Characterisation of Tomographic Adaptive Optics Performance in Realistic Atmospheric Conditions

Abstract

In ground based astronomy at optical and near infrared wavelengths, the Earth's turbulent atmosphere results in aberration of light before it reaches the telescope. Adaptive optics (AO) allows astronomers to partially correct this aberration, resulting in improved resolution and sensitivity. Tomographic AO techniques have been developed that enable correction over large fractions of the sky, or over a wide field of view. These techniques are employed on current world-leading telescopes and are a key part of the instrumentation for future Extremely Large Telescopes (ELTs). An important piece of information for any tomographic AO system is the vertical distribution (profile) of optical turbulence in the atmosphere. The profile is specific to an observing site and varies temporally on timescales from minutes to seasons.

This thesis is concerned with quantifying the effect of changing profile on tomographic AO by using a large dataset of turbulence profiles from ESO Paranal, Chile. Using a novel clustering method, we extract 18 reference profiles from this dataset. These reference profiles can be used in slow Monte Carlo simulation to give a realistic description of the variability of the profile at Paranal.

We validate these profiles alongside other common reference profiles using a fast Fourier tomographic AO simulation. We find that reference profiles composed of the average of many profiles give worse than expected performance in simulation. Using the same simulation and dataset we also assess the impact of sub-optimal tomographic reconstruction, where the profile is sampled with a small number of layers and is temporally evolving.

Finally we present simulation and on-sky measurements from a SCIDAR turbulence profiler operating on a relatively small 0.5~m aperture. These preliminary results show that a small telescope SCIDAR may be viable for site characterisation and monitoring in the context of ELT-scale tomographic AO.