New constraints on Lyman-α opacity with a sample of 62 quasars at z > 5.7

We present measurements of the mean and scatter of the IGM Lyman- opacity at 4. 9 5. 7, the largest sample assembled at these redshifts to date by a factor of two. The sample size enables us to sample cosmic variance at these redshifts more robustly than ever before. The spectra used here were obtained by the SDSS, DES-VHS and SHELLQs collaborations, drawn from the ESI and X-Shooter archives, reused from previous studies or observed specifically for this work. We measure the effective optical depth of Lyman- in bins of 10, 30, 50 and 70 cMpc h-1, construct cumulative distribution functions under two treatments of upper limits on flux and explore an empirical analytic fit to residual Lyman- transmission. We verify the consistency of our results with those of previous studies via bootstrap re-sampling and confirm the existence of tails towards high values in the opacity distributions, which may persist down to z = 5. 2. Comparing our results with predictions from cosmological simulations, we find further strong evidence against models that include a spatially uniform ionizing background and temperature-density relation. We also compare to IGM models that include either a fluctuating UVB dominated by rare quasars or temperature fluctuations due to patchy reionization. Although both models produce better agreement with the observations, neither fully captures the observed scatter in IGM opacity. Our sample of 62 z > 5. 7 quasar spectra opens many avenues for future study of the reionisation epoch.