12.1. b from the IGM Lyman- forest absorption
The gas in the IGM is observed through H I Ly absorption in the spectra of all QSOs. Gunn & Peterson  discussed how redshift produces continuous absorption in the ultraviolet spectra of QSOs. Density fluctuations in the IGM trun this continuous absorption into the Ly absorption lines. The IGM fills the volume of space, and at redshifts z > 1  it contains most of the baryons.
The baryon density is estimated from the total amount of H I absorption, correcting for density fluctuations which change the ionization. The gas is photoionized, recombination times are faster in the denser gas, and hence this gas shows more H I absorption per unit gas. Using the observed ionizing radiation from QSOs, we have a lower limit on the ionizing flux, and hence a lower limit on the ionization of the gas. If the gas is more ionized than this, then we have underestimated the baryon density in the IGM.
Three different groups obtained similar results , , : b > 0.035 h70-2. This seems to be a secure lower limit, but not if the IGM is less ionized than assumed, because there is more neutral gas in high density regions, and these were missing from simulations which lack resolution.
We do not have similar measurements at lower redshifts, because the space based data are not yet good enough, and the universe has expanded sufficiently that simulations are either too small in volume or lack resolution. Cen & Ostriker  have shown that by today, structure formation may have heated most local baryons to temperatures of 105-107 K, which are extremely hard to detect , .