Final Report: Deconvolution of Adaptive Optics Images of Titan, Neptune, and Uranus

Gibbard, S ; Marchis, F

Lawrence Livermore National Laboratory

关键词: Amplification;    99 General And Miscellaneous//Mathematics, Computing, And Information Science;    Asteroids;    Telescopes;    Targets;   

DOI  :  10.2172/15003348
RP-ID  :  UCRL-ID-151346
RP-ID  :  W-7405-ENG-48
RP-ID  :  15003348

美国|英语

来源: UNT Digital Library

【 摘 要 】

This project involved images of Titan, Neptune, and Uranus obtained using the 10-meter W.M. Keck II Telescope and its adaptive optics system. An adaptive optics system corrects for turbulence in the Earth's atmosphere by sampling the wavefront and applying a correction based on the distortion measured for a known source within the same isoplanatic patch as the science target (for example, a point source such as a star). Adaptive optics can achieve a 10-fold increase in resolution over that obtained by images without adaptive optics (for example, Saturn's largest moon Titan is unresolved without adaptive optics but at least 10 resolution elements can be obtained across the disk in Keck adaptive optics images). The adaptive optics correction for atmospheric turbulence is not perfect; a point source is converted to a diffraction-limited core surrounded by a ''halo''. This halo is roughly the size and shape of the uncorrected point spread function one would observe without adaptive optics. In order to enhance the sharpness of the Keck images it is necessary to apply a deconvolution algorithm to the data. Many such deconvolution algorithms exist such as maximum likelihood and maximum entropy. These algorithms suffer to various degrees from noise amplification and creation of artifacts near sharp edges (''ringing''). In order to deconvolve the Keck images I have applied an algorithm specifically developed for observations of planetary bodies, the myopic deconvolution algorithm MISTRAL (''Myopic Iterative STep-preserving Restoration ALgorithm'') (Conan et al. 1998, 2000). MISTRAL was developed by ONERA (Office National d'Etudes et de Recherches Aerospatiales) and has been extensively tested on simulated and real AO observations, including observations of Titan (Coustenis et al.2001), Io (Marchis et al.2002, 2001), and asteroids (Hestroffer et al.2001, Rosenberg et al.2001, Makhoul et al.2001). Compared to more classical methods, MISTRAL avoids noise amplification and ringing artifacts, and better restores the initial photometry (Conan et al.1998). MISTRAL uses a stochastic approach to finding the best image reconstruction, using information about the object and the PSF. The general conclusions from this deconvolution effort is that MISTRAL does an excellent job of enhancing the sharpness of the data and preserving the photometry. The continued use of this algorithm for deconvolution of adaptive optics data is strongly suggested.