Deep-Learning Study of the 21cm Differential Brightness Temperature During the Epoch of Reionization

TL;DR

This paper introduces a CNN-based deep learning method to analyze 21-cm brightness temperature maps, accurately reconstructing the Epoch of Reionization history even with realistic observational noise and resolution limitations.

Contribution

It presents a novel CNN approach for predicting the neutral hydrogen fraction from 21-cm maps, accounting for observational effects, and demonstrating high accuracy in EoR reconstruction.

Findings

CNN accurately predicts neutral hydrogen fraction

Method remains effective with noise and resolution limitations

Potential for future 21-cm tomography surveys

Abstract

We propose a deep learning analyzing technique with convolutional neural network (CNN) to predict the evolutionary track of the Epoch of Reionization (EoR) from the 21-cm differential brightness temperature tomography images. We use 21cmFAST, a fast semi-numerical cosmological 21-cm signal simulator, to produce mock 21-cm maps between $z=6 \sim 13$. We then apply two observational effects into those 21-cm maps, such as instrumental noise and limit of (spatial and depth) resolution somewhat suitable for realistic choices of the Square Kilometre Array (SKA). We design our deep learning model with CNN to predict the sliced-averaged neutral hydrogen fraction from the given 21-cm map. The estimated neutral fraction from our CNN model has a great agreement with its true value even after coarsely smoothing with broad beamsize and frequency bandwidth, and also heavily covered by noise with narrow. Our results have shown that deep learning analyzing method has a large potential to efficiently reconstruct the EoR history from the 21-cm tomography surveys in future.