Imputation-based machine learning for measuring the Fried parameter with a sky monitor, day and night

This study develops a machine learning framework to impute missing Fried parameter values, crucial for assessing atmospheric turbulence's impact on optical systems. By leveraging data from a sky monitor, including atmospheric conditions and imagery, the research addresses the challenge of continuous data collection. The approach, which separately imputes nighttime and daytime values, significantly improves accuracy, offering insights into atmospheric dynamics and aiding in the optimization of optical communications.