Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural Network

The thermal Sunyaev–Zel’dovich (SZ) effect offers a unique probe of the hot and diffuse Universe that could help close the missing baryon problem. Traditional extractions of the SZ effect, however, exhibit systematic noise that may lead to unreliable results. In this work, we provide an alternative...

Full description

Saved in:
Bibliographic Details
Main Authors: Cameron T. Pratt, Zhijie Qu, Joel N. Bregman
Format: Article
Language:English
Published: IOP Publishing 2025-01-01
Series:The Astrophysical Journal
Subjects:
Online Access:https://doi.org/10.3847/1538-4357/adf069
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849230645034745856
author Cameron T. Pratt
Zhijie Qu
Joel N. Bregman
author_facet Cameron T. Pratt
Zhijie Qu
Joel N. Bregman
author_sort Cameron T. Pratt
collection DOAJ
description The thermal Sunyaev–Zel’dovich (SZ) effect offers a unique probe of the hot and diffuse Universe that could help close the missing baryon problem. Traditional extractions of the SZ effect, however, exhibit systematic noise that may lead to unreliable results. In this work, we provide an alternative solution using a 3D attention nested U-Net trained end to end with supervised learning. Our labeled data consist of simulated SZ signals injected into Planck frequency maps, allowing our model to learn how to extract SZ signals in the presence of realistic noise. We implement a curriculum learning scheme that gradually exposed the model to weaker SZ signals. The absence/presence of curriculum learning significantly impacted the amount of bias and variance present in the reconstructed SZ signal. The results from our method were comparable to those from the popular needlet internal linear combination method when evaluated on simulated data as well as real-world SZ signals. We conclude by discussing future avenues for advancing machine learning extractions of SZ signals.
format Article
id doaj-art-36e4fed9a8e54b5b8ef1b9567a367d5e
institution Kabale University
issn 1538-4357
language English
publishDate 2025-01-01
publisher IOP Publishing
record_format Article
series The Astrophysical Journal
spelling doaj-art-36e4fed9a8e54b5b8ef1b9567a367d5e2025-08-21T06:01:08ZengIOP PublishingThe Astrophysical Journal1538-43572025-01-019901210.3847/1538-4357/adf069Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural NetworkCameron T. Pratt0https://orcid.org/0000-0002-6653-8490Zhijie Qu1https://orcid.org/0000-0002-2941-646XJoel N. Bregman2https://orcid.org/0000-0001-6276-9526Department of Astronomy, University of Michigan , Ann Arbor, MI 48109, USA ; campratt@umich.eduDepartment of Astronomy & Astrophysics, The University of Chicago , 5640 S. Ellis Ave., Chicago, IL 60637, USA; Department of Astronomy, Tsinghua University , Beijing, People's Republic of ChinaDepartment of Astronomy, University of Michigan , Ann Arbor, MI 48109, USA ; campratt@umich.eduThe thermal Sunyaev–Zel’dovich (SZ) effect offers a unique probe of the hot and diffuse Universe that could help close the missing baryon problem. Traditional extractions of the SZ effect, however, exhibit systematic noise that may lead to unreliable results. In this work, we provide an alternative solution using a 3D attention nested U-Net trained end to end with supervised learning. Our labeled data consist of simulated SZ signals injected into Planck frequency maps, allowing our model to learn how to extract SZ signals in the presence of realistic noise. We implement a curriculum learning scheme that gradually exposed the model to weaker SZ signals. The absence/presence of curriculum learning significantly impacted the amount of bias and variance present in the reconstructed SZ signal. The results from our method were comparable to those from the popular needlet internal linear combination method when evaluated on simulated data as well as real-world SZ signals. We conclude by discussing future avenues for advancing machine learning extractions of SZ signals.https://doi.org/10.3847/1538-4357/adf069Sunyaev-Zeldovich effectCosmic microwave background radiationAstronomical simulations
spellingShingle Cameron T. Pratt
Zhijie Qu
Joel N. Bregman
Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural Network
The Astrophysical Journal
Sunyaev-Zeldovich effect
Cosmic microwave background radiation
Astronomical simulations
title Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural Network
title_full Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural Network
title_fullStr Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural Network
title_full_unstemmed Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural Network
title_short Supervised Extraction of the Thermal Sunyaev–Zel’dovich Effect with a Three-dimensional Convolutional Neural Network
title_sort supervised extraction of the thermal sunyaev zel dovich effect with a three dimensional convolutional neural network
topic Sunyaev-Zeldovich effect
Cosmic microwave background radiation
Astronomical simulations
url https://doi.org/10.3847/1538-4357/adf069
work_keys_str_mv AT camerontpratt supervisedextractionofthethermalsunyaevzeldovicheffectwithathreedimensionalconvolutionalneuralnetwork
AT zhijiequ supervisedextractionofthethermalsunyaevzeldovicheffectwithathreedimensionalconvolutionalneuralnetwork
AT joelnbregman supervisedextractionofthethermalsunyaevzeldovicheffectwithathreedimensionalconvolutionalneuralnetwork