Speaker
Description
Within the LOFAR EoR KSP team, Gaussian Process Regression (GPR) has been used for foreground subtraction from data, to constrain the Neutral Hydrogen 21-cm signal power spectrum from the Epoch of Reionization. To avoid signal loss due to the misestimation of the covariance kernel for the 21-cm signal, we developed a Machine Learning (ML) trained model, by training on a large variety of N-body + 1D radiative transfer simulations (GRIZZLY). We tested its limitations by exploring a variety of mock datasets. Now I will show the results of implementing it on 10 nights of observational data from the LOFAR telescope at z~9.1.
Additionally, I will showcase a new suite of simulations (N-body with GADGET-4 + Semi-Analytic Modelling with L-Galaxies + 1D radiative transfer with GRIZZLY) exploring a range of cosmologies (varying h0 and sigma8) for a wider training set for the ML model as well as for data interpretation from tighter upper limits from future observations with LOFAR, SKA, etc. I will also show improvements using the Fixed & Paired method to reduce cosmic variance and discuss future improvements via subgrid modelling, allowing us to improve resolution. This suite of simulations also provides mock observations for telescopes like JWST and Euclid, thus providing a platform for multi-wavelength studies.
