During my PhD my work has focused on all aspects of the Multi-site All-Sky CAmeRA (MASCARA), a survey aimed at detecting transiting exoplanets around the brightest stars in the sky by monitoring the entire sky using a station in both the northern and southern hemisphere. In addition I was involved in the related β Pictoris b Ring (bRing) project, aimed at observing the 2017 Hill sphere transit of β Pictoris b.
MASCARA was designed to find transiting exoplanets around the brightest stars, meaning visual magnitudes between 4 and 8. Exoplanets around such bright stars are ideal candidates for high resolution, high signal-to-noise studies aimed at constraining the composition, temperature-pressure profiles and atmospheric circulation of their atmospheres. In particular, high resolution spectroscopic studies have been shown to be able to unambiguously detect the presence of molecules such as water, carbon monoxide and methane but have been limited to the small number of planets orbiting sufficiently bright stars.
MASCARA also presents an opportunity to expand our knowledge of planets orbiting the hotter "early-type" stars. As a result of the higher stellar irradiation such planets are expected to receive, in particular the stronger ultra-violet (UV) radiation, these planets are expected to host chemistry in their atmospheres not seen in planets around colder stars. In addition, it has been shown that the orbital planes of planets around "early-type" stars might not be aligned with the stellar spin-axis as has been found for planets around cooler stars, which might be due to dynamical interactions or be a general feature of star formation. To date, only a handful of planets have been confirmed due to the challenges in determining the masses of planets around hotter stars through radial velocity (RV) measurements. The brightness of the stars targeted by MASCARA allows us to obtain such measurements at the required signal-to-noise more easily.
To date MASCARA has found two new transiting exoplanets in the northern hemisphere with several candidates still awaiting confirmation. MASCARA-1b orbits the A8 star HD 201585 with a period of 2.15 days and has mass and radius of 3.7 Jupiter masses and 1.5 Jupiter radii respectively. MASCARA-2b orbits the A2 star HD 185603 with a period of 3.47 days and has a radius of 2.2 Jupiter radii. The mass of MASCARA-2b has not yet been measured but has been constrained to be less then 15 Jupiter masses. In the southern hemisphere a second MASCARA station, operational since July 2017, is expected to find several more in the near future.
For the MASCARA survey I primarily work on the calibration of the photometric light curves and the search for the periodic signals of transiting exoplanets.
Each MASCARA station produces 500 GB in images of the local sky every night. This data volume is so large that transfer and storage are not feasible. As such the images are processed by the station itself, performing aperture photometry on all stars and transferring the resulting light curves, up to 20 GB per night, to Leiden Observatory for calibration. In order to calibrate the light curves I wrote an algorithm capable of removing the main systematic effects from this large amount of data in python. The algorithm initially accounts for the 3 main systematic effects in the MASCARA light curves: the changing transmission of the optics with position; sinusoidal modulations believed to be a result of to the interline design of the CCD cameras; and changes in the atmosphere. In creating the calibration algorithm the design of MASCARA presented a number of unique challenges not present in other surveys, such as the large field-of-view of the individual cameras (50 x 70 degrees) and the fixed pointings of the cameras.
Uncalibrated MASCARA light curve of HD189733. The different colours indicate data taken by different cameras. The effects of transmission and sinusoidal modulations are clearly visible.
To reconstruct the spatial effects of the transmission and sinusoidal modulations the calibration algorithm uses stars in small declination bins, i.e. those tracing the same path across the CCD every night. Once these spatial corrections have been obtained stars located in the same region of the sky, and thus are affected similarly by the atmosphere, are used to reconstruct the temporal changes in the atmosphere. These spatial and temporal corrections are iteratively refined and applied to the data. Subsequently empirical fits to the individual light curves are used to correct for residual changes with position on the CCD as well as long-term variations in the baseline. In this way I was able to reach percentage level precision on 5 minute time-scales.
Reconstructed transmission profile for the Zenith camera as a
function of position on the CCD. A grid of hour angle angle and
declination is overlaid on top.
For the transit search I use my own version of the box
least-squares algorithm, also written in python. Promising
candidates are selected and manually vetted using external
information on known variable stars as well as estimates of the
stellar radii based on the GAIA parallaxes. For the best candidates
follow-up observations are obtained trough collaborations with
telescopes in both the northern and southern hemispheres.
Box least-squares periodogram and phase-folded light curve used
in the discovery of MASCARA-2b, a planet orbiting the bright star
bRing uses the same optics and CCD cameras as MASCARA to monitor β Pictoris for the 2017 Hill sphere transit of β Pictoris b. However, while MASCARA used a single (long) exposure time bRing used an alternating sequence of short and long exposures to prevent β Pictoris from saturating while also maximizing scientific output on the same stars monitored by MASCARA. In addition, the calibrated light curves had to be available within minutes of the data taking so larger observatories could be triggered when the Hill sphere transit was detected. To achieve this I updated much of the MASCARA software to deal with the alternating exposure times and created a way for the calibration to run alongside the data taking, where it had only been run on a bi-weekly basis for MASCARA.