Removing Radio Frequency Interference from Auroral Kilometric Radiation with Stacked Autoencoders

TL;DR

This paper introduces DAARE, a deep learning autoencoder trained on synthetic data, which effectively removes radio frequency interference from auroral kilometric radiation spectrograms, enhancing data quality for astrophysical research.

Contribution

The study presents a novel denoising autoencoder trained solely on simulated data to clean real AKR observations, outperforming existing methods in RFI removal.

Findings

DAARE achieves 42.2 PSNR on synthetic data.

DAARE improves SSIM by 0.064 over state-of-the-art.

Qualitative tests show effective RFI removal from real observations.

Abstract

Radio frequency data in astronomy enable scientists to analyze astrophysical phenomena. However, these data can be corrupted by radio frequency interference (RFI) that limits the observation of underlying natural processes. In this study, we extend recent developments in deep learning algorithms to astronomy data. We remove RFI from time-frequency spectrograms containing auroral kilometric radiation (AKR), a coherent radio emission originating from the Earth's auroral zones that is used to study astrophysical plasmas. We propose a Denoising Autoencoder for Auroral Radio Emissions (DAARE) trained with synthetic spectrograms to denoise AKR signals collected at the South Pole Station. DAARE achieves 42.2 peak signal-to-noise ratio (PSNR) and 0.981 structural similarity (SSIM) on synthesized AKR observations, improving PSNR by 3.9 and SSIM by 0.064 compared to state-of-the-art filtering and denoising networks. Qualitative comparisons demonstrate DAARE's capability to effectively remove RFI from real AKR observations, despite being trained completely on a dataset of simulated AKR. The framework for simulating AKR, training DAARE, and employing DAARE can be accessed at github.com/Cylumn/daare.