Valuing information from mesoscale forecasts

摘要: The development of meso-gamma scale numerical weather prediction (NWP) models requires a substantial investment in research, and computational resources. Traditional objective verification deterministic model output fails to demonstrate the added value high-resolution forecasts made by such models. It is generally accepted from subjective that these nevertheless have predictive potential for small-scale phenomena extreme events. This has prompted an extensive body research into new techniques scores aimed at developing mesoscale performance measures objectively return on NWP. In this article it argued evaluation information should be essentially connected method used extract direct (DMO). could forecaster, but, given probabilistic nature weather, more likely form statistical post-processing. Using statistics (MOS) traditional scores, approach demonstrated both educational abstraction real world example. MOS incorporates concepts fuzzy verification. weighs different forecast quality as essential extension methods.