Small‐Scale Structure in the Lyα Forest at High Redshift

Abstract

We have obtained spectra of the four components of the system B1422+231 with HST's Faint Object Spectrograph in order to study spatial inhomogeneities and velocity variations in the absorbers which are pierced by the four lines of sight. The four images are well resolved, and the spectra clearly show the system is a lensed quasar; a new redshift based on the Lyα and N V emission lines is measured to be 3.6261. The following are results from the analysis of the four individual lines of sight that probe redshifts 2.9 ≤ z ≤ 3.5 and transverse dimensions 0-0.14 h -->−1100 kpc: (1) In the region between Lyβ and Lyα (less 6000 km s-1 to allow for the proximity effect: 4745-5511 A), the limiting rest equivalent width (1σlim) is measured to be 0.059, 0.052, 0.034, and 0.22 A for components A, B, C, and D, respectively. (2) The total number of 5σW lines found in this region are 83, 85, 102, and 44 for A-D, respectively; for A-C 11, 9, and 13 are identified with metal line systems and the remaining 72, 76, and 89 are assumed to be Lyα lines. Between paired lines of sight A and B to C, and A paired to B, there are 62, 68, and 62 coincidences and 3, 1, and 1, anticoincidences for the Lyα lines. For the metal lines there are 8, 9, and 8 coincidences for AC, BC, and AB, respectively, and each has one anticoincidence. (3) Scatter plots of the equivalent widths for the coincident lines for both Lyα and metals are highly correlated and all pairs match within 100 and 75 km s-1, respectively. (4) Maximum-likelihood analysis limits spatial inhomogeneity for Lyα absorbers in components A, B, and C to no more than ~15% of the equivalent width of an average (W = 2 A) line. This corresponds to a limit on the mean variation in column density, assuming unsaturated lines, of log (ΔNH) = 13.1 cm-2. (5) A method of detecting spatial inhomogeneities independent of continuum-fitting and line-profile measurement is developed yielding an upper limit on the column density difference for unsaturated lines of log (ΔNH) = 12.8. This is approximately a factor of 2 smaller than the limits obtained from the line pairs. (6) The distribution of the velocity differences for AC, BC, and AB for the Lyα lines and the metal line pairs have a mean close to zero and rms ~40 km s-1. Maximum-likelihood analysis gives an upper limit on the variations in velocity of ~40 km s-1 for Lyα absorbers in paired lines of sight. The overall result of this work is a set of stringent limits on the inhomogeneity of low column density hydrogen absorbers on subkiloparsec scales at high redshift.