Abstract
We describe an online method to estimate the wavefront outer scale profile, L0(h), for very large and future extremely large telescopes. The stratified information on this parameter impacts the estimation of the main turbulence parameters [turbulence strength, Cn2(h); Fried's parameter, r0; isoplanatic angle, ?0; and coherence time, t0) and determines the performance of wide-field adaptive optics (AO) systems. This technique estimates L0(h) using data from the AO loop available at the facility instruments by constructing the cross-correlation functions of the slopes between two or more wavefront sensors, which are later fitted to a linear combination of the simulated theoretical layers having different altitudes and outer scale values. We analyse some limitations found in the estimation process: (i) its insensitivity to large values of L0(h) as the telescope becomes blind to outer scales larger than its diameter; (ii) the maximum number of observable layers given the limited number of independent inputs that the cross-correlation functions provide and (iii) the minimum length of data required for a satisfactory convergence of the turbulence parameters without breaking the assumption of statistical stationarity of the turbulence. The method is applied to the Gemini South multiconjugate AO system that comprises five wavefront sensors and two deformable mirrors. Statistics of L0(h) at Cerro Pachón from data acquired during 3 yr of campaigns show interesting resemblance to other independent results in the literature. A final analysis suggests that the impact of error sources will be substantially reduced in instruments of the next generation of giant telescopes.

Abstract
Enhancement and wise archiving of astronomical images require an accurate estimate of the observational Point Spread Function (PSF). Although modelling of the telescope and its optics is a well-understood problem, PSF reconstruction becomes challenging when the observations include adaptive optics (AO) correction. The approach presented in this paper consists in feeding an end-to-end (E2E) simulation of the telescope, the instrument and its environment with the characterized disturbances from the telemetry and AO loop data, in order to produce the estimated PSFs. This method benefits from the developments made in the last years with respect to the estimation of external disturbances during AO correction, such as turbulence profile and its dynamics as well as sensor noise and vibrations characteristics. In particular, characterization of the turbulence profile in terms of strength, C2n (h), and outer scale, L0(h), is considered with an example on on-sky recorded AO telemetry from the GALACSI AO system. Once identified, the internal and external parameters of the observing conditions are used as inputs to carry out E2E simulations of the optical propagation and estimate the PSF. The method can be regarded as a "brute force" approach, as it is highly intensive in computer power; particularly for the ELTs. However, its ability to integrate complex combination of effects from all disturbances and not relying on analytical approximations for the aliasing or fitting errors makes the approach worth of a deeper study. E2E simulations have been used before in PSF reconstruction, but limited to a theoretical modelling of the system. Here, the development of the E2E simulation part is an ongoing work. A simplified AO system similar to the GALACSI WFM is currently simulated to obtain the PSF estimates and illustrate how such approach allows to account for the anisoplanatic effects and for the inuence of the outer scale values.

Abstract
The scientific potential of the ELT will rely on the performance of its AO systems that will require to be perfectly calibrated before and during the operations. The actual design of the ELT will provide a constraining environment for the calibration and new strategies have to be developed to overcome these constraints. This will be particularly true concerning the Interaction Matrix of the system with no calibration source upward M4 and moving elements in the telescope. After a brief presentation of the ELT specificities for the calibration, this communication focuses on the different strategies that have already been developed to get/measure the Interaction Matrix of the system, either based on synthetic models or using on-sky measurements. First tests of these methods have been done using numerical simulations for a simple AO system and a proposition for a calibration strategy of the ELT will be presented.

Abstract
The Pyramid wave-front sensor (WFS) is currently the baseline for several future adaptive optics (AO) systems, including the First light systems planned for the era of Extremely Large Telescopes (ELTs). Extensive investiga-tion into the Pyramid WFS aim to prepare for this new generation of AO systems, characterizing its behavior under realistic conditions and developing experimental procedures to optimise performance. An AO bench at Laboratoire d'Astrophysique de Marseille has been developed to analyze the behavior of the Pyramid and develop the necessary operational and calibration routines to optimize performance. The test bench comprises a Pyramid WFS, an ALPAO 9-9 deformable mirror (DM), a rotating phase screen to simulate atmospheric turbulence and imaging camera. The Pyramid WFS utilizes the low noise OCAM2 camera to image the four pupils and real time control is realized using the adaptive optics simulation software OOMAO (Object Oriented Matlab Adaptive Optics toolbox).1 Here we present the latest experimental results from the Pyramid test bench, including comparison with current Pyramid models and AO simulations. We focus on the calibration of the AO system and testing the impact of non-linear effects on the performance of the Pyramid. The results demonstrate good agreement with our current models, in particular with the addition of more realistic elements: non-common path aberrations and the optical quality of the Pyramid prism.

Abstract
In nights at Cerro Paranal where good seeing and low wind conditions are present, the PSF delivered to the focal plane of the SPHERE instrument has been shown to have significant errors, and have been aptly described as the Low Wind Effect' (LWE). We demonstrate here a method to quantify the LWE using experimental and on-sky data. We find single image phase diversity is a useful tool in quantifying the LWE and can be used to monitor this effect over the course of the night.

Abstract
Minimum-variance control of adaptive optics (AO) systems relies on a stochastic dynamical model of the per-turbation and on models of the components, including loop delays. Resulting LQG controllers have been imple-mented in SCAO and WFAO both on laboratory benches and on-sky. Their efficiency has been recognized in several modes of operation, e.g. I) on-sky control of TT or low-order modes with vibration mitigation (SPHERE, GPI, CANARY, Raven, GeMS, in H2 formulation at the McMath-Pierce solar telescope) ii) full SCAO mode (CANARY) and MOAO mode (CANARY, Raven) and iv) in general it is advocated to control the low-order modes in laser tomography systems (E-ELT HARMONI LTAO, NFIRAOS). We first point out two examples related to VLT AO controllers to illustrate the need for RTC exibility. The implementation of LQG control in the framework of the future ELTs raises many questions related both to real-time control computation and associated parameter updates (at a far lower rate), and to the performance that can be reached compared with simpler control strategies. By gathering many lab and on-sky results, we draw the performance trends observed so far. We then outline some promising research directions for control design and implementations for future ELTs AO systems.