Understanding why the expansion of the universe is getting faster with time is one of the biggest mysteries in cosmology today. Kyle Boone, a fellow at the DiRAC Institute, focuses his research on developing novel statistical methods for astronomy and cosmology. He is particularly interested in using Type Ia supernovae to probe the accelerated expansion of the universe that researchers believe is due to some form of “dark energy”.
Dr. Boone is a DiRAC Postdoctoral Fellow. He joined the DiRAC institute in September 2019 after finishing up his PhD in Physics at the University of California in Berkeley, California. He previously received a Bachelors degree in Engineering Physics at the University of British Columbia in Vancouver, Canada in 2013.
Type Ia supernovae are interesting for cosmology because they can be used as “standard candles” to make a map of the universe. In the late 1990s, with a sample of only 50 Type Ia supernovae, researchers were able to discover that the expansion of the universe was accelerating. Upcoming astronomical surveys such as the Large Synoptic Survey Telescope (LSST) will discover hundreds of thousands of Type Ia supernovae, and will enable researchers to probe the fundamental nature of dark energy and the accelerated expansion of the universe.
Dr. Boone is working on developing statistical techniques to handle this deluge of data. One challenge for LSST is that there will not be enough resources to follow up every transient that is discovered. Dr. Boone developed the software package “avocado” that uses machine learning to identify what these transients are. With the algorithms in this package, Dr. Boone won a Kaggle competition that simulated the challenges that LSST will face out of 1094 different teams. The results of this work will be published in a paper that has been accepted by the Astronomical Journal.
In a series of papers in preparation, Dr. Boone is working on developing better methods to estimate the distances to Type Ia supernovae. Dr. Boone has shown that manifold learning can be used to capture the diversity of Type Ia supernovae that we see in the universe. His results show that our current methods of estimating distances to Type Ia supernovae are highly biased, and could lead surveys such as LSST to make incorrect conclusions about the properties of dark energy.
Dr. Boone is very grateful to the DiRAC institute for warmly welcoming him, and is excited to collaborate with researchers there. The DiRAC institute is heavily involved with several large scale transient surveys, such as ZTF and LSST, and has a wide range of researchers working on applying machine learning to astronomical time series data. Dr. Boone is also an eScience Postdoctoral Fellow, and is interested in collaborating with researchers outside of astronomy who are working on data science problems with similar time series data.