Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural Networks

The phase space of stellar streams is proposed to detect dark substructure in the Milky Way through the perturbations created by passing subhalos—and thus is a powerful test of the cold dark matter paradigm and its alternatives. Using graph convolutional neural network (GCNN) data compression and si...

Full description

Saved in:
Bibliographic Details
Main Authors: Peter Xiangyuan Ma, Keir K. Rogers, Ting S. Li, Renée Hložek, Jeremy J. Webb, Ruth Huang, Julian Meunier
Format: Article
Language:English
Published: IOP Publishing 2025-01-01
Series:The Astrophysical Journal
Subjects:
Online Access:https://doi.org/10.3847/1538-4357/add698
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850106791673724928
author Peter Xiangyuan Ma
Keir K. Rogers
Ting S. Li
Renée Hložek
Jeremy J. Webb
Ruth Huang
Julian Meunier
author_facet Peter Xiangyuan Ma
Keir K. Rogers
Ting S. Li
Renée Hložek
Jeremy J. Webb
Ruth Huang
Julian Meunier
author_sort Peter Xiangyuan Ma
collection DOAJ
description The phase space of stellar streams is proposed to detect dark substructure in the Milky Way through the perturbations created by passing subhalos—and thus is a powerful test of the cold dark matter paradigm and its alternatives. Using graph convolutional neural network (GCNN) data compression and simulation-based inference (SBI) on a simulated GD-1-like stream, we improve the constraint on the mass of a [10 ^8 , 10 ^7 , 10 ^6 ] M _⊙ perturbing subhalo by factors of [11, 7, 3] with respect to the current state-of-the-art density power spectrum analysis. We find that the GCNN produces posteriors that are more accurate (better calibrated) than the power spectrum. We simulate the positions and velocities of stars in a GD-1-like stream and perturb the stream with subhalos of varying mass and velocity. Leveraging the feature encoding of the GCNN to compress the input phase space data, we then use SBI to estimate the joint posterior of the subhalo mass and velocity. We investigate how our results scale with the size of the GCNN, the coordinate system of the input, and the effect of incomplete observations. Our results suggest that a survey with 10× fewer stars (300 stars) with complete 6D phase space data performs about as well as a deeper survey (3000 stars) with only 3D data (photometry, spectroscopy). The stronger constraining power and more accurate posterior estimation motivate further development of GCNNs in combining future photometric, spectroscopic, and astrometric stream observations.
format Article
id doaj-art-44dae0dbbc3e448995992b2e1974579c
institution OA Journals
issn 1538-4357
language English
publishDate 2025-01-01
publisher IOP Publishing
record_format Article
series The Astrophysical Journal
spelling doaj-art-44dae0dbbc3e448995992b2e1974579c2025-08-20T02:38:45ZengIOP PublishingThe Astrophysical Journal1538-43572025-01-0198719610.3847/1538-4357/add698Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural NetworksPeter Xiangyuan Ma0https://orcid.org/0000-0001-8975-3719Keir K. Rogers1Ting S. Li2https://orcid.org/0000-0002-9110-6163Renée Hložek3https://orcid.org/0000-0002-0965-7864Jeremy J. Webb4https://orcid.org/0000-0003-3613-0854Ruth Huang5Julian Meunier6Department of Astronomy , UC Berkeley, 501 Campbell Hall, Berkeley, CA, 94720, USA ; peter_ma@berkeley.edu; Department of Mathematics, University of Toronto , 40 St. George Street, Toronto, ON, M5S 2E4, CanadaDepartment of Physics, Imperial College London , Blackett Laboratory, Prince Consort Road, London, SW7 2AZ, UK; Dunlap Institute for Astronomy and Astrophysics, University of Toronto , 50 St. George Street, Toronto, ON, M5S 3H4, CanadaDavid A. Dunlap Department of Astronomy and Astrophysics, University of Toronto , 50 St. George Street, Toronto, ON, M5S 3H4, CanadaDunlap Institute for Astronomy and Astrophysics, University of Toronto , 50 St. George Street, Toronto, ON, M5S 3H4, Canada; David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto , 50 St. George Street, Toronto, ON, M5S 3H4, CanadaDavid A. Dunlap Department of Astronomy and Astrophysics, University of Toronto , 50 St. George Street, Toronto, ON, M5S 3H4, Canada; Department of Science, Technology and Society, Division of Natural Science, York University , 218 Bethune College, Toronto, ON, M3J 1P3, CanadaDavid A. Dunlap Department of Astronomy and Astrophysics, University of Toronto , 50 St. George Street, Toronto, ON, M5S 3H4, CanadaDavid A. Dunlap Department of Astronomy and Astrophysics, University of Toronto , 50 St. George Street, Toronto, ON, M5S 3H4, CanadaThe phase space of stellar streams is proposed to detect dark substructure in the Milky Way through the perturbations created by passing subhalos—and thus is a powerful test of the cold dark matter paradigm and its alternatives. Using graph convolutional neural network (GCNN) data compression and simulation-based inference (SBI) on a simulated GD-1-like stream, we improve the constraint on the mass of a [10 ^8 , 10 ^7 , 10 ^6 ] M _⊙ perturbing subhalo by factors of [11, 7, 3] with respect to the current state-of-the-art density power spectrum analysis. We find that the GCNN produces posteriors that are more accurate (better calibrated) than the power spectrum. We simulate the positions and velocities of stars in a GD-1-like stream and perturb the stream with subhalos of varying mass and velocity. Leveraging the feature encoding of the GCNN to compress the input phase space data, we then use SBI to estimate the joint posterior of the subhalo mass and velocity. We investigate how our results scale with the size of the GCNN, the coordinate system of the input, and the effect of incomplete observations. Our results suggest that a survey with 10× fewer stars (300 stars) with complete 6D phase space data performs about as well as a deeper survey (3000 stars) with only 3D data (photometry, spectroscopy). The stronger constraining power and more accurate posterior estimation motivate further development of GCNNs in combining future photometric, spectroscopic, and astrometric stream observations.https://doi.org/10.3847/1538-4357/add698Galaxy dynamicsGalaxy dark matter halosNeural networksAstronomical simulations
spellingShingle Peter Xiangyuan Ma
Keir K. Rogers
Ting S. Li
Renée Hložek
Jeremy J. Webb
Ruth Huang
Julian Meunier
Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural Networks
The Astrophysical Journal
Galaxy dynamics
Galaxy dark matter halos
Neural networks
Astronomical simulations
title Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural Networks
title_full Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural Networks
title_fullStr Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural Networks
title_full_unstemmed Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural Networks
title_short Toward Characterizing Dark Matter Subhalo Perturbations in Stellar Streams with Graph Neural Networks
title_sort toward characterizing dark matter subhalo perturbations in stellar streams with graph neural networks
topic Galaxy dynamics
Galaxy dark matter halos
Neural networks
Astronomical simulations
url https://doi.org/10.3847/1538-4357/add698
work_keys_str_mv AT peterxiangyuanma towardcharacterizingdarkmattersubhaloperturbationsinstellarstreamswithgraphneuralnetworks
AT keirkrogers towardcharacterizingdarkmattersubhaloperturbationsinstellarstreamswithgraphneuralnetworks
AT tingsli towardcharacterizingdarkmattersubhaloperturbationsinstellarstreamswithgraphneuralnetworks
AT reneehlozek towardcharacterizingdarkmattersubhaloperturbationsinstellarstreamswithgraphneuralnetworks
AT jeremyjwebb towardcharacterizingdarkmattersubhaloperturbationsinstellarstreamswithgraphneuralnetworks
AT ruthhuang towardcharacterizingdarkmattersubhaloperturbationsinstellarstreamswithgraphneuralnetworks
AT julianmeunier towardcharacterizingdarkmattersubhaloperturbationsinstellarstreamswithgraphneuralnetworks