2004ApJ...605....7D

Absorption between the rest-frame wavelengths of 973 and 1026 Å in quasar spectra arises from two sources (apart from occasional metals): one is due to Lyα absorption by materials at a low redshift, and the other is from Lyβ at a higher redshift. These two sources of absorption are to a good approximation uncorrelated because of their wide physical separation. Therefore, the two-point correlation of absorption in this region of quasar spectra neatly factorizes into two pieces: the Lyβ correlation at high-z and the Lyα correlation at low-z. The latter can be independently measured from quasar spectra at lower redshifts using current techniques. A simple division then offers a way to statistically separate out the Lyβ two-point correlation from the Lyα correlation. Several applications of this technique are discussed. First, since the Lyβ absorption cross section is lower than Lyα by about a factor of 5, the Lyβ forest is a better probe of the intergalactic medium (IGM) at higher redshifts, where Lyα absorption is often saturated. Second, for the same reason, the Lyβ forest allows a better measurement of the equation of state of the IGM at higher overdensities, yielding stronger constraints on its slope when used in conjunction with the Lyα forest. Third, models of the Lyα forest based on gravitational instability make unique predictions for the Lyβ forest, which can be tested against observations. We briefly point out that feedback processes, which affect higher density regions but leave low-density structure intact, may be better constrained by the Lyβ forest. Finally, extending our technique to the higher Lyman series is, in principle, possible but becomes increasingly difficult because of diminishing path lengths.