Leveraging Space-based Data from the Nearest Solar-type Star to Better Understand Stellar Activity Signatures in Radial Velocity Data

Abstract

Stellar variability is a key obstacle in reaching the sensitivity required to recover Earth-like exoplanetary signals using the radial velocity (RV) detection method. To explore activity signatures in Sun-like stars, we present SolAster, a publicly distributed analysis pipeline 10 10 https://tamarervin.github.io/SolAster/ that allows for comparison of space-based measurements with ground-based disk-integrated RVs. Using high-spatial-resolution Dopplergrams, magnetograms, and continuum filtergrams from the Helioseismic and Magnetic Imager aboard the Solar Dynamics Observatory (SDO), we estimate “Sun-as-a-star” disk-integrated RVs due to rotationally modulated flux imbalances and convective blueshift suppression, as well as other observables such as unsigned magnetic flux. Comparing these measurements with ground-based RVs from the NEID instrument, which observes the Sun daily using an automated solar telescope, we find a strong relationship between magnetic activity indicators and RV variation, supporting efforts to examine unsigned magnetic flux as a proxy for stellar activity in slowly rotating stars. Detrending against measured unsigned magnetic flux allows us to improve the NEID RV measurements by ∼20% (∼50 cm s−1 in a quadrature sum), yielding an rms scatter of ∼60 cm s−1 over five months. We also explore correlations between individual and averaged spectral line shapes in the NEID spectra and SDO-derived magnetic activity indicators, motivating future studies of these observables. Finally, applying SolAster to archival planetary transits of Venus and Mercury, we demonstrate the ability to recover small amplitude (<50 cm s−1) RV variations in the SDO data by directly measuring the Rossiter–McLaughlin signals.

Stellar signal components seen in HARPS and HARPS-N solar radial velocities

“…Additionally, satellites like the Solar Dynamical Observatory (SDO, Pesnell et al 2012) continuously observe the solar surface at high-spatial resolution in different wavelengths, but also in velocity space, and comparing what is happening in the RV measurements and simultaneously at the level of the solar surface can give us clues on the origin of the different types of stellar signals (e.g. Haywood et al 2016;Al Moulla et al 2022;Ervin et al 2022).…”

“…Using HARPS-N during the day prevents competing with night-time observations, providing several hours of observations per day, on every clear day, which give us an exceptional sampling to probe stellar signals from minute to year timescales. Additionally, satellites such as the Solar Dynamical Observatory (SDO, Pesnell et al 2012) continuously observe the solar surface at high spatial resolution in different wavelengths, but also in velocity space, and comparing what is happening in the RV measurements and simultaneously at the level of the solar surface can give us clues about the origin of the different types of stellar signals (e.g., Haywood et al 2016;Al Moulla et al 2022;Ervin et al 2022).…”

Revisiting ϵ Eridani with NEID: Identifying New Activity-sensitive Lines in a Young K Dwarf Star

“…Recent work has made significant advances in applying various techniques to characterizing and mitigating the effect of stellar activity on RV measurements. These techniques include employing regular-cadence observing strategies (Jeffers et al 2022), using proxies and models to estimate the effect of stellar activity and subtract it from RV measurements (Ervin et al 2022;Haywood et al 2022), and applying various statistical techniques to the spectra or cross-correlation functions (CCFs) to identify and separate variation caused by different sources of activity (Cretignier et al 2022;Siegel et al 2022;Simola et al 2022;Zhao et al 2022). In addition to these techniques, machine learning has the potential to become a promising new avenue for astronomical data analysis.…”