Climate Model Performance in Simulating Winter Extratropical Cyclones over North America

Extratropical cyclones bring much of the significant weather, including precipitation and winds, to North America, especially during the cool seasons. These cyclones arise from various factors and exhibit different development patterns as well as weather impacts. To evaluate the climate model performance in simulating these cyclones, we use sea level pressure data to track the winter extratropical cyclones entering the contiguous United States at resolutions of T42 and T106 in the ERA5 reanalysis and GFDL SPEAR model simulations. These cyclones are then classified into different types by performing a cluster analysis of the interpolated cyclone trajectories weighted by mean cyclone intensities. Additionally, the cyclones are categorized by the defined US climate regions that the cyclones pass through. Both procedures yield reasonable cyclone types, showing distinct physical features and geographical paths. We assess the cyclone feature statistics for each type, including lifetime and maximum amplitude. Composite structures and precipitation patterns of each cyclone type differ from one another and exhibit both similarities and discrepancies between the reanalysis and model outputs. By examining these discrepancies, the processes responsible for the cyclone statistic biases in the model can be identified. These findings enhance our understanding of SPEAR’s reliability in simulating each type of extratropical cyclones and provide confidence in the model predictions of extratropical precipitation.