In this project the PIs will document regional differences of short-term numerical forecast skill and frequency of large forecast errors for the North American continent and the surrounding ocean areas, and will appraise the differences in skill among several international operational numerical models. The relationship between forecast skill and large-scale flow regime will also be examined. Furthermore, for individual events that experienced major forecast failures, the PIs will explore the effects of model errors in terms of model physics on forecast errors, and the effects of initial condition error by using a multi-analysis ensemble approach. This study will address the questions of whether particular operational forecast models experience fewer forecast failures, whether particular regions experience larger forecast errors, whether there are particular regions that suffer frequent and/or large forecast failure events, and if these large forecast failure events are associated with obvious, large initial condition errors. Graduate students will participate in the research.
