Work

A Statistical Study of Ion-scale Waves at 1 au Using Machine Learning

Year: 2026

Type: article

Abstract: Circularly polarized waves at ion scales are thought to play vital roles in the heating and acceleration of coronal plasma, though the extent of their roles throughout the heliosphere remains an open question. In this work, a 1D convolutional neural network trained to identify waves in magnetic field time-series data is used to classify the Wind spacecraft data set at 1 au for the yr 2005–2024. Approximately 1.8 million intervals out of about one hundred million intervals are found to contain circularly polarized wave signatures with amplitudes >0.1 nT and spacecraft-frame frequencies >0.16 Hz. Wave intervals are observed most often in both the fast, hot wind and in slow Alfvénic wind. The number of wave intervals per day fluctuates at frequencies that closely follow the harmonics of solar rotation, implying a connection between wave occurrence and large-scale solar wind features. Comparison of the measured temperature anisotropy and proton plasma beta in wave intervals to kinetic instability criteria shows a correlation between wave occurrence and the proton cyclotron and mirror instabilities. However, local generation mechanisms associated with core proton temperature anisotropy instability only account for ≈30% of the identified waves.

Source: The Astrophysical Journal

DOI: 10.3847/1538-4357/aea41b