资源环境科技发展态势分析平台

An evaluation of forecasts by ALADIN numerical weather prediction model is performed by comparing synthetic and measured IR10.8 μm (Infrared channel with a wavelength of 10.8 μm) and WV06.2 μm (Water vapour channel with a wavelength of 6.2 μm) brightness temperatures (BTs) by the geostationary satellite Meteosat Second Generation. To calculate synthetic satellite images, we used a radiative transfer model RTTOV v11.3, as implemented in the ALADIN model. The evaluation of the forecasts is performed for summer month June 2020 and winter month February 2021 and for lead times from 1 to 72 h. The model horizontal resolution is 2.3 km and the verification domain covers a large part of Europe. As for verification parameters, we calculated bias, root-mean-square error, correlation coefficient (PC), Fraction Skill Score and a technique that take into account the relative position of the verified objects. We diagnosed relationships, measured by PC, between the forecasted and measured BTs on one side and the forecasted and measured precipitation on the other side. We present and discuss verification results in the view of usefulness for model developers.