Calibration of solar radiation ensemble forecasts using convolutional neural network
<p><span>E</span><span>nsemble </span><span>forecast </span><span>approaches have become state-of-the-art for the quantification of weather forecast uncertainty. </span><span>However</span><span>, ensemble forecasts </span><span>from</span><span> numerical weather prediction models (NWPs) still tend to be biased and underdispersed, </span>hence justifying the use of statistical post-processing techniques <span>to improve forecast skill. </span></p><p>In this study, ensemble forecasts are post-processed using a convolutional neural network (CNN). CNNs are the most popular machine learning tool to deal with images. In our case, CNNs allow to integrate information from spatial patterns contained in NWP outputs.</p><p>We focus on solar radiation forecasts for 48 hours ahead over Europe from the 35-members ARPEGE (M&#233;t&#233;o-France global NWP) and a 512-members WRF (Weather Research and Forecasting) ensembles. We used a U-Net (a special kind of CNN) designed to produce a probabilistic forecast (quantiles) using as ground truth the CAMS (Copernicus Atmosphere Monitoring System) radiation service dataset with a spatial resolution of 0.2&#176;.</p>