Detection of Ly\beta auto-correlations and Ly\alpha-Ly\beta cross-correlations in BOSS Data Release 9

The Lyman-$\beta$ forest refers to a region in the spectra of distant quasars that lies between the rest-frame Lyman-$\beta$ and Lyman-$\gamma$ emissions. The forest in this region is dominated by a combination of absorption due to resonant Ly$\alpha$ and Ly$\beta$ scattering. When considering the 1D Ly$\beta$ forest in addition to the 1D Ly$\alpha$ forest, the full statistical description of the data requires four 1D power spectra: Ly$\alpha$ and Ly$\beta$ auto-power spectra and the Ly$\alpha$-Ly$\beta$ real and imaginary cross-power spectra. We describe how these can be measured using an optimal quadratic estimator that naturally disentangles Ly$\alpha$ and Ly$\beta$ contributions. Using a sample of approximately 60,000 quasar sight-lines from the BOSS Data Release 9, we make the measurement of the one-dimensional power spectrum of fluctuations due to the Ly$\beta$ resonant scattering. While we have not corrected our measurements for resolution damping of the power and other systematic effects carefully enough to use them for cosmological constraints, we can robustly conclude the following: i) Ly$\beta$ power spectrum and Ly$\alpha$-Ly$\beta$ cross spectra are detected with high statistical significance; ii) the cross-correlation coefficient is $\approx 1$ on large scales; iii) the Ly$\beta$ measurements are contaminated by the associated OVI absorption, which is analogous to the SiIII contamination of the Ly$\alpha$ forest. Measurements of the Ly$\beta$ forest will allow extension of the usable path-length for the Ly$\alpha$ measurements while allowing a better understanding of the physics of intergalactic medium and thus more robust cosmological constraints.