This work aims to provide a comprehensive assessment of the U.S. DOE’s Energy Exascale Earth System Model (E3SM) atmospheric simulations over the Southern Ocean and Antarctica. Large changes in atmospheric conditions over Antarctica and the surrounding ocean could have tremendous impacts on the cryosphere system and the cascading effect on the Earth’s climate system could have important implications for the societies. Surface conditions exert direct influence on the cryosphere system as atmospheric forcing. Upper air conditions control the characteristics of surface fields through cloud structure and large-scale dynamics and thermodynamics all the way to lower stratosphere. They also carry signals of remote influences from lower latitudes that can be used to diagnose variabilities at interannual and longer time scales. A model’s performance in simulating the surface and upper air conditions in this region must be evaluated and properly understood in order to be used with confidence for climate change projections that account for the influence of cryosphere system in the Southern high latitudes. In this study, we will evaluate the climatology and seasonality of surface fields important to the cryosphere system from the E3SMv1 DECK (Diagnostic, Evaluation and Characterization of Klima) simulations, including surface winds, precipitation and heat fluxes. Upper air fields will be analyzed in a way to help explain the characteristics of the simulated surface quantities. Skills in simulating interannual variability due to tropical SST oscillation will also be evaluated. Reanalysis products, satellite-based measurements along with available station data will be used for the evaluation. The effect of coupling and resolution increase on the simulated atmospheric fields will be analyzed by including both the uncoupled and coupled DECK simulations, along with additional simulations at high grid-resolution. Preliminary results show that model biases in various fields in uncoupled simulations tend to be amplified after coupling, while high resolution simulations tends to significantly reduce the biases. Selected diagnostics developed in this work will be used as fundamental metrics for evaluating the evolving E3SM model over the Southern polar region.