In this thesis, I present a new technique for measuring the optical aberrations produced by a telescope, with an eye towards future use of these aberration measurements to align wide-field telescopes. This method determines the aberrations by simultaneously fitting a pair of oppositely defocused images to a mostly analytic model. I develop the model and describe its software implementation in detail, and then report on the results of tests with simulated and real data. This technique is able to extract the aberrations from simulated data rapidly and accurately, and it has been used with mixed success to analyze data from the VISTA telescope. With the VISTA data, the algorithm is unable to match small-scale brightness variations in the images. However, it was able to determine aberrations with median accuracies of 0.08 um for coma, 0.08 um for astigmatism, 0.9 um for tilt, and 0.3 um for defocus. It was also quite fast, with an average of 34 iterations until convergence.
展开▼
