Learning the Representations of Moist Convection with Convolutional Neural Networks

Shih-Wen Tsou, Chun-Yian Su, Chien-Ming Wu

arXiv preprint · 23 May 2019

In one sentence. CNN predicts moist convection effects on the environment at grey-zone resolution, using horizontal gradients, as a potential replacement for conventional cumulus parameterization.

Abstract

The representations of atmospheric moist convection in general circulation models have been one of the most challenging tasks due to its complexity in physical processes, and the interaction between processes under different time/spatial scales. This study proposes a new method to predict the effects of moist convection on the environment using convolutional neural networks. With the help of considering the gradient of physical fields between adjacent grids in the grey zone resolution, the effects of moist convection predicted by the convolutional neural networks are more realistic compared to the effects predicted by other machine learning models. The result also suggests that the method proposed in this study has the potential to replace the conventional cumulus parameterization in the general circulation models.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.