A comparison of mesoscale model forecast accuracy using random and a simplified targetting approach

Abstract

The error in mesoscale model forecasts on the West Coast of the United States often depends strongly on the quality of the synoptic scale forecast. Kuypers (2000) demonstrated that small differences in synoptic scale initial analyses due to different random samples of the large scale structure are sufficient to cause large errors in the mesoscale forecast. This dependence of the mesoscale on the synoptic scale is often mirrored in statements like, A good mesoscale forecast requires a good synoptic scale forecast. The method by which a good synoptic scale forecast is achieved is the subject of numerous efforts at improving the observations over the Pacific through targeting of observations.