Predicting the yields of photometric surveys for transiting extrasolar planets

We develop a method for predicting the yield of transiting planets from a photometric survey given the parameters of the survey (nights observed, bandpass, exposure time, telescope aperture, locations of the target fields, observational conditions, and detector characteristics), as well as the underlying planet properties (frequency, period and radius distributions). Using our updated understanding of transit surveys provided by the experiences of the survey teams, we account for those factors that have proven to have the greatest effect on the survey yields. Specifically, we include the effects of the surveys' window functions, adopt revised estimates of the giant planet frequency, account for the number and distribution of main-sequence stars in the survey fields, and include the effects of Galactic structure and interstellar extinction. We approximate the detectability of a planetary transit using a signal-to-noise ratio (S/N) formulation. We argue that our choice of detection criterion is the most uncertain input to our predictions, and has the largest effect on the resulting planet yield. Thus, drawing robust inferences about the frequency of planets from transit surveys will require that the survey teams impose and report objective, systematic, and quantifiable detection criteria. Nevertheless, with reasonable choices for the minimum S/N, we calculate yields that are generally lower, more accurate, and more realistic than previous predictions. As examples, we apply our method to the Trans-Atlantic Exoplanet Survey, the XO survey, and the Kepler mission. We discuss red noise and its possible effects on planetary detections. We conclude with estimates of the expected detection rates for future wide-angle synoptic surveys.