Dr Guillem Anglade-Escude - Project Abstracts
BSc Project Abstracts
Confident detection of super-Earths around nearby stars using Doppler spectroscopy
In the last 20 years, most extra-solar planet candidates have been found using Doppler spectroscopy. After years of technical improvements, Doppler precision is now sufficient to detect planets down to a few Earth-masses in the habitable zone of nearby stars. Confident detection of such small signals requires a good understanding of all sources of noise (random, instrumental, activity induced) and how to include them in the model of the data. The performance of Bayesian, spectral and other data analysis methods will be compared using real and simulated data-sets. This project involves writing and running codes to i) analyze new and archival spectra from the HARPS spectrograph, ii) perform Bayesian Monte Carlo Markov Chain analyses of time-series.
Prerequisites : interest in data-analysis and programming (basic programming training will be provided).
MSci Review Project Abstracts
Searching small exoplanets in publicly available data
Several missions and programs have accumulated many years of high- precision observations. This includes space-based photometry (e.g., Kepler) and ground based spectroscopic measurements (e.g. HARPS spectrograph). Many of these datasets need to be carefully re-analysed to detect small signals that might be missed by automatic search algorithms. When a signal is spotted, several tests muct be applied to verify its statistical significance and likelihood. We will go through the complete process of exo-planet detection on selected targets and fully characterize new planetary systems if found.
Prerequisites : interest in programming and data analysis (some training will be provided).
MSci Research/Investigative Project Abstracts
Measuring accurate mass-radius of super-Earths detected by Kepler
Masses of transiting extra-solar planets can be measured by analyzing the instants of transits in multiplanet systems. The mutual gravitational interactions produce transit-timing variations that encode the masses of the bodies. While most masses from the Kepler mission have been estimated using this method, it is known that non-trivial correlations and the existence of multiple solutions satisfying the data can produce strongly biased results, producing density estimates that are likely unphysical (<0.05 g/cm^3 compared to 1.0 g/cm3 of water). In this project we will use Bayesian Monte Carlo methods combined with numerical integration of planetary orbits to properly explore the space of allowed solutions and produce uniformly derived mass-radius measurements of Earth/super-Earth sized objects detected by the Kepler mission.
Prerequisites : interest in planetary system dynamics and numerical methods.
