\texttt{stella}: Convolutional Neural Networks for Flare Identification in \textit{TESS}

Abstract

Nearby young moving groups are kinematically bound systems of stars that are believed to have formed at the same time. With all member stars having the same age, they provide snapshots of stellar and planetary evolution. In particular, young (< 800 Myr) stars have increased levels of activity, seen in both fast rotation periods, large spot modulation, and increased flare rates (Ilin, Schmidt, Davenport, & Strassmeier, 2019; Zuckerman, Song, & Bessell, 2004). Flare rates and energies can yield consequences for the early stages of planet formation, particularly with regards to their atmospheres. Models have demonstrated that the introduction of superflares (> 5% flux increase) are able to irreparably alter the chemistry of an atmosphere (Venot, Rocchetto, Carl, Roshni Hashim, & Decin, 2016) and expedite atmospheric photoevaporation (Lammer et al., 2007). Thus, understanding flare rates and energies at young ages provides crucial keys for understanding the exoplanet population we see today.