We detect exoplanets by monitoring large parts of the sky and attempting to determine if the light from a star periodically dims for a short time. This short "dip" in the light received may be caused by a small planetary-sized object moving in its orbit in front of the star as seen from our vantage point on Earth. This is a so-called exoplanet transit signal. The length and shape of this transit signal can be related to the radius and orbital period of the object passing in front of the star. Extensive follow-up observations are usually carried out to determine if the object detected in this way has a radius and mass consistent with that of a planet. Exoplanet detection by this method has proven to be enormously successful, and is responsible for a large fraction of all discovered exoplanets over the past twenty years.
HATNet telescopes are located at the Fred Lawrence Whipple Observatory (FLWO) at Mount Hopkins in Arizona, USA (5 telescopes), and at the Mauna Kea Observatory in Hawaii, USA (2 telescopes). The large separation in longitude allows us to seamlessly monitor the sky over the better part of 24 hours, reducing our susceptibility to false-positive signals caused by interruptions in observing. We use off-the-shelf 200-mm f/1.8 lenses attached to large format CCD cameras as the basis of our instruments, coupled with purpose-built robust telescope mounts and dome enclosures. All operations are automated; our telescopes are fully robotic and make decisions about when and what to observe based on weather conditions and pre-programmed observing priorities.
-
The HATNet telescope mount, camera, and lens. Photo credit: Gaspar Bakos. -
HATNet telescopes at FLWO, open for servicing during the day. Photo credit: Gaspar Bakos. -
HATNet telescopes at Mauna Kea Observatory. Photo credit: Gaspar Bakos. -
HATNet telescopes at FLWO operating at dusk. Photo credit: Gaspar Bakos.
We and our collaborators have developed sophisticated algorithms to tease out such planetary transit signals from large time-series datasets, including the Trend Filtering Algorithm (TFA), and the Box Least-Squares (BLS) algorithm to find periodic signals in noisy data. An open-source software package written by HATNet team-member Joel Hartman implementing these and other useful algorithms for astrophysical time-series analysis is available for download.