arXiv subjects

Szabolcs Márka

Publications and source records attributed to Szabolcs Márka.

At least 19 recordsLinked to original sources

arXiv2026

Artificial Precision Timing Array: bridging the decihertz gravitational-wave sensitivity gap with clock satellites

Gravitational-wave astronomy has developed enormously over the last decade, with the first detections and continuous development across broad frequency bands. However, the decihertz range has largely been left out of this development. Gravitational waves in this band are emitted by some of the most enigmatic sources, including intermediate-mass binary black hole mergers, early inspiraling compact binaries$\unicode{x2014}$whose mergers are seen by Earth-based detectors$\unicode{x2014}$, and possibly primordial gravitational waves. To tap this exciting band, we propose the construction of a detector based on pulsar timing principles, the Artificial Precision Timing Array (APTA). We envision APTA as a solar system array of artificial ``pulsars''$\unicode{x2014}$precision-time-reference-carrying satellites that emit periodic electromagnetic signals towards Earth or another satellite constellation receiver location. In this fundamental study, we estimate the clock precision needed for gravitational-wave detection with APTA. Our results suggest that 6 satellites and a clock relative uncertainty of $10^{-18}$ at 1~s of averaging, which is currently attainable with ground-based atomic clocks, would be sufficient for APTA to reach pristine sensitivity in the decihertz band and observe $10^3\unicode{x2013}10^4$ $\mathrm{M}_\odot$ black hole mergers and the early inspiral of heavy LIGO-Virgo-KAGRA sources. Future clock and oscillator technologies realistically expected in the next decade(s) would enable the detection of an increasingly diverse set of sources, allowing APTA to reach a better sensitivity than other detector concepts proposed for the decihertz band. This work opens up a new area of research into designing and constructing gravitational-wave detectors relying on principles used successfully in pulsar timing.

astro-ph.IM↗

arXiv2025

Galactic forcing increases origination of marine microplankton

The continuous flux of Galactic cosmic rays that bombard Earth's atmosphere creates ionizing radiation that can damage the DNA of living organisms. While this radiation on Earth is relatively constant in the short term, large and long-scale fluctuations are expected with a period of $\sim 63.5$ million years. As the Solar System moves above or below the Galactic plane during its oscillatory motion about the Galactic center, the Galactic magnetic shielding weakens, allowing more cosmic rays to reach Earth and trigger mutations in organisms. We identify a significant correlation (weighted global p-value: $1.25\times 10^{-4}$, or $3.72σ$) between the Solar System's Galactic oscillations and the origination of marine zoo- and phyto-microplankton genera over the Phanerozoic. When we restrict the analysis to time intervals during which all four groups coexisted, a post-trial significance of $4.52σ$ emerges. Our findings suggest that changes in biodiversity have been significantly influenced by long-term Galactic forcing.

astro-ph.HE↗

arXiv2025

Fast, Accurate Manifold Denoising by Tunneling Riemannian Optimization

Learned denoisers play a fundamental role in various signal generation (e.g., diffusion models) and reconstruction (e.g., compressed sensing) architectures, whose success derives from their ability to leverage low-dimensional structure in data. Existing denoising methods, however, either rely on local approximations that require a linear scan of the entire dataset or treat denoising as generic function approximation problems, often sacrificing efficiency and interpretability. We consider the problem of efficiently denoising a new noisy data point sampled from an unknown $d$-dimensional manifold $M \in \mathbb{R}^D$, using only noisy samples. This work proposes a framework for test-time efficient manifold denoising, by framing the concept of "learning-to-denoise" as "learning-to-optimize". We have two technical innovations: (i) online learning methods which learn to optimize over the manifold of clean signals using only noisy data, effectively "growing" an optimizer one sample at a time. (ii) mixed-order methods which guarantee that the learned optimizers achieve global optimality, ensuring both efficiency and near-optimal denoising performance. We corroborate these claims with theoretical analyses of both the complexity and denoising performance of mixed-order traversal. Our experiments on scientific manifolds demonstrate significantly improved complexity-performance tradeoffs compared to nearest neighbor search, which underpins existing provable denoising approaches based on exhaustive search.

cs.LG↗

arXiv2025

Enhancing the significance of astrophysical events with multimessenger coincidences

Coincident multimessenger observations of cosmic sources can offer numerous benefits, especially when used in the context of synergistic astrophysics. One significant advantage is enhancing the detection significance of separate detectors by correlating their data and assuming joint emission. We have formulated an approach for updating the Bayesian posterior probability of an astrophysical origin, namely $p_{\rm astro}$, relying on multimessenger coincidences assuming an emission model. The description is applicable to any combination of messengers. We demonstrated the formalism for the gravitational waves and high-energy neutrinos case. Applying our method to the public data of candidate coincident high-energy neutrinos with subthreshold gravitational-wave triggers, we found that in the case of highly energetic neutrino coincidences, $p_{\rm astro}$ can increase from approximately $\sim 0.1$ to $\sim 0.9$. The amount of improvement depends on the assumed joint emission model. If models are trusted, the marked improvement makes subthreshold detections much more confident. Moreover, the model dependency can also be used to test the consistency of different models. This work is a crucial step toward the goal of uniting all detectors on equal footing into a statistically integrated, Earth-sized observatory for comprehensive multimessenger astrophysics.

astro-ph.HE↗

arXiv2023

Gravitational-wave matched filtering on a quantum computer

State of the art quantum computers have very limited applicability for accurate calculations. Here we report the first experimental demonstration of qubit-based matched filtering for a detection of the gravitational-wave signal from a binary black hole merger. With our implementation on noisy superconducting qubits, we obtained a similar signal-to-noise ratio for the binary black hole merger as achievable with classical computation, providing evidence for the utility of qubits for practically relevant tasks. The algorithm we invented for this application is a Monte Carlo algorithm which uses quantum and classical computation together. It provides a quasi-quadartic speed-up for time-domain convolution, similar to achievable with fast Fourier transform.

quant-ph↗

arXiv2023

Gamma-ray burst precursors from tidally resonant neutron star oceans: potential implications for GRB 211211A

Precursor emission has been observed seconds to minutes before some short gamma-ray bursts. While the origins of these precursors remain unknown, one potential explanation relies on the resonance of neutron star pulsational modes with the tidal forces during the inspiral phase of a compact binary merger. In this paper, we present a model for short gamma-ray burst precursors which relies on tidally resonant neutron star oceans. In this scenario, the onset of tidal resonance in the crust-ocean interface mode corresponds to the ignition of the precursor flare, possibly through the interaction between the excited neutron star ocean and the surface magnetic fields. From just the precursor total energy, the time before the main event, and a detected quasi-periodic oscillation frequency, we may constrain the binary parameters and neutron star ocean properties as never before. Our model can immediately distinguish neutron star-black hole mergers from binary neutron star mergers without gravitational wave detection. We apply our model to GRB 211211A, the recently detected long duration short gamma-ray burst with a quasi-periodic precursor, and explore the parameters of this system within its context. The precursor of GRB 211211A is consistent with a tidally resonant neutron star ocean explanation that requires an extreme-mass ratio NSBH merger and a high mass neutron star. While difficult to reconcile with the gamma-ray burst main emission and associated kilonova, our results constrain the possible precursor generating mechanisms in this system. A systematic study of short gamma-ray burst precursors with the model presented here can test precursor origin and could probe the possible connection between gamma-ray bursts and neutron star-black hole mergers.

astro-ph.HE↗

arXiv2023

TpopT: Efficient Trainable Template Optimization on Low-Dimensional Manifolds

In scientific and engineering scenarios, a recurring task is the detection of low-dimensional families of signals or patterns. A classic family of approaches, exemplified by template matching, aims to cover the search space with a dense template bank. While simple and highly interpretable, it suffers from poor computational efficiency due to unfavorable scaling in the signal space dimensionality. In this work, we study TpopT (TemPlate OPTimization) as an alternative scalable framework for detecting low-dimensional families of signals which maintains high interpretability. We provide a theoretical analysis of the convergence of Riemannian gradient descent for TpopT, and prove that it has a superior dimension scaling to covering. We also propose a practical TpopT framework for nonparametric signal sets, which incorporates techniques of embedding and kernel interpolation, and is further configurable into a trainable network architecture by unrolled optimization. The proposed trainable TpopT exhibits significantly improved efficiency-accuracy tradeoffs for gravitational wave detection, where matched filtering is currently a method of choice. We further illustrate the general applicability of this approach with experiments on handwritten digit data.

cs.LG↗

arXiv2023

The Timing System of LIGO Discoveries

LIGO's mission critical timing system has enabled gravitational wave and multi-messenger astrophysical discoveries as well as the rich science extracted. Achieving optimal detector sensitivity, detecting transient gravitational waves, and especially localizing gravitational wave sources, the underpinning of multi-messenger astrophysics, all require proper gravitational wave data time-stamping. Measurements of the relative arrival times of gravitational waves between different detectors allow for coherent gravitational wave detections, localization of gravitational wave sources, and the creation of skymaps. The carefully designed timing system achieves these goals by mitigating phase noise to avoid signal up-conversion and maximize gravitational wave detector sensitivity. The timing system also redundantly performs self-calibration and self-diagnostics in order to ensure reliable, extendable, and traceable time stamping. In this paper, we describe and quantify the performance of these core systems during the latest O3 scientific run of LIGO, Virgo, and KAGRA. We present results of the diagnostic checks done to verify the time-stamping for individual gravitational wave events observed during O3 as well as the timing system performance for all of O3 in LIGO Livingston and LIGO Hanford. We find that, after 3 observing runs, the LIGO timing system continues to reliably meet mission requirements of timing precision below 1 $μ$s with a significant safety margin.

astro-ph.IM↗

arXiv2023

Multi-messenger Emission from Tidal Waves in Neutron Star Oceans

Neutron stars in astrophysical binary systems represent exciting sources for multi-messenger astrophysics. A potential source of electromagnetic transients from compact binary systems is the neutron star ocean, the external fluid layer encasing a neutron star. We present a groundwork study into tidal waves in neutron star oceans and their consequences. Specifically, we investigate how oscillation modes in neutron star oceans can be tidally excited during compact binary inspirals and parabolic encounters. We find that neutron star oceans can sustain tidal waves with frequencies between $0.01-20$ Hz. Our results suggest that tidally resonant neutron star ocean waves may serve as a never-before studied source of precursor electromagnetic emission prior to neutron star-black hole and binary neutron star mergers. If accompanied by electromagnetic flares, tidally resonant neutron star ocean waves, whose energy budget can reach $10^{46}$ erg, may serve as early warning signs ($\gtrsim 1$ minute before merger) for compact binary mergers. Similarly, excited ocean tidal waves will coincide with neutron star parabolic encounters. Depending on the neutron star ocean model and a flare emission scenario, tidally resonant ocean flares may be detectable by Fermi and NuSTAR out to $\gtrsim 100$ Mpc with detection rates as high as $\sim 7$ yr$^{-1}$ for binary neutron stars and $\sim0.6$ yr$^{-1}$ for neutron star-black hole binaries. Observations of emission from neutron star ocean tidal waves along with gravitational waves will provide insight into the equation of state at the neutron star surface, the composition of neutron star oceans and crusts, and neutron star geophysics.

astro-ph.HE↗

arXiv2022

Boosting the Efficiency of Parametric Detection with Hierarchical Neural Networks

Gravitational wave astronomy is a vibrant field that leverages both classic and modern data processing techniques for the understanding of the universe. Various approaches have been proposed for improving the efficiency of the detection scheme, with hierarchical matched filtering being an important strategy. Meanwhile, deep learning methods have recently demonstrated both consistency with matched filtering methods and remarkable statistical performance. In this work, we propose Hierarchical Detection Network (HDN), a novel approach to efficient detection that combines ideas from hierarchical matching and deep learning. The network is trained using a novel loss function, which encodes simultaneously the goals of statistical accuracy and efficiency. We discuss the source of complexity reduction of the proposed model, and describe a general recipe for initialization with each layer specializing in different regions. We demonstrate the performance of HDN with experiments using open LIGO data and synthetic injections, and observe with two-layer models a $79\%$ efficiency gain compared with matched filtering at an equal error rate of $0.2\%$. Furthermore, we show how training a three-layer HDN initialized using two-layer model can further boost both accuracy and efficiency, highlighting the power of multiple simple layers in efficient detection.

astro-ph.IM↗

arXiv2022

Architectural Optimization and Feature Learning for High-Dimensional Time Series Datasets

As our ability to sense increases, we are experiencing a transition from data-poor problems, in which the central issue is a lack of relevant data, to data-rich problems, in which the central issue is to identify a few relevant features in a sea of observations. Motivated by applications in gravitational-wave astrophysics, we study the problem of predicting the presence of transient noise artifacts in a gravitational wave detector from a rich collection of measurements from the detector and its environment. We argue that feature learning--in which relevant features are optimized from data--is critical to achieving high accuracy. We introduce models that reduce the error rate by over 60% compared to the previous state of the art, which used fixed, hand-crafted features. Feature learning is useful not only because it improves performance on prediction tasks; the results provide valuable information about patterns associated with phenomena of interest that would otherwise be undiscoverable. In our application, features found to be associated with transient noise provide diagnostic information about its origin and suggest mitigation strategies. Learning in high-dimensional settings is challenging. Through experiments with a variety of architectures, we identify two key factors in successful models: sparsity, for selecting relevant variables within the high-dimensional observations; and depth, which confers flexibility for handling complex interactions and robustness with respect to temporal variations. We illustrate their significance through systematic experiments on real detector data. Our results provide experimental corroboration of common assumptions in the machine-learning community and have direct applicability to improving our ability to sense gravitational waves, as well as to many other problem settings with similarly high-dimensional, noisy, or partly irrelevant data.

cs.LG↗

arXiv2022

Detecting and Diagnosing Terrestrial Gravitational-Wave Mimics Through Feature Learning

As engineered systems grow in complexity, there is an increasing need for automatic methods that can detect, diagnose, and even correct transient anomalies that inevitably arise and can be difficult or impossible to diagnose and fix manually. Among the most sensitive and complex systems of our civilization are the detectors that search for incredibly small variations in distance caused by gravitational waves -- phenomena originally predicted by Albert Einstein to emerge and propagate through the universe as the result of collisions between black holes and other massive objects in deep space. The extreme complexity and precision of such detectors causes them to be subject to transient noise issues that can significantly limit their sensitivity and effectiveness. In this work, we present a demonstration of a method that can detect and characterize emergent transient anomalies of such massively complex systems. We illustrate the performance, precision, and adaptability of the automated solution via one of the prevalent issues limiting gravitational-wave discoveries: noise artifacts of terrestrial origin that contaminate gravitational wave observatories' highly sensitive measurements and can obscure or even mimic the faint astrophysical signals for which they are listening. Specifically, we demonstrate how a highly interpretable convolutional classifier can automatically learn to detect transient anomalies from auxiliary detector data without needing to observe the anomalies themselves. We also illustrate several other useful features of the model, including how it performs automatic variable selection to reduce tens of thousands of auxiliary data channels to only a few relevant ones; how it identifies behavioral signatures predictive of anomalies in those channels; and how it can be used to investigate individual anomalies and the channels associated with them.

astro-ph.IM↗

arXiv2022

Probing the Dark Solar System: Detecting Binary Asteroids with a Space-Based Interferometric Asteroid Explorer

With the inception of gravitational wave astronomy, astrophysical studies using interferometric techniques have begun to probe previously unknown parts of the universe. In this work, we investigate the potential of a new interferometric experiment to study a unique group of gravitationally interacting sources within our solar system: binary asteroids. We present the first study into binary asteroid detection via gravitational signals. We identify the interferometer sensitivity necessary for detecting a population of binary asteroids in the asteroid belt. We find that the space-based gravitational wave detector LISA will have negligible ability to detect these sources as these signals will be well below the LISA noise curve. Consequently, we propose a 4.6 AU and a 1 AU arm-length interferometers specialized for binary asteroid detection, targeting frequencies between $10^{-6}$ and $10^{-4}$ Hz. Our results demonstrate that the detection of binary asteroids with space-based gravitational wave interferometers is possible though very difficult, requiring substantially improved interferometric technology over what is presently proposed for space-based missions. If that threshold can be met, an interferometer may be used to map the asteroid belt, allowing for new studies into the evolution of our solar system.

astro-ph.EP↗

arXiv2022

Generalized Approach to Matched Filtering using Neural Networks

Gravitational wave science is a pioneering field with rapidly evolving data analysis methodology currently assimilating and inventing deep learning techniques. The bulk of the sophisticated flagship searches of the field rely on the time-tested matched filtering principle within their core. In this paper, we make a key observation on the relationship between the emerging deep learning and the traditional techniques: matched filtering is formally equivalent to a particular neural network. This means that a neural network can be constructed analytically to exactly implement matched filtering, and can be further trained on data or boosted with additional complexity for improved performance. Moreover, we show that the proposed neural network architecture can outperform matched filtering, both with or without knowledge of a prior on the parameter distribution. When a prior is given, the proposed neural network can approach the statistically optimal performance. We also propose and investigate two different neural network architectures MNet-Shallow and MNet-Deep, both of which implement matched filtering at initialization and can be trained on data. MNet-Shallow has simpler structure, while MNet-Deep is more flexible and can deal with a wider range of distributions. Our theoretical findings are corroborated by experiments using real LIGO data and synthetic injections, where our proposed methods significantly outperform matched filtering at false positive rates above $5\times 10^{-3}\%$. The fundamental equivalence between matched filtering and neural networks allows us to define a "complexity standard candle" to characterize the relative complexity of the different approaches to gravitational wave signal searches in a common framework. Finally, our results suggest new perspectives on the role of deep learning in gravitational wave detection.

astro-ph.IM↗

arXiv2021

Characterizing the observation bias in gravitational-wave detections and finding structured population properties

The observed distributions of the source properties from gravitational-wave detections are biased due to the selection effects and detection criteria in the detections, analogous to the Malmquist bias. In this work, this observation bias is investigated through its fundamental statistical and physical origins. An efficient semi-analytical formulation for its estimation is derived which is as accurate as the standard method of numerical simulations, with only a millionth of the computational cost. Then, the estimated bias is used for unmodelled inferences on the binary black hole population. These inferences show additional structures, specifically two peaks in the joint mass distribution around binary masses $\sim10$ M$_\odot$ and $\sim30$ M$_\odot$. Example ready-to-use scripts and some produced datasets for this method are shared in an online repository.

gr-qc↗

arXiv2021

An Archival Search for Very-High-Energy Counterparts to Sub-Threshold Neutron-Star Merger Candidates

The recent discovery of electromagnetic signals in coincidence with gravitational waves from neutron-star mergers has solidified the importance of multimessenger campaigns for studying the most energetic astrophysical events. Pioneering multimessenger observatories, such as the LIGO/Virgo gravitational wave detectors and the IceCube neutrino observatory, record many candidate signals that fall short of the detection significance threshold. These sub-threshold event candidates are promising targets for multimessenger studies, as the information provided by these candidates may, when combined with time-coincident gamma-ray observations, lead to significant detections. In this contribution, I describe our use of sub-threshold binary neutron star merger candidates identified in Advanced LIGO's first observing run (O1) to search for transient events in very-high-energy gamma rays using archival observations from the VERITAS imaging atmospheric Cherenkov telescope array. I describe the promise of this technique for future joint sub-threshold searches.

astro-ph.HE↗

arXiv2021

How to search for multiple messengers -- a general framework beyond two messengers

Quantification of the significance of a candidate multi-messenger detection of cosmic events is an emerging need in the astrophysics and astronomy communities. In this paper we show that a model-independent optimal search does not exist, and we present a general Bayesian method for the optimal model-dependent search, which is scalable to any number and any kind of messengers, and applicable to any model. In the end, we demonstrate it through an example for a joint gravitational wave, high-energy neutrino, short gamma-ray burst event search; which has not been examined heretofore.

astro-ph.HE↗

arXiv2021

Search for Black Hole Merger Families

The origin, environment, and evolution of stellar-mass black hole binaries are still a mystery. One of the proposed binary formation mechanisms is manifest in dynamical interactions between multiple black holes. A resulting framework of these dynamical interactions is the so-called hierarchical triple merger scenario, which happens when three black holes become gravitationally bound, causing two successive black hole mergers to occur. In such successive mergers, the black holes involved are directly related to each other, and hence this channel can be directly tested from the properties of the detected binary black hole mergers. Here we present a search for hierarchical triple mergers among events within the GWTC-1 and GWTC-2 catalogs of LIGO/Virgo, the eccentric localization of GW190521 and those found by the IAS-Princeton group. The search includes improved statistical quantification that also accounts for black hole spins. We perform our analysis for different upper bounds on the mass distribution of first generation BHs. Our results demonstrate the importance of the mass distributions' properties for constraining the hierarchical merger scenario. We present the individually significant merger pairs. The search yields interesting candidate families and hints of its future impact.

astro-ph.HE↗