Multi-Conjugate Adaptive Optics (MCAO) holds the promise of moderate to large adaptively compensated field of view with uniform image quality. This paper is a first effort to analyse the fundamental limitations of such systems, and that are mainly related to the finite number of deformable mirrors and guide stars. We demonstrate that the ultimate limitation is due to the vertical discretization of the correction. This effect becomes more severe quite rapidly with increasing compensated field of view or decreasing wavelength, but does not depend at first order on the telescope aperture. We also discuss limitations associated with the use of laser guide stars and ELT related issues.

The Gemini Multi-conjugate adaptive optics System (GeMS) at the Gemini South telescope in Cerro Pachón is the first sodium-based multi-Laser Guide Star (LGS) adaptive optics system. It uses five LGSs and two deformable mirrors to measure and compensate for atmospheric distortions. The GeMS project started in 1999, and saw first light in 2011. It is now in regular operation, producing images close to the diffraction limit in the near infrared, with uniform quality over a field of view of two square arcminutes. The present paper (I) is the first one in a two-paper review of GeMS. It describes the system, explains why and how it was built, discusses the design choices and trade-offs, and presents the main issues encountered during the course of the project. Finally, we briefly present the results of the system first light.

A modified method for maximum-likelihood deconvolution of astronomical adaptive optics images is presented. By parametrizing the anisoplanatic character of the point-spread function (PSF), a simultaneous optimization of the spatially variant PSF and the deconvolved image can be performed. In the ideal case of perfect information, it is shown that the algorithm is able to perfectly cancel the adverse effects of anisoplanatism down to the level of numerical precision. Exploring two different modes of deconvolution (using object bases of pixel values or stellar field parameters), we then quantify the performance of the algorithm in the presence of Poissonian noise for crowded and noncrowded stellar fields.

This paper focuses on two main categories of the multiconjugate adaptive optics (MCAO) parameter space for performance optimization: the geometrical configuration of guide stars and deformable mirrors (DMs) , and the wavefront reconstructors. From the first category it is shown how, for a fixed reconstructor and imaging wavelength, the performance metrics with a few important exceptions improve with an increasing number of i) DMs, ii) actuators per DM and iii) guide stars. The metrics are seen to degrade with i) an increasing field of compensation and ii) DM conjugation altitude mismatch with the significant turbulent atmospheric layers. In the second category, this study also compares the performance with a fixed MCAO configuration using the least-square estimator (LSE) and the maximum a posteriori estimator (MAP) for wavefront reconstruction. The MAP is shown to perform significantly better than the LSE at low or intermediate signal-to-noise ratios (SNRs) , and somewhat better even in the absence of noise due to its a priori knowledge of the phase statistics.

The way in which the sequence of phase correction impinges on performance in multiconjugate adaptive optics systems is described. When multiple phase modulators with different conjugate ranges are used, the conjugate images of the phase modulators in the atmosphere must be reimaged in the reverse order post focus, and adaptive phase correction applied in this sequence for perfect amplitude and phase cancellation. Performing the correction without relay optics results in residual amplitude and phase aberrations. It is shown, by Monte Carlo simulations of Fresnel propagation, that the effects of wave optical propagation become nonnegligible at visible wavelengths and for large air masses.

In this paper, we provide an overview of the adaptive optics (AO) program for the Thirty Meter Telescope (TMT) project, including an update on requirements; the philosophical approach to developing an overall AO system architecture; the recently completed conceptual designs for facility and instrument AO systems; anticipated first light capabilities and upgrade options; and the hardware, software, and controls interfaces with the remainder of the observatory. Supporting work in AO component development, lab and field tests, and simulation and analysis is also discussed. Further detail on all of these subjects may be found in additional papers in this conference.

Abstract.A method is presented for estimating the long exposure point spread function (PSF) degradation due to tilt anisoplanatism in a laser-guide-star-based multiconjugate adaptive optics systems from control loop data. The algorithm is tested in numerical Monte Carlo simulations of the separately driven low-order null-mode system, and is shown to be robust and accurate with less than 10% relative error in both H and K bands down to a natural guide star (NGS) magnitude of m R = 21, for a symmetric asterism with three NGS on a 30 arcsec radius. The H band limiting magnitude of the null-mode system due to NGS signal-to-noise ratio and servo-lag was estimated previously to m R = 19. At this magnitude, the relative errors in the reconstructed PSF and Strehl are here found to be less than 5% and 1%, suggesting that the PSF retrieval algorithm will be applicable and reliable for the full range of operating conditions of the null-mode system.

A data set of adaptive optics images has been analyzed in order to study the effects of angular anisoplanatism and to characterize Mauna Kea nighttime turbulence. The data set consists of a selection of Galactic center images obtained with the adaptive optics instrument Hokupa'a and the near-infrared camera QUIRC on the Gemini-North 8 m telescope. Using the Strehl ratio and the FWHM as tracers for anisoplanatism in the images, it is possible to draw conclusions about the effective turbulence height by model-fitting synthetic h data. From the relatively small statistical sample, we obtain a median km, with 10th and 90th percentiles h p 3.5 being 2.2 and 5.6 km. The implications of these findings for a turbulence-conjugated adaptive optics system are addressed. We also find the Zernike modal equivalent of Hokupa'a to be , i.e., on the average slightly N ≈ 18 better than 4 radial degrees.