Decoding the coupling of mesoscale cloud patterns and meteorology using machine learning techniques

Sammanfattning

Climate model projections of future climate scenarios play a critical role in informing decision making and policy contexts. This study focuses primarily on marine low-level mesoscale clouds, which are poorly represented in Earth System Models (ESMs). Using machine learning tools, couplings between cloud properties and the prevailing meteorological conditions could be found. Several artificial neural networks were implemented and analyzed to solve this problem. The results show that the predictive performance of machine learning models is affected by many things, such as the data used, configuration, and computational resources. In addition, comparative and factor analysis provides the foundation to build further models that can be used to help ESMs better represent mesorscale clouds