The Statistics and Structure of Subseasonal Midlatitude Variability in NASA GSFC GCMs

Abstract

Abstract A comprehensive analysis of midlatitude intraseasonal variability in extended integrations of NASA GSFC general circulation models (GCMs) is conducted. This is approached by performing detailed intercomparisons of the representation of the storm tracks and anomalous weather regimes occurring during wintertime in the Atmospheric Model Intercomparison Project (AMIP)-type simulations of both the NASA–NCAR and a version of the Aries model used in NASA’s Seasonal-to-Interannual Prediction Project (NSIPP) model. The model-simulated statistics, three-dimensional structure, and dynamical characteristics of these phenomena are diagnosed and directly compared to parallel observational analyses derived from NCEP–NCAR reanalyses. A qualitatively good representation of the vertical structure of intraseasonal eddy kinetic energy (EKE) is provided by both models with maximum values of EKE occurring near 300 hPa. The main model shortcoming is an underestimation of EKE in the upper troposphere, especially for synoptic eddies in the NSIPP model. Nonetheless, both models provide a reasonable representation of the three-dimensional structure and dynamical characteristics of synoptic eddies. Discrepancies in the storm-track structures simulated by the models include an anomalous local minimum over the eastern Pacific basin. However, both GCMs faithfully reproduce the observed Pacific midwinter storm-track suppression. Interestingly, the NSIPP model also produces a midwinter suppression feature over the Atlantic storm track in association with the anomalously strong upper-level jet stream simulated by NSIPP in this region. The regional distribution of anomalous weather regime events is well simulated by the models. However, substantial structural differences exist between observed and simulated events over the North Pacific region. In comparison to observations, model events are horizontally more isotropic, have stronger westward vertical tilts, and are more strongly driven by baroclinic dynamics. The structure and dynamics of anomalous weather regimes occurring over the North Atlantic region are qualitatively better represented by the models. The authors suggest that model deficiencies in representing the zonally asymmetric climatological-mean flow field (particularly the magnitude and structure of the Pacific and Atlantic jet streams) help contribute to model shortcomings in (i) the strength and seasonal variability of the storm tracks and (ii) dynamical distinctions in the maintenance of large-scale weather regimes.