Thesis

Advanced statistical data analysis methods for the detection of other Earths

• Call:

  IDPASC Portugal - PHD Programme 2017

• Academic Year:

  2017 / 2018

• Domain:

  Astrophysics

• Supervisor:

  Pedro Viana

• Co-Supervisor:

  Nuno Santos

• Institution:

  Universidade do Porto

• Host Institution:

  Institute of Astrophysics and Space Sciences

• Abstract:

  Advanced statistical data analysis methods will be used to detect and characterize exoplanets through the radial velocity technique, in particular planets with Earth-like mass and orbital period, moving around a star similar to the Sun. This will require the identification of planetary-induced radial velocity signals that could be an order of magnitude lower in amplitude than confounding signals, namely those produced by stellar activity, which is beyond what is achievable by existing algorithms. For such purpose, deep learning methods based on Gaussian Processes will be used to disentangle the effects on stellar spectra of orbital planetary motion, stellar activity and instrumental/telluric effects, by jointly taking into account the information contained in any number of time-series, extracted from spectroscopic and photometric data. This project is essential for the full use of the capabilities of third-generation spectrographs like ESPRESSO@VLT, on which Portugal has heavily invested and which will become operational in 2018.