Welcome to my personal webpage.
I am a third year PhD student at the Kapteyn Astronomical Institute and Bernoulli Institute for Mathematics and Computer Science in Groningen. Prior to that, I completed my Master's degree at Universität Heidelberg with the project "Nuclear reactions in astrophysical plasmas" under the supervision of Dr. Adriana Pálffy.
Currently, I am working on the project "Scalable algorithms to process massive datasets from radio astronomy" under the supervision of Prof. dr. Léon Koopmans, Dr. Michael Wilkinson and Dr. André Offringa. My research focuses on analysing LOFAR data with statistical methods.
LOFAR (Low Frequency ARray) is a new generation radio interferometer that covers the unexplored low-frequency range from 10 to 240 MHz. Among its key science projects, LOFAR-EoR (Epoch of Reionisation) is set to detect the spectral fluctuations of the redshifted neutral hydrogen (HI) 21-cm signal. The EoR is a watershed period in the history of the Universe when the predominantly neutral intergalactic medium (IGM) was ionised by the formation of the first luminous sources such as stars, galaxies and quasars. The EoR provides an opportunity to study the first generation of astrophysical objects. Detection of the 21-cm signal is a major challenge because our measured signal is strongly dominated by astronomical foregrounds, instrumental errors, etc. The errors introduced by the instruments and ionospheres are corrected by calibration, and the astronomical foregrounds are subtracted by sky model after calibration and imaging. The first limits on the EoR HI power spectra from LOFAR were published in 2017. These results show that, after calibration and foreground subtraction, the noise level in the power spectra is higher than thermal noise a factor of 2-3. Possible sources of this noise can be the incompleteness of the sky model, ionospheric effects and calibration. Our research focuses on finding sources of this noise. In particular, we are investigating the correlation between calibration and noise level using statistical methods and machine learning techniques. Finding the source of the noise will improve the chances of detecting the 21-cm signal.