Les Trente Glorieuses: 29 years of helioseismic observations with the Luminosity Oscillations Imager

Institut d’Astrophysique Spatiale, UMR8617, Université Paris-Saclay 91405 Orsay Cedex, France

^★ Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Received: 20 September 2025
Accepted: 13 November 2025

Abstract

Context. The Luminosity Oscillations Imager (LOI) of the Variability of Solar Irradiance and Gravity Oscillations (VIRGO) instrument on board the SoHO mission has been operating for almost 30 years.

Aims. The goal of this article is to report on the observation of p-mode parameters throughout these 30 years, which cover two solar cycles, using the LOI scientific and guiding pixels as well as the guiding high-voltage housekeeping. The article also provides a definitive description of the data for the final SoHO archive. The LOI time series analysed here starts on 1 April 1996 0:00 TAI and ends on 31 March 2025 23:59 TAI with a 60 s cadence.

Methods. I use the level 0 data of the LOI to compute level 2 corrected time series for each pixel taking into account engineering calibration, the presence of attractors (locked values), and the SoHO-Sun distance, the spacecraft roll. I used the scientific pixels to derive Sun-as-a-star signals, individual (l, m) signals, east-west and north-south difference signals, guiding pixel signals and east-west and north-south high-voltage signals to compute 29 one-year power spectra. A 29-year time series of the Sun-as-a-star signal was also used to push the detection of p modes to low frequency. The power spectra were globally fitted (fully over the p-mode envelope) using maximum likelihood estimators using a C++ code with two versions of the code depending on the presence of l = 0 modes (progFIT and guiFIT). The Fourier spectra for the individual (l, m) signals are locally fitted (around each p-mode) using maximum likelihood estimators using an IDL code for l ≤ 3.

Results. I report on the effect of solar activity upon mode frequencies, linewidths, heights, and energy rates. I report on the variation as a function of frequency for frequency, a₂ coefficient and linewidth changes, as well as the average over the degree and the frequency of these changes. Using the 29-year time series, I report on the frequencies, linewidths and mode heights fitted with progFIT. Using the collapsogram technique, I also report on the detection of modes below 1600 μHz, making the lowest frequencies detected with an instrument observing the Sun in intensity. I also report on the detection of p modes in the high-voltage and guiding pixel signals with a mode height about five to ten times larger than is observed in the Sun-as-a-star signal for l = 1. The ratios of the observed mode visibilities for the different signals are provided following a calibration of the size of the guiding pixels. While the visibility ratios for the signals excluding the limb are in good agreement with theory, those covering the solar limb are in strong disagreement.