On the ground, a handful of transit surveys have been successful in detecting planets amongst which WASP and HAT stand as the most prolific, with an average rate of more than one planet discovered every 3 weeks. The main advantage of ground-based surveys is their ability to search the whole sky and hence to find planets orbiting bright stars. Current surveys have, within their design limitations, mapped almost the whole sky and found hundreds of short period giant planets.
The first dedicated space transit mission, CoRoT, was successfully launched in December 2006. It was a pioneer in its use of ultra-precision photometry with high sampling rate. The satellite primarily observed two fields of view each 4 square degrees – once per year. Each field counted typically a few thousand dwarf stars with magnitudes ranging from roughly 11 to 16. After 5 years of operations, CoRoT has discovered a few dozen exoplanets, among them the first planet identified as having a rocky core: CoRoT-7-b.
Two years after CoRoT, the launch of the Kepler mission has turned out to be a landmark in transiting planet searches. With its capability to measure continuously brightness variations of about 100,000 solar-like stars to an unprecedented accuracy of order 20 ppm in a single field of view approximately 100-square degrees, it has outperformed CoRoT. Kepler found many thousands of transiting planetary candidates, some of them with radii as small as the Earth and many multiple transiting systems. With these discoveries Kepler has provided the community with a large uniform database of potential planetary systems enabling the derivation of distribution functions for planetary orbits, radii and hierarchical structure of systems (Borucki et al. 2011). Comparing Kepler to CoRot points out the importance of staring at a target for long time in order to obtain accurate photometry needed for the detection of shallow transits. In five years, enough transit events for every star will have been observed to average over the stellar noise and to increase the signal-to-noise of the transit detections. Typical stellar noise of 20 ppm averaged over six hours needs to be co-added > 5 times to detect (S/Ntransit>10) an Earth size transit toward a G star (80 ppm depth).
Most of the low mass planets (about Neptune mass and smaller than Neptune) are in multiple-systems with mutual inclinations less than a few degrees (Fabrycky et al. 2012). From the Kepler data it was found that about 20% of the transiting candidates have at least one additional transiting companion (Batalha et al. 2012). A small fraction of these multi-planet systems show clear signs of being in orbital resonance and thus will exhibit predictable transit timing variations (TTVs) which can provide important information about the planetary system, including the discovery of very low mass planets detectable in no other way. Yet any multi-planet system will exhibit variations in eclipse timings due to mutual gravitational interactions and this too can be used to estimate the planet masses. These alternative methods have been used on some planetary systems to get the mass of some planets with unfortunately somewhat large uncertainties. In any case, it requires long observation periods to add enough transit events to constrain a TTV signal variation (typically 100 transits). Both CoRoT and Kepler have been successful in reaching their design goals. However, it is revealing that despite this, only two rocky planets have been identified for certain (CoRoT-7 and Kepler-10). This paucity of the most interesting targets is related to the faintness of the target stars. The need to stare at a given field for a long time (in order not to miss a transit) as well as to have large numbers of targets in a given field of view (to maximize the chance of detection) dictated that both CoRoT and Kepler would search for transits toward stars typically between V ~ 13-16 magnitude (see Fig. 4). Measuring sufficiently precise radial velocities for stars this faint in order to obtain a reliable detection from an Earth mass planet is virtually impossible. The example of CoRoT-7 shows that with HARPS spectrograph it is possible to measure the mass of small planets in the super-Earth domain, located on short period orbit for stars brighter than V ~ 11 magnitude. Similar measurements for planets with longer orbital periods on fainter stars, typical of Kepler candidates, would require a prohibitive amount of telescope time.
In total Kepler has found a dozen of the smallest transiting planetary candidates orbiting stars brighter than 11th magnitude with only a fraction of them offering hope for an accurate planet mass determination with the recently installed HARPS-North facility. CHEOPS – by targeting bright stars located anywhere on the sky – will not suffer from these limitations and will provide a uniquely large sample of small planets with well-measured radii, enabling robust bulk density estimates needed to test theories of planet formation and evolution.
Future missions: JWST, PLATO and TESS
The James Webb Space Telescope (JWST, launch scheduled for October 2018) will provide powerful capabilities to study transiting planets (Deming et al. 2011). All four of its infrared instruments (NIRCam, NIRISS, NIRSPEC, and MIRI) will attempt transit observations. While no other facility from ground or space will match the sensitivity of JWST in the infrared from 1 to 28 microns, JWST does not operate in the visible and was not designed for precision photometric stability. NIRCam, NIRISS, and NIRSPEC will provide a number of modes to obtain light curves as well as spectroscopic data from 1-5 microns to the stability enabled by the platform (currently unknown, but speculated to be 100 ppm as found by HST and Spitzer). MIRI will enable studies from 5-28 microns at a variety of resolving powers. The infrared opens a unique science by allowing secondary eclipse observations thereby revealing planetary spectra in thermal emission. JWST will be able to measure a handful of transit light curves at high spectral resolution (100 < R < 1500).
The design philosophy, wavelength range and observing strategy makes CHEOPS extremely complementary to JWST. The fact that they will be in operation at the same time (assuming a JWST launch in October 2018) is another reason to explore in detail synergistic science beyond the obvious such as broadband photometry providing radii versus detailed spectrophotometry providing chemical abundances.
Beyond JWST, CHEOPS will also provide exciting targets for any future space mission with IR spectroscopic capability.
At present, the list of potential future missions dedicated to exoplanets includes PLATO (ESA) and TESS (NASA):
PLATO, next ESA M3 launch slot, has significantly more ambitious science goals but with a number of overlaps. Not being able to point at a given location on the sky, PLATO will not follow-up known targets but rather conduct a blind search like Kepler. However, its larger sky coverage translates in targets that are, on average, significantly brighter than Kepler’s. Significant spectroscopic follow-up observations will be needed to obtain precise masses. Hence, PLATO does not make the same use of the large efforts already invested in spectroscopic surveys as does CHEOPS.
TESS is the next NASA EXPLORER mission, its launch expected by mid-2017. Obvious synergies may be expected with TESS, for example, CHEOPS can observe interesting targets identified by TESS to get more precise radii.
The next generation of extremely large telescopes is about to be built. In Europe the E-ELT (ESO) with its 39 m diameter will have spatial resolution and sensitivity sufficient to detect the faint signatures of small planets orbiting very nearby stars. High signal-to-noise high resolution IR spectroscopy could enable phase-modulated (relative) measurements of close-in planets to analyse their atmospheres. Both such capabilities would be complementary to the CHEOPS program goals.