Development of the CHORD Galaxy Search Strategy

Abstract

This thesis presents my contribution to the CHORD galaxies science case. I helped develop an algorithm that can automatically pick out galaxies from CHORD driftscan data. The method used is a matched filter, and it acts on spatial data and spectral data.

On the spatial side, it searches for point sources. Because CHORD is a highly redundant interferometer, it suffers from spatial aliasing. I present a program that is able to predict the severity of spatial aliasing. It predicts that integrating over periods of time is required to break the alias degeneracy, and that "dithering" CHORD (rotating it by a couple degrees) further helps in reducing the alias issue. I offer a framework for estimating the probability of spatial alias confusion.

On the frequency side, I present a method of running the matched filter quickly. CHORD frequency data undergoes a process called upchannelization, which would distort the shape of a galaxy profile. I show how this can be accounted for without incurring a large time-cost penalty.

Lastly, I discuss how a full matched filter program would be put together, and implications that my research has on selecting search parameters for a future CHORD galaxy survey.