Verification of mesoscale forecasts in the 2014 Olympic Games region. Part II: Preliminary results of diagnostic evaluation of quality and calibration of the forecasts by the COSMO-RU2 model

Abstract

Spatial (polygonal) averaging of verification scores of the mesoscale forecast over a complex mountainous region may result in not very informative conclusions, when the number of stations turnes out to be incomparable with the number of the model grid points. In such situation, the approach of diagnostic verification makes the forecast evaluation not only more informative but also more useful for operational forecasters and model developers, A description of principles of diagnostic verification is given, verification results of the surface air temperature in the mountain cluster of the Sochi-2014 Olympic Games are presented, and necessary conditions for the possible calibration of the operational forecast are discussed with account of accumulated conditional statistics.