The overall goal of this project is to determine the predictability of large-scale atmospheric patterns on timescales of one to two weeks and rapidly developing midlatitude storms on timescales of one to two days. Sources of error in predicting these phenomena using numerical models can come from the model itself or from errors in the initial data used at the start of the prediction. This research will attempt to distinguish between the two, examining cases of rapidly developing storms (explosive cyclogenesis) and persistent periods of high amplitude atmospheric wave patterns known as atmospheric blocks. The work represents a continuation of research begun under a previous grant. The principal investigator will employ a more sophisticated and more complete version of a general circulation model than he used previously. This research ultimately should contribute to a better understanding of atmospheric dynamics and to improving short and medium range weather prediction.
