Evolution of Ly clouds: observational biases due to line crowding
X. D. Liu, Bernard J. T. Jones · Monthly Notices of the Royal Astronomical Society · 1988
The intrinsic properties and the evolution with redshift of Lyman-α absorption line clouds is estimated by counting lines as a function of redshift. We simulate spectra, with and without photon counting noise, in order to see what biases there are in deriving cloud properties. The line samples and the line counts are biased by the crowding of the lines. The bias results in an underestimate of the evolutionary rate exponent γ, |$\frac{dN}{dz}\approx N_0(1+z)^\gamma$| Noise in the data increases this bias. We believe that the actual value is at least as large as γ ≈ 2, and may be as large as γ ≈ 3. The systemic error in determining γ could certainly be larger than the quoted formal errors. The small size of the line sample is a major factor in the uncertainty in determining γ and this can only be overcome by combining data from many lines-of-sight. An exponential distribution exp (|$-W/W^\ast_0$|) of intrinsic line equivalent widths W produces an apparent plateau in the distribution of inferred column densities. Both the crowding bias and the noise enhance this redshift dependent plateau and results in an overestimate of the scale of the distribution |$W^\ast_0$|.