Title: The primordial deuterium abundance at z = 2.504 from a high signal-to-noise spectrum of Q1009+2956

Abstract: The spectrum of the $z_{\rm em} = 2.63$ quasar Q1009+2956 has been observed extensively on the Keck telescope. The Lyman limit absorption system $z_{\rm abs} = 2.504$ was previously used to measure D/H by Burles & Tytler using a spectrum with signal to noise approximately 60 per pixel in the continuum near Ly {\alpha} at $z_{\rm abs} = 2.504$. The larger dataset now available combines to form an exceptionally high signal to noise spectrum, around 147 per pixel. Several heavy element absorption lines are detected in this LLS, providing strong constraints on the kinematic structure. We explore a suite of absorption system models and find that the deuterium feature is likely to be contaminated by weak interloping Ly {\alpha} absorption from a low column density H I cloud, reducing the expected D/H precision. We find D/H = $2.48^{+0.41}_{-0.35}\times10^{-5}$ for this system. Combining this new measurement with others from the literature and applying the method of Least Trimmed Squares to a statistical sample of 15 D/H measurements results in a "reliable" sample of 13 values. This sample yields a primordial deuterium abundance of (D/H)$_{\rm p} = (2.545 \pm 0.025)\times10^{-5}$. The corresponding mean baryonic density of the Universe is $\Omega_{\rm b}h^2 = 0.02174\pm0.00025$. The quasar absorption data is of the same precision as, and marginally inconsistent with, the 2015 CMB Planck (TT+lowP+lensing) measurement, $\Omega_{\rm b}h^2 = 0.02226\pm0.00023$. Further quasar and more precise nuclear data are required to establish whether this is a random fluctuation.