Inpainting: A powerful interpolation technique for helio- and asteroseismic data

TL;DR

This paper demonstrates that inpainting algorithms effectively interpolate gaps in helio- and asteroseismic data, preserving spectral features and enabling accurate seismic analysis despite data discontinuities.

Contribution

The study introduces and validates an inpainting-based interpolation method tailored for helio- and asteroseismic datasets, improving data continuity and spectral fidelity.

Findings

Inpainting preserves the power spectrum of seismic data.

Interpolated data yields seismic inferences consistent with original data.

Method effectively handles non-random data gaps.

Abstract

In Helio- and asteroseismology, it is important to have continuous, uninterrupted, data sets. However, seismic observations usually contain gaps and we need to take them into account. In particular, if the gaps are not randomly distributed, they will produce a peak and a series of harmonics in the periodogram that will destroy the stellar information. An interpolation of the data can be good solution for this problem. In this paper we have studied an interpolation method based on the so-called 'inpainting' algorithms. To check the algorithm, we used both VIRGO and CoRoT satellite data to which we applied a realistic artificial window of a real CoRoT observing run to introduce gaps. Next we compared the results with the original, non-windowed data. Therefore, we were able to optimize the algorithm by minimizing the difference between the power spectrum density of the data with gaps and the complete time series. In general, we find that the power spectrum of the inpainted time series is very similar to the original, unperturbed one. Seismic inferences obtained after interpolating the data are the same as in the original case.