[Paper Review] Evolution of the Lyman-alpha forest from high to low redshift

R. Riediger, Patrick Petitjean+1 more|arXiv (Cornell University)|Sep 14, 1997

Galaxies: Formation, Evolution, Phenomena3 citations

TL;DR

This study uses N-body simulations of a CDM model to investigate the redshift evolution of the Lyman-alpha forest from z=5 to z=0, assuming baryons trace dark matter. It finds that the number density of absorption lines with log N(HI) > 12 remains nearly constant from z=4 to z=1, and at z=0, the line densities are ~400, 100, and 20 per unit redshift for log N(HI) > 12, 13, and 14, respectively, indicating most forest lines are weak and not tightly correlated with galaxies.

ABSTRACT

We study the evolution with redshift, from z=5 to z=0, of the Lyman-alpha forest in a CDM model using numerical simulations including collisionless particles only. The baryonic component is assumed to follow the dark matter distribution. We distinguish between two populations of particles: Population P{s} traces the filamentary structures of the dark matter, evolves slowly with redshift and, for N(HI)>10^14 cm^-2, dominates the number density of lines at z<3; most of population P{u} is located in underdense regions and for the same column densities, disappears rapidly at high redshift. We generate synthetic spectra from the simulation and show that the redshift evolution of the Lyman-alpha forest (decrement, N(HI) distribution) is well reproduced over the whole redshift range for Omega_b*h^2=0.0125 and J_-21=0.1 at z=3 where J_-21 is the UV background flux intensity in units of 10^-21 erg cm^-2 s^-1 Hz^-1 sr^-1. The total number of lines with N(HI)>10^12 cm^-2 remains approximately constant from z=4 to z=1. At z=0, the number density of lines per unit redshift with log N(HI)>12, 13, 14 is of the order of 400, 100, and 20 respectively. Therefore, at low redshift, if most of the strong (w_r>0.3 Å) lines are expected to be associated with galaxies, the bulk of the Lyman-alpha forest however should have lower equivalent width and should not be tightly correlated with galaxies.

Motivation & Objective

• To understand the redshift evolution of the Lyman-alpha forest across cosmic time.
• To model how the distribution and number density of HI absorption lines change from high to low redshift.
• To assess the role of dark matter structures in shaping the forest's properties.
• To determine whether the Lyman-alpha forest is correlated with galaxies at low redshift.
• To test the consistency of simulated forest properties with observed decrement and column density distributions.

Proposed method

• Simulates a CDM model using collisionless particles to trace dark matter distribution.
• Assumes baryons follow the dark matter distribution, with no radiative cooling or feedback.
• Classifies particles into two populations: P{s} in filamentary structures and P{u} in underdense regions.
• Generates synthetic spectra from the simulation to compare with observed forest properties.
• Uses a UV background flux intensity J_-21 = 0.1 at z=3 to model photoionization.
• Compares simulated decrement and N(HI) distribution with observations across z=0 to z=5.

Experimental results

Research questions

• RQ1 How does the number density of Lyman-alpha forest lines evolve from z=5 to z=0?
• RQ2 What is the contribution of filamentary (P{s}) versus underdense (P{u}) regions to the forest at different redshifts?
• RQ3 To what extent are strong Lyman-alpha lines at low redshift associated with galaxies?
• RQ4 Can a CDM model with baryons tracing dark matter reproduce the observed column density distribution and decrement?
• RQ5 What is the predicted line density at z=0 for different column density thresholds?

Key findings

• The total number density of lines with N(HI) > 10^12 cm^-2 remains approximately constant from z=4 to z=1.
• At z=0, the number density of lines per unit redshift is ~400 for log N(HI) > 12, ~100 for log N(HI) > 13, and ~20 for log N(HI) > 14.
• Population P{s}, associated with filamentary dark matter structures, dominates the line count at z<3 for N(HI) > 10^14 cm^-2.
• Population P{u}, located in underdense regions, rapidly disappears at high redshift for the same column densities.
• The simulated redshift evolution of the decrement and N(HI) distribution matches observations well for Omega_b*h^2 = 0.0125 and J_-21 = 0.1 at z=3.
• Most Lyman-alpha forest lines at low redshift have low equivalent width and are not tightly correlated with galaxies.