Abstract
The CO Mapping Array Project (COMAP) is a line intensity mapping experiment that aims to map the distribution of star-forming carbon monoxide (CO) in the Universe from the Epoch of Galaxy Assembly to the Epoch of Reionization, for the target redshifts z = 2.4–3.4 and z = 6–8, respectively. In 2022, COMAP presented the first ever constraint on the three dimensional CO(1–0) clustering power spectrum. Two years later, the COMAP Season 2 results improved on these limits by an order of magnitude. Despite these recent successes, approximately 75% of the Season 2 map data was discarded, primarily due to concerns of systematic errors. As COMAP approaches the sensitivities necessary to claim a first CO detection, increasing data retention will be of great importance. A deeper understanding of the COMAP systematics could result in regaining lost data by leading to more informed data cuts and data cleaning strategies.
The aim of this work is to aid this effort by studying the behavior of modeled systematic effects throughout the data pipeline. To do this, I have developed the first physically motivated systematic effect simulation framework for COMAP, centering three of the most concerning systematic effects in the experiment: the ground-pickup emission, “start-of-scan” effect, and the “turn-around” effect, as well as a model for white noise. The ground-pickup effect is caused by far sidelobe pickup of ground emission from the landscape surrounding the telescope, while the standing wave effects “start-of- scan” and “turn-around” form in the optic cavities of the instrument at the beginning of each scan and as the pointing changes directions, respectively. The ground-pickup model produced in this work is completely physically motivated, while the standing wave models are a mix of first principles modeling and an empirical by-eye estimation of wavelengths and amplitudes. These studies are used to assess the possibility of recovering data from our discarded map scales, the effectiveness of the map-domain PCA filter, the elevation split method used to generate the cross-elevation power spectra, and the origins of the standing wave systematic effects.
This work presents in total eight simulations, consisting of individual systematic effect simulations with and without noise, one “full simulation” containing all the modeled systematics and noise simultaneously, and a final pure white noise simulation. These simulations are processed through the COMAP data pipeline, and the resulting data products are compared to the corresponding data products from the COMAP Season 2 analysis. I present results that reproduce the broad morphology of the standing wave effects in the map domain, discover a new mechanism for constant standing waves propagating through the COMAP pipeline, corroborate many of the COMAP data-cut choices, and find that the map-domain PCA filter effectively removes the modeled standing wave residuals and ground-pickup residuals.
Supervisors:
- Ingunn Kathrine Wehus, Professor at Institute of Theoretical Astrophysics, UiO
- Hans Kristian Eriksen, Professor at Institute of Theoretical Astrophysics, UiO
- Jonas Gahr Sturtzel Lunde, Postdoctoral Fellow at Institute of Theoretical Astrophysics, UiO
- Nils-Ole Stutzer, Postdoctoral Fellow at Institute of Theoretical Astrophysics, UiO
Intern. assessor: Tiago Pereira, Professor at Institute of Theoretical Astrophysics, UiO
Extern. assessor: Håvard Tveit Ihle, Researcher at Forsvarets forskningsinstitutt