Conclusions

In this chapter Fourier filters were developed in order to suppress the noise in the short exposure images. These improved the image FWHM which could be obtained using faint reference stars, and helped to slightly suppress the noise in the final Lucky Exposures images.

The images obtained by applying the Lucky Exposures method to data taken with the low noise CCDs were generally of high quality. It is clear that the Lucky Exposures method can provide a substantial improvement in resolution over conventional imaging.

Measurements of the isoplanatic angle on the night of 2001 July 26 showed it to be $\sim30$ $as$. This is substantially higher than the typical value achieved for I-band adaptive optics. Analysis of further observations will be required in order to determine whether this result is representative of the summer seeing conditions at the NOT .

The Lucky Exposures method was found to work successfully using reference stars as faint as $I=15.9$. This magnitude limit is expected to be further increased with the use of back-illuminated CCDs and an anti-reflection coated dewar window.

Based on the limiting magnitude for the Lucky Exposures method, the isoplanatic angle measured on the night of 2001 July 26, and star counts from the literature we would expect the sky coverage for the Lucky Exposures technique to be about $25\%$ at I-band. This is substantially better than that achieved with natural guide star adaptive optics at this wavelength.

Analysis of data taken on M13 suggest that accurate astrometry will be possible in crowded fields if the plate scale and image distortions can be suitably determined and if the charge transfer efficiency of the detector is good. Some of the data presented shows evidence of problems with charge transfer efficiency, and it will be important to address this for future observing runs.