Π-ML: A dimensional analysis-based machine learning parameterization of optical turbulence in the atmospheric surface layer
This work addresses the need for accurate optical turbulence prediction in atmospheric surface layers, which is crucial for developing reliable free-space optical communication systems, representing an incremental improvement with a novel hybrid approach.
The paper tackled the problem of modeling optical turbulence strength (C_n^2) for free-space optical communication by proposing a physics-informed machine learning method based on dimensional analysis and gradient boosting, achieving high performance with an R^2 of 0.958±0.001 on out-of-sample data.
Turbulent fluctuations of the atmospheric refraction index, so-called optical turbulence, can significantly distort propagating laser beams. Therefore, modeling the strength of these fluctuations ($C_n^2$) is highly relevant for the successful development and deployment of future free-space optical communication links. In this letter, we propose a physics-informed machine learning (ML) methodology, $Π$-ML, based on dimensional analysis and gradient boosting to estimate $C_n^2$. Through a systematic feature importance analysis, we identify the normalized variance of potential temperature as the dominating feature for predicting $C_n^2$. For statistical robustness, we train an ensemble of models which yields high performance on the out-of-sample data of $R^2=0.958\pm0.001$.