How do upper level Rossby wave packets affect medium-range forecast errors?

Upper level Rossby wave packets (RWPs) have frequently been linked to high impact weather events, including being precursors to explosively deepening surface cyclones and heavy precipitation events. Previous studies have shown that short-range (1-3 days)forecast errors disperse and grow like linear RWPs. Both medium- (3-7 days)and short- range forecast uncertainties also can develop and propagate like linear RWPs. However, to date it is still not clear how RWPs affect the development of forecast errors, although there have been suggestions that in some cases, large forecast uncertainties developed simultaneously with the initiations of RWPs.

Medium-range GFS forecast errors of 300 hPa geopotential over the eastern North America and western Atlantic oceans (verification region) have been examined. Occasional forecast skill busts or “dropout” cases are first defined based on the magnitude of the root-mean-squared error of 300-hPa geopotential height forecast as well as the anomaly correlation coefficient between forecasts and analyses over the verification region. Extended Empirical Orthogonal Function (EEOF) analysis has been applied to diagnose RWPs amplitudes (computed based on a Hilbert Transform technique) for dropout cases. Results suggest that in many cases, enhanced coherent wave packets propagating along different paths upstream are associated with large medium-range forecast errors over the verification region. The distribution and propagation of absolute forecast errors have also been investigated using compositing analysis. The results show that forecast errors are often larger and grow rapidly in the presence of stronger RWPs. These results suggest that RWPs have significant impacts on the development and propagation of forecast errors in medium range forecasts.