deep21: a Deep Learning Method for 21cm Foreground Removal

TL;DR

This paper presents a deep learning approach using a UNet CNN architecture to effectively remove foreground contamination from 21cm intensity mapping data, enabling more accurate cosmological analysis.

Contribution

The study introduces a novel deep CNN method trained on simulations for foreground removal in 21cm maps, outperforming traditional PCA techniques and providing uncertainty estimates.

Findings

Achieves within 10% of true clustering statistics across scales

Reduces prediction variance by over an order of magnitude on small scales

Outperforms PCA in accuracy for small radial scales

Abstract

We seek to remove foreground contaminants from 21cm intensity mapping observations. We demonstrate that a deep convolutional neural network (CNN) with a UNet architecture and three-dimensional convolutions, trained on simulated observations, can effectively separate frequency and spatial patterns of the cosmic neutral hydrogen (HI) signal from foregrounds in the presence of noise. Cleaned maps recover cosmological clustering statistics within 10% at all relevant angular scales and frequencies. This amounts to a reduction in prediction variance of over an order of magnitude on small angular scales ($\ell > 300$), and improved accuracy for small radial scales ($k_{\parallel} > 0.17\ \rm h\ Mpc^{-1})$ compared to standard Principal Component Analysis (PCA) methods. We estimate posterior confidence intervals for the network's prediction by training an ensemble of UNets. Our approach demonstrates the feasibility of analyzing 21cm intensity maps, as opposed to derived summary statistics, for upcoming radio experiments, as long as the simulated foreground model is sufficiently realistic. We provide the code used for this analysis on Github https://github.com/tlmakinen/deep21 as well as a browser-based tutorial for the experiment and UNet model via the accompanying http://bit.ly/deep21-colab Colab notebook.