arXiv subjects

Tjitske Starkenburg

Publications and source records attributed to Tjitske Starkenburg.

At least 19 recordsLinked to original sources

arXiv2026

Modeling Globular Cluster Stellar Streams with a Basis-Expansion N-body Code

Globular cluster stellar streams probe galaxy-formation processes and can potentially reveal the distribution of dark matter in galaxies. In many theoretical studies, streams are modeled with particle-spray or direct N-body codes. But particle-spray methods abstract away the internal dynamics of the progenitor by making strong assumptions about the escape physics, while direct N-body is prohibitively expensive for realistic (N>10^5) systems. In this paper, we present the stream-modeling capabilities of KRIOS, a new basis-expansion N-body code for collisional stellar dynamics, that bridges this runtime vs. accuracy gap. We show that KRIOS reproduces NBODY6++GPU cluster models, and their associated streams, more accurately than particle spray in a fraction of the NBODY6++GPU wall-clock time. We then compare KRIOS to various particle-spray methods on 10 orbits similar to known Milky Way streams. The morphology and kinematics of these streams most disagree when the progenitor is tightly bound to the host, as these systems are often subject to stronger tidal forces. Finally, we discuss which elements of the progenitor physics are most important for modeling stellar streams and how these might be incorporated into particle-spray methods.

astro-ph.GA↗

arXiv2026

Evidence for the First Globular Cluster Stellar Stream beyond the Milky Way

The dark matter content of ultra-diffuse galaxies is the subject of considerable debate. Stellar streams, which form when a host galaxy tidally strips stars from an orbiting stellar system, provide a powerful technique to constrain the dark matter content of external galaxies. The stripped stars form long, thin leading and trailing tidal arms that persist for billions of years. Stellar streams from globular clusters are particularly sensitive probes of dark matter halos and substructure. Globular cluster streams are expected to exist in a variety of host galaxy types, but so far, they have only been observed in the Milky Way. We present evidence for the first extragalactic globular cluster stellar stream, identified in deep Hubble Space Telescope imaging of the ultra-diffuse galaxy, UGC9050-Dw1. The stream's morphology, colour, and apparent association with a compact source support the globular cluster progenitor interpretation observationally, and we reproduce the observed surface brightness with simulated globular cluster stellar populations. We use generative stream modelling, which fits dynamical models directly to the stream morphology, to constrain the mass of the progenitor and present the first stream-based halo constraint for an ultra-diffuse galaxy. The stream models point to a globular cluster origin and suggest a massive dark matter host halo. By extending the reach of globular cluster stream analysis to external galaxies, this work opens a new chapter in dark matter science.

astro-ph.GA↗

arXiv2026

Machine Learning Methods for Stellar Collisions. I. Predicting Outcomes of SPH Simulations

Stellar collisions can occur frequently in dense cluster environments, and play a crucial role in producing exotic phenomena from blue stragglers in globular clusters to high-energy transients in galactic nuclei. Successive collisions and mergers of massive stars could also lead to the formation of massive black holes, serving as seeds for supermassive black hole in the early universe. While analytic fitting formulae exist for predicting collision outcomes, they do not generalize across different energy scales or stellar evolutionary phases. Smoothed particle hydrodynamics (SPH) simulations are often used to compute the outcomes of stellar collisions, but, even at low resolution, their computational cost makes running on-the-fly calculations during an $N$-body simulation quite challenging. Here we present a new grid of $27,720$ SPH calculations of main-sequence star collisions, spanning a wide range of masses, ages, relative velocities, and impact parameters. Using this grid, we train machine learning models to predict both collision outcomes (merger vs disruption, or flyby) and final remnant masses. We compare the performance of nearest neighbors, support vector machines, and neural networks, achieving classification balanced accuracy of $98.4\%$, and regression relative errors as low as $0.11\%$ and $0.15\%$ for the final stars $1$ and $2$, respectively. We make our trained models publicly available as part of the package collAIder, enabling rapid predictions of stellar collision outcomes in $N$-body models of dense star cluster dynamics.

astro-ph.HE↗

arXiv2026

ShuffleFlow: Scalable Posterior Inference for Bayesian Inverse Imaging

Variational inference (VI) is a powerful method for principled posterior inference for scientific inverse imaging. VI learns the posterior distribution, often with a flow-based network, which can cheaply generate posterior samples upon optimization, and can flexibly incorporate score-based or classic priors. However, its application to large-scale image reconstruction is severely hindered by the poor scalability of the flow-based networks. In this work, we introduce ShuffleFlow, a scalable VI framework to address this challenge. Our method breaks down the problem into three parts: a pixel-unshuffling-based image coordinate sampler, a neural field as feature encoder, and a conditional normalizing flow (CNF) as posterior estimator. Specifically, our framework partitions an image into a stack of sub-images with pixel-unshuffling and uses a shared CNF to model the joint distribution of the sub-image stack. We condition the CNF on the output of a neural field, which embeds feature vectors corresponding to pixel-unshuffling sample locations to capture spatial structures, and share the flow's latent variable across the channels to model their correlations. We demonstrate our method's effectiveness and efficiency on both linear and nonlinear imaging inverse problems, and show its ability to more rapidly generate a high-sample-count posterior than diffusion samplers.

cs.CV↗

arXiv2026

Dancing Streams In Merging Halos: Stellar Streams in a MW--LMC-like merger

Stellar streams -- formed from tidally stripped globular clusters or dwarf galaxies -- are sensitive tracers of a galaxy's accretion history and gravitational potential. While numerous streams are known in the Milky Way (MW), the formation and evolution of stellar streams have been primarily studied in isolated settings. The impact of subsequent galaxy interactions on stellar streams remains largely unexplored. Understanding merger-induced effects is however, crucial given the accretion of the Large Magellanic Cloud (LMC) onto the MW, and the fact that for example M31 and Cen A have experienced recent mergers. We analyze the detailed evolution of 1024 stellar streams during a complete MW--LMC-like merger, systematically varying initial stream properties and considering various orbits for the infalling perturber. We find that an MW--LMC mass-ratio merger significantly alters stellar stream properties, including energy, angular momentum, orbit, and morphology. Some streams exhibit dramatic morphological changes or develop complex substructures, while others see substantial shifts in energy and/or angular momentum, or re-orient their orbital plane. Interestingly, strong morphological alterations do not necessarily correlate with large changes in energy or orbit. A few streams split apart with parts moving to different orbits, appearing disconnected in position and kinematics despite their common origin. Strong effects correlate with close encounters between stream particles and the infalling perturber at various times during the merger. Our findings highlight the considerable impact of significant accretion events on the properties of stellar streams, and the challenge to recover the initial orbits of streams from their appearance at the present-day. Visualizations of the detailed evolution of all 1024 stellar streams are available at https://dancingstreamsinmerginghalos.github.io.

astro-ph.GA↗

arXiv2026

Opportunities in AI/ML for the Rubin LSST Dark Energy Science Collaboration

The Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) will produce unprecedented volumes of heterogeneous astronomical data (images, catalogs, and alerts) that challenge traditional analysis pipelines. The LSST Dark Energy Science Collaboration (DESC) aims to derive robust constraints on dark energy and dark matter from these data, requiring methods that are statistically powerful, scalable, and operationally reliable. Artificial intelligence and machine learning (AI/ML) are already embedded across DESC science workflows, from photometric redshifts and transient classification to weak lensing inference and cosmological simulations. Yet their utility for precision cosmology hinges on trustworthy uncertainty quantification, robustness to covariate shift and model misspecification, and reproducible integration within scientific pipelines. This white paper surveys the current landscape of AI/ML across DESC's primary cosmological probes and cross-cutting analyses, revealing that the same core methodologies and fundamental challenges recur across disparate science cases. Since progress on these cross-cutting challenges would benefit multiple probes simultaneously, we identify key methodological research priorities, including Bayesian inference at scale, physics-informed methods, validation frameworks, and active learning for discovery. With an eye on emerging techniques, we also explore the potential of the latest foundation model methodologies and LLM-driven agentic AI systems to reshape DESC workflows, provided their deployment is coupled with rigorous evaluation and governance. Finally, we discuss critical software, computing, data infrastructure, and human capital requirements for the successful deployment of these new methodologies, and consider associated risks and opportunities for broader coordination with external actors.

astro-ph.IM↗

arXiv2025

Picture an Astronomer: Best Practices for Retaining Talent in Astrophysics

Women are consistently underrepresented in astrophysics yet are simultaneously subject to disproportionate attrition at every career stage. This disparity between demonstrated efficacy in job performance and ultimate career outcome was the primary motivation for the Picture an Astronomer series, which included both targeted public outreach to increase representation of women in astrophysics and high-level, solution-oriented discussions among professional astronomers. In March 2025, more than 200 astronomers came together in a hybrid-format symposium focused on the state of the field for female scientists, combining scientific exchange with discussions of policies and practices to strengthen retention of talent in the field. This white paper is the result of those discussions, offering a wide range of recommendations developed in the context of gendered attrition in astrophysics but which ultimately support a healthier climate for all scientists alike.

astro-ph.IM↗

arXiv2025

LIMFAST. IV. Learning High-Redshift Galaxy Formation from Multiline Intensity Mapping with Implicit Likelihood Inference

By opening up new avenues to statistically constrain astrophysics and cosmology with large-scale structure observations, the line intensity mapping (LIM) technique calls for novel tools for efficient forward modeling and inference. Implicit likelihood inference (ILI) from semi-numerical simulations provides a powerful setup for investigating a large model parameter space in a data-driven manner, therefore gaining significant recent attention. Using simulations of high-redshift 158$μ$m [CII] and 88$μ$m [OIII] LIM signals created by the LIMFAST code, we develop an ILI framework in a case study of learning the physics of early galaxy formation from the auto-power spectra of these lines or their cross-correlation with galaxy surveys. We leverage neural density estimation with normalizing flows to learn the mapping between the simulated power spectra and parameters that characterize the physics governing the star formation efficiency and the $\dotΣ_{\star}$-$Σ_\mathrm{g}$ relation of high-redshift galaxies. Our results show that their partially degenerate effects can be unambiguously constrained when combining [CII] with [OIII] measurements to be made by new-generation mm/sub-mm LIM experiments.

astro-ph.GA↗

arXiv2025

Learning the Universe: Cosmological and Astrophysical Parameter Inference with Galaxy Luminosity Functions and Colours

We perform the first direct cosmological and astrophysical parameter inference from the combination of galaxy luminosity functions and colours using a simulation based inference approach. Using the Synthesizer code we simulate the dust attenuated ultraviolet-near infrared stellar emission from galaxies in thousands of cosmological hydrodynamic simulations from the CAMELS suite, including the Swift-EAGLE, IllustrisTNG, Simba & Astrid galaxy formation models. For each galaxy we calculate the rest-frame luminosity in a number of photometric bands, including the SDSS $\textit{ugriz}$ and GALEX FUV & NUV filters; this dataset represents the largest catalogue of synthetic photometry based on hydrodynamic galaxy formation simulations produced to date, totalling >200 million sources. From these we compile luminosity functions and colour distributions, and find clear dependencies on both cosmology and feedback. We then perform simulation based (likelihood-free) inference using these distributions to constrain $Ω_m$, $σ_8$, and four parameters controlling the strength of stellar and AGN feedback. Both colour distributions and luminosity functions provide complementary information on certain parameters when performing inference. We achieve constraints on the stellar feedback parameters, as well as $Ω_m$ and $σ_8$. The latter is attributable to the fact that the photometry encodes the star formation-metal enrichment history of each galaxy; galaxies in a universe with a higher $σ_8$ tend to form earlier and have higher metallicities, which leads to redder colours. We find that a model trained on one galaxy formation simulation generalises poorly when applied to another, and attribute this to differences in the subgrid prescriptions, and lack of flexibility in our emission modelling. The photometric catalogues are publicly available at: https://camels.readthedocs.io/

astro-ph.GA↗

arXiv2025

Breaking Down the $\textsf{CosmoGEMS}$: Toward Modeling and Understanding Globular Cluster Stellar Streams in a Fully Cosmological Context

Next-generation surveys are expected to uncover thousands of globular cluster (GC) stellar streams, motivating the need for a theoretical framework that produces realistic GC streams in a fully cosmological, Milky Way-like environment. We present $\textsf{CosmoGEMS}$, a star-by-star cosmological GC stream framework that self-consistently links small-scale cluster physics with large-scale Galactic dynamics. The initial phase-space positions of stream stars are informed by post-processed GC populations within the FIRE cosmological simulation. Escaped stars are orbit-integrated from their time of escape to the present day in a time-evolving Galactic potential extracted from the same simulation using a basis function expansion. We explore two example streams on different orbits. One forms a long, thin stream with a velocity dispersion consistent with Milky Way GC streams. However, it exhibits a clump and orbital-phase-dependent misalignments due to the evolving potential. The other stream develops both a thin component and a diffuse, shell-like structure, similar to features observed in streams like Jhelum. These results highlight the power of fully cosmological models in producing realistic stream morphologies and kinematics. Unlike idealized simulations, our models naturally incorporate time-dependent changes in the progenitor's orbit, including orbital plane evolution, which significantly affects stream structure. This challenges common assumptions in stream-finding algorithms and interpretation. $\textsf{CosmoGEMS}$ provides a key step toward connecting future stellar stream observations with the physics of globular cluster evolution and hierarchical galaxy formation in a cosmological context.

astro-ph.GA↗

arXiv2025

ChemZz I: Comparing Oxygen and Iron Abundance Patterns in the Milky Way, the Local Group and Cosmic Noon

Our understanding of the chemical evolution of galaxies has advanced through measurements from both distant galaxies across redshift, and our own Milky Way (MW). To form a comprehensive picture, it is essential to unify these constraints, placing them on a common scale and parlance and to understand their systematic differences. In this study, we homogenize oxygen and iron measurements from star-forming galaxies at Cosmic Noon ($z{\sim}2-3$) with resolved stellar abundances from the Local Group. The MW is divided into four components, assuming the outer halo is dominated by debris from the Gaia-Sausage-Enceladus (GSE) progenitor. After converting all abundances to a common Solar scale, we identify clear $α$- and iron-enhancement trends with mass in the $z{\sim}2-3$ galaxies and find good agreement between these galaxies and the MW high-$α$ disc in [O/Fe] vs. [Fe/H]. We also find excellent agreement between the [O/Fe] trends seen in the MW high- and low-$α$ discs with O-abundances seen in old and young planetary nebulae in M~31 respectively, supporting the existence of $α$-bimodality in the inner regions of M~31. Finally, we use globular cluster ages to project the MW and GSE back in time to $z{\sim}3$ and find that their estimated mass, oxygen and iron abundances are strikingly consistent with the mass-metallicity relation of star-forming galaxies at $z{\sim}3$. In the future, increased transparency around the choice of Solar scale and abundance methodology will make combining chemical abundances easier -- contributing to a complete picture of the chemical evolution of all galaxies.

astro-ph.GA↗

arXiv2025

Learning the Universe: physically-motivated priors for dust attenuation curves

Understanding the impact of dust on the spectral energy distributions (SEDs) of galaxies is crucial for inferring their physical properties and for studying the nature of interstellar dust. We analyze dust attenuation curves for $\sim 6400$ galaxies ($M_{\star} \sim 10^9 - 10^{11.5}\,M_{\odot}$) at $z=0.07$ in the IllustrisTNG50 and TNG100 simulations. Using radiative transfer post-processing, we generate synthetic attenuation curves and fit them with a parametric model that captures known extinction and attenuation laws (e.g., Calzetti, MW, SMC, LMC) and more exotic forms. We present the distributions of the best-fitting parameters: UV slope ($c_1$), optical-to-NIR slope ($c_2$), FUV slope ($c_3$), 2175 Angstrom bump strength ($c_4$), and normalization ($A_{\rm V}$). Key correlations emerge between $A_{\rm V}$ and the star formation rate surface density $Σ_{\rm SFR}$, as well as the UV slope $c_1$. The UV and FUV slopes ($c_1, c_3$) and the bump strength and visual attenuation ($c_4, A_{\rm V}$) exhibit robust internal correlations. Using these insights from simulations, we provide a set of scaling relations that predict a galaxy's median (averaged over line of sight) dust attenuation curve based solely on its $Σ_{\rm SFR}$ and/or $A_{\rm V}$. These predictions agree well with observed attenuation curves from the GALEX-SDSS-WISE Legacy Catalog despite minor differences in bump strength. This study delivers the most comprehensive library of synthetic attenuation curves for local galaxies, providing a foundation for physically motivated priors in SED fitting and galaxy inference studies, such as those performed as part of the Learning the Universe Collaboration.

astro-ph.GA↗

arXiv2025

Streams, Shells, and Substructures in the Accretion-Built Stellar Halo of NGC 300

We present deep optical observations of the stellar halo of NGC 300, an LMC-mass galaxy, acquired with the DEEP sub-component of the DECam Local Volume Exploration survey (DELVE) using the 4 m Blanco Telescope. Our resolved star analysis reveals a large, low surface brightness stellar stream ($M_{V}\sim-8.5$; [Fe/H] $= -1.4\pm0.15$) extending more than 40 kpc north from the galaxy's center. We also find other halo structures, including potentially an additional stream wrap to the south, which may be associated with the main stream. The morphology and derived low metallicities of the streams and shells discovered surrounding NGC 300 are highly suggestive of a past accretion event. Assuming a single progenitor, the accreted system is approximately Fornax-like in luminosity, with an inferred mass ratio to NGC 300 of approximately $1:15$. We also present the discovery of a metal-poor globular cluster ($R_{\rm{proj}}=23.3$~kpc; $M_{V}=-8.99\pm0.16$; [Fe/H] $\approx-1.6\pm0.6$) in the halo of NGC 300, the furthest identified globular cluster associated with NGC 300. The stellar structures around NGC 300 represent the richest features observed in a Magellanic Cloud analog to date, strongly supporting the idea that accretion and subsequent disruption is an important mechanism in the assembly of dwarf galaxy stellar halos.

astro-ph.GA↗

arXiv2024

Prospects for Detecting Gaps in Globular Cluster Stellar Streams in External Galaxies with the Nancy Grace Roman Space Telescope

Stellar streams form through the tidal disruption of satellite galaxies or globular clusters orbiting a host galaxy. Globular cluster streams are exciting since they are thin (dynamically cold) and, therefore sensitive to perturbations from low-mass subhalos. Since the subhalo mass function differs depending on the dark matter composition, these gaps can provide unique constraints on dark matter models. However, current samples are limited to the Milky Way. With its large field of view, deep imaging sensitivity, and high angular resolution, the upcoming Nancy Grace Roman Space Telescope (Roman) presents a unique opportunity to increase the number of observed streams and gaps significantly. This paper presents a first exploration of the prospects for detecting gaps in streams in M31 and other nearby galaxies with resolved stars. We simulate the formation of gaps in a Palomar-5-like stream and generate mock observations of these gaps with background stars in M31 and the foreground Milky Way stellar fields. We assess Roman's ability to detect gaps out to 10 Mpc through visual inspection and with the gap-finding tool ${\texttt{FindTheGap}}$. We conclude that gaps of $\approx 1.5$ kpc in streams that are created from subhalos of masses $\geq5 \times 10^6$M$_{\odot}$ are detectable within a 2-3 Mpc volume in exposures of 1000s to 1 hour. This volume contains $\approx 150$ galaxies, including $\approx 8$ galaxies with luminosities $>10^{9}~$L$_{\odot}$. Large samples of stream gaps in external galaxies will open up a new era of statistical analyses of gap characteristics in stellar streams and help constrain dark matter models.

astro-ph.GA↗

arXiv2023

Code Comparison in Galaxy Scale Simulations with Resolved Supernova Feedback: Lagrangian vs. Eulerian Methods

We present a suite of high-resolution simulations of an isolated dwarf galaxy using four different hydrodynamical codes: {\sc Gizmo}, {\sc Arepo}, {\sc Gadget}, and {\sc Ramses}. All codes adopt the same physical model which includes radiative cooling, photoelectric heating, star formation, and supernova (SN) feedback. Individual SN explosions are directly resolved without resorting to sub-grid models, eliminating one of the major uncertainties in cosmological simulations. We find reasonable agreement on the time-averaged star formation rates as well as the joint density-temperature distributions between all codes. However, the Lagrangian codes show significantly burstier star formation, larger supernova-driven bubbles, and stronger galactic outflows compared to the Eulerian code. This is caused by the behavior in the dense, collapsing gas clouds when the Jeans length becomes unresolved: gas in Lagrangian codes collapses to much higher densities than in Eulerian codes, as the latter is stabilized by the minimal cell size. Therefore, more of the gas cloud is converted to stars and SNe are much more clustered in the Lagrangian models, amplifying their dynamical impact. The differences between Lagrangian and Eulerian codes can be reduced by adopting a higher star formation efficiency in Eulerian codes, which significantly enhances SN clustering in the latter. Adopting a zero SN delay time reduces burstiness in all codes, resulting in vanishing outflows as SN clustering is suppressed.

astro-ph.GA↗

arXiv2022

Hot-mode accretion and the physics of thin-disk galaxy formation

We use FIRE simulations to study disk formation in z~0, Milky Way-mass galaxies, and conclude that a key ingredient for the formation of thin stellar disks is the ability for accreting gas to develop an aligned angular momentum distribution via internal cancellation *prior* to joining the galaxy. Among galaxies with a high fraction (>70%) of their young stars in a thin disk (h/R~0.1) we find that: (i) hot, virial-temperature gas dominates the inflowing gas mass on halo scales (>~20 kpc), with radiative losses offset by compression heating; (ii) this hot accretion proceeds until angular momentum support slows inward motion, at which point the gas cools to T~10^4 K or less; (iii) prior to cooling, the accreting gas develops an angular momentum distribution that is aligned with the galaxy disk, and while cooling transitions from a quasi-spherical spatial configuration to a more flattened, disk-like configuration. We show that the existence of this "rotating cooling flow" accretion mode is strongly correlated with the fraction of stars forming in a thin disk among a sample of 17 z~0 galaxies spanning a halo mass range of 10^10.5 solar masses to 10^12 solar masses, or a stellar mass range 10^8 solar masses to 10^11 solar masses. Notably, galaxies with a thick disk or irregular morphology do not undergo significant angular momentum alignment of gas prior to accretion and show no correspondence between halo gas cooling and flattening. Our results suggest that rotating cooling flows (or, more generally, rotating subsonic flows) that become coherent and angular momentum-supported prior to accretion onto the galaxy are likely a necessary condition for the formation of thin, star-forming disk galaxies in a LambdaCDM universe.

astro-ph.GA↗

arXiv2022

The In-situ Origins of Dwarf Stellar Outskirts in FIRE-2

Extended, old, and round stellar halos appear to be ubiquitous around high-mass dwarf galaxies ($10^{8.5}<M_\star/M_\odot<10^{9.6}$) in the observed universe. However, it is unlikely that these dwarfs have undergone a sufficient number of minor mergers to form stellar halos that are composed of predominantly accreted stars. Here, we demonstrate that FIRE-2 (Feedback in Realistic Environments) cosmological zoom-in simulations are capable of producing dwarf galaxies with realistic structure, including both a thick disk and round stellar halo. Crucially, these stellar halos are formed in-situ, largely via the outward migration of disk stars. However, there also exists a large population of "non-disky" dwarfs in FIRE-2 that lack a well-defined disk/halo and do not resemble the observed dwarf population. These non-disky dwarfs tend to be either more gas poor or to have burstier recent star formation histories than the disky dwarfs, suggesting that star formation feedback may be preventing disk formation. Both classes of dwarfs underscore the power of a galaxy's intrinsic shape -- which is a direct quantification of the distribution of the galaxy's stellar content -- to interrogate the feedback implementation in simulated galaxies.

astro-ph.GA↗

arXiv2021

Deep Hubble Space Telescope Observations of GW170817: Complete Light Curves and the Properties of the Galaxy Merger of NGC 4993

We present the complete set of {\it Hubble Space Telescope} imaging of the binary neutron star merger GW170817 and its optical counterpart AT 2017gfo. Including deep template imaging in F814W, F110W, F140W, and F160W at 3.4 years post-merger, we re-analyze the full light curve of AT 2017gfo across 12 bands from 5--1273 rest-frame days after merger. We obtain four new detections of the short $γ$-ray burst (GRB) 170817A afterglow from 109--170 rest-frame days post-merger. These detections are consistent with the previously observed $β=-0.6$ spectral index in the afterglow light curve with no evidence for spectral evolution. We also analyze our limits in the case of novel late-time optical and IR emission signatures, such as a kilonova afterglow or infrared dust echo, but find our limits are not constraining in these contexts. We use the new data to construct deep optical and infrared stacks, reaching limits of $M=-6.3$ to $-4.6$ mag, to analyze the local environment around AT 2017gfo and low surface brightness features in its host galaxy NGC 4993. We rule out the presence of any globular cluster at the position of AT 2017gfo to $2.3 \times 10^{4} L_{\odot}$, including those with the reddest $V-H$ colors. Finally, we analyze the substructure of NGC 4993 in deep residual imaging, and find shell features which extend up to 71.8\arcsec\ (14.2 kpc) from the center of the galaxy. We find that the shells have a cumulative stellar mass of $6.3\times10^{8} M_{\odot}$, roughly 2% the total stellar mass of NGC 4993, and mass-weighted ages of $>$3 Gyr. We conclude that it was unlikely the GW170817 progenitor system formed in the galaxy merger, which based on dynamical signatures and the stellar population in the shells mostly likely occurred 220--685 Myr ago.

astro-ph.HE↗