arXiv subjects

Nabeel Rehemtulla

Publications and source records attributed to Nabeel Rehemtulla.

30 records · Page 2Linked to original sources

arXiv2025

Applying multimodal learning to Classify transient Detections Early (AppleCiDEr) I: Data set, methods, and infrastructure

Modern time-domain surveys like the Zwicky Transient Facility (ZTF) and the Legacy Survey of Space and Time (LSST) generate hundreds of thousands to millions of alerts, demanding automatic, unified classification of transients and variable stars for efficient follow-up. We present AppleCiDEr (Applying Multimodal Learning to Classify Transient Detections Early), a novel framework that integrates four key data modalities (photometry, image cutouts, metadata, and spectra) to overcome limitations of single-modality classification approaches. Our architecture introduces (i) two transformer encoders for photometry, (ii) a multimodal convolutional neural network (CNN) with domain-specialized metadata towers and Mixture-of-Experts fusion for combining metadata and images, and (iii) a CNN for spectra classification. Training on ~ 30,000 real ZTF alerts, AppleCiDEr achieves high accuracy, allowing early identification and suggesting follow-up for rare transient spectra. The system provides the first unified framework for both transient and variable star classification using real observational data, with seamless integration into brokering pipelines, demonstrating readiness for the LSST era.

astro-ph.IM↗

arXiv2025

Transfer Learning for Transient Classification: From Simulations to Real Data and ZTF to LSST

Machine learning has become essential for automated classification of astronomical transients, but current approaches face significant limitations: classifiers trained on simulations struggle with real data, models developed for one survey cannot be easily applied to another, and new surveys require prohibitively large amounts of labelled training data. These challenges are particularly pressing as we approach the era of the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST), where existing classification models will need to be retrained using LSST observations. We demonstrate that transfer learning can overcome these challenges by repurposing existing models trained on either simulations or data from other surveys. Starting with a model trained on simulated Zwicky Transient Facility (ZTF) light curves, we show that transfer learning reduces the amount of labelled real ZTF transients needed by 95% while maintaining equivalent performance to models trained from scratch. Similarly, when adapting ZTF models for LSST simulations, transfer learning achieves 94% of the baseline performance while requiring only 30% of the training data. These findings have significant implications for the early operations of LSST, suggesting that reliable automated classification will be possible soon after the survey begins, rather than waiting months or years to accumulate sufficient training data.

astro-ph.IM↗

arXiv2025

$\texttt{BTSbot}$: A Multi-input Convolutional Neural Network to Automate and Expedite Bright Transient Identification for the Zwicky Transient Facility

The Bright Transient Survey (BTS) relies on visual inspection ("scanning") to select sources for accomplishing its mission of spectroscopically classifying all bright extragalactic transients found by the Zwicky Transient Facility (ZTF). We present $\texttt{BTSbot}$, a multi-input convolutional neural network, which provides a bright transient score to individual ZTF detections using their image data and 14 extracted features. $\texttt{BTSbot}$ eliminates the need for scanning by automatically identifying and requesting follow-up observations of new bright ($m\,<18.5\,\mathrm{mag}$) transient candidates. $\texttt{BTSbot}$ outperforms BTS scanners in terms of completeness (99% vs. 95%) and identification speed (on average, 7.4 hours quicker). See Rehemtulla et al. 2024, ApJ, 972, 7R for the full BTSbot publication

astro-ph.IM↗

arXiv2025

A Massive Black Hole 0.8 kpc from the Host Nucleus Revealed by the Offset Tidal Disruption Event AT2024tvd

Tidal disruption events (TDEs) that are spatially offset from the nuclei of their host galaxies offer a new probe of massive black hole (MBH) wanderers, binaries, triples, and recoiling MBHs. Here we present AT2024tvd, the first off-nuclear TDE identified through optical sky surveys. High-resolution imaging with the \textit{Hubble Space Telescope} shows that AT2024tvd is $0.914\pm 0.010^{\prime\prime}$ offset from the apparent center of its host galaxy, corresponding to a projected distance of $0.808\pm 0.009$ kpc at $z=0.045$. Chandra and VLA observations support the same conclusion for the TDE's X-ray and radio emission. AT2024tvd exhibits typical properties of nuclear TDEs, including a persistent hot UV/optical component that peaks at $L_{\rm bb}\sim 6\times 10^{43}\,{\rm erg\,s^{-1}}$, broad hydrogen lines in its optical spectra, and delayed brightening of luminous ($L_{\rm X,peak}\sim 3\times 10^{43}\,{\rm erg\,s^{-1}}$), highly variable soft X-ray emission. The MBH mass of AT2024tvd is $10^{6\pm1}\,M_\odot$, at least 10 times lower than its host galaxy's central black hole mass ($\gtrsim 10^8\,M_\odot$). The MBH in AT2024tvd has two possible origins: a wandering MBH from the lower-mass galaxy in a minor merger during the dynamical friction phase or a recoiling MBH ejected by triple interactions. Combining AT2024tvd with two previously known off-nuclear TDEs discovered in X-rays (3XMM J2150 and EP240222a), which likely involve intermediate-mass black holes in satellite galaxies, we find that the parent galaxies of all three events are very massive ($\sim 10^{10.9}\,M_\odot$). This result aligns with expectations from cosmological simulations that the number of offset MBHs scales linearly with the host halo mass.

astro-ph.GA↗

arXiv2025

The BTSbot-nearby discovery of SN 2024jlf: rapid, autonomous follow-up probes interaction in an 18.5 Mpc Type IIP supernova

We present observations of the Type IIP supernova (SN) 2024jlf, including spectroscopy beginning just 0.7 days ($\sim$17 hours) after first light. Rapid follow-up was enabled by the new $\texttt{BTSbot-nearby}$ program, which involves autonomously triggering target-of-opportunity requests for new transients in Zwicky Transient Facility data that are coincident with nearby ($D<60$ Mpc) galaxies and identified by the $\texttt{BTSbot}$ machine learning model. Early photometry and non-detections shortly prior to first light show that SN 2024jlf initially brightened by $>$4 mag/day, quicker than $\sim$90% of Type II SNe. Early spectra reveal weak flash ionization features: narrow, short-lived ($1.3 < τ~\mathrm{[d]} < 1.8$) emission lines of H$α$, He II, and C IV. Assuming a wind velocity of $v_w=50$ km s$^{-1}$, these properties indicate that the red supergiant progenitor exhibited enhanced mass-loss in the last year before explosion. We constrain the mass-loss rate to $10^{-4} < \dot{M}~\mathrm{[M_\odot~yr^{-1}]} < 10^{-3}$ by matching observations to model grids from two independent radiative hydrodynamics codes. $\texttt{BTSbot-nearby}$ automation minimizes spectroscopic follow-up latency, enabling the observation of ephemeral early-time phenomena exhibited by transients.

astro-ph.HE↗

arXiv2025

The La Silla Schmidt Southern Survey

We present the La Silla Schmidt Southern Survey (LS4), a new wide-field, time-domain survey to be conducted with the 1 m ESO Schmidt telescope. The 268 megapixel LS4 camera mosaics 32 2k$\times$4k fully depleted CCDs, providing a $\sim$20 deg$^2$ field of view with $1''$ pixel$^{-1}$ resolution. The LS4 camera will have excellent performance at longer wavelengths: in a standard 45 s exposure the expected 5$σ$ limiting magnitudes in $g$, $i$, $z$ are $\sim$21.5, $\sim$20.9, and $\sim$20.3 mag (AB), respectively. The telescope design requires a novel filter holder that fixes different bandpasses over each quadrant of the detector. Two quadrants will have $i$ band, while the other two will be $g$ and $z$ band and color information will be obtained by dithering targets across the different quadrants. The majority (90%) of the observing time will be used to conduct a public survey that monitors the extragalactic sky at both moderate (3 d) and high (1 d) cadence, as well as focused observations within the Galactic bulge and plane. Alerts from the public survey will be broadcast to the community via established alert brokers. LS4 will run concurrently with the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST). The combination of LS4+LSST will enable detailed holistic monitoring of many nearby transients: high-cadence LS4 observations will resolve the initial rise and peak of the light curve while less-frequent but deeper observations by LSST will characterize the years before and after explosion. Here, we summarize the primary science objectives of LS4 including microlensing events in the Galaxy, extragalactic transients, the search for electromagnetic counterparts to multi-messenger events, and cosmology.

astro-ph.IM↗

arXiv2025

CCSNscore: A multi-input deep learning tool for classification of core-collapse supernovae using SED-Machine spectra

Supernovae (SNe) come in various flavors and are classified into different types based on emission and absorption lines in their spectra. SN candidates are now abundant with the advent of large systematic sky surveys like the Zwicky Transient Facility (ZTF), however, the identification bottleneck lies in their spectroscopic confirmation and classification. Fully robotic telescopes with dedicated spectrographs optimized for SN follow-up have eased the burden of data acquisition. However, the task of classifying the spectra still largely rests with the astronomers. Automating this classification step reduces human effort and can make the SN type available sooner to the public. For this purpose, we have developed a deep-learning based program for classifying core-collapse supernovae (CCSNe) with ultra-low resolution spectra from the SED-Machine spectrograph on the Palomar 60-inch telescope. The program consists of hierarchical classification task layers, with each layer composed of multiple binary classifiers running in parallel to produce a reliable classification. The binary classifiers utilize RNN and CNN architecture and are designed to take multiple inputs to supplement spectra with $g$- and $r$-band photometry from ZTF. On non-host-contaminated and good quality SEDM spectra ("gold" test set) of CCSNe, CCSNscore is ~94% accurate in distinguishing between hydrogen-rich (Type II) and hydrogen-poor (Type Ibc) CCSNe. With light curve input, CCSNscore classifies ~83% of the gold set with high confidence (score $\geq 0.8$ and score-error $<0.05$), with ~98% accuracy. Based on SNIascore's and CCSNscore's real-time performance on bright transients ($m_{pk}\leq18.5$) and our reporting criteria, we expect ~0.5% (~4) true SNe Ia to be misclassified as SNe Ibc and ~6% (~17) of true CCSNe to be misclassified between Type II and Type Ibc annually on the Transient Name Server.

astro-ph.IM↗

arXiv2024

Dramatic rebrightening of the type-changing stripped-envelope supernova SN 2023aew

Multi-peaked supernovae with precursors, dramatic light-curve rebrightenings, and spectral transformation are rare, but are being discovered in increasing numbers by modern night-sky transient surveys like the Zwicky Transient Facility (ZTF). Here, we present the observations and analysis of SN 2023aew, which showed a dramatic increase in brightness following an initial luminous (-17.4 mag) and long (~100 days) unusual first peak (possibly precursor). SN 2023aew was classified as a Type IIb supernova during the first peak but changed its type to resemble a stripped-envelope supernova (SESN) after the marked rebrightening. We present comparisons of SN 2023aew's spectral evolution with SESN subtypes and argue that it is similar to SNe Ibc during its main peak. P-Cygni Balmer lines are present during the first peak, but vanish during the second peak's photospheric phase, before H$α$ resurfaces again during the nebular phase. The nebular lines ([O I], [Ca II], Mg I], H$α$) exhibit a double-peaked structure which hints towards a clumpy or non-spherical ejecta. We analyze the second peak in the light curve of SN 2023aew and find it to be broader than normal SESNe as well as requiring a very high $^{56}$Ni mass to power the peak luminosity. We discuss the possible origins of SN 2023aew including an eruption scenario where a part of the envelope is ejected during the first peak which also powers the second peak of the light curve through SN-CSM interaction.

astro-ph.HE↗

arXiv2024

The Zwicky Transient Facility Bright Transient Survey. III. $\texttt{BTSbot}$: Automated Identification and Follow-up of Bright Transients with Deep Learning

The Bright Transient Survey (BTS) aims to obtain a classification spectrum for all bright ($m_\mathrm{peak}\,\leq\,18.5\,$mag) extragalactic transients found in the Zwicky Transient Facility (ZTF) public survey. BTS critically relies on visual inspection ("scanning") to select targets for spectroscopic follow-up, which, while effective, has required a significant time investment over the past $\sim5$ yr of ZTF operations. We present $\texttt{BTSbot}$, a multi-modal convolutional neural network, which provides a bright transient score to individual ZTF detections using their image data and 25 extracted features. $\texttt{BTSbot}$ is able to eliminate the need for daily human scanning by automatically identifying and requesting spectroscopic follow-up observations of new bright transient candidates. $\texttt{BTSbot}$ recovers all bright transients in our test split and performs on par with scanners in terms of identification speed (on average, $\sim$1 hour quicker than scanners). We also find that $\texttt{BTSbot}$ is not significantly impacted by any data shift by comparing performance across a concealed test split and a sample of very recent BTS candidates. $\texttt{BTSbot}$ has been integrated into Fritz and $\texttt{Kowalski}$, ZTF's first-party marshal and alert broker, and now sends automatic spectroscopic follow-up requests for the new transients it identifies. During the month of October 2023, $\texttt{BTSbot}$ selected 296 sources in real-time, 93% of which were real extragalactic transients. With $\texttt{BTSbot}$ and other automation tools, the BTS workflow has produced the first fully automatic end-to-end discovery and classification of a transient, representing a significant reduction in the human-time needed to scan. Future development has tremendous potential for creating similar models to identify and request follow-up observations for specific types of transients.

astro-ph.IM↗

arXiv2023

Direct detection of supernova progenitor stars with ZTF and LSST

The direct detection of core-collapse supernova (SN) progenitor stars is a powerful way of probing the last stages of stellar evolution. However, detections in archival Hubble Space Telescope images are limited to about one per year. Here, we explore whether we can increase the detection rate by using data from ground-based wide-field surveys. Due to crowding and atmospheric blurring, progenitor stars can typically not be identified in pre-explosion images alone. Instead, we combine many pre-SN and late-time images to search for the disappearance of the progenitor star. As a proof of concept, we implement our search for ZTF data. For a few hundred images, we achieve limiting magnitudes of about 23 mag in the g and r band. However, no progenitor stars or long-lived outbursts are detected for 29 SNe within z<0.01, and the ZTF limits are typically several magnitudes less constraining than detected progenitors in the literature. Next, we estimate progenitor detection rates for the Legacy Survey of Space and Time (LSST) with the Vera C. Rubin telescope by simulating a population of nearby SNe. The background from bright host galaxies reduces the nominal LSST sensitivity by, on average, 0.4 mag. Over the ten-year survey, we expect the detection of about 50 red supergiant progenitors and several yellow and blue supergiants. The progenitors of SNe Ib and Ic are detectable if they are brighter than -4.7 mag or -4.0 mag in the LSST i band, respectively. In addition, we expect the detection of hundreds of pre-SN outbursts depending on their brightness and duration.

astro-ph.SR↗

arXiv2023

Overview of the DESI Milky Way Survey

We describe the Milky Way Survey (MWS) that will be undertaken with the Dark Energy Spectroscopic Instrument (DESI) on the Mayall 4m telescope at the Kitt Peak National Observatory. Over the next 5 yr DESI MWS will observe approximately seven million stars at Galactic latitudes |b|>20 degrees, with an inclusive target selection scheme focused on the thick disk and stellar halo. MWS will also include several high-completeness samples of rare stellar types, including white dwarfs, low-mass stars within 100pc of the Sun, and horizontal branch stars. We summarize the potential of DESI to advance understanding of Galactic structure and stellar evolution. We introduce the final definitions of the main MWS target classes and estimate the number of stars in each class that will be observed. We describe our pipelines for deriving radial velocities, atmospheric parameters, and chemical abundances. We use ~500,000 spectra of unique stellar targets from the DESI Survey Validation program (SV) to demonstrate that our pipelines can measure radial velocities to ~1 km/s and [Fe/H] accurate to ~0.2 dex for typical stars in our main sample. We find the stellar parameter distributions from ~100 sq. deg of SV observations with >90% completeness on our main sample are in good agreement with expectations from mock catalogs and previous surveys.

astro-ph.GA↗

arXiv2022

Non-Parametric Spherical Jeans Mass Estimation with B-splines

Spherical Jeans modeling is widely used to estimate mass profiles of systems from star clusters to galactic stellar haloes to clusters of galaxies. It derives the cumulative mass profile, M(<r), from kinematics of tracers of the potential under the assumptions of spherical symmetry and dynamical equilibrium. We consider the application of Jeans modeling to mapping the dark matter distribution in the outer reaches of the Milky Way using field halo stars. We present a novel non-parametric routine for solving the spherical Jeans equation by fitting B-splines to the velocity and density profiles of halo stars. While most implementations assume parametric forms for these profiles, B-splines provide non-parametric fitting curves with analytical derivatives. Our routine recovers the mass profiles of equilibrium systems with flattened haloes or a stellar disc and bulge excellently (<~ 10% error at most radii). Tests with non-equilibrium, Milky Way-like galaxies from the Latte suite of FIRE-2 simulations perform quite well (<~ 15% error for r <~ 100 kpc). We also create observationally motivated datasets for the Latte suite by imposing selection functions and errors on phase space coordinates characteristic of Gaia and the DESI Milky Way Survey. The resulting imprecise and incomplete data require us to introduce an MCMC-based subroutine to obtain deconvolved density and velocity dispersion profiles from the tracer population. With these observational effects taken into account, the accuracy of the Jeans mass estimate remains at the level 20% or better.

astro-ph.GA↗