English

{\Pi}-ML: A dimensional analysis-based machine learning parameterization of optical turbulence in the atmospheric surface layer

Atmospheric and Oceanic Physics 2023-08-11 v2 Machine Learning

Abstract

Turbulent fluctuations of the atmospheric refraction index, so-called optical turbulence, can significantly distort propagating laser beams. Therefore, modeling the strength of these fluctuations (Cn2C_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\Pi-ML, based on dimensional analysis and gradient boosting to estimate Cn2C_n^2. Through a systematic feature importance analysis, we identify the normalized variance of potential temperature as the dominating feature for predicting Cn2C_n^2. For statistical robustness, we train an ensemble of models which yields high performance on the out-of-sample data of R2=0.958±0.001R^2=0.958\pm0.001.

Keywords

Cite

@article{arxiv.2304.12177,
  title  = {{\Pi}-ML: A dimensional analysis-based machine learning parameterization of optical turbulence in the atmospheric surface layer},
  author = {Maximilian Pierzyna and Rudolf Saathof and Sukanta Basu},
  journal= {arXiv preprint arXiv:2304.12177},
  year   = {2023}
}
R2 v1 2026-06-28T10:15:58.075Z