A Refined Model of Convectively-Driven Flicker in Kepler Light Curves

Abstract

Light curves produced by the Kepler mission demonstrate stochastic brightness fluctuations (or "flicker") of stellar origin which contribute to the noise floor, limiting the sensitivity of exoplanet detection and characterization methods. In stars with surface convection, the primary driver of these variations on short (sub-eight-hour) timescales is believed to be convective granulation. In this work, we improve existing models of this granular flicker amplitude, or $F_8$, by including the effect of the Kepler bandpass on measured flicker, by incorporating metallicity in determining convective Mach numbers, and by using scaling relations from a wider set of numerical simulations. To motivate and validate these changes, we use a recent database of convective flicker measurements in Kepler stars, which allows us to more fully detail the remaining model--prediction error. Our model improvements reduce the typical misprediction of flicker amplitude from a factor of 2.5 to 2. We rule out rotation period and strong magnetic activity as possible explanations for the remaining model error, and we show that binary companions may affect convective flicker. We also introduce an "envelope" model which predicts a range of flicker amplitudes for any one star to account for some of the spread in numerical simulations, and we find that this range covers 78% of observed stars. We note that the solar granular flicker amplitude is lower than most Sun-like stars. This improved model of convective flicker amplitude can better characterize this source of noise in exoplanet studies as well as better inform models and simulations of stellar granulation.