Previous Article Next Article Table of Contents
A Report on WIYN Image Quality (1Mar95) (from KPNO, NOAO Newsletter No. 41, 1 March 1995) Since June 1994, the WIYN Project staff has been making fairly regular delivered image quality (DIQ) measurements. Most of these measurements were made by our observing technicians, Alex Macdonald and Bridget Watts, but Dave Sawyer, the WIYN Site Manager, and Nick Roddier, the NOAO engineer with primary responsibility for the WIYN active optics control, also significantly contributed to this effort. A histogram of these measurements is presented below. [Figure not included] As the figure illustrates, the median WIYN DIQ between June 1994 and January 1995 was 0.7" FHWM. This is true whether all measurements are considered or only those with exposure times equal to or greater than 10 seconds (i.e., the hatched measurements). Furthermore, 24% (18/74) of the total dataset values and 18% (6/34) of the reduced dataset values are 0.6" FWHM or better. All measurements were made with a cooled science grade CCD with 0.2" pixels in the R-band. The fields imaged were typically at airmasses less than 1.05 (i.e., at zenith distances less than 25 degrees). For ease of use, image sizes were measured using the IRAF task IMEXAMINE. The measured FWHM was rounded up to the nearest tenth of an arcsec before being included in the figure. The WIYN DIQ was usually measured immediately after making a delivered wavefront measurement and tuning the WIYN active optics system, if the measured wavefront indicated that the latter was necessary. These measurements were not guided. In fact, many of the longer exposure images show evidence for slight tracking errors and small deviations from ideal focus. When tracking and focus is brought under better control, the median DIQ will probably be reduced, perhaps by as much as 0.1" FWHM. Qualitatively, the typical WIYN stellar image does not look purely Gaussian. Rather, it has a tighter core sitting on top of slightly broader wings, reminiscent of a Moffat function. This raises the interesting question: how does one quantify DIQ? Different measurement techniques give different answers. For example, the FWHM calculated by the IRAF task PSFMEASURE, now is typically 0.1" larger than the FWHM calculated by IMEXAMINE. On the other hand, a fitted Moffat function can have a FWHM which is 0.1" or smaller than the IMEXAMINE FWHM. This is demonstrated by the image published in the September 1994 NOAO Newsletter (p. 3). A Moffat function kindly fitted by Peter Stetson (DAO) had a mean FWHM of 0.45", which is the image size reported in the Newsletter. The PSF FWHM measured by IMEXAMINE, however, was 0.6". The WIYN Project will continue to use IMEXAMINE since it is easy to use and appears to deliver a conservative DIQ estimate. Over the next six months, the Project intends to study in detail how DIQ is affected by such things as primary mirror temperature and shape, dome venting, and telescope zenith distance. Dave Silva
Previous Article Next Article Table of Contents