Constraining Inflation with BICEP/Keck and SPT
Inflation generically predicts a background of stochastic gravitational waves. In the standard cosmological model LCDM, these primordial gravitational waves (PGW) are the only source to the odd-parity (curl) B-mode polarization on the Cosmic Microwave Background (CMB) at the epoch of recombination. Therefore, we can learn much about inflation by constraining or potentially detecting primordial gravitational waves. In this talk, I will present two frontiers that advance the search for primordial gravitational waves: instrumentation of BICEP3 and delensing with SPTpol. I will conclude with the future outlook of such search with CMB-S4. In the first part of the talk, I will discuss the design and performance of BICEP3 — the path-finder of BICEP/Keck program’s Stage-3 effort. The BICEP/Keck telescopes are small aperture (~0.5m) cryogenic refractors that focus on measuring the degree-angular scale feature of the PGW-generated B modes. BICEP3 has ~10x optical throughput compared to BICEP2 due to a larger field-of-view and aperture size. As a result, we adopt new designs in BICEP3 for thermal filtering, optical elements selection, and detector modules packing. We deployed BICEP3 in the Austral summer of 2014-2015 and upgraded the instrument after the first season of testing and observations. I will present the improvements of the noise levels due to this upgrade. The second part of my talk will focus on delensing SPTpol data. Gravitational lensing of primordial (curl-free) E modes generates B modes. These lensing B modes are a source of foreground contamination to detecting the PGW-generated B modes. We can characterize and reduce the lensing B-mode contribution through a technique called “delensing.” I will present results of a first demonstration of delensing on polarization data. In this work, we delens B-mode maps from multi-frequency SPTpol observations of a 100 deg^2 patch of sky by subtracting a lensing B template constructed from two components: SPTpol E-mode maps and a lensing potential map formed from a Herschel 500 μm map of the CIB. In addition, we build and use a suite of realistic simulations to study the current limitations and expected future improvements in delensing with implications for future experiments.