The Modulation of Atmospheric Rivers by Circulation Regimes in the ECMWF Coupled Model: Fidelity and Resolution Dependence



Korendyke, Mary

Journal Title

Journal ISSN

Volume Title



The large scale circulation of the atmosphere, as characterized by circulation regimes, has a significant influence on storminess and precipitation extremes. One type of structure known to be related to precipitation is atmospheric rivers, regions of intense poleward moisture flux. This raises the question of what spatial resolution is required to accurately simulate the circulation dependence of atmospheric rivers. This paper investigates the impact of enhanced resolution on the simulation of circulation regimes over the Pacific North American region during boreal winter, and the regime dependence of atmospheric river patterns. In particular we examine the realism of ensemble forecasts made with the ECMWF model at three different resolutions (100km, 31km, 16km). The regimes are determined from the k-means clustering method; our results focus on two choices for the number of clusters: 4 and 5. We create regime composites of z500 and u250 anomalies, then calculate spatial correlation, differences in regime persistence distribution, and annual maximum episode persistence to compare reanalysis with model simulations. Composites for regime-associated atmospheric rivers were also computed, and spatial correlations calculated. We find that regimes are better simulated in going from low to medium resolution, but that does not apply in going from the medium to high resolution. `We find that generally regime persistence is well represented, with some exceptions. We also find that atmospheric rivers typically show good spatial correlation with their assigned regimes, and that the correlation is somewhat improved in going from the low to medium resolution despite a consistent undercounting of rivers in the model. Future work is needed to determine if the model loses moisture as the forecast length increases.



Circulation regimes, Atmospheric rivers, Resolution dependence, Metis, ECMFW