Joint Analysis Techniques for Line Intensity Mapping, with
Application to the CO Mapping Array Project
Author: Dunne, Delaney Anne
Year: 2026
Degree: Dissertation (Ph.D.)
Advisors: Cleary, Kieran A.; Hallinan, Gregg W.
Committee Members: Steidel, Charles C.; Hopkins, Philip F.; Bock, James J.; El-Badry, Kareem J.; Cleary, Kieran A.; Hallinan, Gregg W.
Option: Astrophysics
DOI: 10.7907/r11t-4p78
Abstract
Line intensity mapping (LIM) is an emerging observational technique with broad applications in cosmology and galaxy evolution. Rather than cataloging galaxies as discrete objects, LIM experiments measure the spatial fluctuations of galaxy populations across large three-dimensional volumes. Thus, LIM is able to measure the contribution to the overall spectral-line luminosity from all galaxies in a population, even the very faintest. LIM began with the 21 cm neutral hydrogen line but has recently been expanding to other tracers, such as CO, [CII], Hα, and Lyα. The first generations of dedicated experiments targeting these tracers are beginning to come online. In order to draw science conclusions from their observations, these experiments must address challenges in the form of sensitivity and systematic errors.
In this thesis, I develop an analysis technique designed to mitigate these challenges by stacking LIM data on the positions of galaxies in an external galaxy catalog. I use a custom-designed simulation pipeline to produce simultaneous mock observations of a LIM experiment and galaxy survey covering the same galaxy population, and use these simulations to build a theoretical understanding of this stacking statistic. I find that cosmological clustering of galaxies around the objects in the reference galaxy catalog drives the stacked signal.
I also use the stack to draw conclusions about molecular gas in the period preceding the peak of cosmic star formation, using data from the CO Mapping Array Project (COMAP), the leading CO LIM experiment. I use eBOSS and DESI quasars, as well as Lyα emitters from HETDEX, as external reference catalogs. I do not detect any cosmic CO emission in the COMAP stacks, and I find a moderate tension between the resulting upper limits and the brightest models of the stacked emission. This tension could be attributable to non-linear radiative transfer or clustering properties of the galaxy catalogs, but it could also be due to models for cosmic CO overpredicting the average CO luminosity z~3.
Finally, I search for spectral-line emission from axionic dark matter by stacking the DESI and HETDEX galaxies over a range of rest frequencies. I constrain the axion decay rate Γ and the axion-photon coupling constant |gaγγ|. In the axion mass range probed by COMAP, I establish the tightest constraints available on an axion model of dark matter.