Pulsar-Appropriate Clock Statistics

Abstract : Because pulsar phase, spin rate, and spin-down rate cannot be determined a priori, a pulsar's possible contribution to a time scale is best measured using a statistic insensitive to these quantities. One such statistic is sigma sub z (tau) (Matsakis, Taylor, and Eubanks 1996, hereafter [1]), which is based upon the third-order polynomial variations of the clock phases and named in analogy to sigma sub y (tau). It is similar to measures based upon the variance of the third differences, but more robust in the presence of data gaps. About 6 years of data, from MJD 47752 to 50343 (18 Aug 89-17 Sep 96), were used to determine the sigma sub z (tau) of cesium and maser standards in a variety of imperfect but probably acceptable ways; directions for improved reductions are also indicated. Because of the considerably improved stability of cavity-tuned masers and Hewlett-Packard Model 5071 cesium standards, the best results use only the data taken since MJD 49400 (17 Feb 94), by which time most of the new standards had been acquired. An upper limit to the time variation of the fine structure constant is also given.