Title of thesis
Extreme adaptive optics and focal plane speckle minimisation techniques : application to the SPHERE+/VLT instrument
Composition of the jury
- Jean-Pierre Véran (University of Victoria) - Rapporteur
- Julien Girard (STScI) - Rapporteur
- Sonia Fornasier (LIRA, Université Paris Cité) - Examiner
- Arielle Bertrou-Cantou (GMTO) - Examiner
- Norbert Hubin (ESO) - Examiner
- Raphaël Galicher (LIRA, Université Paris-Cité) - Examiner, thesis supervisor
- Fabrice Vidal (LIRA, CNRS) - Guest member, co-supervisor
Résumé
Directly detecting the light from an exoplanet enables spectroscopic characterisation of its atmosphere or surface. But imaging an exoplanet is an instrumental challenge for two reasons. The planet is very close to its host star on the celestial vault, at an angle of less than one second, and the star is much brighter than the planet, from a factor of ten thousand for the brightest planets to ten billion for an exo-Earth. Astronomers use a coronagraph, an optical component that blocks the light from the star but allows the light from the planet to pass through to the detector. To function optimally, a coronagraph requires a flat incident wave surface, in other words one without aberrations, which is not the case with terrestrial telescopes that observe through atmospheric turbulence. This is why ground-based exoplanet imagers are equipped with adaptive optics (AO) upstream of the coronagraph. The aim is to correct the aberrations introduced by the Earth’s atmosphere to minimise the residual stellar intensity in the coronagraphic image.
The SPHERE instrument, installed at the Very Large Telescope (VLT, Chile) since 2014, is one of the most powerful exoplanet imagers today, detecting planets a hundred thousand times fainter than their star at a distance of 300 milliarcseconds. Its AO system, called SAXO, is equipped with a Shack-Hartmann wave surface analyser (WSA) and a deformable mirror, operating in a closed loop at a maximum frequency of 1.38 kHz. To overcome the current limitations (ASO sensitivity and loop speed), the SPHERE upgrade project, called SPHERE+, plans to improve SAXO into SAXO+ by adding a second correction loop downstream of the first existing loop (SAXO). Currently in the design phase, the second loop is faster, up to 3 kHz, and uses a near-infrared pyramid-type ASO. Once in closed loop, the performance of SAXO+ will be limited by the differential aberrations between the pyramid analysis channel and the coronagraph channel. In the final image, these aberrations show up as speckles. The aim is to correct these bright spots, which are similar to the image of an exoplanet.
During my thesis, I carried out numerical simulations for the SPHERE+ instrument. I used the end-to-end simulation tool Compass, developed at LIRA, for which I implemented a module for calculating coronagraphic images. In the first part of my thesis, I launched a campaign of simulations exploring the important parameters of SAXO+, which consists of two asynchronous loops working in different spectral bands. I define the optimal parameters as those that minimise the residual stellar intensity in the coronagraphic images. I show that SAXO+ improves the performance of the system by at least a factor of 10 compared with SAXO. For different magnitudes and atmospheric conditions, I give the optimum frequencies for each loop and the optimum gain for the first loop. I also quantify the impact of the modulation radius of the pyramid on the performance of the AO. In the second part of the thesis, I explore two speckle correction methods. The first is phase compensation of differential aberrations in the pupil plane. The second method is a third correction loop, measuring the static electric field in the coronagraphic images and retroacting on the AO loop to minimise speckle intensity. In each case, the pyramid ASO measurement must be biased to add a static phase to the coronagraph channel. I then implement a method of calibrating the optical gains of the pyramid to quantify its non-linearities. I show that each of the two speckle suppression methods significantly improves the performance of SAXO+ and I conclude on the benefits or otherwise of using optical gains depending on atmospheric conditions.
