Exploring the Neptune Desert with NGTS and Machine Learning

The sub-Jovian desert is marked by a scarcity of short period Neptune-sized exoplanets. While theoretical work has begun to explain the distinct triangular shape of the desert via migration and photo-evaporation (e.g. Owen+18), a robust set of well characterised planets both in and near the desert boundaries is still critical for such efforts. In this proposal, we aim to add new exoplanets to this sample using the Next Generation Transit Survey (NGTS). This will be done by improving the efficiency of the survey using machine learning techniques, targeting planets which no other survey is currently able to detect, and leveraging NGTS’s unique capabilities to follow-up promising candidates from the Transiting Exoplanet Survey Satellite (TESS) mission. Finally, we aim to improve our understanding of the desert by conducting a statistically rigorous study of the relative occurrence rates of different planets on its upper-boundary. The proposal will significantly advance our understanding of the mechanisms which form the desert, helping to shed light on how planets evolve over time.