Skip to main content

37 Summary

Keywords

Big Bang cosmology, CMB polarization, POLARBEAR, LiteBIRD, Crosstalk

37 Summary

In part I, a brief overview of the standard model of Big Bang cosmology is given. Unresolved problems such as fine-tuning problems can be resolved by introducing cosmic inflation. Inflation is generally believed to be true but has not been experimentally or observationally verified. Due to the extreme high energy involved, observation of the CMB radiation offers one of the best chances to falsify inflationary models.

In part II, the physics of the CMB is briefly presented, showing the anisotropies in the intensity and polarization of the CMB hold a wealth of information to be extracted to better characterize and understand the universe. In particular, the primordial \(B\)-mode holds the key to falsify inflationary models.

In part III, an overview of CMB observations is given. It ranges from the intrinsic properties of the CMB that affect experimental designs, to the mathematics of CMB data analysis, to the computational aspects including simulations and systematics mitigations, to the statistical aspects which enable us to extract information of the universe from the data. It concludes with the current measurements of the CMB and what to expect next.

In part IV, the POLARBEAR experiment is introduced. Some of the key details on existing data analysis in POLARBEAR are given, such as the null-test framework, mapmaking method including CRHWP demodulation, and power-spectra estimation method.

Building on top of that, part V shows original research that builds upon the observed data and existing data analysis framework. In particular, a measurement of the high-\(\ell\) \(B\)-mode in the range of 600–3000 using 3rd–5th seasons of POLARBEAR data is presented. It started with the expected sensitivity that should have been known before taking the observation. It also pointed out the major challenges experienced in this research including computational efficiency, reproducible research, and null-test failures. Among other details, the solutions to these major obstacles are presented. It also presented improved statistical techniques, including the definition of null-statistics and its combined statistics, a full-Bayesian likelihood procedure, and an improved formula to calculate the analytical statistical uncertainty. The limitation of the null-test framework is documented where some are general and some are specific to POLARBEAR. In the end, the lensing amplitude is detected with \(P(A_\text{lens} > 0) = {99.6\,\%}\).

In part VI, the crosstalk systematic is studied in detail. An analytical model on the map-domain is developed to predict the effect of crosstalk on the power-spectra given a set of conditions. A TOAST operator is developed to enable time-domain simulations of crosstalk systematics in the framework of the software Time Ordered Astrophysics Scalable Tools (TOAST). The two predictions are compared and show a high degree of agreement in the example of the LiteBIRD experiment. Exploiting the analytical behavior in the model, a hardware based mitigation technique is developed to minimize its effect on the 2pt-power-spectra. This mitigation technique is general as long as the set of conditions is fulfilled, and a simple rule of thumb is presented that is easy to use for experimentalists. Given a new focal plane design, the effect can be predicted without time-consuming time-domain simulations for fast prototyping. One limitation of this model and mitigation technique is the requirement of the HWP.