Repository logo

Searching for exoplanets using the transit method.

Thumbnail Image



Journal Title

Journal ISSN

Volume Title



We present a study designed to detect transiting exoplanets in Kepler light curve data. We developed an exoplanet detection algorithm based on modelling transit light curves and fitting the models to light curve data using a chi-square minimization approach in order to identify exoplanets and estimate their properties such as orbital period, planetary radius and semi- major axis (orbital radius) from the best t parameters of the model. We applied our algorithm to a blind sample of Kepler mission data consisting of approximately 4500 stars. The selection criteria for the blind sample were Tstar < 6000 K, Rstar < 1R and 13:5 < Kepler Magnitude < 14. The blind sample contained 70 known exoplanets. Our algorithm detected 50 of the 70 known exoplanets in the blind sample. We found that our algorithm was effective in detecting exoplanets with planet-star radius ratios greater than 0.01 (k > 0:01) and/or exoplanets with radii greater than 2:5R , as well as short-period exoplanets (p < 90 days). Twenty four of the exoplanets in the blind sample were from multi-planetary systems and, in these cases, we found our algorithm first fits for the largest transit depth and/or (subsequently) for the shortest orbital period. We did not find any potentially habitable exo- planets in our blind sample. This is not unexpected as, of more than 3400 exoplanets found to date after surveying upward of 500 000 stars, only 52 exoplanets are considered potentially habitable to varying degrees i.e. 1.5% of all exoplanets found to date are considered potentially habitable.


Master of Science in Applied Mathematics, University of KwaZulu-Natal, Westville, 2017.