Finding planets around other stars in our galaxy is extremely challenging because they hardly emit any light of their own, which in most cases is completely negligible when compared with the brightness of their parent star. It is often the case that extrasolar planets are discovered from detecting the effect they produce on their parent stars, such as a the gravitational reflex motion that induces periodic variations on the stellar radial velocity. The SuperWASP project is a UK led project that consists of two fully automated astronomical observatories designed to search for planetary transits. It does this by monitoring the brightness of every star it detects in its 61 square degree field of view (over all 8 cameras) and looking for the characteristic periodic dip caused by a planet passing in front of its host star. This method is extremely important, since when combined with radial velocity measurements it breaks the degeneracy between the mass and inclination of the orbit, whilst measuring the planetary radius. Our understanding of planetary structure and formation relies heavily on the study of such systems.
The SuperWASP project employs relatively inexpensive "off-the-shelf" equipment to achieve its objectives. It is comprised of two observatories, one located at the Roche de Los Muchachos observatory, La Palma, Spain (Northern Hemisphere) and one at the South African Astronomical Observatory, Sutherland, South Africa (Southern Hemisphere). Each observatory consists of an array of 8 Canon wide-field lenses backed with high-quality Andor 2048 x 2048 thinned CCDs using e2v chips (technical information). Using a custom designed software pipeline and the ability to survey the majority of the available sky, it is responsible for the discovery of around 70 confirmed transiting planets, making it the most successful project of its kind to date. The relatively bright magnitude range of its targets also means that they are easier to follow-up and perform detailed studies.
My contribution to this project has come in the form of a study of the noise sources affecting the SuperWASP lightcurves. The results from studies of this kind usually yield improvements to the software pipeline and potential upgrades to the instrument. This analysis was performed based on results from the production of detector maps, which are the result of averaging the residuals of every measurement made by a camera in CCD coordinates. These provide with a measurement of CCD features that have not been corrected for by the software pipeline and calibration stages, whilst proving to be a useful source of information on the evolution of the chips themselves. I have found that the current flat fielding strategy is introducing a source of systematic noise due to the wavelength dependent nature of the CCDs, and therefore should be revised. Results from quality control analysis have also revealed important clues about whether hardware improvements and software upgrades have had an impact on the overall quality of the photometry.
In addition to the analysis, I have been a part of the candidate eyeballing stages, which forms a crucial part of finding good potential candidates for follow-up. The automated pipeline contains a transit search algorithm that looks for periodic dips in the photometric lightcurves, but the vast majority of these are false positives. Examples of such false signals include eclipsing binaries of several kinds (standard, grazing and low mass), variable stars and periodic systematic noises, and so every candidate has to be manually inspected before any follow-up is performed.
The Next Generation Transit Survey is a new initiative to build a robotic observatory designed to detect transits of Neptune-sized planets around bright stars. It has recently secured full funding and authorisation to be commissioned at ESO's Paranal Observatory in Chile, which is widely considered to be the best available location on Earth for astronomical observations. Building from previous experience with the SuperWASP project, this project has similar goals and uses a similar approach, but targets a different range of stars. It is optimised to perform 1mmag precision photometry on K and early M stars, thereby allowing smaller planets to be found. It uses a large format 2Kx2K cameras developed by Andor, in Belfast, with enhanced sensitivity in the 600 to 900nm range, and Astro Systeme telescopes with custom modifications to improve our desired performance.
This project has deployed a prototype telescope at the Roche de los Muchachos observatory in La Palma, Spain, in 2009 and 2010, making use of a 1Kx1K CCD chip. A careful analysis of the data collected during this period has successfully demonstrated that the key performance issues have been resolved, and that 1mmag precision is achieved for stars brighter than V=13, thereby fulfilling our science requirements. The final instrument will consist of 12 telescope on individual mounts, which will be on sky in 2013.
My contributions to this project have been varied. I have developed a calibration pipeline for the prototype data to be analysed and performed a part of the CCD testing stage prior to the prototype commissioning. I have also been closely involved with the data quality analysis and noise budget models required to test the instruments limits.
The largest contribution has been to develop a simulation software to help towards optimising the choice of sky areas to observe, in order to maximise the chances of finding transiting exoplanets. This code takes into account several factors, such as weather, night time, lunar proximity and contribution to the sky background, galactic plane/center proximity, blending and airmass. It is also capable of calculating the probability of finding planets on given coordinates based on potential restrictions. This is crucial in order to optimise the amount of coverage that is necessary on a given field and thereby improving the performance of the project.
ULTRACAM is a 3 channel high-speed camera for astronomical photometry. It offers the chance of performing high precision photometry on 3 simultaneous optical bands with negligible readout time, due to its frame transfer CCDs. It has been used in multiple several meter class telescopes, such as ESOs Very Large Telescope and the William Herschell Telescope in La Palma. It was designed primarily to resolve the eclipses of fast orbiting binaries, but we have used it to perform transmission studies of extrasolar planets.
During planetary transits a portion of the light from the host star passes through the atmosphere of planet, therefore a spectrum of a planetary system whilst in transit should be a combination of the stellar spectrum and the planets' atmospheric spectrum. If one measures the stellar spectrum outside of transit, a simple comparison should show the atmospheric features. Similarly, by comparing the depth of a transit in different bands, a measure of the opacity of the atmosphere at several wavelengths can provide clues towards understanding the larger scale spectral profile of the planet's envelope. The multiband capabilities of ULTRACAM are ideal to attempt this. A choice of filter combinations allows us to test several predicted scenarios and probe for the presence of elements such as NaI and TiO.
We have been awarded several nights to use this instrument on ESO's NTT telescope, La Silla, Chile. We observed transits of planets WASP-15b and WASP-17b with a range of filter combinations to test for several scenarios. These are planets with large atmospheric scale heights, which would enhance any planet radius difference. Analysis is ongoing. We have also been awarded one night to attempt the detection of the secondary eclipse of WASP-33b in the z band. This planet orbits a pulsating A star, and we planned to use the g band photometry to decorrelate the pulsations. Unfortunately, clouds affected this observation, but a feasibility study can still be made.
The SuperWASP South telescope in South Africa.
The NGTS prototype telescope in La Palma, Spain.
ULTRACAM on the WHT, La Palma, Spain.