Using Rotation to Explore Stellar Binarity and Period Prediction Techniques
Talk, The 245th Meeting of the American Astronomical Society, iPoster Sessions, Washington, DC, USA
Stars rotate, and measuring that rotation is critical to understanding the star’s history and future evolution. First, I will discuss the utility of rapid rotation for finding close companions to binary systems using APOGEE DR17. We identify 4,751 binary systems with rapid rotation, radial velocity scatter, and evidence of only one set of spectral lines. By assuming tidal synchronization, we estimate minimum masses for companions to 4,288 of these systems. Second, I will present a method for stellar rotation period prediction using a convolutional neural network on ASAS-SN light curves. We feed the network a transformed 2D frequency-time representation of the light curve and train the network using quiescent light curves injected with simulated rotation models. Finally, we present the efficacy of the network at period prediction and discuss ASAS-SN systematics that make period prediction difficult.