Exploring the potential of neural networks to predict statistics of solar wind turbulence

TL;DR

This study demonstrates that small neural networks can effectively predict large-scale fluctuations in solar wind turbulence time series with missing data, outperforming simple imputation methods in high missing data scenarios.

Contribution

The paper introduces the use of neural networks to predict turbulence statistics in solar wind data, showing their advantages over traditional imputation methods for large missing data fractions.

Findings

Neural networks outperform mean imputation and linear interpolation at high missing data levels.

Small neural networks can predict large-scale fluctuation amplitudes effectively.

Performance varies when capturing both shape and amplitude of turbulence statistics.

Abstract

Time series datasets often have missing or corrupted entries, which need to be ignored in subsequent data analysis. For example, in the context of space physics, calibration issues, satellite telemetry issues, and unexpected events can make parts of a time series unusable. Various approaches exist to tackle this problem, including mean/median imputation, linear interpolation, and autoregressive modeling. Here we study the utility of artificial neural networks (ANNs) to predict statistics, particularly second-order structure functions, of turbulent time series concerning the solar wind. Using a dataset with artificial gaps, a neural network is trained to predict second-order structure functions and then tested on an unseen dataset to quantify its performance. A small feedforward ANN, with only 20 hidden neurons, can predict the large-scale fluctuation amplitudes better than mean imputation or linear interpolation when the percentage of missing data is high. Although, they perform worse than the other methods when it comes to capturing both the shape and fluctuation amplitude together, their performance is better in a statistical sense for large fractions of missing data. Caveats regarding their utility, the optimisation procedure, and potential future improvements are discussed.