Abstract
Abstract We present a new application of deep learning to reconstruct the cosmic microwave background (CMB) temperature maps from images of the microwave sky and to use these reconstructed maps to estimate the masses of galaxy clusters. We use a feed-forward deep-learning network, mResUNet, for both steps of the analysis. The first deep-learning model, mResUNet-I, is trained to reconstruct foreground and noise-suppressed CMB maps from a set of simulated images of the microwave sky that include signals from the CMB, astrophysical foregrounds like dusty and radio galaxies, instrumental noise as well as the cluster’s own thermal Sunyaev–Zel’dovich signal. The second deep-learning model, mResUNet-II, is trained to estimate cluster masses from the gravitational-lensing signature in the reconstructed foreground and noise-suppressed CMB maps. For SPTpol-like noise levels, the trained mResUNet-II model recovers the mass for 104 galaxy cluster samples with a 1σ uncertainty Δ M 200 c est / M 200 c est = 0.108 and 0.016 for input cluster mass M 200 c true = 10 14 M ⊙ and 8 × 1014 M ⊙, respectively. We also test for potential bias on recovered masses, finding that for a set of 105 clusters the estimator recovers M 200 c est = 2.02 × 10 14 M ⊙ , consistent with the input at 1% level. The 2σ upper limit on potential bias is at 3.5% level.
Highlights
The number density of galaxy clusters is a promising approach to constrain cosmological models, especially those affecting late-time structure growth (e.g. Mantz et al 2008; Vikhlinin et al 2009; Hasselfield et al 2013; Planck Collaboration et al 2016a; de Haan et al 2016; Bocquet et al 2019; Costanzi et al 2019)
We present a new application of deep learning to reconstruct the cosmic microwave background (CMB) temperature maps from the images of microwave sky, and to use these reconstructed maps to estimate the masses of galaxy clusters
We find that the fractional mass uncertainties from the deep learning method are factors of 1.3 to 1.45 higher than maximum likelihood estimator (MLE) estimates in the compared mass range
Summary
The number density of galaxy clusters is a promising approach to constrain cosmological models, especially those affecting late-time structure growth (e.g. Mantz et al 2008; Vikhlinin et al 2009; Hasselfield et al 2013; Planck Collaboration et al 2016a; de Haan et al 2016; Bocquet et al 2019; Costanzi et al 2019). We reconstruct the CMB temperature maps from the simulated images of microwave sky maps This is done by training the mResUNet-I network to learn CMB features and mitigate the foreground (tSZ and astrophysical) signals as well as instrumental noise. We use the reconstructed CMB temperature maps and mResUNet-II network to estimate the underlying mass for individual galaxy clusters. The mResUNet-II network is trained to extract lensing features from CMB temperature maps After training these models for CMB reconstruction and mass estimation, we test the robustness of the process by using external hydrodynamical simulations of galaxy clusters. The central mass and the 1-σ uncertainty is calculat√ed as the median and the standard deviation divided by N (standard error), where N is the number of clusters, respectively
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
