Predictive capability of machine learning algorithms for reconstructing high-level cloud parameters based on lidar observations

Abstract

The paper focuses on machine learning algorithms used to predict backscattering phase matrix (BSPM) elements of high-level clouds based on meteorological observations. Several machine learning methods, such as random forest, support vector, and linear regression, are used to detect the relationship between meteorological parameters and BSPM elements. It is shown that the random forest algorithm provides the most accurate predictions compared to other models. Despite a relatively small amount of the initial data, these methods have a good potential for their use in analyzing complex atmospheric interactions.