Postdoc position in exoplanets & statistics

Applications are invited for a 2(+1)-year Postdoctoral position in the exoplanet group at the Department of Astronomy of the University of Geneva, to start Fall 2020. Candidates from underrepresented groups in Astronomy are strongly encouraged to apply. All materials should be received by February the 20th for full consideration.

The successful applicant will work as a member of the SCORE research team led by Prof. Xavier Dumusque, and will join the vibrant exoplanet group at the department of astronomy, composed of about 50 PhD students and scientists, specialised in radial-velocity measurements, transit photometry, exoplanet atmosphere characterization, direct imaging as well as instrumentation. The exoplanet group is also part of the Switzerland-wide network PlanetS (http://nccr-planets.ch/), specialised in all different fields of planetary sciences with a large focus on exoplanets.

Overview
The ERC-funded SCORE project (Signal Correction to Reveal other Earths) aims at probing the main limitations to the detection of Earth-like planets using extreme precision radial-velocity measurements. The SCORE project relies on the unprecedented data obtained by two solar telescopes, connected to the HARPS and HARPS-N spectrographs, which allow to obtain sub-meter-per-second radial-velocity measurements of the Sun every possible day. Analysing such data is crucial to understand and find correction techniques to mitigate the impact of the different astrophysical and instrumental signals perturbing radial-velocity measurements and preventing the detection of other Earths. Any new technique developed in this framework will then be tested on the extensive RV data sets available in Geneva, coming from in-house instruments like CORALIE, HARPS, HARPS-N, ESPRESSO and NIRPS

Postdoc project: Time-series analysis of solar spectra and radial-velocities
With more than 200’000 solar spectra from HARPS and HARPS-N, it becomes possible to use robust statistical and machine learning techniques to extract and separate the complex stellar signal observed in radial-velocity measurements from tiny planetary signals. The goal of the selected candidate will be to develop techniques based on neural networks, PCA, ICA or any other statistical or relevant machine learning techniques to analyse time-series of spectra and model stellar signals. The Postdoc will also investigate the use of Gaussian Processes, autoregressive (integrated) moving average or any other relevant modelling techniques to mitigate the correlated noise induced by stellar activity.

Candidates will be evaluated on:
1) experience with statistical and machine learning algorithms,
2) experience with time-series analysis, 
3) experience with high-resolution spectroscopy and
4) commitment to fostering an inclusive research environment.

The duration of the appointment is initially for two years, with a possible prolongation for a third year depending on performance and available funding. Interested candidates should submit via email to [email protected], in a single PDF, the following material:
1) a cover letter that summarises the candidate's experience in the different evaluation criteria listed above, as well as the candidate’s motivation to pursue a Postdoc (no more than two pages)
2) a CV with publications
3) three recommendation letters to be sent directly to [email protected].