Evaluating Arctic Lower Tropospheric Stability in Transpose AMIP Models

Previous research has shown that the CMIP3 models have an Arctic lower troposphere that is more stable than reanalysis products, and that this bias may lead to an unrealistic Arctic feedback. Basic questions of why these models drift toward a more stable lower troposphere in the Arctic have not been thoroughly investigated. What season is the bias more likely? What atmospheric level or levels is causing the bias? In this analysis, we examine the lower tropospheric stability in the Transpose AMIP models. Transpose AMIP models are a collection of models, which were run in forecast mode for a selected time periods. Biases in fast physics processes are easier to diagnosis in forecast runs because the dynamics are prescribed. Lower tropospheric stability is examined across the Arctic domain with analysis products, and in more detail using in-situ data at the North Slope of Alaska. We show that models Arctic lower tropospheric stability bias is greater during the winter months, and that the biases are largely due to models' surface temperature. Last, we show that the models' winter surface temperature relates to the resilience of the models' Arctic clouds.