Exploring excess variance in LOFAR-EoR using machine learning
Allison Man (firstname.lastname@example.org)
*All are welcome to this event!
The detection of the 21 cm signal from the Epoch of Reionisation (EoR) is challenging due to the strong astrophysical foregrounds, radio frequency interference (RFI), and ionospheric and instrumental effects. Most if not all observations of the 21 cm signal at high redshifts show so-called “excess variance” in their power spectrum, well beyond what would be expected based on the thermal noise limit. Understanding the sources of the excess is crucial for improving upper limits on the EoR signal.
In this talk, I will describe possible sources of the excess variance in LOFAR-EoR and I will introduce a general-purpose data analysis tool, the self-organising attribute maps, to analyse features in residual sky images more efficiently. Also, I will describe how this method can be applied to CHIME data for data quality diagnostics.
I am a postdoctoral research fellow at the University of British Columbia. Currently, I am working in 21 cm cosmology on CHIME (The Canadian Hydrogen Intensity Mapping Experiment) data analysis and map-making pipeline. Prior to that, I completed my PhD degree at Kapteyn Institute and Bernoulli Institute in Groningen under the supervision of Prof. dr. Léon Koopmans, Dr. Michael Wilkinson and Dr. André Offringa with the project ''Hunting elusive excess variance in big LOFAR data''.
View Hyoyin's personal webpage here
See Hyoyin speak on "Exploring sources of excess noise in detecting 21cm signal with LOFAR" (Youtube) here