The need for high resolution optical imaging

Ever since Galileo first pointed a simple refracting telescope at the heavens in 1609 and resolved the Jovian system, astronomers have wished for higher and higher resolution imaging instruments. Early telescopes were limited by the accuracy with which large lenses could be figured. The development of reflecting telescopes by James Gregory and Isaac Newton lead to a rapid increase in the resolution available to astronomers. With the work of Thomas Young in the 19th Century, astronomers realised that the resolution of their telescopes was limited by the finite diameter of the mirror used. This limit was set by the wave properties of light and meant that large, accurately figured mirrors would be required in order to obtain higher resolution. Well figured telescopes with larger aperture diameters were constructed, but the improvement in resolution was not as great as had been expected. The resolution which could be obtained varied with the atmospheric conditions, and it was soon realised that Earth's atmosphere was degrading the image quality obtained through these telescopes.

For much of the 20th Century, the blurring effect of the atmosphere (known as atmospheric ``seeing'') limited the resolution available to optical astronomers. This degradation in image quality results from fluctuations in the refractive index of air as a function of position above the telescope. The image of an unresolved (i.e. essentially point-like) star is turned into a dancing pattern of ``speckles''. An example short exposure image from such a pattern is shown in Figure 1.1. In order to obtain better atmospheric seeing conditions, telescopes were constructed at high altitudes on sites where the air above the telescope was particularly stable. Even at the best observatory sites the atmospheric seeing conditions typically limit the resolution which can be achieved with conventional astronomical imaging to about $0.5$ $arcseconds$ ($0.5$ $as$) at visible wavelengths.

Figure 1.1: A K-band $140$ $ms$ exposure image obtained at the $10$ $m$ Keck I telescope showing a typical speckle pattern produced by atmospheric seeing. The image is plotted using a negative greyscale to highlight the fainter features. The pixel scale of $0.0206$ $as$ $pixel^{-1}$ was set by the Keck facility Near Infra-Red Camera (NIRC) instrument. This image is taken from data kindly provided by Peter Tuthill.

Studies of short exposure images obtained through atmospheric seeing by Antoine Labeyrie in 1970 (Labeyrie, 1970) indicated that information about the high resolution structure of an astronomical object could be obtained from these short exposures despite the perturbing influence of the atmosphere. A number of imaging techniques were developed based on his approach, most involving fast frame-rate cameras (essentially high performance motion picture or video cameras) situated at the telescope focus. This thesis discusses one of these techniques in detail, that of Lucky Exposures . The Lucky Exposures method was first discussed in depth by David Fried in 1978 (Fried, 1978), and the first experimental results followed in the 1980s. The optimum performance for the technique was not achieved during those observations, partly due to the camera equipment available at the time and partly due to the approach used for the data analysis. This thesis presents more recent results which demonstrate the enormous potential of the technique.

The effects of atmospheric seeing are qualitatively similar throughout the visible and near infra-red wavebands. At large telescopes the long exposure image resolution is generally slightly higher at longer wavelengths, and the timescale for the changes in the dancing speckle patterns is substantially lower. This would argue for the use of long wavelengths in experimental studies of these speckle patterns (although short wavelengths are of equal astronomical interest). The high cost of sensitive imaging detectors which operate at wavelengths longer than $\sim 1$ $\mu m$ makes them less appealing for studies of imaging performance, so the results presented in later chapters of this thesis will be restricted to wavelengths shorter than $\sim 1$ $\mu m$. The cameras used for my work are sufficiently fast to accurately sample the atmosphere at the wavelengths used. The approaches developed in this thesis could equally be applied to longer wavelengths given suitable detectors and telescopes, broadening the astronomical potential of the method substantially.