Improving machine learning-based weather forecast post-processing with clustering and transfer learning

Machine learning has been widely applied in numerical weather prediction, but the incorporation of new observational sites into models trained on stations with long historical records remains a cha...