UC Berkeley

Molecular gas, observed through tracers such as CO rotational transitions, is a vital component of galactic evolution and star formation. Recent detections of the CO molecule in massive galaxies at high redshift have demonstrated its existence in the early Universe, and have motivated its use as a means of exploring large-scale structure and as a probe of galaxy evolution in the early Universe. But many questions about molecular gas and the evolution of galaxies in the early Universe still remain: its distribution at high redshift understood is so poorly that theoretical models of the mean abundance of CO for the first several billion years of cosmic history span orders of magnitude. Direct detection of molecular gas in galaxies at these redshifts have only found the largest and most luminous of galaxies in the early Universe, whereas the bulk of the molecular gas is expected to be in the unseen masses of smaller galaxies. While difficult to detect individually, these smaller galaxies are likely detectable as an integrated ensemble with the technique of ``intensity mapping". This technique, similar to those employed by HI epoch of reionization experiments, utilizes measurements of different 3D Fourier modes to construct a power spectrum.

In this thesis, I present results from the CO Power Spectrum Survey (COPSS), an intensity mapping experiment performed with the Sunyaev Zel'dovich Array (SZA). I present power spectrum constraints from of the first and second phase of this project, utilizing an archival dataset (covering 44 fields in 1400 hours observing time) and a 5000-hour observing campaign (covering 19 fields) with the SZA. With these data, we are capable of observing CO(1-0) emission arising from $z=2.3{-}3.3$, surveying a volume of more than ten million cubic megaparsecs. With this measurement, we place the first-ever constraints on the CO autocorrelation power spectrum, and place constraints on the CO(1-0) galaxy luminosity function and the cosmic molecular gas density at $z\sim3$.

I also present a series of simulations designed to probe the impact of cosmic variance, continuum foregrounds, and systematic errors on our measurement with the SZA. I will also present simulations designed to probe the challenges that will face intensity mapping experiments with the Yuan-Tseh Lee Array (YTLA). Finally, I will present some preliminary work on CO(3-2) and [CII] intensity mapping experiments.