Applying Machine Learning in Numerical Weather and Climate Modeling Systems

Abstract

In this paper major machine learning (ML) tools and the most important applications developed elsewhere for numerical weather and climate modeling systems (NWCMS) are reviewed. NWCMSs are briefly introduced. The most important papers published in this field in recent years are reviewed. The advantages and limitations of the ML approach in applications to NWCMS are briefly discussed. Currently, this field is experiencing explosive growth. Several important papers are published every week. Thus, this paper should be considered as a simple introduction to the problem.

Keywords

• machine learning
• numerical weather modeling
• numerical climate modeling
• post-processing
• neural networks
• deep learning