Thursday, 13 January 2005
Diagnosis of medium-range predictability enhancement associated with anomalous zonal flow over western North America during winter
This investigation analyzes medium-range predictability enhancement during winter cases of anomalous, upper troposphere zonal flows over western North America. Time correlations based upon a 50-year record of reanalyses suggest that winters with anomalously strong zonal winds are wetter over the region, while years with anomalously weak zonal winds are relatively drier. Forecasts are selected based upon anomalously weak and strong zonal flows during January. Results from 15-day simulations using a variety of operational and research global model configurations are presented to diagnose the predictability of precipitation and large scale features. Model forecasts of precipitation accumulation delineate qualitatively between wet and dry events at both 5 and 10 days. Anomaly correlations of the geopotential height field reveal useful predictability for some ensembles extending to 9.5 days. Uniform resolution forecasts are compared with two model configurations that employ rotated, variable resolution. Uniform and variable resolution forecasts maintain representative precipitation into the second week over the western U.S. The rotated variable resolution simulations provide more precipitation detail. Diagnostics and model simulations of a small number of extreme events suggest that flow modifications associated with ambient flows exist over the orography during the winter season, and that a predictable regional response may be present to about 10 days. The persistence of the anomalies may also contribute to the improved model performance in certain cases. Improved performance may be related to the large inertia of the flow in wet events, and to the persistence and increased predictability of initial, large scale anomalies in both wet and dry events. Present conclusions are limited by the small case sampling, which will be expanded in future investigations.
Supplementary URL: