The primary aim of this project is to determine the factors that influence false positive detection, and perhaps more specifically how the introduction of new or updated datasets influences this.

Between the years 2009 and 2018, the Kepler Space Telescope continuously examined a roughly square-shaped region of the sky within the Cygnus constellation in search of exoplanets, or planets belonging to solar systems other than our own. Designed to utilize the transit method to detect these planets, it used an array of charge-coupled devices (CCDs) to record dips in brightness of stars, which would be indicative of a planet transiting the star (Basri, 2005). An automated system called Robovetter was used to vet Kepler Objects of Interest (KOIs) that exhibited planet-like transit signals using a procedural, rules-based process (Mullally, 2016). This produced a preliminary catalogue of possible planets; however, it did not always assign the correct disposition to some KOIs due to its inability to consistently account for phenomena that can closely mimic a typical planetary transit. This has led to a large set of false positive objects.

Such phenomena include eclipsing binaries, both as standalone objects and in the background of much nearer stars due to chance alignment, and which account for a significant percentage of false positive objects (Koch, 2006). As two stars orbit one another, if they are aligned with the observer, they produce periodic dips in brightness similar to a planet. However, unlike a typical planetary transit, which exhibits a roughly trapezoidal light curve (Li, 2019), the eclipsing binary pair is characterized by brightness dips that alternate between two consistent depths on a periodic, odd/even basis. This arises from the fact that both stars that eclipse each other emit light, and together emit the most light when they are not eclipsing; the two depths result from common differences in luminosity between the two stars, thus causing different amounts of light to be blocked on an alternating basis. The transit depth of such a system is typically far greater than that of a planet. When this eclipsing binary pair is closely aligned in the sky with a closer, single star, the light from the nearer star “washes out” the transit signal of the binary pair. This brings the depth of the transit much closer to that of a typical planet, and makes a now much shallower secondary eclipse difficult to discern.

Literature Cited