It was only 17 years ago that the first planet outside of our own solar system was detected in the form of 51 Pegasi b. This planet is unlike anything in our own solar system. In fact, this planet was the first representative of a class of planets later known as “hot Jupiters”– gas giants that orbit very close to their parent star.
Since then several hundreds of exoplanets have been discovered – the vast majority using the radial velocity method, as with 51 Pegasi b, or the transit methods, for
instance in the form of the Kepler mission. Both of these two methods have very significant biases, i.e. it is much easier to detect high mass planets in close orbits.
With these two methods it is hard to detect planets in an exo-solar system with a structure similar to our own solar system; specifically, it is hard to detect Earth-like planets in Earth-like orbits. It is presently unknown how common such planets are in our galaxy.
There are a few other known methods for detecting exoplanets which have very different bias patterns. This thesis has been divided into two parts, treating two of
these other methods. Part I is dedicated to the method of gravitational microlensing, a method that utilises the lensing effect of light bend in the gravitational of stars
to detect perturbations in said gravitational field, which can be caused by bound planets. So far the discovery of 16 exoplanets detected with gravitational microlensing have been published.
The discovery rate with this method is low because of the lack of dedicated resources for this method, but this will change in the near future with the completion of several global telescope networks like SONG, Korean Microlensing Telescope Network (KMTNet) and the Las Cumbres Global Telescope network.
The gravitational microlensing method is also hampered by biases and degeneracies in the detections, but the method is in fact most sensitive to planets in the so-called
lensing zone located approximately one AU from the parent star in the typical lensing configuration.
To first order, the mass of a detected planet with this method is translated to the time duration of the signal, not the signal amplitude, rendering it critical to sample
ongoing events very densely in time to detect Earth-mass planets. The lower limit of planet mass that will give rise to a signal is set by the angular size of the source
which illuminates the lensing system. It can be shown that in the crowded fields where microlensing is observed, the primary obstacle for detecting Earth-mass planets is the crowding, rendering it hard to extract accurate photometry from faint sources at seeing limited resolutions. As all the sources tend to be at approximately the same distance, namely in the center of the galaxy, there is a direct relation between the magnitude of the source and its angular diameter.
To detect Earth-mass planets it is therefore crucial to increase the resolution when observing microlensing. Because of the dense sampling required it is infeasible to
utilise the traditional, but very expensive, methods for improving resolution, i.e. adaptive optics and space telescopes.
One appealing solution is the technique of “lucky imaging”, made feasible by the advent of the electron multiplying CCD, or EMCCD. This type of CCD is not limited
by readout noise because of cascade amplification of the photo electrons, but this type of amplification adds other sources of noise that need to be addressed; specifically, the amplification adds extra shot noise.
This thesis presents the method of gravitational microlensing and the technique of lucky imaging. A detailed analysis of the noise in a typical EMCCD is presented along with two methods for recovering the traditional shot noise
limit.
Furthermore, the photometric capabilities of the EMCCD and the lucky imaging technique is assessed. The improvement in resolution from the lucky imaging technique has been discussed extensively, but the photometric ability and stability has not been addressed in the literature. To this end, modifications to the traditional procedure for reducing CCD images is presented along with a discussion of how to optimally utilise lucky imaging, specifically for microlensing observations, by combining it with difference image analysis (DIA).
Part II is dedicated to the transit timing variation (TTV) method. In a transiting system additional planets can be revealed via their gravitational effects on the transiting
planet. This would result in telltale systematic deviations in the mid-transit times from a linear ephemeris. We present 11 high-precision photometric transit observations
of the transiting super-Earth planet GJ1214 b. Combining these data with observations from other authors, we investigate the ephemeris for possible signs of TTV using a Bayesian approach.