| Summary: | A period of inflationary, accelerated expansion in the early Universe -- while a radical extrapolation from understood physics -- naturally solves the horizon, flatness, and monopole problems of standard cosmology. Models of inflation generically predict tensor perturbations to the metric, which would create a stochastic gravitational-wave background and imprint B-mode (curl-type) polarization in the cosmic microwave background (CMB). Measurement of primordial B modes at degree angular scales would be direct evidence for inflation; the amplitude, parametrized by the tensor-to-scalar ratio r, would indicate its energy scale. Since 2006, the BICEP/Keck Array CMB polarization experiments located at the Amundsen-Scott South Pole Station have been mapping ~1% of low-foreground sky at degree angular scales using small-aperture, on-axis refracting telescopes. By deploying thousands of detectors at the multiple frequencies necessary to separate the CMB from Galactic foregrounds, we have produced the tightest constraints to date on inflationary gravitational waves: r < 0.07 and sigma(r) = 0.024 as of October 2015. The BICEP/Keck Array telescopes measure polarization by differencing co-located, orthogonally polarized detectors. The most prominent systematic effect in the polarization measurement is leakage of the bright CMB temperature sky into the much fainter polarization due to mismatched beam shapes. In this dissertation, we use high-fidelity in situ calibration data to measure beam shapes and quantify the level of temperature-to-polarization leakage expected in the final BK15 multifrequency CMB maps after filtering. We discuss the degree of cancellation expected in the leaked maps resulting from averaging over many detectors, observation angles, and frequencies. Finally, we comment on how such leakage levels will scale in next-generation experiments with hundreds of thousands of detectors. Astronomy cosmic background radiation; cosmology; gravitational waves; inflation; instrumentation; polarimeters; optics
|
|---|