Search for: All records

Abstract The follow-up of gravitational-wave events by wide-field surveys is a crucial tool for the discovery of electromagnetic counterparts to gravitational wave sources, such as kilonovae. Machine learning tools can play an important role in aiding search efforts. We have developed a public tool to predict kilonova light curves using simulated low-latency alert data from the International Gravitational Wave Network during observing runs 4 (O4) and 5 (O5). It uses a bidirectional long-short-term memory model to forecast kilonova light curves from binary neutron star and neutron star–black hole mergers in the Zwicky Transient Facility (ZTF) and Rubin Observatory’s Legacy Survey of Space and Time filters. The model achieves a test mean squared error (MSE) of 0.12 for ZTF filters and 0.23 for Rubin filters, calculated by averaging the squared error over all time steps, filters, and light curves in the test set. We evaluate the performance of the model against merger events followed-up by the ZTF partnership during O4c. We also analyze the effect of incorporating constraints on physical features such as ejecta mass. Using ejecta mass, the performance of the model improves to an MSE of 0.1 for ZTF filters and 0.15 for Rubin filters. Our model is publicly available and can help to add important information to help plan follow-up of candidate events discovered by current and next-generation public surveys. 

Abstract Tidal disruption events (TDEs) have traditionally been discovered in optical sky surveys through targeted searches of nuclear transients. However, it is expected that some TDEs will occur outside the galaxy nucleus, arising from wandering black holes (BHs) originating in galaxy mergers. Here, we present observations of TDE 2025abcr, the first optical TDE discovered in the outskirts of a host galaxy. The TDE was identified by a custom “off-nuclear” implementation of the machine learning classifiertdescore, which classifies new ZTF transients based on their lightcurves. Follow-up observations confirm that TDE 2025abcr is a TDE-H+He, occurring 9 $\stackrel{\text{″}}{.}$ 5 (9.3 kpc projected distance) from the nucleus of a massive galaxy (M_⋆= 10^11.18±0.03M_⊙) with a central BH mass of 10^8.82±0.65M_⊙. TDE 2025abcr itself was likely disrupted by a much lighter BH (10^6.09±0.53M_⊙, as estimated with peak luminosity scaling relations). The BH was either dynamically ejected from the nucleus or lies at the center of a very faint tidally stripped dwarf galaxy undergoing a minor merger. Late-time observations of TDE 2025abcr could confirm the origin of this apparent “wandering” BH. The rate of highly offset (≳3 kpc) TDEs can be constrained to <10% of the nuclear TDE rate, but our discovery implies that many dozens of similar sources will be detected by the Vera C. Rubin Observatory each year with resolvable offsets. 

Abstract Modern wide-field time-domain surveys facilitate the study of transient, variable and moving phenomena by conducting image differencing and relaying alerts to their communities. Machine learning tools have been used on data from these surveys and their precursors for more than a decade, and convolutional neural networks (CNNs), which make predictions directly from input images, saw particularly broad adoption through the 2010s. Since then, continually rapid advances in computer vision have transformed the standard practices around using such models. It is now commonplace to use standardized architectures pre-trained on large corpora of everyday images (e.g., ImageNet). In contrast, time-domain astronomy studies still typically design custom CNN architectures and train them from scratch. Here, we explore the effects of adopting various pre-training regimens and standardized model architectures on the performance of alert classification. We find that the resulting models match or outperform a custom, specialized CNN like what is typically used for filtering alerts. Moreover, our results show that pre-training on galaxy images from Galaxy Zoo tends to yield better performance than pre-training on ImageNet or training from scratch. We observe that the design of standardized architectures are much better optimized than the custom CNN baseline, requiring significantly less time and memory for inference despite having more trainable parameters. On the eve of the Legacy Survey of Space and Time and other image-differencing surveys, these findings advocate for a paradigm shift in the creation of vision models for alerts, demonstrating that greater performance and efficiency, in time and in data, can be achieved by adopting the latest practices from the computer vision field. 

Abstract In 2026 March, the Einstein Probe (EP) discovered its nearest (z= 0.0343) Fast X-ray Transient (FXT), EP 260321a, the first EP FXT to provide a strong match to expectations for X-ray “shock breakout” (SBO) emission. Here, we present our multiwavelength follow-up campaign of EP 260321a and its broad-line Type Ic (Ic-BL) supernova (SN) counterpart, SN 2026gzf. We show that our radio follow-up extending over 5.8–54.5 days post-FXT rules out an on-axis jet counterpart of isotropic-equivalent kinetic energyE_K ≳ 10⁴⁹ erg for circumburst densitiesn > 10⁻²cm⁻³and assuming microphysical parametersϵ_e = ϵ_B = 0.1. Our radio data also constrain a median mass-loss rate of $\dot{M}\lesssim 1.2\times 10^{-5}{M}_{\odot }{\mathrm{yr}}^{-1}$ for a Wolf–Rayet progenitor. In addition, we derive SN 2026gzf’s properties, including⁵⁶Ni mass, diffusion timescale, and expansion velocities, from our ∼nightly cadence optical data and compare them with those of optically discovered Type Ic-BL SNe, finding that SN 2026gzf is well within the 90% confidence interval across all properties. We further fit SN 2026gzf’s light curve and determine that combined emission from both interaction with CSM and⁵⁶Ni radioactive decay provides the best fit with plausible model parameters. Finally, using the rate of Ic-BL SNe from the Zwicky Transient Facility Bright Transient Survey and assuming all Type Ic-BL SNe produce EP 260321a-like FXTs, we infer an expected rate of EP-detected SBOs of 4.4–16 yr⁻¹. This is inconsistent at the 90% confidence level with current EP detection rates, potentially indicating that most Type Ic-BL SNe produce less luminous X-ray SBO signals compared to EP 260321a. 

Abstract We present early (≲1 yr) multiwavelength observations and analysis of six luminous fast blue optical transients (LFBOTs). We identified these LFBOTs in Zwicky Transient Facility survey data from their fast light-curve evolution (t_1/2≤ 12 days), blue colors at peak brightness (g − r≤ −0.2 mag), a visible host galaxy, high optical luminosity (M_g < −20 mag), and an X-ray and/or radio detection. With the exception of AT2024aehp (ZTF24abygbss), these transients exhibit peaks in their 10 GHz radio light curves att_rest ≈ 50–100 days, with peak radio luminosities ranging from 10³⁸to 10⁴⁰erg s⁻¹. Modeling the radio emission as synchrotron radiation indicates a fast (v= 0.1–0.3c) shock in a dense (n_e ≈ 10³–10⁴cm⁻³) medium. The X-ray emission varies by ≈2 orders of magnitude in luminosity (10⁴²–10⁴⁴erg s⁻¹) att_rest ∼ 20 days. Analysis of the host-galaxy photometry and spectroscopy for each transient shows that LFBOTs are predominantly nonnuclear (a few kiloparsecs offset) with star-forming host galaxies of stellar masses = 10⁹–10¹¹M_⊙. Unlike all other LFBOTs to date, AT2024aehp exhibited a luminous (M < −19 mag) plateau in the optical light curve and its 6–15 GHz radio emission brightened by over an order of magnitude fromt_rest ≈ 70 to 130 days. The mostly consistent radio behavior between these LFBOTs implies a similar circumburst medium, leading us to prefer a progenitor scenario in which mass is lost in a consistent way shortly prior to the terminal event. 

Abstract Among the various classes of fast optical transients (FOTs), kilonovae (KNe), which can emerge as a result of neutron star mergers, are extremely challenging to observe because of not only the rapid timescale on which they fade (on the order of days), but also due to the relative scarcity of their occurrence. This scarcity is compounded by the large number of other FOTs that may initially resemble the characteristic rise of a KNe. While these objects can be ruled out as candidate KNe by taking spectroscopy, a method of confidently ruling out transients based on photometric analysis alone would be incredibly valuable. We describe the compilation of various “imposter” transients, including a plurality of IIb SNe, and investigate a number of comparative metrics by which one might be able to remove transients from consideration without the use of spectroscopy. We provide a list of these objects and their classifications as well as a glossary of the transient types included in the sample. 

ABSTRACT We study AT2024kmq and AT2024lhc, two tidal disruption events (TDEs) with blue featureless spectra associated with high-mass black holes ($$M_{\rm BH}\sim 10^8\, \mathrm{ M}_\odot$$). Both events show optical precursors consistent with shock dissipation from stream self-intersection. Their X-ray emission is luminous ($$L_{\rm X}\sim 10^{44}\, {\rm erg\, s^{-1}}$$), highly variable (with minimum observed variability time-scales of 1.3 and 4.8 h for factor of ∼3 flux changes), long-lasting (>1 yr), emerging no later than the optical peak, and well characterized by power laws with $$1.7\lt \Gamma \lt 3$$ (where $$f_\nu \propto \nu ^{1-\Gamma }$$). The X-ray properties and radio non-detections support a compact corona ($$\lesssim 10 r_{\rm g}$$) producing Comptonized X-ray emission. Using all published featureless TDEs, we find statistically significant bimodality in the distribution of their peak ultraviolet/optical blackbody luminosities and radii. We assemble a comparison TDE sample with early-time X-ray observations with eROSITA, in which we find different $$M_{\rm BH}$$ distributions in TDEs with different X-ray spectral evolution properties: low-mass black holes ($$M_{\rm BH} \sim 10^6 \mathrm{ M}_\odot$$) remain soft ($$\Gamma \gt 4$$) within $$t\lesssim 2$$ yr, intermediate masses ($$\sim 10^7 \mathrm{ M}_\odot$$) transition from soft to hard at $$\sim$$1 yr, while high masses ($$\sim 10^8 \mathrm{ M}_\odot$$) are hard ($$1.5\lt \Gamma \lesssim 3$$) from the outset. We interpret this result as evidence that the soft-to-hard state transition in TDEs occurs at the critical threshold of $$\dot{M}_{\rm acc} \sim 0.03 \dot{M}_{\rm Edd}$$ (similar to X-ray binaries), using the fact that the transition time-scale predicted by simple disc theory scales with black hole mass as $$t_{\rm tr}\propto M_{\rm BH}^{-3/4}$$. 

ABSTRACT The identification of extragalactic fast optical transients (eFOTs) as potential multimessenger sources is one of the main challenges in time-domain astronomy. However, recent developments have allowed for probes of rapidly evolving transients. With the increasing number of alert streams from optical time-domain surveys, the next paradigm is building technologies to rapidly identify the most interesting transients for follow-up. One effort to make this possible is the fitting of objects to a variety of eFOT light curve models such as kilonovae and γ-ray burst (GRB) afterglows. In this work, we describe a new framework designed to efficiently fit transients to light curve models and flag them for further follow-up. We describe the pipeline’s workflow and a handful of performance metrics, including the nominal sampling time for each model. We highlight as examples ZTF20abwysqy, the shortest long gamma-ray burst discovered to date, and ZTF21abotose, a core-collapse supernova initially identified as a potential kilonova candidate. 

Abstract The Bright Transient Survey (BTS) aims to obtain a classification spectrum for all bright (m_peak≤ 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 ∼5 yr of ZTF operations. We presentBTSbot, a multimodal convolutional neural network, which provides a bright transient score to individual ZTF detections using their image data and 25 extracted features.BTSbotis able to eliminate the need for daily human scanning by automatically identifying and requesting spectroscopic follow-up observations of new bright transient candidates.BTSbotrecovers all bright transients in our test split and performs on par with scanners in terms of identification speed (on average, ∼1 hr quicker than scanners). We also find thatBTSbotis not significantly impacted by any data shift by comparing performance across a concealed test split and a sample of very recent BTS candidates.BTSbothas been integrated intoFritzandKowalski, ZTF’s first-party marshal and alert broker, and now sends automatic spectroscopic follow-up requests for the new transients it identifies. Between 2023 December and 2024 May,BTSbotselected 609 sources in real time, 96% of which were real extragalactic transients. WithBTSbotand 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. 

Abstract Spectroscopic observations of nine cataclysmic variables that have been postulated to contain magnetic white dwarfs were obtained to further characterize their classifications, orbital parameters, inclinations, and/or accretion properties. Zwicky Transient Facility (ZTF) and Transiting Exoplanet Survey Satellite (TESS) data were also used when available. This information enables these systems to be useful in global population and evolution studies of close binaries. Radial velocity curves were constructed for eight of these systems, at various states of accretion. High-state spectra of ZTF0548+53 reveal strong Heiiemission, large radial velocity amplitudes, as well as cyclotron harmonics yielding a magnetic field strength of 50 MG, confirming this as a polar system. Analysis of TESS data reveals an orbital period of 92.1 minutes. High-state spectra of SDSS0837+38 determine a period of 3.18 hr, removing the ambiguity of periods found during the low state, and showing this is a regular polar and not a pre-polar system. The ZTF light curve of CSS0026+24 shows a total eclipse with a period of 122.9 minutes, and features indicative of two accretion poles. A new, remarkably large spin-to-orbit ratio is found for ZTF1631+69 (0.61), making it, along with 2011+60 (=Romanov V48), likely stream-accreting intermediate polars. ZTF data reveal the presence of ∼2 mag low states in ZTF1631+69, and along with McDonald Observatory 2.1 m and TESS light curves, confirm a grazing eclipse that is deepest at a narrow subset of beat phases. The TESS data on PTF12313+16 also indicate a partial eclipse. Analysis of ZTF data on SDSS1626+33 reveals a period of 3.17 hr and suggests the presence of a partial eclipse.