Dr. Ian Sullivan is awarded an NSF Astronomy and Astrophysics Postdoctoral Fellowship to carry out a program of research and education at the University of Washington (UW). Dr. Sullivan will conduct experiments employing two types of measurements to study the epoch of reionization when the first stars and galaxies formed--the Cosmic Infrared Background Experiment (CIBER), which will measure the fluctuations in the cosmic near-infrared background (NIRB) from the first galaxies, and the Murchison Widefield Array (MWA), which will map the structure of neutral hydrogen (HI) as it is ionized by those galaxies. The best measurements to date of the epoch of reionization exist from measurements of the cosmic microwave background, which indicate that the optical depth due to Thomson scattering is 0.089 +/- 0.030. Additional measures are needed, however, to reveal a more complete picture of reionization and, consequently, the history of the universe. Using two such measures, which incorporate very different technologies and are susceptible to different systematics, will, taken together, gather complementary information and serve as important cross-checks on each measurement.
For CIBER Dr. Sullivan has led the fabrication of the instruments and analysis of the imager data; under this award he will continue that analysis, leading to the first definitive measurement of NIRB fluctuations from a targeted mission. CIBER consists of four infrared instruments on board a NASA sounding rocket, which has already achieved one successful flight. Over the course of the next few years, CIBER will launch an additional three times, culminating in a long-duration non-recoverable flight and followed by a complete redesign of the instrument with two more launches. The four instruments of CIBER include two spectrometers to measure the absolute brightness of the Infrared background from 700-1800 nm, and two wide-field imagers to measure NIRB fluctuations at 1000 nm and 1600 nm.
Dr. Sullivan will work concurrently on the MWA with Dr. Miguel Morales at UW. The MWA is a radio interferometer that operates in the frequency range of 80-300 MHz, which is tuned to the 21-cm line of HI from the redshift ranges appropriate to the epoch of reionization. Since the MWA measures a redshifted line of radiation over a wide range of frequencies, it is possible to map the HI distribution over the course of reionization. The instrument is currently under construction in the radio-quiet zone of Western Australia, and Dr. Sullivan will assist with the instrument deployment and commissioning. Dr. Sullivan will thus gain sufficient familiarity with the instrument to undertake the precision data analysis necessary to remove foregrounds and instrumental effects. Finally, through an active role in both CIBER and MWA collaborations, Dr. Sullivan will cross-correlate data from the instruments where they observe the same field to look for the anti-correlations implied by their origins.
Dr. Sullivan will also initiate a series of hands-on workshops for high school students, designed to expose them to scientific research. Using the CIBER project as a model, students in the program will learn about astrophysics, design a small science experiment, rigorously test and calibrate it, and launch it on a model rocket. The students will assemble a poster based on the results of their project for public display and will write independent reports of the experience. These reports may be used for college application essays, which will encourage them to pursue a degree in science.
Precision analysis of very large sets of data is the heart of modern radio astrophysics, and absolutely essential for a detection of the radio imprint of the epoch of reionization. The epoch of reionization began when the first stars formed, and the ultraviolet light from these young, massive, and hot stars ionized the surrounding Hydrogen gas. As reionization progressed, these ionized regions around the first stars expanded, until eventually they began to overlap, leaving the entire universe almost completely ionized. We can detect redshifted 21-cm radiation from that neutral Hydrogen gas cloud pervading the early universe, but it takes great sensitivity and sophisticated analysis and modeling to recover it from behind billions of years of intervening foregrounds. Modern radio telescopes achieve high sensitivity with many antennas, and achieve precision analysis through detailed modeling and simulation of the response pattern of every antenna element, but all at the cost of vastly greater data volumes and demands on computational resources. We have developed Fast Holographic Deconvolution (FHD) to address these challenges and achieve efficient yet precise analysis to extract the background radio signal from the epoch of reionization. Fast Holographic Deconvolution is a new algorithm that rethinks the traditional imaging equations for interferometry, by performing many calculations up front to construct the full covariance matrix of the instrument. The covariance matrix of a modern radio telescope is very large, but when stored as a sparse matrix it can be used to calculate an accurate model for deconvolution in a matter of seconds, even on a desktop computer. Additionally, with FHD the raw visibility data is not needed after being gridded once, which permits massive data compression and more efficient data transmission and storage. FHD is now being used to process all of the data from the full 128 antenna Murchison Widefield Array (MWA), and the upcoming Hydrogen Epoch of Reionization Array (HERA) has been designed to make optimal use of the new capabilities and efficiencies offered by FHD.
