The effect of red noise on planetary transit detection

Abstract

Since the discovery of short-period exoplanets a decade ago, photometric surveys have been recognized as a feasible method to detect transiting hot Jupiters. Many transit surveys are now under way, with instruments ranging from 10-cm cameras to the Hubble Space Telescope. However, the results of these surveys have been much below the expected capacity, estimated in the dozens of detections per year. One of the reasons is the presence of systematics (``red noise'') in photometric time series. In general, yield predictions assume uncorrelated noise (``white noise''). In this paper, we show that the effect of red noise on the detection threshold and the expected yields cannot be neglected in typical ground-based surveys. We develop a simple method to determine the effect of red noise on photometric planetary transit detections. This method can be applied to determine detection thresholds for transit surveys. We show that the detection threshold in the presence of systematics can be much higher than with the assumption of white noise, and obeys a different dependence on magnitude, orbital period and the parameters of the survey. Our method can also be used to estimate the significance level of a planetary transit candidate (to select promising candidates for spectroscopic follow-up). We apply our method to the OGLE planetary transit search, and show that it provides a reliable description of the actual detectionthreshold with real correlated noise. We point out in what way the presence of red noise could be at least partly responsible for the dearth of transiting planet detections from existing surveys, and examine some possible adaptations in survey planning and strategy. Finally, we estimate the photometric stability necessary to the detection of transiting ``hot Neptunes''.