Deeplearning-based approach to improving numerical weather forecasts

• Datum: 2023
• Hauptverfasser: Doroshenko, А.Yu., Shpyg, V.M., Kushnirenko, R.V.
• Format: Artikel
• Sprache: Ukrainisch
• Veröffentlicht: PROBLEMS IN PROGRAMMING 2023
• Schlagworte: deep learning; numerical weather prediction; COSMO; 2m temperature; UDC 51:681.3.06
• Online Zugang: https://pp.isofts.kiev.ua/index.php/ojs1/article/view/585
• Назва журналу: Problems in programming

This paper briefly describes the history of numerical weather prediction development. The difficulties, which occur in the modelling of atmospheric processes, their nature and possible ways of their mitigation, are described. It also indicates alternative methods of improving the quality of meteorol...