Cosmological constraints from low redshift 21 cm intensity mapping with machine learning

Abstract

The future 21 cm intensity mapping observations constitute a promising way to trace the matter distribution of the Universe and probe cosmology. Here we assess its capability for cosmological constraints using as a case study the BINGO radio telescope, that will survey the Universe at low redshifts (0.13 < z < 0.45). We use neural networks (NNs) to map summary statistics, namely, the angular power spectrum (APS) and the Minkowski functionals (MFs), calculated from simulations into cosmological parameters. Our simulations span a wide grid of cosmologies, sampled under the ΛCDM scenario, {Ωc, h}, and under an extension assuming the Chevallier-Polarski-Linder (CPL) parameterization, {Ωc, h, w0, wa}. In general, NNs trained over APS outperform those using MFs, while their combination provides 27% (5%) tighter error ellipse in the Ωc − h plane under the ΛCDM scenario (CPL parameterization) compared to the individual use of the APS. Their combination allows predicting Ωc and h with 4.9% and 1.6% fractional errors, respectively, which increases to 6.4% and 3.7% under CPL parameterization. Although we find large bias on wa estimates, we still predict w0 with 24.3% error. We also confirm our results to be robust to foreground contamination, besides finding the instrumental noise to cause the greater impact on the predictions. Still, our results illustrate the capability of future low redshift 21 cm observations in providing competitive cosmological constraints using NNs, showing the ease of combining different summary statistics.

Deep Learning Voigt Profiles. I. Single-Cloud Doublets

“…Then, real-world data are presented to the network, which predicts the characteristic properties in the real data. Successful applications include measuring galactic star formation rates (e.g., Delli Veneri et al 2019;Simet et al 2021;Euclid Collaboration et al 2023;Santos-Olmsted et al 2023), metallicities (e.g., Liew-Cain et al 2021), stellar masses and redshifts (e.g., Bonjean et al 2019;Wu & Boada 2019;Surana et al 2020), masses of galaxy clusters (e.g., Ntampaka et al 2015), cosmological parameters from weak lensing (e.g., Gupta et al 2018), and large-scale structure formation (e.g., He et al 2018), identifying reionization sources (e.g., Hassan et al 2019) and the duration of reionization (e.g., La Plante & Ntampaka 2019), and constraining cosmological parameters (e.g., Fluri et al 2019;Ribli et al 2019;Hassan et al 2020;Matilla et al 2020;Ntampaka et al 2020;Ntampaka & Vikhlinin 2022;Andrianomena & Hassan 2023;Bengaly et al 2023;Lu et al 2023;Novaes et al 2024;Qiu et al 2023). Recently, Monadi et al (2023) applied Gaussian processes to Sloan Digital Sky Survey (SDSS) DR12 quasar spectra to detect C IV absorbers and measure their VP parameters.…”