Deep Learning the Intergalactic Medium using Lyman-alpha Forest at $ 4 \leq z \leq 5$

TL;DR

This paper introduces a deep learning method using Bayesian neural networks to reconstruct the thermal history of the intergalactic medium from Lyman-alpha forest data at redshifts 4 to 5, enabling precise constraints with minimal sightlines.

Contribution

The study develops a novel deep Bayesian neural network approach to accurately infer IGM temperature and density from Lyman-alpha spectra, improving constraints over traditional methods.

Findings

Successfully predicts IGM temperature and density with high fidelity.

Achieves constraints on T0 with ~1000 K uncertainty from a single sightline.

Provides more stringent slope constraints of the temperature-density relation.

Abstract

Unveiling the thermal history of the intergalactic medium (IGM) at $4 \leq z \leq 5$ holds the potential to reveal early onset HeII reionization or lingering thermal fluctuations from HI reionization. We set out to reconstruct the IGM gas properties along simulated Lyman-alpha forest data on pixel-by-pixel basis, employing deep Bayesian neural networks. Our approach leverages the Sherwood-Relics simulation suite, consisting of diverse thermal histories, to generate mock spectra. Our convolutional and residual networks with likelihood metric predicts the Ly$\alpha$ optical depth-weighted density or temperature for each pixel in the Ly$\alpha$ forest skewer. We find that our network can successfully reproduce IGM conditions with high fidelity across range of instrumental signal-to-noise. These predictions are subsequently translated into the temperature-density plane, facilitating the derivation of reliable constraints on thermal parameters. This allows us to estimate temperature at mean cosmic density, $T_{\rm 0}$ with one sigma confidence $\delta T_{\rm 0} \sim 1000{\rm K}$ using only one $20$Mpc/h sightline ($\Delta z\simeq 0.04$) with a typical reionization history. Existing studies utilize redshift pathlength comparable to $\Delta z\simeq 4$ for similar constraints. We can also provide more stringent constraints on the slope ($1\sigma$ confidence interval $\delta {\rm \gamma} \lesssim 0.1$) of the IGM temperature-density relation as compared to other traditional approaches. We test the reconstruction on a single high signal-to-noise observed spectrum ($20$ Mpc/h segment), and recover thermal parameters consistent with current measurements. This machine learning approach has the potential to provide accurate yet robust measurements of IGM thermal history at the redshifts in question.