arXiv subjects

J. Read

Publications and source records attributed to J. Read.

10 recordsLinked to original sources

arXiv2025

Why the Northern Hemisphere Needs a 30-40 m Telescope and the Science at Stake: Resolved Stellar Populations Studies in M31 and its Satellites

A 30 m class optical/near-IR telescope in the Northern Hemisphere, equipped for diffraction-limited imaging and high-resolution, multi-object spectroscopy of faint stars, would enable a transformational investigation of the formation and evolution of M31 and its satellite system - on par with what Gaia, the HST, and other major photometric and spectroscopic facilities have achieved for the Milky Way (MW) and its satellites. The unprecedented detail obtained for our home system has reshaped our understanding of the assembly of the MW disk, halo, and bulge, and that of its satellites, which now serve as a benchmark for galaxy formation and evolution models. Extending this level of insight to the M31 system - that of the nearest massive spiral and the only one for which such a comprehensive, resolved stellar population study is feasible - will allow us to address a fundamental question: how representative is the MW and its satellite system within the broader context of galaxy evolution?

astro-ph.IM↗

arXiv2025

A Bayesian approach to time-domain Photonic Doppler Velocimetry

Photonic Doppler Velocimetry (PDV) is an established technique for measuring the velocities of fast-moving surfaces in high-energy-density experiments. In the standard approach to PDV analysis, a short-time Fourier transform (STFT) is used to generate a spectrogram from which the velocity history of the target is inferred. The user chooses the form, duration and separation of the window function. Here we present a Bayesian approach to infer the velocity directly from the PDV oscilloscope trace, without using the spectrogram for analysis. This is clearly a difficult inference problem due to the highly-periodic nature of the data, but we find that with carefully chosen prior distributions for the model parameters we can accurately recover the injected velocity from synthetic data. We validate this method using PDV data collected at the STAR two-stage light gas gun at Sandia National Laboratories, recovering shock-front velocities in quartz that are consistent with those inferred using the STFT-based approach, and are interpolated across regions of low signal-to-noise data. Although this method does not rely on the same user choices as the STFT, we caution that it can be prone to misspecification if the chosen model is not sufficient to capture the velocity behavior. Analysis using posterior predictive checks can be used to establish if a better model is required, although more complex models come with additional computational cost, often taking more than several hours to converge when sampling the Bayesian posterior. We therefore recommend it be viewed as a complementary method to that of the STFT-based approach.

physics.plasm-ph↗

arXiv2022

Expanding RIFT: Improving performance for GW parameter inference

The Rapid Iterative FiTting (RIFT) parameter inference algorithm provides a framework for efficient, highly-parallelized parameter inference for GW sources. In this paper, we summarize essential algorithm enhancements and operating point choices for the RIFT iterative algorithm, including choices used for analysis of LIGO/Virgo O3 observations. We also describe other extensions to the RIFT algorithm and software ecosystem. Some extensions increase RIFT's flexibility to produce outputs pertinent to GW astrophysics. Other extensions increase its computational efficiency or stability. Using many randomly-selected sources, we assess code robustness with two distinct code configurations, one designed to mimic settings as of LIGO O3 and another employing several performance enhancements. We illustrate RIFT's capabilities with analysis of selected events.

gr-qc↗

arXiv2021

A survey of Monte Carlo methods for noisy and costly densities with application to reinforcement learning and ABC

This survey gives an overview of Monte Carlo methodologies using surrogate models, for dealing with densities which are intractable, costly, and/or noisy. This type of problem can be found in numerous real-world scenarios, including stochastic optimization and reinforcement learning, where each evaluation of a density function may incur some computationally-expensive or even physical (real-world activity) cost, likely to give different results each time. The surrogate model does not incur this cost, but there are important trade-offs and considerations involved in the choice and design of such methodologies. We classify the different methodologies into three main classes and describe specific instances of algorithms under a unified notation. A modular scheme which encompasses the considered methods is also presented. A range of application scenarios is discussed, with special attention to the likelihood-free setting and reinforcement learning. Several numerical comparisons are also provided.

cs.LG↗

arXiv2020

Fluctuations in galactic bar parameters due to bar-spiral interaction

We study the late-time evolution of the central regions of two Milky Way-like simulations of galaxies formed in a cosmological context, one hosting a fast bar and the other a slow one. We find that bar length, R_b, measurements fluctuate on a dynamical timescale by up to 100%, depending on the spiral structure strength and measurement threshold. The bar amplitude oscillates by about 15%, correlating with R_b. The Tremaine-Weinberg-method estimates of the bars' instantaneous pattern speeds show variations around the mean of up to ~20%, typically anti-correlating with the bar length and strength. Through power spectrum analyses, we establish that these bar pulsations, with a period in the range ~60-200 Myr, result from its interaction with multiple spiral modes, which are coupled with the bar. Because of the presence of odd spiral modes, the two bar halves typically do not connect at exactly the same time to a spiral arm, and their individual lengths can be significantly offset. We estimated that in about 50% of bar measurements in Milky Way-mass external galaxies, the bar lengths of SBab type galaxies are overestimated by ~15% and those of SBbc types by ~55%. Consequently, bars longer than their corotation radius reported in the literature, dubbed "ultra-fast bars", may simply correspond to the largest biases. Given that the Scutum-Centaurus arm is likely connected to the near half of the Milky Way bar, recent direct measurements may be overestimating its length by 1-1.5 kpc, while its present pattern speed may be 5-10 km/s/kpc smaller than its time-averaged value.

astro-ph.GA↗

arXiv2015

Scalable Multi-Output Label Prediction: From Classifier Chains to Classifier Trellises

Multi-output inference tasks, such as multi-label classification, have become increasingly important in recent years. A popular method for multi-label classification is classifier chains, in which the predictions of individual classifiers are cascaded along a chain, thus taking into account inter-label dependencies and improving the overall performance. Several varieties of classifier chain methods have been introduced, and many of them perform very competitively across a wide range of benchmark datasets. However, scalability limitations become apparent on larger datasets when modeling a fully-cascaded chain. In particular, the methods' strategies for discovering and modeling a good chain structure constitutes a mayor computational bottleneck. In this paper, we present the classifier trellis (CT) method for scalable multi-label classification. We compare CT with several recently proposed classifier chain methods to show that it occupies an important niche: it is highly competitive on standard multi-label problems, yet it can also scale up to thousands or even tens of thousands of labels.

stat.ML↗

arXiv2013

Global Network of Optical Magnetometers for Exotic (GNOME): Physics Novel scheme for exotic physics searches

A novel experimental scheme enabling investigation of transient exotic spin couplings is discussed. The scheme is based on synchronous measurements of optical-magnetometer signals of several devices operating in magnetically shielded environments in distant locations ($\gtrsim100$ km). Although signatures of such exotic couplings may be present in the signal from a single magnetometer, it would be challenging to distinguish them from noise. By analyzing the correlation between signals from multiple, geographically separated magnetometers, it is not only possible to identify the exotic transient but also to investigate its nature. The ability of the network to probe presently unconstrained physics beyond the Standard Model is examined by considering the spin coupling to stable topological defects (e.g., domain walls) of axion-like fields. In the spirit of this research, a brief ($\sim 2$ hours) demonstration experiment involving two magnetometers located in Krak\'ow and Berkeley ($\sim 9000$ km separation) is presented and discussion of the data-analysis approaches that may allow for the transient signal identification is provided. The prospects of the network are outlined in the last part of the paper.

physics.atom-ph↗

arXiv2009

Gravitational waves from neutron stars: Promises and challenges

We discuss different ways that neutron stars can generate gravitational waves, describe recent improvements in modelling the relevant scenarios in the context of improving detector sensitivity, and show how observations are beginning to test our understanding of fundamental physics. The main purpose of the discussion is to establish promising science goals for third-generation ground-based detectors, like the Einstein Telescope, and identify the various challenges that need to be met if we want to use gravitational-wave data to probe neutron star physics.

astro-ph.SR↗

arXiv2009

Detecting the Milky Way's Dark Disk

In LambdaCDM, massive satellites are dragged into the disk-plane by dynamical friction where they dissolve into a stellar thick disk and a more massive dark matter disk. The distinctive kinematics of the dark disk matches the stars that also entered in the satellites. The lower velocities of the dark disk with respect to the Earth enhances detection rates at low recoil energy. For WIMP masses > 50 GeV/c^2, the detection rate increases by up to a factor of 3 in the 5 - 20 keV recoil energy range. Comparing this with rates at higher energy will improve constraints on the WIMP mass, particularly for masses > 100 GeV/c^2. The annual modulation signal of the dark disk is significantly boosted and its modulation phase is shifted by ~3 weeks relative to the dark halo. The variation of the observed phase with recoil energy can also be used to determine the WIMP mass once the dark disk properties are fixed by future astronomical surveys. The constraints on the WIMP interaction cross section from current experiments improve by factors of 1.4 to 3.5 when a typical contribution from the dark disk is included.

astro-ph↗

arXiv2008

Signatures of the Milky Way's Dark Disk in Current and Future Experiments

In hierarchical structure formation models of disk galaxies, a dark matter disk forms as massive satellites are preferentially dragged into the disk-plane where they dissolve. Here, we quantify the importance of this dark disk for direct and indirect dark matter detection. The low velocity of the dark disk with respect to the Earth enhances detection rates in direct detection experiments at low recoil energy. For WIMP masses M_{WIMP} >~ 50 GeV, the detection rate increases by up to a factor of 3 in the 5 - 20 keV recoil energy range. Comparing this with rates at higher energy is sensitive to M_{WIMP}, providing stronger mass constraints particularly for M_{WIMP}>~100 GeV. The annual modulation signal is significantly boosted by the dark disk and the modulation phase is shifted by ~3 weeks relative to the dark halo. The variation of the observed phase with recoil energy determines M_{WIMP}, once the dark disk properties are fixed by future astronomical surveys. The low velocity of the particles in the dark disk with respect to the solar system significantly enhances the capture rate of WIMPs in the Sun, leading to an increased flux of neutrinos from the Sun which could be detected in current and future neutrino telescopes. The dark disk contribution to the muon flux from neutrino back conversion at the Earth is increased by a factor of ~5 compared to the SHM, for rho_d/rho_h=0.5.

astro-ph↗