A Bayesian method for detecting stellar flares

• DOI: 10.1093/mnras/stu1889
• Publication Date: 2014-10-18T05:05:00Z
• Authors: M. Pitkin; D. Williams; L. Fletcher; S. D. T. Grant
• Keywords: False alarm; Flare

We present a Bayesian-odds-ratio-based algorithm for detecting stellar flares in light-curve data. assume are described by model which there is rapid rise with half-Gaussian profile, followed an exponential decay. Our signal also contains polynomial background required to fit underlying variations the data, could otherwise partially mimic flare. characterize false alarm probability and efficiency of this method under assumption that any unmodelled noise data Gaussian, compare it simpler thresholding based on used Walkowicz et al. find our has significant increase detection low signal-to-noise ratio (S/N) flares. For conservative can detect 95 per cent S/N less than 20, as compared 25 method. test how well Gaussian holds applying selection ‘quiet’ Kepler stars. As example we have applied stars Quarter 1 The finds 687 flaring total 1873 after vetos been applied. these made preliminary characterizations their durations S/N.