite
their possible climatic importance, little attention has been paid to this cloud regime by either
the climate modeling or observation community.
-Polar regions.
Little is known about the prevailing cloud types in the polar regions. Current wisdom suggests
that the cloud distribution is bimodal, with boundary layer cloudiness controlled by surface
heat and moisture fluxes and midlevel cloudiness due to poleward moisture transport by midlatitude storms (Curry and Ebert, 1992). Multilayer cloudiness may be common, as well as
"cloudless" ice crystal precipitation that complicates attempts to compare data sets. Because
of the difficulty in distinguishing snow lice-covered surfaces from clouds, traditional satellite
detection techniques have large uncertainties. The ISCCP seasonal cycle of polar cloudiness
disagrees completely with that inferred by surface observers, and the ERBE algorithm gets the
incorrect sign of cloud forcing at these latitudes. Consequently, most GCM modelers pay no attention to their polar cloudiness, absent an accepted validation standard. But snow lice-albedo
feedback and polar amplification of climate warming, central to the overall question of climate
sensitivity, may well be determined by climate changes in polar clouds.
b) Regional dynamic variability effects.
Since clouds are to a great extent tracers of the general circulation, it is logical to approach
the question of regional climate feedbacks from the standpoint of dynamics. Systematic trends
in specific components of the seasonal or annual mean dynamics, such as the baroclinic wave
issue discussed above, are one example. But the current climate also varies considerably on
interannual time scales, as a result of ocean-atmosphere interactions. Might the regional pattern
of climate changes in a warmer world bear any resemblance to the preferred modes of interannual
variability of the current climate?
The most direct example of observed interannual variability is the eastward shift in tropical
Pacific SST that occurs during ENSO. Under non-ENSO conditions, rising motion, deep convection, and heavy precipitation characterize the west Pacific warm pool region and the maritime
continent, while subsidence, low stratus, and drier conditions prevail in the east Pacific and
along the west coast of South America. ENSO warming of the central and east Pacific shifts
the Walker circulation eastward and causes regional climate anomalies on either side of the
Pacific, e.g., drought in Australia and Indonesia and flood conditions over South America.
It is not clear how the tropical ocean circulation will change in response to greenhouse gas
forcing. But there are purely energetic reasons for anticipating that longitudinal Pacific SST
gradients might weaken as climate warms, causing the regional pattern of climate change in
the tropics to resemble that for the warm phase of ENSO. First, because of the prevailing trade
winds near the equator and attendant ocean upwelling off the west coast of South America, the
ocean mixed layer and thermocline are shallower in the east Pacific than in the west. Thus,
the effective ocean heat capacity is lower in the east, and the eastern ocean surface should