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

arXiv: https://arxiv.org/abs/2304.12177 · HTML: https://mlsanity.com/paper/2304.12177/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 1 of its 27 references that are in the map.
- First posted on arXiv 24 Apr 2023.

## Builds on (most cited first, 1 of 1)

- [FLAML: A Fast and Lightweight AutoML Library](https://mlsanity.com/paper/1911.04706/) (2019)

## Built on by (most cited first, 1 of 1)

- [PSTNet: Physically-Structured Turbulence Network](https://mlsanity.com/paper/2603.07957/) (2026)

## Abstract (arXiv)

Turbulent fluctuations of the atmospheric refraction index, so-called optical turbulence, can significantly distort propagating laser beams. Therefore, modeling the strength of these fluctuations (Cn2) 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, Pi-ML, based on dimensional analysis and gradient boosting to estimate Cn2. Through a systematic feature importance analysis, we identify the normalized variance of potential temperature as the dominating feature for predicting Cn2. For statistical robustness, we train an ensemble of models which yields high performance on the out-of-sample data of R2=0.958±0.001.