Instruments and Statistical Tools to Study Supermassive Black Holes at Event Horizon Scales

Abstract

Supermassive black holes are known to exist at the center of galaxies, including the one at our Milky Way galaxy, Sagittarius A* (Sgr A*). To observationally study very near environment of the black hole close to its event horizon, we need an Earth-size telescope operating at millimeter wavelengths. In this dissertation, I developed the very-long-baseline interferometry (VLBI) receiver for the South Pole Telescope (SPT). The receiver enables the Event Horizon Telescope (EHT) VLBI array to achieve the baseline that can resolve the apparent size of Sgr A*. However, understanding physics around the black hole from the Sgr A* observation is challenging due to its rapid variability. In this thesis, I also studied statistical methods to compare the variable Sgr A* observations with physically motivated, time-dependent, black hole simulations using a Bayesian framework. In Chapter 2, I present the development of the VLBI receiving system for the SPT. Since the SPT was built for the cosmic microwave background observation, it was solely equipped with the multi-pixel, wide-field, bolometric receiver. I designed and assembled the coherent receiver working at 1.3 and 0.87 mm, and installed the system, including the VLBI recording setup at the South Pole. In Chapter 3, I report the vector beam measurement of the SPT VLBI receiver. Due to the receiver design inside the dewar, it was essential to analyze the beam measurement accounting for the tilted geometry and characterize the beam propagation relative to the optical components. In Chapter 4, I report the first VLBI experiment result at the SPT. We observed Centaurus A (Cen A) and successfully detected the interferometric fringe. Although the observation focused on the demonstration of the VLBI capability at the telescope, the ~7000 km baseline between the South Pole and Chile provided the highest resolution observation of Cen A published to date. In Chapter 5, I introduce the Bayesian technique to compare VLBI observation of the black holes such as Sgr A* to time-varying general relativistic magnetohydrodynamics (GRMHD) simulation models. The method statistically takes the variabilities of both observation and the models into account and has the power to perform parameter estimation and quantitative model comparison. I show the application of the method using synthetic data generated from the simulation as well as the three-station EHT data from its early stage. Finally, in Chapter 6, I summarize the work presented in this dissertation.