Mesoscale predictability under various synoptic regimes

Abstract. Numerical model experiments using slightly rotated terrain are compared to gauge the sentivity of mesoscale forecasts to small perturbations that arise due to small synoptic-scale wind direction errors relative to topographic features. The surface and above surface wind speed errors, as well as the precipitation forecast errors, are examined for a landfalling cold front that occurred during the California Landfalling Jets (CALJET) experiment. The slight rotation in the terrain results in nearly identical synoptic-scale forecasts, but result in substantial forecast errors on the mesoscale in both wind and precipitation. The largest mesoscale errors occur when the front interacts with the topography, which feeds back on the frontal dynamics to produce differing frontal structures, which, in turn, result in mesoscale errors as large as 40% (60%) of the observed mesoscale variability in rainfall (winds). This sensitivity differs for the two rotations and a simple average can still have a substantial error. The magnitude of these errors is very large given the size of the perturbation, which raises concerns about the predictability of the detailed mesoscale structure for landfalling fronts.