The Impact of Large-scale Circulation Regimes on Extreme Weather over Pacific and North America

摘要

Global weather and climate models have difficulty in accurate simulation of storm- related quantities such as precipitation, atmospheric rivers, and storm track covariances. On the other hand, models generally show reasonable skill in at least simulating large-scale atmospheric circulation patterns. The circulation regimes, which are the preferred large- scale circulation states, strongly influence the extreme weather over the United States. The goals of this research are: i) to determine how well these preferred patterns of the large-scale circulation impact the storm-related quantities, in particular extreme precipitation and ii) to determine if the observed relations are well simulated by a modern forecast model, namely that of the European Centre for Medium-Range Forecasts (ECMWF).;Circulation regimes are identified from a k-means cluster analysis applied to running 5-day means of the combined anomalies of 500 hPa geopotential height and the 250 hPa zonal wind over the extended Pacific-North America region. Five regimes are identified: the Pacific Wavetrain (WT), Arctic High (AH), Arctic Low (AL), Alaskan Ridge (AR) and Pacific Trough (PT). I show that the circulation regimes alter the geographical pattern of the frequency of occurrence of atmospheric rivers and storm tracks over the region. In addition, the regime-specific geographical patterns of precipitation anomaly are examined. For each regime I identify the regions for which extreme precipitation or intra-seasonal drought are more likely to occur.;The ECMWF atmosphere-ocean coupled forecasts realistically simulate the circulation regimes and associated storm-related quantities, with high consistency among ensemble members. These results go beyond previous work in examining the many indicators of storminess in relation to circulation regimes, and provide the foundation for using these regimes as a potential predictive tool for forecasting extreme weather.