Abstract
The fidelity of a prerelease Energy Exascale Earth System Model (E3SM) in reproducing atmospheric blocking is evaluated with ∼100‐ and ∼25‐km horizontal resolution within ensembles of simulations with Atmospheric Modeling Intercomparison Protocol‐type active atmosphere and land surface configurations. This evaluation is conducted via a hybrid bidimensional blocking index based on geopotential height. The lower‐resolution model correctly reproduces the spatial distribution of blocking frequency maxima over the Northern Hemisphere. However, it overestimates the blocking frequency over western North America, the Pacific Northeast, and Eastern Europe regions and largely underestimates blocking frequency over the North Atlantic. The high‐resolution model significantly reduces the bias over the North Pacific region, particularly over western North America, but the biases over the North Atlantic blocking sector mostly persist. A diagnosis of the mean flow reveals that the subtropical jet is displaced poleward in the high‐resolution model with a jet core speed that is generally more realistic. A time scale analysis of eddies suggests that synoptic eddies associated with the blocking events are increased in the high‐resolution configuration. The discrepancy between the models implies that both synoptic and low‐frequency eddies play significant roles in determining North Pacific blocking, whereas North Atlantic blocking is mainly driven by low‐frequency eddies. The characteristic time scale associated with blocking events is also investigated, and it shows that both model resolutions produce events that are overly persistent. However, the bias in the high‐resolution model is reduced, which reflects more efficient energy dispersion with higher resolution.
