Deep learning the astrometric signature of dark matter substructure

Vattis, Kyriakos; Toomey, Michael W.; Koushiappas, Savvas M.

doi:10.1103/physrevd.104.123541

Cited by 10 publications

(6 citation statements)

References 90 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the data obtained from many cosmological observations can be processed as two-and three-dimensional images, CNNs are well-suited to several aspects of cosmological simulations and data analysis. For example, some current CNN applications in cosmology include producing full-sky CMB simulations [79], identification of HII regions in reionization [80], analysis of dark matter substructure [81,82], and cosmic velocity field reconstruction [83].…”

Section: Resunet-cmb Architecture and Methodsmentioning

confidence: 99%

Reconstructing cosmic polarization rotation with ResUNet-CMB

Guzman

Meyers

2022

J. Cosmol. Astropart. Phys.

View full text Add to dashboard Cite

Cosmic polarization rotation, which may result from parity-violating new physics or the presence of primordial magnetic fields, converts E-mode polarization of the cosmic microwave background (CMB) into B-mode polarization. Anisotropic cosmic polarization rotation leads to statistical anisotropy in CMB polarization and can be reconstructed with quadratic estimator techniques similar to those designed for gravitational lensing of the CMB. At the sensitivity of upcoming CMB surveys, lensing-induced B-mode polarization will act as a limiting factor in the search for anisotropic cosmic polarization rotation, meaning that an analysis which incorporates some form of delensing will be required to improve constraints on the effect with future surveys. In this paper we extend the ResUNet-CMB convolutional neural network to reconstruct anisotropic cosmic polarization rotation in the presence of gravitational lensing and patchy reionization, and we show that the network simultaneously reconstructs all three effects with variance that is lower than that from the standard quadratic estimator nearly matching the performance of an iterative reconstruction method.

show abstract

Section: Resunet-cmb Architecture and Methodsmentioning

confidence: 99%

Reconstructing cosmic polarization rotation with ResUNet-CMB

Guzman

Meyers

2022

J. Cosmol. Astropart. Phys.

View full text Add to dashboard Cite

show abstract

“…For this task, we use a CNN, specifically EfficientNet (Tan & Le 2019), as our base architecture. This is the same type of architecture that has achieved top performance in previous applications to lensing data sets (Alexander et al 2020b(Alexander et al , 2020cVattis et al 2020). More generally, CNNs are known to outperform other methods of classification for strong gravitational lenses (Metcalf et al 2019).…”

Section: Domain Adaptationmentioning

confidence: 99%

“…A particularly well-suited signature that can be used to distinguish among dark matter models is the morphology and distribution of its substructure within dark matter halos. Some promising directions for inferring the properties of substructure include tidal streams (Ngan & Carlberg 2014;Bovy 2016;Carlberg 2016;Erkal et al 2016;Benito et al 2020;Shih et al 2021) and astrometric observations (Feldmann & Spolyar 2015;Sanderson et al 2016;Van Tilburg et al 2018;Mishra-Sharma et al 2020;Vattis et al 2020;Mishra-Sharma 2022;Pardo & Doré 2021). A particularly sensitive probe is strong gravitational lensing (Buckley & Peter 2018;Drlica-Wagner et al 2019;Simon et al 2019), to which we restrict ourselves in this paper.…”

Section: Introductionmentioning

confidence: 99%

Domain Adaptation for Simulation-based Dark Matter Searches with Strong Gravitational Lensing

Alexander,

Gleyzer,

Parul

et al. 2023

ApJ

Self Cite

View full text Add to dashboard Cite

The identity of dark matter has remained surprisingly elusive. While terrestrial experiments may be able to nail down a model, an alternative method is to identify dark matter based on astrophysical or cosmological signatures. A particularly sensitive approach is based on the unique signature of dark matter substructure in galaxy–galaxy strong lensing images. Machine-learning applications have been explored for extracting this signal. Because of the limited availability of high-quality strong lensing images, these approaches have exclusively relied on simulations. Due to the differences with the real instrumental data, machine-learning models trained on simulations are expected to lose accuracy when applied to real data. Here domain adaptation can serve as a crucial bridge between simulations and real data applications. In this work, we demonstrate the power of domain adaptation techniques applied to strong gravitational lensing data with dark matter substructure. We show with simulated data sets representative of Euclid and Hubble Space Telescope observations that domain adaptation can significantly mitigate the losses in the model performance when applied to new domains. Lastly, we find similar results utilizing domain adaptation for the problem of lens finding by adapting models trained on a simulated data set to one composed of real lensed and unlensed galaxies from the Hyper Suprime-Cam. This technique can help domain experts build and apply better machine-learning models for extracting useful information from the strong gravitational lensing data expected from the upcoming surveys.

show abstract

“…This requires studying perturbations to the precise Gaia astrometric solution. As with the photometric case, these perturbations can arise in the time-domain [297,298,299] or may be revealed through higher-point correlations across a broader statistical sample [300]. The hope is that these methods of astrometry can provide a census of dark compact objects.…”

Section: Patterns In Milky Way Field Starsmentioning

confidence: 99%

The Milky Way, Coming into Focus: Precision Astrometry Probes its Evolution, and its Dark Matter

Gardner,

McDermott,

Yanny

2021

Preprint

View full text Add to dashboard Cite

The growing trove of precision astrometric observations from the Gaia space telescope and other surveys is revealing the structure and dynamics of the Milky Way in ever more exquisite detail. We summarize the current status of our understanding of the structure and the characteristics of the Milky Way, and we review the emerging picture: the Milky Way is evolving through interactions with the massive satellite galaxies that stud its volume, with evidence pointing to a cataclysmic past. It is also woven with stellar streams, and observations of streams, satellites, and field stars offer new constraints on its dark matter, both on its spatial distribution and its fundamental nature. The recent years have brought much focus to the study of dwarf galaxies found within our Galaxy's halo and their internal matter distributions. In this review, we focus on the predictions of the cold dark matter paradigm at small mass scales through precision astrometric measurements, and we summarize the modern consensus on the extent to which small-scale probes are consistent with this paradigm. We note the discovery prospects of these studies, and also how they intertwine with probes of the dynamics and evolution of the Milky Way in various and distinct ways.

show abstract

Deep learning the astrometric signature of dark matter substructure

Cited by 10 publications

References 90 publications

Reconstructing cosmic polarization rotation with ResUNet-CMB

Reconstructing cosmic polarization rotation with ResUNet-CMB

Domain Adaptation for Simulation-based Dark Matter Searches with Strong Gravitational Lensing

The Milky Way, Coming into Focus: Precision Astrometry Probes its Evolution, and its Dark Matter

Contact Info

Product

Resources

About