Constraining the Variation of the Fine Structure Constant with Observations of Narrow Quasar Absorption Lines

Abstract

Attempts to measure the variability of the fine structure constant alpha over cosmological time, using spectra of high redshift quasars have produced conflicting results. We use the Many Multiplet (MM) method with Mg II and Fe II lines on very high signal-to-noise, high resolution (R = 72,000) Keck HIRES spectra of eight narrow quasar absorption systems and consider both systematic uncertainties in spectrograph wavelength calibration and also velocity offsets introduced by internal velocity structure. We find no significant change in alpha, Delta(alpha)/alpha =(0.43 +/- 0.34)x 10^(-5), in the redshift range z = 0.7 - 1.5 (statistical and systematic). Scatter in measurements of Delta(alpha)/alpha arising from absorption line structure can be considerably larger than assigned statistical errors even for apparently simple and narrow absorption systems. We find a null result of Delta(alpha)/alpha = (-0.59 +/- 0.55) x 10^(-5) in a system at z = 1.7382 using Cr II, Zn II and Mn II, whereas using Cr II and Zn II in a system at z = 1.6614 we find a systematic velocity trend which, interpreted as a shift in alpha, would imply Delta(alpha)/alpha = (1.88 +/- 0.47) x 10^(-5) (statistical plus systematic). This latter result is almost certainly caused by varying ionic abundances in subcomponents of the line. We conclude that spectroscopic measurements of quasar absorption lines are not yet capable of unambiguously detecting variation in alpha using the MM method.