Does the Fine Structure Constant Vary? a Detailed Investigation into Systematic Effects

We have previously presented 5.7 σ evidence for a smaller α at redshifts 0.2 < zabs < 3.7 from a sample of 128 Keck/HIRES quasar absorption systems: α/α = (−0.57 ± 0.10) × 10−5. A non-zero α/α manifests itself as a distinct pattern of shifts in the measured absorption line wavelengths. The statistical error is now small: we do detect small line shifts in the HIRES data. Whether these shifts are due to systematic errors or due to real variation in α is now the central question. Here we summarize the two potentially most important systematic effects: atmospheric dispersion and isotopic abundance evolution. Previously, these have been difficult to quantify/model but here we find that neither of them can explain our results. Furthermore, the HIRES spectra themselves contain no evidence for these effects. Independent measurements of α/α with a different telescope and spectrograph are now crucial if we are to rule out or confirm the present evidence for a variable α.