Reduction of adaptive optics images

Adaptive optics (AO) now provides superb, diffraction limited images from ground-based telescopes. AO involves a very complex technology, but, in modern systems such as PUEO at CFHT, this complexity is hidden behind easy, user friendly interfaces, making the AO observations simple and efficient. The data reduction process, however, is quite difficult. The main reason is that the AO correction is always only partial, so that even if the AO point spread function (PSF) has a central core whose width is usually limited by the diffraction of the collecting aperture, a significant fraction of the light remains scattered far away from this central core. The presence of this large halo makes the image difficult to interpret qualitatively and quantitatively, unless proper data processing, i.e. deconvolution, is undertaken to remove it. This problem is made even worse by the fact that the AO PSF significantly varies in time las the observing conditions, e.g, seeing, change) and in space (anisoplanatism). The goal of this paper is to familiarize the reader with these AO specific data processing problems, to review the work that has been done in this area over the last few years, and to suggest ways to improve the scientific output of AO. The latter involves a careful preparation of the observations, an adequate data reduction toolbox, and the availability of AO specific data, which must be provided by the AO system and archived with each AO image.