Renee Hlozek

University of Toronto
Project Title: Optimal classification of radio transients
Industry Partner: Thoth Technology Inc.
Platform: Blue Gene/Q, Cloud Analytics, Large Memory System

Digital Media

When we look up at the dark night sky, we have a sense for how the brightness of stars and galaxies gives us information about their nature. Yet while we cannot see them with our eyes, signals in the radio wavelengths are coming from the sky both day and night.

This `radio sky’ provides us with a wealth of information. That deluge of information provides us with the challenge of separating out and making useful the data we receive. Much like deciphering the different noises we hear when we are walking on a crowded street when listening for the brief call of some exotic bird, we need to develop sophisticated techniques to identify and classify radio transient signals.

New studies have discovered new types of radio transients that produce large radio pulses, opening up a whole new era of radio astronomy. Bright (loud) and rapidly varying radio transients have been discovered in the past decade, and groups around the world are searching for these radio signals in order to understand where they come from. Are the galactic in origin or do they come from the distant universe? What can they tell us about the properties of our local environment?

With the SOSCIP TalentEdge Fellow, we will develop methodology to classify these radio transients into different groups/types in order to reveal their nature and origin. While this is beneficial to understanding astronomical signals, it has a more commercial application: the classification of different RFI signals used for communication. Through a collaboration with Thoth Technology Inc., already a major partner in radio astronomy research in Canada, we will use Canada’s largest radio telescope, the impressive 46m dish at the Algonquin Observatory in northern Ontario, to search for these rapidly varying radio signals. We apply our techniques to the radio data and we develop tools to fully classify bright radio transients. These tools will be readily of use for both astrophysical and man-made transient signals.