The icehouse effect is a hypothesized polar climate trend toward cooling (or lack of warming) in response to greenhouse warming of adjacent lower latitudes. When greenhouse warmed air from lower latitudes moves over ice and snow, it generates a stronger, more stable, capping inversion than in a parallel case without greenhouse warming. Because the degree of decoupling between vertically adjacent air masses is directly dependent on the strength of the inversion, the capping inversion acts somewhat analogously to the walls and roof of the icehouse of generations past. What is inside the icehouse, namely the cold polar atmospheric boundary layer (ABL) air, is preserved by the "insulation" or decoupling, provided by the warm air aloft.
Observations over the Arctic Ocean have shown an unexpected lack of any detect able surface warming trend over the past 40 years. This finding strongly contra- dicts climate model predictions that polar regions should show the strongest effect of greenhouse warming. It also stands in contrast to the consensus reached by the Intergovernmental Panel on Climate Change (IPCC), that human caused greenhouse warm ing is now detectable globally. One might ask: Are these Arctic observations wrong? Or, if right, is there a plausible physical explanation for them?
The published observations mentioned above used about 50,000 soundings over the Arctic Ocean. Here I present a novel analysis of ALL available Arctic rawinsonde data north of 65N--a total of more than 1.1 million soundings. The analysis confirms the previously published result: There is indeed a slight climate cooling trend in the vast majority of the data. Importantly, there are also select conditions (very strong and very weak stability of the ABL) which show a consistent, strong Arctic warming trend. It is the juxtaposition of these warming and cooling trends which defines a unique "icehouse signature" for which an explanation can be sought.
I constructed a sensitive computer model to explore this signature. It models an air column with very fine vertical resolution (on the order of 10m). It also includes a theoretical innovation--the direct physical interaction of radiative exchange with the ABL turbulence, which recognizes a correlation between eddy tem- perature fluctuations, their radiative cooling response, and their vertical motion. This new model enabled detailed experimentation with the radiative-turbulent ABL processes under the influence of regional-scale ice-ocean temperature contrasts. The model readily reproduces the correct signature of the observed cooling and warming trends, and thus offers a tentative explanation for it, pending confirmation by independent research: Head-to-head comparison between initially identical doubled CO2 and present day air columns show that colder surface temperature evolves in the doubled CO2 atmosphere over snow which lies within a few thousand km of a green- house-warmed open ocean. In simple terms, the greater surface cooling occurs in the doubled CO2 case precisely because the air above the surface is warmer. This warmer air produces a stronger, more stable "lid" over the ABL, leaving a shallower, turbulently-mixed layer in contact with the snow. Because this shallower layer contains less mass, its response to clear sky radiative heat loss in autumn and winter (when there is no sun) is a faster temperature decrease.
The model also demonstrates that Arctic haze and ice crystals in the ABL--a particularly common phenomenon in autumn and winter when the previously published results showed statistically significant climate cooling trends--act to strengthen and prolong the icehouse cooling. This is because hazy air emits more radiation and thus loses heat more effectively.
By strengthening the temperature contrast between polar regions and adjacent lower latitudes, the icehouse effect may contribute to the apparent strengthening of storms along the polar front, which has been reported in published research.