Validace předpovědního modelu atmosféry v meso-měřítku

Abstract

In this thesis, we are aimed at various physical principles of convective parametrization. First, traditional parametrization schemes are described. This schemes are based on hypothesis of quasi-equilibrium between large-scale processes and convective activity and suggestion about negligible mesh fraction of convection in grid-box. But, passing to resolution in the range below 10 km, the deep convection si partly subgrid, partly resolved. A new generation of prediction with high horizontal resolution is now introduced. This model has new prognostic convective scheme with new description of interaction between convective mass transport and microphysical computation of clouds and precipitation. A prognostic scheme has been used for the vertical velocity of udraft and downdraft and more elaborate microphysics, include effect of autoconversion, evaporation and sedimentation of precipitation. These models are expected to forecast many atmospheric events much more precisely (e.g. strong convection). Validation of model in experiments at different resolutions is the main task of this work. Comparsion of new developed model with model currently used, sensitivity study of new components, are presented.