Constraining Inflation Models with the BICEP/Keck B-mode Experiments

Abstract

Modern observational Cosmology is highly developed and the observable Universe on a large scale is currently well described by the standard LCDM model. The model has a small number of free parameters and describes how the Universe evolved from a hot, dense and homogeneous state with primordial perturbations that are Gaussian, adiabatic and close to scale-invariant. The inflation paradigm extends the LCDM model and interprets the initial conditions as natural consequences of a hypothesized exponential expansion. However, generic inflationary models make a prediction that has not yet been observed: the existence of a background of primordial gravitational waves (PGWs), which would leave an imprint of B-mode polarization in the Cosmic Microwave Background (CMB). Measuring the degree-scale B-mode polarization therefore emerges as one of the most promising methods to detect or set limits on PGWs. For the past decade, the BICEP/Keck collaboration has been operating a series of telescopes at the Amundsen-Scott South Pole Station optimized for B-mode observation. These telescopes are compact refracting polarimeters mapping about 2% of the sky under the exceptionally stable and transparent atmosphere of the Geographical South Pole. They observe at a broad range of frequencies to separate the cosmological signals from polarized Galactic synchrotron and thermal dust emission which dominate at the two ends of the microwave regime. In this dissertation, we discuss the BICEP/Keck experimental progress in two major areas. We first present the "BK18 analysis" utilizing data collected up to the 2018 observing season, in conjunction with selected WMAP and Planck polarization maps. The analysis particularly exploits new data from (1) the 3-year BICEP3 map, the current deepest CMB polarization measurement at the foreground-minimum 95 GHz; and (2) the Keck 220 GHz map which has a higher signal-to-noise ratio on the dust foreground than the Planck 353 GHz map. The likelihood analysis of the BB auto- and cross-spectra of the maps reduces the experimental uncertainty on the tensor-to-scalar ratio r to sigma(r)=0.009, and the inference of r from our baseline model is tightened to r=0.014+0.010-0.011 and r<0.036 at 95% confidence, the most powerful constraints on PGWs to date. These B-mode spectra hence provide unprecedented power to discriminate among popular classes of inflation theories in the r-ns plane — the natural inflation models and the monomial power law inflation models are now strongly disfavored by the data. We subsequently discuss the instrument development of BICEP Array, a BICEP3-style multiple-frequency (30/40/95/150/220/270 GHz) telescope succeeding Keck from 2020. We focus on the construction of a novel mount and the corresponding pointing model for the telescope. The performance of both are validated by the first robust detection of cosmological polarization signals at 40 GHz in our observation field. Preliminary forecasts show that the constraint can be improved to sigma(r) < 0.003 using the upcoming BICEP3, BICEP Array and SPT-3G data up to 2027.