During my PhD, I worked as part of the SPECULOOS (Search for habitable Planets EClipsing ULtra-cOOl Stars) team to optimise the detection potential of Earth-sized, habitable-zone planets transiting ultra-cool dwarfs (UCDs) from the ground. These cool, red dwarfs are small, abundant, and predicted to host a wealth of terrestrial planets. Within our current detection capabilities, these planets are also the most favourable candidates to search for traces of life. Our main goal is to provide the most promising candidates for future atmospheric characterisation by the James Webb Space Telescope (JWST). For this thesis, I focused on removing both the astrophysical and non-astrophysical sources that limit the detection of rocky planets and on constraining these planets’ frequency.

Firstly, I developed the SSO Pipeline, a photometric pipeline specific for the SPECULOOS Project and the ultra-cool objects that it observes. This automated pipeline carries out image reduction, cross-matches with preexisting stellar catalogues, and performs precise aperture and differential photometry every night. With the SSO Pipeline, I address both the instrumental and atmospheric contamination that can prevent detecting small planets. I tackle the first and second-order effects of the Earth’s rapidly-varying atmosphere on photometric observations through specialised photometric treatment. Specifically, this pipeline removes the photometric impact of water vapour absorption by implementing a novel, first-principles correction. I optimise the performance of this pipeline through rigorous quality checking and demonstrate SPECULOOS’s unprecedented precision for UCDs. I showed that SPECULOOS is reaching its survey goals in regularly achieving the required precisions to detect temperate, terrestrial planets.

Secondly, due to the enhanced stellar activity of UCDs, accounting for photometric variability is an essential step in detecting rocky planets and in understanding whether these planets can initiate and sustain life. My work assesses whether this activity is beneficial or detrimental to the habitability of planetary systems around UCDs, and probes their complex magnetic behaviour. In performing a flaring and rotation study of SPECULOOS’s targets, I found that the activity of the coolest stars would be too low to initiate abiogenesis or to sustain an Earth-like biosphere on their planets. By studying the activity of these targets, I mitigate the astrophysical sources of contamination in our lightcurves.

Finally, I established bounds on the unexplored planetary populations around UCDs, by quantifying the detection potential of SPECULOOS for planets like TRAPPIST-1b. I developed a transit-search pipeline and performed transit injection-recovery tests to measure its detection efficiency. I concluded that worlds like TRAPPIST-1b, very short-period, rocky planets, are rare for these stellar hosts; otherwise, we would have already detected a small number. Several more years of SPECULOOS observations will be needed to confirm these results, and potentially, to find another TRAPPIST-1 system.