Search

In this thesis we study the necessary methods to perform a transit search for extrasolar planets. We apply these methods to search for planets in one of the fields of the LAIWO project: the Cygnus-Lyra field (Laiwo VI). We describe the problems that systematic effects can introduce for precise relative photometry at the millimagnitude level (~3mmag). Ways to minimize and quantify this correlated noise are also described. We test the weaknesses and strenghts of two transit detection algorithms (TDA) namely the Box fitting algorithm (BLS) and the TRUFAS algorithm using archive data from the OGLE project and simulations of the first year of the Pan-Planets survey. These projects are similar in terms of telescope size and field of view to the LAIWO survey. We have found that the main limitations of the BLS algorithm are the transit depth and correlated noise (Red Noise). The TRUFAS detection efficiency correlates with the number of points in transit and the number of transits present in the light curve, and, its detection efficiency is low (less than ~ 50%) for these type of ground-based observations. Finally, we create from the LAIWO data light curves which are suitable to detect planets among the stars brighter than R = 16.5 mag. We have found 31 eclipsing binaries and 18 light curves that have transits consistent with a planet. Of these detections, 3 eclipsing binaries and 8 planet candidates were independently found by the KEPLER survey. Of the 10 newly discovered transiting planets, 3 are promising to justify follow-up confirmation studies, which are always necessary to probe the planetary nature of a transiting companion.