Understanding stellar activity-induced radial velocity jitter using simultaneous K2 photometry and HARPS RV measurements

TL;DR

This study uses simultaneous K2 photometry and HARPS RV measurements to analyze stellar activity's impact on radial velocity jitter, identifying F8 as a key activity proxy and evaluating modeling techniques for RV prediction.

Contribution

It provides the first simultaneous high-precision photometric and RV dataset for multiple stars, assessing correlations and modeling methods across different activity levels.

Findings

Strong correlations between activity indicators and RV jitter in very active stars.

F8 index remains a reliable proxy for RV jitter across activity levels.

FF' and SOAP2.0 models predict RV jitter well in active stars, less so in quieter stars.

Abstract

One of the best ways to improve our understanding of the stellar activity-induced signal in radial velocity (RV) measurements is through simultaneous high-precision photometric and RV observations. This is of prime importance to mitigate the RV signal induced by stellar activity and therefore unveil the presence of low-mass exoplanets. The K2 Campaign 7 and 8 field-of-views were located in the southern hemisphere, and provided a unique opportunity to gather unprecedented simultaneous high precision photometric observation with K2 and high-precision RV measurements with the HARPS spectrograph to study the relationship between photometric variability and RV jitter. We observed nine stars with different levels of activity; from quiet to very active. We probe the presence of any meaningful relation between measured RV jitter and the simultaneous photometric variation, and also other activity indicators (e.g. BIS, FWHM, $logR'_{HK}$, and F8), by evaluating the strength and significance of the correlation between RVs and each indicator. We found that for the case of very active stars, strong and significant correlations exist between almost all the observables and measured RVs; however, for lower activity levels the correlations become random. Except for the F8 which its strong correlation with RV jitter persists over a wide range of stellar activity level, and thus our result suggests that F8 might be a powerful proxy for activity induced RV jitter. Moreover, we examine the capability of two state-of-the-art modeling techniques, namely the FF' method and SOAP2.0, in accurately predicting the RV jitter amplitude using the simultaneous photometric observation. We found that for the very active stars both techniques can reasonably well predict the amplitude of the RV jitter, however, at lower activity levels the FF' method underpredicts the RV jitter amplitude.