Impact of instrumental systematic errors on fine-structure constant measurements with quasar spectra

We present a new supercalibration' technique for measuring systematic distortions in the wavelength scales of high resolution spectrographs. By comparing spectra of solar twin’ stars or asteroids with a reference laboratory solar spectrum, distortions in the standard thorium–argon calibration can be tracked with $\sim$10 m s${}^{-1}$ precision over the entire optical wavelength range on scales of both echelle orders ($\sim$50–100 \AA) and entire spectrographs arms ($\sim$1000–3000 \AA). Using archival spectra from the past 20 years we have probed the supercalibration history of the VLT–UVES and Keck–HIRES spectrographs. We find that systematic errors in their wavelength scales are ubiquitous and substantial, with long-range distortions varying between typically $\pm$200 m s${}^{-1}$ per 1000 \AA. We apply a simple model of these distortions to simulated spectra that characterize the large UVES and HIRES quasar samples which previously indicated possible evidence for cosmological variations in the fine-structure constant, $\alpha$. The spurious deviations in $\alpha$ produced by the model closely match important aspects of the VLT–UVES quasar results at all redshifts and partially explain the HIRES results, though not self-consistently at all redshifts. That is, the apparent ubiquity, size and general characteristics of the distortions are capable of significantly weakening the evidence for variations in $\alpha$ from quasar absorption lines.