Tools for Measuring the Cosmic History of Molecular Gas

Abstract

The coldest, densest phase of the interstellar medium plays a critical role in the evolution of galaxies. It is only in this phase, where the gas is primarily molecular, that interstellar material can collapse and form new stars. Spectral line emission from the carbon monoxide (CO) molecule is the favored diagnostic of molecular gas masses, and decades of measurements have revealed a tight connection between the abundance of molecular gas and the rate of star formation. Most star forming galaxies reside along tight sequence in the stellar mass-star formation rate plane, which is shaped and regulated by the abundance of gas and the efficiency with which it is converted to stars. These relationships extend to high redshift where significantly larger gas reservoirs drove main sequence star formation rates an order of magnitude higher than we observe today. While the close connection between molecular gas and star formation in galaxies is now firmly established, many details remain hotly debated or completely unknown. Refining and advancing our understanding of galaxy evolution today and throughout cosmic history requires careful development of new theoretical, analytical, and observational tools for studying cold gas. This thesis represents a collection of three such tools, each developed in response to specific, open questions with bearing on ongoing and future research in the field. How should we design surveys to efficiently measure the molecular gas abundance in the high redshift universe, with minimal bias? In Part I, I combine results from large-scale cosmological simulations with galaxy property scaling relations to model the gas content of high redshift galaxies in mock millimeter deep field surveys. I use these mocks to explore how survey depth and volume affect measurements of the CO luminosity function and the total gas mass of the universe. I provide a tool for estimating the relative contributions of sampling uncertainty (due to small number statistics) and field-to-field variance (due to small survey areas) and show that without careful selection of fields and survey geometry, the precision of future surveys will be limited by the latter effect. Can we study molecular gas abundances over significantly larger volumes with existing millimeter facilities? Line intensity mapping is an alternative approach to studying the cosmic abundance of molecular gas. In large, spectrally resolved surveys, line emission from CO (or other species) in undetected galaxies contributes excess fluctuations that can be identified statistically. This approach promises to survey large volumes while simultaneously measuring the gas content in galaxies below the detection limits of targeted surveys. However, extracting astrophysical information from the noise-dominated intensity maps will require careful control of systematics and sophisticated statistical and modeling tools. In Part II, I discuss one such tool – cross-correlation between millimeter-wave intensity maps and optical galaxy catalogs. Using data from the CO Power Spectrum Survey, I demonstrate the power of cross-correlation to remove signal contamination. I measure the cross-power spectrum and use results from Part I to model and account for systematic uncertainties. I set an upper limit on the molecular gas abundance at z~2, providing a factor of two improvement over prior constraints on the clustering term of the CO power spectrum. How should observations of molecular gas using different spectral lines of the CO molecule be compared? The CO molecule produces a ladder of rotational transitions at millimeter and submillimeter wavelengths. The transitions at shorter wavelengths arise in warmer and denser gas than the ground state transition, complicating their use as a proxy for the bulk of molecular gas. In Part III, I present the results of the Arizona Molecular ISM Survey with the SMT (AMISS), a multi-line study of 177 nearby galaxies designed to measure how the three lowest energy CO lines relate to one another and the physical conditions of molecular gas. Using this sample, I determine “typical” luminosity ratios between these lines, then proceed to explore how they vary systematically as a function of galaxy properties. I show that the CO(1–0) and CO(2–1) lines are not interchangeable tracers for the study of the connection between gas mass and star formation, and provide an empirical prescription for translating between the luminosities of the two lines. I also present an extensive set of calibration measurements for the two Arizona Radio Observatory telescopes, which were required to minimize calibration uncertainties and observing inefficiencies in AMISS.