Searching for transits in data with long time baselines and poor sampling

摘要

Aims. The standard method of searching parameter space for transits is ill-suited to data sets with long time baselines and poor temporal coverage, such as that anticipated from Gaia. In this paper, we present an alternative method for identifying transit candidates is such data, one focusing on finding periodicity in high S/N outliers. Methods. We describe a technique for testing a small number of flux measurements for periodicity and consistency with an origin in a transit with a constant change in flux and test their performance with Monte Carlo simulations. To complement this, we also include a description of a statistical method to analyze the distribution of these measurements to determine if they are normally distributed around a constant, reduce flux consistent with a planetary transits. Results. Large numbers of light curves can be quickly scanned for transit signatures with minimal loss in effectiveness for data sets with long time baselines and poor temporal coverage, where one observation per transit is the norm by testing for periodicity and analyzing their distribution. Conclusions. If the noise characteristics of the data set and the intrinsic noise of the individual stars are understood, this method focusing on statistical outliers is nearly equivalent to the standard method of scanning parameter space and significantly faster, if the signal ??? noise, the individual transits are sampled no more than once and a periodicity test is employed. Moreover, the test for a transit origin can eliminate additional false positives.