Modeling Galaxy Spectra for Lyα Tomography

Mapping the dark matter distribution at high redshifts is important for understanding how the cosmic neighborhood around galaxies affected their development. The distribution of neutral hydrogen (H I) follows that of dark matter on large scales and can be observed in the Lyα forest. Recently, Lyα tomography has made it feasible to create three-dimensional maps of Lyα absorption using a network of Lyman break galaxy (LBG) spectra, a technique that is currently being implemented by the COSMOS Lyα Mapping and Tomography Observations (CLAMATO) survey and the Lyα Tomography IMACS Survey (LATIS). Mapping the Lyα absorption requires separating galaxies’ intrinsic spectra in the Lyα forest from the absorption imprinted by foreground H I. Current surveys use at most one degree of freedom to model line strengths in the intrinsic spectra, but a more flexible method would allow for increased data retention and possibly a reduction of noise in the maps. We apply principal component analysis (PCA) to 2627 LBG spectra at redshift z = 2.2-3.2. With four degrees of freedom, we are able to explain 96% of the variation in the spectra. This method will allow for improved modeling for Lyα tomography surveys and other applications.