rapid-hotwiring2026

Abstracts

Multiwavelength view of the most luminous fast blue optical transient - AT2024wpp

By: Nayana A.J.

Type: Talk

Abstract: Wide-field surveys are populating the phase space of explosive transients with new short-time-scale events. Interpreting these events requires rapid follow-up observations across multiple wavelengths, particularly for constraining their explosion mechanisms, energetics, and environments. In this talk, I will present results from our recent multiwavelength campaign on the most luminous fast blue optical transient, AT2024wpp, drawing on observations across the UV, optical, IR, X-ray, and radio bands. I will describe how these data informed our understanding of the event’s evolution and helped place it within the broader population of FBOTs. I will then discuss what multiwavelength observations across several FBOTs have revealed about this class as a whole, emphasizing both common trends and notable diversity.

Co-Authors: Raffaella Margutti Ryan Chornock Eli Wiston

The shadowing of the Einstein Probe by the Zwicky Transient Facility

By: Tomas Ahumada

Type: Poster

Abstract: Unlike the well-characterized population of gamma-ray bursts (GRBs), the origin of soft X-ray flashes (XRFs) remains poorly understood. Starting in March 2025, the Zwicky Transient Facility (ZTF) began shadowing Einstein Probe WXT observations to search for optical counterparts to EP-detected transients by crossmatching ZTF alerts with EP localization regions. Using automated tools to determine the ZTF schedule and crossmatch unfiltered EP events with ZTF alerts, along with human vetting to assess spatial and temporal consistency, this program has enabled the identification of multiple plausible optical counterparts to EP XRFs, including subthreshold events later confirmed as real transients. This work aims to describe the types of sources found during the ZTF campaign and to quantify the rate of optical counterpart detections to XRFs, as well as to describe the coordinated wide-field strategy that has led to multiple discoveries.

Co-Authors: Mansi Kasliwal, Chichuan Jin, Weimin Yuan, Lin Yan, Theophile Jegou du Laz, Dongyue Li

The first Rubin Target-of-Opportunity search for a neutron star merger candidate counterpart

By: Shreya Anand

Type: Talk

Abstract: The last part of LIGO-Virgo-KAGRA (LVK)’s fourth observing run (O4c) yielded an unexpected gravitational-wave candidate type: subsolar mass neutron star (NS) merger. On November 12, 2025, LVK discovered the subsolar mass NS merger candidate S251112cm, localized to 1681 sq. deg., at a distance of 93 +/- 27 Mpc, with a false alarm rate of 1 per 4 years. Given the exceptional nature of this merger event, we triggered Target-of-Opportunity (ToO) follow-up observations with the NSF-DOE Vera C. Rubin Observatory of the event, starting two days after the merger, following recommendations from the Rubin ToO advisory board and Rubin Observatory leadership. We obtained four epochs of observations between 2025-11-14 and 2025-11-22, covering ~39% of the skymap probability over a sky region of ~850 sq. deg. in g- and i-bands, to a depth of ~24.4 mag (g) and ~23.6 mag (i). Due to the lack of templates available in the region, we used a custom image subtraction pipeline devised to create DECam templates for use with LSSTCam images. In my talk, I will describe the results of the candidate filtering and vetting that ensued, to identify any plausible counterpart candidates to S251112cm from a large volume of ~8.7 million alerts from the observing campaign. The unprecedented breadth and depth achieved in this observing campaign by Rubin signals that it will be one of the most powerful discovery engines for counterparts to gravitational-wave mergers in the future.

TRACE-T: Automating rapid radio follow-up of transients

By: Gemma Elizabeth Anderson

Type: Talk

Abstract: As we enter a golden era of transient astrophysics with upcoming facilities such as SKAO and Rubin, we require innovative and automated observing systems to properly exploit these exciting data streams. In this talk, I will present the Transient RApid-response using Coordinated Event Triggering (TRACE-T) software suite, which automates rapid (rapid-response) follow-up of transients using radio telescopes. Such observations target the earliest radio signals emitted by transient events detected by multi-messenger facilities or output via Rubin event brokers. TRACE-T includes a web application that enables the user to select, receive and parse transients streams, which then go through user-defined decision logic. The user has full oversight of all automated observing requests sent to their target telescopes. TRACE-T has built-in front-ends for performing rapid-response triggered observations with the Australia Telescope Compact Array (ATCA) and the Murchison Widefield Array (MWA). ATCA and MWA are Australian instruments and are two of only three radio telescopes worldwide currently capable of performing rapid-response observations. Given its modularised core functionality, TRACE-T can be easily expanded to trigger other radio telescopes. I will also present results that used TRACE-T to trigger ATCA observations of Swift-detected gamma-ray bursts (GRBs), which have resulted in the very earliest radio detections of GRBs to date, providing insight into their central engines. With multiple transient phenomena motivating rapid radio follow-up, TRACE-T provides an exciting solution for triggering radio telescopes.

Constraining Cas A’s Shock Break-Out with IR Echoes

By: Rodrigo Angulo

Type: Poster

Abstract: Cassiopeia A is a well-studied supernova remnant and one of the youngest remnants in the Milky Way with the supernova occurring in the late 1600s. First infrared (IR) echoes (Krause et al. 2005) and then scattered light echoes of Cas A were found (Rest 2008), which revealed that the supernova was a type IIb (Krause et al. 2008). Further analysis of the IR echoes showed that the EUV-UV radiation of the shock break-out and/or shock cooling of Cas A was the source of the IR echoes (Dwek and Arendt 2008). With JWST, we have now obtained a set of images which shows an intricate and rich set of previously unseen and unresolved substructure in the Cas A IR echoes. I will present how with this new data set and detailed forward modeling, we are able to get much improved constraints on the physical parameters of the shock break-out and cooling of the Cas A SN.

Co-Authors: Armin Rest, Jacob Jencson, Joshua Peek

Hydrogen-rich gap transients in optical and infrared time-domain surveys

By: Andrea Antoni

Type: Talk

Abstract: “Gap" transients are broadly defined as transients whose peak luminosities fall within the gap between classical novae and supernovae. A subset of gap transients involves the partial or complete ejection of the hydrogen envelope of a star with low velocities (100s-1000 km/s), including luminous red novae, intermediate luminosity red transients, and failed supernovae. Often, these hydrogen-rich gap transients can only be detected in the infrared and even those transients that begin in the optical shift to the infrared for decades due to ongoing dust formation in the ejecta. I will give a brief overview of the theoretical and observational properties of these systems. lorem

GOATS: An end-to-end time-domain and multi-messenger astronomy platform

By: Dan Avner

Type: Talk

Abstract: A necessary element for a successful time-domain and multi-messenger astronomy campaign is rapid response of the whole follow-up ecosystem. Various state-of-the-art services, e.g., alert brokers, software for triggering telescopes, modern data reduction and analysis frameworks, etc. were developed to facilitate this. Each serves a specific goal, but there is a severe lack of coordination between these services, which unfortunately leads to a high entry barrier---currently an investigator needs to handle the various software tools and interfaces. A solution to this is building an end-to-end system.

One such system is being released at Gemini/NOIRLab called the Gemini Observation and Analysis of Targets System (GOATS). GOATS is a platform that unifies the end-to-end chain: target ingestion, Target-of-Opportunity (ToO) observation submission to Gemini via the Gemini Program Platform (GPP), automated data retrieval from supported archives, in-browser data reduction with live status and results, lightweight image analysis, and publishing the data products to external services, so investigators can move from alert to calibrated data products within a single web application.

We discuss the major components, design, and the practical challenges of linking many services into one responsive workflow, including building a browser extension that turns alerts into targets, translating a CLI/Python data reduction pipeline into an interactive web UI, streaming progress from long-running downloads and reductions, and submitting ToOs through the new GPP. By turning a complex, multi-service landscape into a dependable, easy-to-use system, GOATS lowers the barrier to rapid follow-up and lets researchers spend less time wiring tools together and more time doing science.

Co-Authors: Monika Soraisam, monika.soraisam@noirlab.edu, Gemini Observatory/NOIRLab Bill Vacca, bill.vacca@noirlab.edu, Gemini Observatory/NOIRLab Bryan Miller, bryan.miller@noirlab.edu, Gemini Observatory/NOIRLab Andrew Stephens, andrew.stephens@noirlab.edu, Gemini Observatory/NOIRLab

Multiwavelength observations of supernovae reveal the necessity of early-time radio and X-ray followup

By: Raphael Baer-Way

Type: Poster

Abstract: In recent years, improved instrumentation and dedicated follow-up have revealed that the final life stages of the progenitors of core-collapse supernovae (SNe) can be marked by drastic changes in mass loss. In particular, radio and X-ray observations that trace the interaction between the ejecta and circumstellar material (CSM) provide unique insight into the fundamental nature of progenitors through direct estimates of mass-loss rates and other shock properties. Early-time (within 0-4 weeks) radio and X-ray observations of crucial SNe, namely stripped-envelope SNe and other exotic objects that do not take tens to hundreds of days to peak in the radio/X-rays, are key to understanding where the CSM begins and how the mass loss changes in these objects. We present radio and X-ray data of three SNe for which early-time observations were pivotal to diagnosing their full mass-loss history. Early (0-50 days) radio/X-ray observations of SN 2023mut, a type IIb supernova, reveal nearby CSM at moderate density that gives way to higher-density CSM with a slower-moving shock, suggesting pulsations or an exotic binary history. Early to late-time radio and X-ray observations of SN 2019yvr, a type Ib which transitioned to show narrow hydrogen lines at optical wavelengths, reveal that the CSM was formed from relatively constant mass loss in the last 40 years pre-explosion, implying that these hydrogen profiles appeared as a function of geometry rather than significant changes in CSM density. Early X-ray observations of SN 2025coe, a Calcium-Strong Transient (CaST), show clear evidence for interaction with high-density confined CSM that now seems to potentially be a fundamental characteristic of the CaST subclass, but which makes understanding these objects more difficult. These results emphasize why early-time observations of stripped-envelope-related SNe at high radio frequencies and X-ray wavelengths are needed for a complete understanding of circumstellar environments and progenitors.

Co-Authors: Poonam Chandra, Maryam Modjaz

Time-Domain Science with the Vera C. Rubin Observatory

By: Eric Christopher Bellm

Type: Talk

Abstract: As it begins its ten-year survey, the NSF-DOE Vera C. Rubin Observatory is turning on a long-anticipated stream of transient, variable, and moving objects of unprecedented scale. I will review the design and status of the Rubin alert stream before discussing what to expect in the next two years of early operations. I will emphasize open questions and scientific and technical opportunities.

Chasing Cosmic Rarities: Intelligent Agents and SBI Physical Modeling for rapid Time-Domain analysis

By: Clecio Bom

Type: Talk

Abstract: Large-scale time-domain surveys such as the Rubin Observatory’s Legacy Survey of Space and Time (LSST) generate millions of nightly alerts, requiring automated and scalable approaches for real-time transient discovery. We present AI-SCoPE, an intelligent platform that integrates autonomous AI agents, large language models (LLMs), and fast physical inference to support alert filtering, ranking, and follow-up prioritization. AI-SCoPE agents analyze alert streams using supervised classification, anomaly detection, and probabilistic modeling within an event-driven architecture. LLMs provide contextual reasoning and cross-referencing with archival surveys, catalogs, and literature, enabling efficient human-in-the-loop discovery. The system incorporates simulation-based inference techniques to rapidly infer physical parameters directly from evolving light curves. AI-SCoPE is designed to optimize the discovery and characterization of rare and fast-evolving transients in the LSST era.

ALeRCE and the Evolving Ecosystem of Time-Domain Alert Brokers: Lessons from First Alerts

By: Guillermo Cabrera-Vives

Type: Talk

Abstract: The Rubin Observatory is ushering in the operational era of large-scale alert streams, where brokers are no longer prototypes but critical scientific infrastructure. ALeRCE is already ingesting Rubin alerts and providing public access to enriched data products, offering a unique vantage point on these early operational challenges. In this talk, I will discuss how ALeRCE has responded to the first alerts from Rubin, including the services currently provided. These early experiences highlight the need for continuous adaptation across the full broker stack, from data ingestion and feature extraction to machine learning models. I will conclude by reflecting on the implications of this evolving landscape for the next generation of time-domain infrastructure.

Hot-wiring the Roman Sky: Resolving the Undersampling Crisis with Fast IMCOM for Real-Time Discovery

By: Kaili Cao

Type: Talk

Abstract: The Nancy Grace Roman Space Telescope will revolutionize time-domain astrophysics by providing Hubble-like resolution over a field of view 100 times larger. However, this high-resolution capability comes with a significant technical challenge: the Wide Field Instrument (WFI) is natively undersampled. Standard image coaddition and regridding methodologies, such as Drizzle or Lanczos interpolation, inadvertently "bake in" aliasing artifacts and non-stationary point spread functions (PSFs). In the context of the Roman Alerts Promptly from Image Differencing (RAPID) pipeline, these artifacts may manifest as high-amplitude "dipole" residuals in difference imaging, significantly inflating false-discovery rates and masking rare transients.

In this talk, I introduce Fast IMCOM, a novel linear image reconstruction algorithm designed to meet the rigorous fidelity requirements of the Roman mission while operating within the strict one-hour latency budget of the RAPID Project Infrastructure Team (PIT). Fast IMCOM improves upon the original IMCOM formalism by separating the regridding and coaddition steps. This two-step architecture allows for an order-of-magnitude reduction in computational costs (from ~100M to <10M core-hours for the HLIS) and a drastic decrease in storage requirements (from ~1.5 PB to ~0.2 PB) via "reference-on-demand" workflows.

For time-domain observations, Fast IMCOM ensures the construction of deep, stable reference templates with user-defined, stationary PSFs. This provides the mathematically ideal input for difference imaging algorithms like ZOGY, eliminating the "PSF mismatch" term and stabilizing detection significance maps. I will present lessons from 1D experiments and 2D extensions, demonstrating how "double resonance" dither handling achieves near-perfect reconstruction of host galaxies in the High Latitude Time Domain Survey and maintains the astrometric rigidity required for the Galactic Bulge Time Domain Survey. Fast IMCOM represents the critical infrastructure needed to transform Roman’s undersampled raw data into a clean, alert-ready sky.

Scaling RAPID: HATS-Format Products and LSDB Workflows for the Dynamic Sky

By: Neven Caplar

Type: Talk

Abstract: LSDB is a LINCC Frameworks project built to enable large-scale, cross-survey analysis of all-sky time-domain datasets. It provides a unified, high-performance data model centered on HATS: a parquet-based, spatially sharded layout that supports fast cone searches, scalable cross-matching, and efficient bulk operations. We are formalizing this approach through an ongoing international IVOA standardization effort in collaboration with STScI, IPAC, CDS, CADC, and ESA.

LSDB and the HATS libraries are already deployed within major survey pipelines. The Roman calibration pipeline (romancal) relies on HATS for high-throughput spatial indexing, and we work closely with Rubin pipelines to identify issues, integrate fixes, and drive reprocessing campaigns.

A major focus of this contribution is the HATS implementation of the Rubin Prompt Product Database—the fully public, nightly database that is a strict superset of the Rubin alert stream. Delivering this dataset in HATS format enables bulk access, local parallel processing, and user-defined function execution at scale. Through the nested-pandas interface, users can efficiently construct and analyze large light-curve collections without bespoke infrastructure.

I will present lessons from operating LSDB on real survey data, the scaling and usability challenges encountered as datasets grow, and early community feedback. I will connect these experiences to the needs of RAPID, highlighting how standardized HATS-based distribution and LSDB-style workflows will accelerate multi-wavelength discovery and low-latency transient science.

Co-Authors: Melissa DeLucchi, Doug Branton, Sandro Campos, Andrew Connolly, Derek Jones, Jeremy Kubica, Olivia Lynn, Konstantin Malanchev, Rachel Mandelbaum, Sean McGuire

Chasing early TDEs with SOAR and Gemini

By: Jonathan Carney

Type: Poster

Abstract: Tidal Disruption Events (TDEs) occur when an unlucky star wanders close enough to a massive black hole to be torn apart by tidal forces, producing a luminous flare that evolves over the course of weeks. Early classification and multi-wavelength follow up enables testing of competing theories for the emission mechanism of TDEs. Pre-peak TDE lightcurves closely resemble those of type Ia supernovae, which occur at ~100x the rate of TDEs, causing most spectroscopic observations of TDEs to occur after maximum-light. We present our campaign to spectroscopically classify TDE candidates observed by the Zwicky Transient Facility using SOAR and Gemini. Our program enabled spectroscopic observations for 10 of the 28 TDEs reported to TNS in the last 12 months. Over the course of the program we transitioned from heavily using alert-filter-driven human scanning to near-exclusive use of machine-learning-aided follow-up selection with the TDEScore and NEEDLE models, with >90% of ZTF TDEs now being found via machine learning. We discuss implications for target selection in the deeper but lower-cadence regime of Rubin, the high-cadence regime of Argus, and the deep, high-cadence, narrow-field regime of Roman.

Co-Authors: Igor Andreoni, Robert Stein, Akash Anumarlapudi, James Freeburn, Sarah Teague, Nicholas Law

Transient hunting in large-sky surveys: using neural processes to tackle sparse, multi-wavelength light curves

By: Siddharth Nitin Chaini

Type: Talk

Abstract: The sparse multiband light curves from large-sky surveys like LSST are crucial for informing follow-up observations and enabling the discovery of novel astronomical events. However, a challenge is the analysis of these light curves, with their long observational gaps and differing cadences over filters. While Gaussian Process (GP) regression is a common method for interpolation, it requires specifying a kernel a priori, making a strong assumption about the object's temporal nature that may not hold for diverse transients. However, the sheer volume of LSST light curves enables a data-driven approach. Neural Processes (NPs) are a family of models that learn distributions over functions, combining the probabilistic framework of GPs with the scalability and flexibility of neural networks. In this talk, I will introduce the neural process family of models for light curve interpolation, their effects on downstream tasks such as classification and anomaly detection, and how they can aid in searching for transients in this new era of time-domain astronomy.

Co-Authors: Federica B. Bianco

From common envelope evolution to luminous red novae

By: Zhuo Chen

Type: Poster

Abstract: The ultimate unbinding of the envelope during the plunge-in phase of common envelope evolution involves complex physical processes that remain poorly understood, yet likely give rise to luminous red novae (LRNe). In this talk, I introduce a one-dimensional radiation hydrodynamic model of common envelope ejecta evolved using the code Guangqi. I will discuss the critical role radiation and gas pressure gradients play in the unbinding process. This model is capable of reproducing the light curves of LRNe and other transients, such as AT2019zhd. In the era of the Vera C. Rubin Observatory, such fast and accurate forward and inverse models are essential for understanding the origins and outcomes of CEE.

Co-Authors: Zhuo Chen

The Northern sky every second: Survey design and science plans for the Argus Array

By: Hank Corbett

Type: Talk

Abstract: The Argus Array is a 1200-telescope, two-color optical survey facility under construction for first light in 2027. Its telescopes work together to form an 8m-class instrument with an 8,000 sq. deg. field of view, building the first deep, high-speed movie of the northern sky. Argus will provide real-time public transient alerts at coadded cadences ranging from 1 second to weeks, along with images and long-term light curves. Due to its all-sky field of view, Argus will contribute continuous optical coverage to the emerging multi-wavelength and multi-messenger alert ecosystem and serve as an optical backstop to targeted follow-up programs. I will discuss how continuous on-sky coverage enables prompt detection of GW optical counterparts, GRBs, and early-time supernova emission—potentially before external triggers are issued—at depths and timescales practical for rapid spectroscopic follow-up.

Co-Authors: Hank Corbett, Nicholas Law, Alan Vasquez Soto, and the Argus Array Science Collaboration

Exoplanet Transit Prospects with Rubin/LSST: Cadence, Cadence, Cadence

By: Suber Donald Corley

Type: Poster

Abstract: The Vera C. Rubin Observatory’s Legacy Survey of Space and Time (LSST) promises transformative sky coverage and cadence, but what are the realistic prospects for discovering transiting exoplanets in such a survey? We present a forward-looking feasibility study that bridges exoplanet science with Rubin/LSST’s unique observing strategy. Using a comprehensive collection of synthetic stellar systems that spans diverse spectral types, planetary radii, orbital periods, and distances, we embed planet transits into Rubin/LSST-like light curves and apply cadence-aware transit detectability metrics. Our framework accounts for photometric uncertainties, transit depths, and the scheduled survey windows dictated by the Wide Fast Deep (WFD) and Deep Drilling Fields (DDF) surveys, enabling an apples-to-apples comparison across survey configurations.

Under the current ten-year plan, Rubin/LSST’s multi-day revisit pattern and limited intra-night sampling markedly constrain transit recovery. Even in optimistic scenarios, the overall yield is vanishingly small relative to the Galaxy’s exoplanet inventory. WFD fields contribute negligible detectable yield, while DDFs offer modest recoveries; neither configuration approaches the reach of space-based missions. Cadence constraints—particularly long gaps between transits and sparse multi-night coverage—dominate detection probabilities, underscoring the need for cadence optimization to unlock transit science with Rubin/LSST.

Our study quantifies how survey design choices translate into exoplanet yield, providing a baseline for evaluating modified observing strategies. If Rubin/LSST evolves toward cadence enhancements or dedicated exoplanet programs, the potential for statistically meaningful transit discoveries could rise substantially. This work lays a quantitative foundation for those design decisions and invites community collaboration to maximize exoplanet detections with Rubin/LSST.

Co-Authors: S. Corley, E.D. Feigelson, C. Caceres, B. Cuevas-Otahola, A.B. Kovacevic

The Faint and the Furious: What are the Lowest Mass Stars that Explode as Core-Collapse Supernovae?

By: Kaustav Kashyap Das

Type: Poster

Abstract: Core-collapse supernovae (CC SNe), the explosive deaths of massive stars, play a pivotal role in galactic chemical evolution, star formation, and the creation of neutron stars or black holes. However, the fate of stars in the ~8–12 solar mass range remains poorly understood. These stars occupy the critical boundary between those that form neutron stars and those that end as white dwarfs. Despite comprising ~50% of massive stars that explode, such events are rarely observed, likely due to their connection with faint, hard-to-detect low-luminosity SNe. I will present results from the Zwicky Transient Facility (ZTF) Census of the Local Universe, the largest volumetric SN survey to date, focusing on the landscape of low-luminosity CC SNe. By examining candidate SNe in this mass range, I will evaluate whether they can account for the missing SNe population and provide insights into the fate of these stars. I will summarize the implications for the next generation of time-domain surveys: how to most effectively use those observations to connect the final stages of binary stellar evolution and the birth of neutron stars and black holes to the observed characteristics of supernovae.

Gravitational-wave Alert System Enabling Multi-messenger Astronomy Follow-up of Fermi and Swift using RAVEN

By: Michael Davis

Type: Poster

Abstract: The RAVEN (Realtime Automated VOEvent Network) pipeline is a low-latency gravitational-wave (GW) alert system designed to enable rapid and reliable multi-messenger follow-up of compact binary coalescences. RAVEN ingests triggers from multiple independent GW search pipelines operating during LIGO–Virgo–KAGRA observing runs and performs automated coincidence analysis with high-energy observatories, including Fermi Gamma-ray Burst Monitor (GBM). By combining temporal and spatial information across instruments, RAVEN identifies candidate GW–gamma-ray associations in near real time and disseminates alerts to the broader astronomy community through standardized channels.

During the most recent observing period, RAVEN operated continuously through the end of the O4 run with stable performance and no major disruptions, demonstrating robustness in a production low-latency environment. Recent development has focused on improving coincidence validation, alert reliability, and simulated signal injection campaigns to characterize sensitivity and false-alarm behavior across astrophysically motivated source populations. These efforts have enabled improved understanding of the system’s sensitivity to different source populations and alert configurations.

Looking ahead, RAVEN is being prepared for extended low-latency operations during upcoming observing periods. By incorporating RAPID and other external alerts alongside GW and gamma-ray triggers, RAVEN aims to provide a unified low-latency platform for cross-facility alert correlation, prioritization, and dissemination. Planned work includes refining injection distributions to better represent astrophysically motivated source populations, improving alert vetting and metadata products, and enhancing scalability for increased trigger rates. Together, these developments will strengthen RAVEN’s role as a critical infrastructure component for multi-messenger astronomy, maximizing the scientific return of coordinated multi-messenger observations.

Co-Authors: Michael Coughlin

The Roman Research Nexus: A Science Platform for Roman Data Analysis

By: Gisella De Rosa

Type: Talk

Abstract: In this talk, I will present the Roman Research Nexus, the science platform developed to support analysis of data from the Nancy Grace Roman Space Telescope. All Roman data are archived and accessible through MAST, with the data hosted in AWS. The Nexus co-locates computational tools with these data, enabling users to explore and work with Roman survey products at scale without the need to move large datasets or manage local infrastructure.

I will place the Nexus within the broader ecosystem of time-domain astronomy infrastructure, focusing on its role in enabling scientific analysis of Roman data products and the challenges of building platforms that can operate at the scale of next-generation surveys.

The Modernization of the InterPlanetary Network

By: Courey Elliott

Type: Poster

Abstract: The InterPlanetary Network (IPN) is a consortium which utilizes detections of short-duration gamma-ray transients by distant detectors in order to find the precise location of gamma-ray bursts, magnetar flares, and other sources. Given the revolutionary time-domain capabilities coming online across the electromagnetic spectrum and beyond, the IPN is a key piece of the global astrophysics ecosystem, both in providing locations for other telescopes to follow-up and providing archival all-sky coverage to search for gamma-ray counterparts to externally identified sources. Historically the operation of the IPN has been manual, which is laborious and induces unnecessary delay. We are in the process of automating the operation of the IPN including tracking the location of multiple spacecraft, gathering data from multiple access points, analyzing the data to classify and localize the transient, and sending alerts to the wider community through the GCN. This system must handle heterogenous data with vastly different arrival timescales, unpredictable data drops, and limited information from instruments on-board planetary and heliophysics spacecraft. In this poster we will give a brief overview of the plans for modernizing the IPN, including how these difficult problems are being handled.

Co-Authors: Courey Elliott, Eric Burns, Suman Bala, Cori Fletcher, Adam Goldstein, Rachel Hamburg, Oindabi Mukherjee, Emily Reily

Toward a Kilonova Foundation Model in the Rubin Era: Low-Latency Multimodal Classification and Parameter Estimation [Late Abstract]

By: Felipe Fontinele Nunes

Type: Talk

Abstract: Multimessenger observations of neutron star mergers are key to understanding heavy-element production and dense-matter physics. In the Rubin era, the time-domain alert stream will be on the order of 10^7 alerts per night and dominated by supernovae and other transients, with kilonovae and GRB afterglows a fast-fading minority; this class imbalance and volume may strain traditional pipelines and risk delaying or missing the counterparts needed for rapid follow-up and joint gravitational-wave and electromagnetic analysis. Building on prior multimodal approaches for early transient classification (Junell et al., 2025), we tailor the method to these rare events and introduce a unified model that combines difference-image triplets (science, reference, difference) with multiband light curves. Within a single transformer-based architecture, a classifier filters kilonovae from the alert stream, and a conditional normalizing flow uses the images and light curves together as context to return calibrated posterior distributions for ejecta properties such as mass, velocity, and composition. Training uses simulated kilonova samples from state-of-the-art radiative-transfer modeling together with real survey data for non-kilonova classes (e.g., supernovae and AGN), with matched synthetic stamps injected into realistic host-galaxy templates. Compared with traditional pipelines used as a baseline, the framework yields accurate classifications and well-calibrated posteriors without the runtime costs of slower methods. The end-to-end pipeline operates at low latency so vetting and inference proceed in a single pass, delivering posteriors ready for follow-up and for combination across events. By enabling timely promotion of candidates together with uncertainty-aware inference at scale, this work supports population studies that constrain r-process yields and neutron-star structure and moves the field toward a general-purpose foundation model for kilonova science.

Roman hIgh-redshift transient SciencE (RISE): Enabling a Non-Cosmology Time-Domain Program

By: Ori Fox

Type: Talk

Abstract: High-redshift supernovae (SNe) are emerging as key probes of stellar evolution, galaxy environments, and fundamental physics in the early universe, with recent JWST discoveries hinting at evolving SN properties, rates, and progenitors at z>2. Progress is currently limited by small, heterogeneous samples and sparse temporal coverage, motivating the need for a large, uniform, well-cadenced, and spectroscopically confirmed high-redshift transient sample. We present the Roman hIgh-redshift transient SciencE (RISE) program, a new Roman WFI initiative designed to deliver to the community deep stacked imaging, image subtraction, and faint alert streams that extend beyond single-epoch detections. RISE is an augmentation of the RAPID project, pushing sensitivity deeper by 1 - 2 magnitudes, enabling the detections and detailed studies of high-redshift transients. It will also provide post-alert characterization and prioritization tools, including forced-photometry light curves, host-galaxy photometric redshifts, color and light-curve constraints, and machine-learning-based classification, enabling efficient identification of the most compelling high-z transients for follow-up. In combination with JWST and Rubin Observatory observations, RISE will unlock a statistically robust view of transient populations across cosmic time and open a new frontier in high-redshift time-domain astrophysics.

Enabling High-Redshift Transient Discovery with Roman: Infrastructure, Pipelines, and JWST-Informed Prioritization

By: Ori Fox

Type: Talk

Abstract: The Roman hIgh-redshift transient SciencE (RISE) program is developing the end-to-end technical infrastructure required to enable deep, wide-field time-domain discovery and characterization beyond single-epoch alerting. Building on the existing Roman RAPID pipeline, RISE introduces stacked image coaddition and subtraction at multiple cadences, faint alert streams, and scalable post-alert characterization and prioritization tools optimized for high-redshift transients. We present the current state of available Roman infrastructure and outline the critical technical work required from post-alert characterization onward, including forced-photometry light-curve construction, host-galaxy photometric redshift estimation, color–magnitude phase-space filtering, light-curve fitting, and machine-learning–based ranking of candidates for follow-up. A central design feature of RISE is the use of empirically validated JWST high-redshift transient samples as a training and truth set, enabling ML approaches that remain robust in the sparse, heterogeneous regime expected for Roman high-z alerts. We discuss the implementation of explainable boosting machines and variational autoencoders to support interpretable decision-making, anomaly detection, and scalable prioritization across thousands of alerts. This presentation focuses on the concrete infrastructure, data products, and algorithmic developments needed to transition Roman from alert generation to actionable science, highlighting remaining technical gaps and a roadmap toward a fully operational, community-facing high-redshift transient ecosystem.

Co-Authors: David Coulter, Armin Rest, Justin Pierel, Lin Yan, Mansi Kasliwal, Schuyler Van Dyk, Ben Rusholme, Jacob Jensson, Ashish Mahabal, Robert Quimby, Yuhan Yao, Robert Stein, Lou Strolger, Melissah Shahbandeh, Matthew Siebert, Takashi Moriya

The Open mulTiwavelength Transient Event Repository (OTTER)

By: Noah Franz

Type: Talk

Abstract: Multiwavelength modeling and analysis of astrophysical transients is essential for understanding the physics of these events. Currently, such analyses are challenging as multiwavelength observations of transients are inhomogeneous and spread across multiple sources. To address this issue, we present the Open mulTiwavelength Transient Event Repository (OTTER), a publicly available catalog of published transient event metadata and photometry. For this initial release, we focus on ingesting data related to tidal disruption events, and we currently include ~118,000 observations spanning 240 TDEs. Unlike previous efforts, our data schema is optimized for the storage of multiwavelength photometric datasets spanning the entire electromagnetic spectrum, from radio to X-rays. The dataset is stored using a document database structure to maximize flexibility while maintaining a level of data standardization. A web application provides an easy way to view the existing data and upload new datasets. A RESTful API, and related Python package, are available for bulk programmatic access of the data. We are currently working to add other transient data (SNe, SLSN, GRBs, etc.), and hope to add transient spectroscopy in the future. We plan to use this database infrastructure for a number of future projects including population studies of various transient events, machine learning with Rubin/Roman datasets, and light curve modeling. The dataset, web application, and API are publicly available at https://otter.idies.jhu.edu and we welcome contributions and feedback as we build our user base.

Co-Authors: Kate D Alexander, Sebastian Gomez, Collin T Christy, Tanmoy Laskar, Sjoert van Velzen, Nicholas Earl, Suvi Gezari, Mitchell Karmen, Raffaella Margutti, Jeniveve Pearson, V. Ashley Villar, Ann I Zabludoff

Disentangling the Dynamic Sky: A Modular AI Framework for Distinguishing Extrasolar Transients from Solar System Science and Satellites

By: Pablo Garcia-Martin

Type: Talk

Abstract: As the Vera C. Rubin Observatory and Roman Space Telescope open their eyes on the dynamic universe, they face a sky teeming with activity. The challenge for next-generation software infrastructure is no longer just detecting changes, but rapidly classifying them: distinguishing genuine extrasolar transients from the "noise" of the growing artificial satellite population, while simultaneously harvesting serendipitous Solar System objects (SSOs).

In this talk, I present a constellation of modular deep-learning tools designed to solve this classification challenge across diverse observational regimes. We demonstrate that robustly identifying artificial interlopers and SSOs is not merely a data-cleaning task, but a prerequisite for high-purity time-domain science. Our framework includes StreakMind, an AI-based pipeline for ground-based detection of satellites and SSOs; the Hubble Asteroid Hunter, which leverages AI to mine archival space-based data for asteroids and artificial objects; and a novel machine learning application for CHEOPS that identifies satellite glints in high-precision photometry.

Crucially, we address the challenge of scaling these tools to future missions through domain adaptation. We discuss methods to reinforce transfer learning between observatories—applying lessons from current ground-based surveys to LSST and space-based archives to Roman. To overcome the bottleneck of manual labeling, we demonstrate the efficacy of mixing synthetic training sets with real data, creating robust models ready to operate.

Our results highlight the dual value of this infrastructure. We quantify satellite contamination by classifying >12,000 trails in CHEOPS and HST data, while simultaneously "hot-wiring" archives to reveal the Solar System population hidden behind the noise—reporting the discovery of >1,000 previously unknown asteroids (down to sub-kilometer sizes). I will discuss how integrating these mission-agnostic AI modules into future alert brokers will optimize discovery, ensuring that the "firehose" of alerts is accurately filtered to deliver pure samples of asteroids, satellites, and astrophysical extrasolar transients.

Co-Authors: Rafael Carrillo, René Duffard, Javier Romero, Nicolás Morales, Luis Gonçalves, Stephan Hellmich, Nicolas Billot and Bruno Merín.

BOOM/Babamul: synergistic brokering over the next decade

By: Matthew Graham

Type: Talk

Abstract: BOOM/Babamul is an open source, Rust-based alert brokering infrastructure built on the capable foundations of ZTF but designed to operate in a fully multistream environment. Currently serving joint LSST-ZTF alerts, we will showcase the features that will allow it to integrate upcoming streams at minimal cost and unlock the science that comes from a multifaceted transient system.

The Algebraic-to-Physical Alphabet: Coupling Morphological Phase Dynamics and Spectral Geometry for Roman TDA Transient Classification

By: Roberto Guenzani

Type: Poster

Abstract: This study proposes a Unified Detection Framework for Roman TDA Transients that builds on a compact, runnable prototype inspired by the Zwicky Transient Facility (ZTF) Bright Transient Survey (BTS). The prototype implements the Landing–Drilling Synchronism Detection model, where controlled mechanical or electromagnetic perturbations—analogous to landing/drilling events—generate synchronized, multi-modal transients detectable across seismic, electromagnetic, and photometric domains. Using ZTF-like light-curve data, a pipeline can simulate this coupling by injecting temporally correlated perturbations into synthetic photometric series, reproducing shock-like coherence patterns. The results demonstrate that phase-locked transients can be statistically isolated through Hilbert-envelope extraction and cross-spectral coherence, validating the capacity of synchronized forcing to amplify specific spectral modes. The theoretical foundation employs a triadic algebraic-to-physical mapping from Non Linear Ginzburg-Landau-Schroedinger equation as an “Alphabet” linking spatial morphology (Cahn–Hilliard phase separation), periodic wave propagation (Hill parametric equation), and spectral correlation (Bochner’s theorem). In this framework, each perturbation functions as a time-tagged parametric excitation, generating Floquet-type instabilities and narrowband amplification that encode information about underlying geometric and spectral structures. Such coordinates can be used as Subjective reference in the new scenario of Coronagraph coupling, so that, when scaled to the Roman TDA and RAPID pipelines, this model provides a commutative architecture unifying mechanical, electromagnetic, and photometric domains. Multi-modal synchronism acts as a physical transducer layer, establishing priors for machine-learning classification of Roman transients. The next step is to implement the pipeline on real ZTF and Roman precursor datasets, correlating simulated drilling-induced transients with NIR light curves (0.48–2.3 μm) from the RAPID alert stream. By coupling theoretical phase-field dynamics with real observational data, the framework offers a novel transducer-based methodology for detecting coherent transient responses, potentially revealing new classes of synchronized, planetary-scale phenomena in the time-variable infrared sky such as Faraday Inversion Nodes.

Utilizing DESI as a Transient Discovery Machine

By: Xander James Hall

Type: Talk

Abstract: Wide field surveys are discovering extragalactic transients at immense rates, that already overwhelm spectroscopic follow-up capabilities. In the era of the Vera C. Rubin's Legacy Survey of Space and Time, it will become impossible to achieve spectroscopic completeness. In order to combat this growing problem in the field, we have turned to the Dark Energy Spectroscopic Instrument (DESI) as a new solution for large scale transient classification. With it's 5000 fiber optic cables, DESI is able to observe over 100,000 galaxies every single night. By utilizing the numerous "spare fibers" in any given DESI field, we can spectroscopically observe dozens of transients. Utilizing a search for legacy transients within DESI, we have demonstrated that DESI has the potential to observe transients down to ~22 magnitude. Furthermore, DESI will be unphased by the given on sky transient density proposed by Rubin. By utilizing a DECam survey (DESIRT), as a test bed, we can see that the percent completeness of transients observes remain constant despite an order of magnitude increase in targets. Here we present the first year results of this experiment and its ongoing performance in concert with the Vera C. Rubin Observatory, and demonstrate some of the science with Tidal Disruption Events, Infant Transients, and host subtraction that can be performed with DESI spectroscopy.

Co-Authors: Antonella Palmese (CMU) John Banovetz (LBNL)

Access Multi-Messenger Alert Data with the Pitt-Google Alert Broker

By: Christopher Hernandez

Type: Talk

Abstract: The Pitt-Google Alert Broker is designed to provide global public access to astronomical alert stream data using tools that address the challenges our field currently faces in data transport, processing, storage, and access. In this tutorial, we will explore the pittgoogle-client, a Python library developed by our team for accessing data served by our broker through the Google Cloud Platform and demonstrate how users can optimize transient discovery with multi-wavelength and multi-messenger alert streams using the services our broker provides.

Decoding the Diversity of Type IIb Supernovae and their Progenitors with ZTF.

By: K-Ryan Hinds

Type: Poster

Abstract: Type IIb supernovae (SNe IIb) offer a unique window into the evolution of partially stripped massive stars before core-collapse, connecting hydrogen-rich Type II to hydrogen-poor stripped-envelope explosions. Their evolution - from early hydrogen features to helium-dominated spectra - and their frequent double-peaked light curves arise from rapid shock-cooling emission of a low-mass, extended hydrogen envelope, followed by the emergence of 56Ni-powered emission.

We present a homogeneous sample of 131 SNe IIb (59 exhibiting two distinct peaks), assembled from Zwicky Transient Facility (ZTF). The cadence and multi-band coverage from ZTF enable systematic recovery of shock cooling flashes and consistent measurements of both peaks, including rise times, peak luminosities, color evolution, and light-curve widths. We identify significant correlations between shock-cooling and radioactive peak luminosities, as well as between peak brightness and light-curve width, allowing us to map the diversity of SN IIb photometric evolution across a broadening observational parameter space, and to connect early-time progenitor structure to later-time photometric behavior.

To interpret these observables in terms of their progenitor systems, we model binary-stripping pathways by evolving interacting binaries with MESA and exploding the pre-SN structures with SNEC across a uniform grid of progenitor parameters (e.g., progenitor radius, explosion energy, ejecta mass, and 56Ni mass). We compare the observed light curves to the hydrodynamical model grids, yielding informed distributions over progenitor parameters. A central outcome of this work is a physically motivated framework for distinguishing SNe IIb from fast blue optical transients and kilonovae in early-time surveys, while placing SNe IIb in context with other energetic explosions.

Co-Authors: Avinash Singh (OKC), Christoffer Fremling (Caltech), Jesper Sollerman (OKC), Kaustav Das (Caltech), Mansi Kasliwal (Caltech), Dan Perley (LJMU).

Relativistic Stellar Transients with ULTRASAT

By: Anna Ho

Type: Talk

Abstract: Relativistic stellar transients are systems in which accretion or rotation of a newborn compact object drives a powerful outflow. For decades, the only firmly established example was long-duration gamma-ray bursts (GRBs), thought to represent the case in which black-hole accretion drives a narrow ultra-relativistic jet. However, in recent years the landscape has broadened dramatically, because discovery methods have expanded from solely gamma-ray satellites to include time-domain surveys at other wavelengths. The physical connections between emerging classes---GRBs, “orphan” afterglows, low-luminosity GRBs (LLGRBs), and luminous fast blue optical transients (LFBOTs)---are uncertain, but likely include differences in end-stage stellar evolution, compact-object accretion, and jet properties. In my talk, I will focus on the impact that ULTRASAT---an upcoming wide-field, high-cadence UV survey---will have on this field. ULTRASAT will increase the discovery rate of relativistic transients outside the gamma-ray band by 1-2 orders of magnitude, and provide high-cadence UV light curves for each class. The infrastructure needed for rapid UV transient discovery will enable spectroscopy of fast transients with future UV missions (UVEX). High-cadence UV light curves will provide unique diagnostics of rapid outflows (e.g., jet cocoon emission) and progenitor structure.

The Multimessenger Treasure TROVE: A Tool for Rapid Object Vetting and Examination

By: Griffin Hosseinzadeh

Type: Talk

Abstract: As it did for GW170817, multimessenger astronomy has the potential to revolutionize our physical understanding of astronomical transients by associating gamma-rays, x-rays, gravitational waves (GWs), neutrinos, and high-energy particles with their UV, optical, and infrared counterparts. However, as improvements to sensitivity and fields of view allow us to probe larger volumes (in all wavelengths and messengers), we are no longer limited by survey speed in covering even poorly localized multimessenger events, but rather by sorting through the large number of false-positive transients in the localization region, which can number in the hundreds for a typical GW event. This has motivated us to build the Treasure TROVE, a public, open-source web platform for vetting optically discovered transients and ranking their likelihood of association with various types of multimessenger events, thus prioritizing them for spectroscopic follow-up. Building on the infrastructure from our GW follow-up project, SAGUARO, the Treasure TROVE ingests all public transients in a multimessenger localization region; crossmatches them with potential host galaxies, variable point sources (stars and AGN), and minor planets; aggregates photometry to investigate light-curve evolution; and calculates a score, updated in real time, that assesses the probability of association. This powerful tool will be made available for public use via both a web interface and an API, allowing the community to more efficiently coordinate follow-up and its results. Our analysis of previous GW observing runs has shown that this approach can save or reprioritize more than half the time currently spent on multimessenger follow-up, including with the world’s largest optical and infrared facilities.

The Future of SN Ia Cosmology with the Nancy Grace Roman Space Telescope

By: Rebekah Hounsell

Type: Talk

Abstract: The Nancy Grace Roman Space Telescope will conduct a monumental time-domain survey focused on Type Ia supernovae (SNe Ia). Roman's combination of wide-area, wavelength coverage, and observational depth will produce the most comprehensive and high-quality SN Ia dataset of this generation. This exceptional dataset will significantly advance our understanding of dark energy. In this talk, I will present the work of the Supernova Cosmology Project Infrastructure team, including the tools, data products, and resources we are developing to enable successful SN Ia science investigations for the broader community.

The Time Domain Ecosystem

By: Andy Howell

Type: Talk

Abstract: I will review our efforts to overhaul our time domain communications infrastructure to improve ground and space-based responses to transient events. These include the Target and Observation Manager (TOM) Toolkit, AEON+, HEROIC, Hopkootch/HERMES, and Treasure Map. The TOM Toolkit is a framework for obtaining and managing observations, including a database, a UI and API, and mechanisms for requesting, obtaining, reducing, and analyzing data. AEON+ is an effort to streamline the automatic requesting of data from telescopes using a common API. HEROIC is a centralized repository of observatory and instrument availability, including telescope pointings. Hopskotch and HERMES are a messaging service blending human and machine readable information, enabling the real-time sharing of data of all types. Finally, Treasure Map is a way to coordinate followup of poorly localized triggers (such as gravitational wave events), allowing users to report and visualize their telescope pointings. All of these systems work together to enable faster responses to transient triggers, ensuring earlier data and better scientific outcomes.

Joint analysis of ZTF and TESS data: Constraints on early-time shock-cooling emission and late-time flaring

By: Rahul Jayaraman

Type: Talk

Abstract: Continuous, high-cadence monitoring of the sky is key to uncovering rarely-observed, short-timescale phenomena in explosive transients, especially at early and late times. This talk will focus on how we leverage the 200-s cadence and 2304 deg^2 field of view of the Transiting Exoplanet Survey Satellite, in conjunction with the Zwicky Transient Facility (ZTF) and upcoming surveys such as the Vera Rubin Observatory's Legacy Survey of Space and Time (LSST), to systematically search for early-time shock cooling emission in core-collapse supernovae and late-time flaring (on timescales of minutes) from central engine-powered transients, such as luminous fast blue optical transients (LFBOTs). I will present our constraints on and characterization of early-time excess emission in Types Ib, Ic, and II supernovae from ZTF’s Bright Transient Survey (BTS), a magnitude-limited survey in which all transients brighter than 19th magnitude in r-band are spectroscopically classified. I will discuss the relative strengths and drawbacks of various difference imaging approaches for TESS when applied to the BTS core-collapse supernova sample. I will also discuss our limits on flares from LFBOTs and other central engine-powered transients, as well as the implications of our results for the rates and duty cycles of such flaring. Finally, I will discuss synergies between ZTF, TESS, and Rubin/LSST for bright, fast transients. In particular, early-time, high-cadence observations from TESS of a large sample of supernovae and LFBOTs can help uncover their underlying physics.

Co-Authors: Anna Ho Michael Fausnaugh Ruslan Konno Eran Ofek

Enabling Time-Domain Science for Roman with RAPID

By: Jacob Jencson

Type: Talk

Abstract: RAPID (Roman Alerts Promptly from Image Differencing) is a Project Infrastructure Team for the Nancy Grace Roman Space Telescope. RAPID will enable a wide suite of time-domain astrophysics, identified as a high priority in the Astro2020 Decadal Survey, for the Roman community. Based at Caltech, the RAPID team is leveraging experience in building similar services for dedicated time-domain surveys, particularly the Zwicky Transient Facility. We will provide four services for Roman: rapid, low-latency image differencing of every new Roman image from a reference image; a prompt public alert stream of all Roman transient and variable candidates; source-matched light curve files for every identified Roman candidate; and, an on-demand forced-photometry service for the photometric history at any observed sky location. In this talk, I will provide an introduction to the RAPID project, an overview of our planned community deliverables, and the latest updates on our pipeline and infrastructure developments.lorem

Low-latency Forecasts of Kilonova Light Curves for Rubin and ZTF

By: Natalya Johnson

Type: Talk

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 (LSTM) 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.

Stellar mergers and common-envelope transients in the Roman-Rubin era

By: Viraj Karambelkar

Type: Poster

Abstract: Stellar mergers following common-envelope evolution in binary systems produce dusty, low-luminosity transients called Luminous Red Novae (LRNe). These transients provide important insights about the elusive phase of common-envelope evolution in massive binaries, and may play an important role in the cosmic dust budget. However, our current understanding of these subluminous, dusty transients is limited by the sensitivity of current surveys. In this talk, I will describe results from systematic searches for massive stellar mergers using the Zwicky Transient Facility, follow-up observations of these dusty transients with JWST, and infrared searches for the dustiest mergers using the Wide-field Infrared Transients Explorer (WINTER) and the NEOWISE missions. I will describe the insights that these searches have provided about binary star evolution, the onset and outcomes of common-envelope evolution, and the potential role of stellar mergers as a previously overlooked source of cosmic dust. Drawing on the lessons learnt from these optical and infrared searches, I will outline strategies for upcoming LRNe-searches with Rubin and Roman, that are poised to increase their discovery rate by orders of magnitude, revolutionizing our understanding of binary star evolution.

TDEs as a probe of supermassive black hole evolution across cosmic time: redshift-dependent TDE rates in LSST, Roman, and JWST

By: Mitchell Karmen

Type: Poster

Abstract: In measuring the mass distribution of supermassive black holes (SMBHs) over cosmic time, the <10^8 M_sun population is the worst-constrained at z>1. However, this population is also the most sensitive to SMBH seed and growth models, and may distinguish between them. In this talk, I forecast the TDE rate in LSST, Roman, and wide-field JWST surveys given multiple SMBH mass function models, and show that the redshift-dependent rate of TDEs is very sensitive to the SMBH mass function. Additional factors that impact the TDE rate also evolve with redshift-- namely, increasing galaxy nuclear densities, enhanced galaxy-galaxy merger rates, dust obscuration, and a possible top-heavy IMF at early cosmic times. I model all four of these effects' impact on the TDE rates and find a new prediction for the volumetric TDE rate that increases with redshift until a maximum at cosmic noon, after which the black hole mass function dominates the TDE rate which decreases with redshift. I provide a basic methodology for measuring the BHMF given a TDE rate in a flux-limited survey, which can be applied to optimize TDE science during survey design.

Co-Authors: Suvi Gezari

AARTFAAC 2.0: Filling the Gaps in Low Frequency Radio Phase Space

By: Catheirne Kelley

Type: Talk

Abstract: Recently, a new population of bright, seconds to minutes duration radio transients have been discovered, for example the long period transient discovered by de Ruiter et al. (2024). These sources were previously missed or overlooked by traditional search methods due to significant computational limitations processing radio data. The LOw-Frequency ARray (LOFAR) is undergoing a major upgrade to LOFAR2.0, which includes a new, faster, fully automated transient detection pipeline enabling us to search vast quantities of data for more of these events.

In addition, we are building a new Amsterdam ASTRON Radio Transient Facility and Analysis Center (AARTFAAC) 2.0, a whole visible sky, low frequency radio transient instrument comprising 12 LOFAR 2.0 Dutch stations (Shulevski et al. 2020). AARTFAAC 2.0 will increase the bandwidth of the original system by an order of magnitude to 30 MHz and produce whole sky images with 1 second cadence. AARTFAAC2.0 will have a ~400 TB storage buffer, which will allow for data to be stored for roughly 24 hours. Meaning, AARTFAAC2.0 will be able to provide observations of multi-messenger transients covering timescales from hours before the event to hours afterwards. This is particularly interesting as coherent, low frequency radio emission is expected prior to and following compact binary mergers (Rowlinson et al. 2019).

Beyond the application to triggering events, AARTFAAC2.0 is an excellent tool for finding transients on timescales of seconds to minutes. For example, we have detected a 30 Jy Galactic transient with several inhomogeneous pulses that have durations between 10-60 seconds. This event, along with several other bright and long duration transients were found using the original AARTFAAC system. Considering the extensive pipeline improvements and the upgrade to AARTFAAC2.0, we expect to find more exciting transients on timescales of seconds to minutes.

Co-Authors: Antonia Rowlinson

OMEGA - the pathfinder for LSST follow-up

By: Katarzyna Z Kruszynska

Type: Talk

Abstract: While Rubin Observatory's LSST is expected to discover thousands of microlensing events, the average cadence is likely to be too low to characterize planetary and stellar binary events from survey data alone.

Rubin's real-time alerts will enable us to trigger follow-up at other facilities to provide higher cadence photometry to constrain light curve features and spectroscopy to determine source properties.

The OMEGA Key Project is a pathfinder follow-up program for microlensing across the Milky Way. Building on the infrastructure of Rubin's alert brokers, the project's Target and Observation Manager can conduct follow-up observations at scale.

Observations of targets from existing surveys are mostly conducted by 1~m telescopes, and occasionally with the 2~m facilities, which are sufficient for currently operating surveys that produce microlensing alerts. However, Rubin's limiting magnitude in the Galactic Plane goes beyond the capabilities of small-scale observatories.

I will present efforts to build infrastructure for gravitational microlensing follow-up observations in the LSST era. For this purpose, I will focus on two Target and Observation Managers that will be used to select, prioritize, and request follow-up on LSST microlensing events: the Microlensing Observing Platform and the Microlensing Observing Program Manager, which is an in-kind contribution from the University of Heidelberg. Additionally, I will discuss useful target selection and prioritization methods, as well as software designed to streamline these algorithms. Our goal is to improve our understanding of gravitational microlensing events in the upcoming LSST era.

Co-Authors: Rachel Street, Etienne Bachelet, Markus Hundertmark, Yannis Tsapras

AT 2024wpp in UV to NIR: The Unprecedented Evolution and Properties of a Luminous Fast Blue Optical Transient

By: Natalie LeBaron

Type: Poster

Abstract: Luminous Fast Blue Optical Transients (LFBOTs) are a new class of transients noted for their rapid rise to peak, extreme luminosities, blue colors, and X-ray and/or radio emission. LFBOT X-ray luminosities suggest that a central engine powered by accretion onto a compact object supplies the energy needed to produce such emission, but due their short evolution timescales, low discovery rates, and typically large distances, the scarcity of well-observed objects has prevented strong constraints on their intrinsic nature. Here we present multiwavelength observations of AT 2024wpp, the third LFBOT to be well sampled in the UV through NIR. AT 2024wpp is the most luminous LFBOT observed to date in both UV (M_peak = -22.98) and bolometrically (L_peak ~ 10^45 erg/s), exceeding that of the prototypical AT 2018cow by a factor of ~5--10. Similar to other LFBOTs, AT 2024wpp’s optical spectra were initially featureless blue continua indicating large expansion velocities of >0.2c and temperatures in excess of 30,000 K. After ~35d, the spectra show clear spectroscopic features indicating the presence of H and He at velocities of both v = 0 and a blueshifted component at v = -6400 km/s. The emergence of these features coincides with when L_X ~ L_UVOIR, similar to behavior observed in AT 2018cow. AT 2024wpp further shows the second NIR excess observed in an LFBOT at +30 d.

SN2025wny: a Hydrogen-poor Superluminous Supernova at z ≈ 2

By: Maggie Li

Type: Poster

Abstract: The properties of stars in the early universe remain poorly understood. At high redshift, lower metallicities and potentially altered binary fractions are expected to significantly affect stellar evolution and end-of-life outcomes, but direct observations of individual stars are largely inaccessible due to their extreme luminosity distances. Hydrogen-poor superluminous supernovae (SLSNe-I), which preferentially occur in high-redshift galaxies and are over 100 times brighter than classical core-collapse supernovae (CCSNe), provide a powerful means of overcoming this limitation. However, SLSNe comprise less than 1% of all SN discoveries and current studies are strongly biased toward local (z < 1) events, leaving their physics and possible cosmic evolution poorly constrained. Here, we present SN2025wny, a SLSN-I at z = 2.011 and the most distant spectroscopically confirmed CCSN to date. SN2025wny is strongly lensed by two foreground galaxies, producing five resolved images and enhancing its observed brightness by a factor of ~30 - 50. This unique lensing configuration provides a once-in-a-lifetime opportunity to obtain detailed observations of a high redshift SLSN-I, from the photospheric to nebular phases. We report spectroscopy and photometry covering the rest-frame UV and optical during the photospheric phase, providing initial constraints on the ejecta composition and explosion physics. These observations, alongside continued monitoring into the nebular phase, will establish the first robust spectral templates of a high-redshift SLSN-I, providing an essential reference for identifying and analyzing high-redshift SLSNe in the upcoming LSST and Roman era.

Co-Authors: Lin Yan, Yujing Qin, Joel Johansson, Ariel Goobar, Suhail Dhawan, Daniel Perley, Avinash Singh, Steve Schulze, Ragnhild Lunnan

Now in Color! The 20-year Light Curves of Eta Carinae's Great Eruption from Its Light Echoes

By: Xiaolong Li

Type: Poster

Abstract: The Great Eruption (GE) of Eta Carinae in the mid-1800s was a spectacular astronomical event, has become the archetype for understanding unstable very massive stars, eruptive mass loss, luminous blue variables, and supernova impostors (e.g., Smith et al. 2014). Prior to the discovery of light echoes, the only observations of Eta Carinae's historical eruption were visual estimates of its brightness and approximate colors (Smith & Frew 2011).

Light echoes (LEs), reflections of light from transients off interstellar dust, offer us the opportunity to re-observe Eta Carinae's eruption with modern instrumentation (e.g., Rest et al. 2012, Smith et al. 2018). Due to light travel delays, LEs from different dust locations capture the eruption at different phases, when combined, offering us a complete and continuous view of the event.

Here we present the complete 20-year light curve of the GE in four photometric filters (griz) derived from 10 years of observations of its LEs. Compared to the historical record, we confirm the three major peaks in 1838 (1st), 1843 (2nd) and 1845 (3rd). We find that between the 2nd and 3rd peak of the GE, Eta Car reddens by more than 1 magnitude, while also increasing by 2 magnitudes in i-band. This extreme reddening suggests a significant expanding and cooling process, consistent with a change in stellar type from G to K (a drop of 1500 K). This was also seen in the light echo spectra of 3rd peak (Prieto et al. 2014). Furthermore, we find a 4th previously unknown peak, occurring during the 1850 plateau, which coincides with the onset of explosive event. Together, our findings provide unprecedented detail on the evolution of the GE, enabling direct association with spectroscopic observations (e.g., Prieto et al. 2014, Smith et al. 2018a,b) and providing new constraints on the physics driving this extreme eruption.

Co-Authors: Rodrigo Angulo, Armin Rest, Jacob Jencson, Jennifer Andrews, Nathan Smith, Charlie Kilpatrick, et al.

BHTOM: The Optical Follow-up Engine for Multi-Messenger Astronomy

By: Joysankar Majumdar

Type: Poster

Abstract: The era of multi-messenger astronomy is defined by a diverse and rapid stream of transient alerts. The sky is now continuously monitored across the high-energy spectrum from X-rays to Very High Energy gamma rays, alongside non-electromagnetic parties including gravitational waves and high-energy neutrinos. A critical challenge in this field is the effective association of these often poorly localized, interesting transients with their optical counterparts which is a necessary step for physical characterization. To bridge the gap between these complex detections and optical follow-up, we present Black Hole Target and Observation Manager (BHTOM, bhtom.space). BHTOM serves as a centralized engine capable of ingesting this wide variety of transients, including standard optical transients from large-scale surveys, and rapidly coordinating follow-up observations. BHTOM leverages a heterogeneous global network of about 162 telescopes for optical observations, including an amateur community of observers. It employs an automated pipeline that ingests raw data and delivers standardized science-ready photometry. We observe a wide range of phenomena, from rapid transients, supernovae, and TDEs, to slowly varying quasars and variable stars, requiring long-term monitoring. This discussion will demonstrate how BHTOM serves as a unified interface for the transient sky, enabling the community to seamlessly convert multi-messenger alerts into comprehensive optical light curves.

Co-Authors: Łukasz Wyrzykowski (Astronomical Observatory, University of Warsaw; Astrophysics Division, National Centre for Nuclear Research) Mariusz Gromadzki (Astronomical Observatory, University of Warsaw)

BHTOM: The Optical Follow-up Engine for Multi-Messenger Astronomy

By: Joysankar Majumdar

Type: Talk

Machine learning and scalable analysis of light curves

By: Konstantin Malanchev

Type: Talk

Abstract: This talk reviews key applications of large-scale light-curve analysis, with an emphasis on classification and anomaly-detection pipelines. Over the past few years, machine-learning approaches have gained significant traction, driven by the Zwicky Transient Facility (ZTF) and by preparations for next-generation surveys such as the Vera C. Rubin Observatory LSST and the Roman Space Telescope. I will discuss methodological and practical challenges specific to irregular and multi-band time-series data, including feature representation and evaluation in regimes with limited or evolving labels. Particular attention will be given to the gap between research prototypes and production-grade machine-learning systems, addressing issues such as scalability, reproducibility, model monitoring, and concept drift. Finally, the talk will contrast analysis strategies for "static" data releases with those required for real-time alert streams, highlighting implications for anomaly detection and rapid classification.

LightCurveLynx: A community-driven time-series modeling toolkit for LSST

By: Alex Malz

Type: Poster

Abstract: Light curve simulations are essential for the time-domain astronomy community to prepare for and to conduct at-scale analyses of LSST data, enabling exploration of the detectability of novel classes, validation of data processing pipelines, and both implementing and executing advanced analysis techniques such as likelihood-free inference. The LSST Interdisciplinary Network for Collaboration and Computing (LINCC) Frameworks team is developing a new software package, LightCurveLynx, to address the need for an accessible, scalable, and extensible light curve simulation suite to meet the needs of the time-domain astronomy community in fully exploiting LSST data for discovery and physical insights. This presentation showcases the capabilities of the LightCurveLynx codebase as a common simulation software infrastructure enabling sophisticated and efficient modeling across all science foci.

Co-Authors: Konstantin Malanchev, Mi Dai, Jeremy Kubica

HEROIC: A Realtime Observatory Coordination System

By: Curtis McCully

Type: Talk

Abstract: We have entered a new era in time domain astronomy where exciting transients are being discovered more rapidly and in larger numbers than ever before. This is partially due to better hardware, but without improved infrastructure, the science being done today would not be possible. Discovery is only the first step: other facilities that provide increased time sampling, cover other wavelengths, or, in the case of neutrinos or gravitational waves, detectors provide an entirely different messenger, giving us a more complete picture of the astronomical phenomenon. To catch these transients before they have time to evolve, we must coordinate between many observing facilities. In those valuable minutes, an observer must identify which facilities are available and have the necessary capabilities, and then trigger them with their pre-approved programs. To solve the former issue, the Scalable Cyber Infrastructure for Multi-Messenger Astronomy (SCIMMA) has developed the Hop-Enabled Rapid Observatory Information and Coordination (HEROIC) System. HEROIC provides a centralized system for observatories to report their current status, instrument updates, and planned, upcoming outages. I will present HEROIC, including the web-based visualization and query options for checking on which instruments can observe and when. I will also provide examples of how the backend API can be used to automate observing workflows, sometimes removing the human from the loop entirely, allowing us to obtain follow-up observations faster, opening evolutionary phases that were previously not possible.

Co-Authors: Gautham Narayan Andy Howell

Volumetric Rates of Type Ia Supernovae in DES and forecasts for Roman using solely photometric data

By: Cole Meldorf

Type: Poster

Abstract: The use of Type Ia Supernovae (SNe Ia) as standardizable candles has led to the discovery of the expanding universe and they continue to contribute to our understanding of Dark Energy. A better understanding of the volumetric rate of SNe Ia will allow for improved survey design to detect as many of these objects as possible, as well as shed light on the nature of their progenitors. In this work, we expand upon a technique for measuring the volumetric rate of SN Ia began in an unpublished Ph.D. thesis. In this method, two simulations are generated using the supernova simulation program SNANA: a simulation of the total number of supernovae in the spacetime volume covered by the survey, as well as a simulation of the number of supernovae that would be detected and pass quality cuts in this survey. Using these simulations, one can infer an efficiency and fit a true volumetric rate to data. Looking to the future, the gigantic and deep samples of SNe Ia we expect from the Vera Rubin Observatory's Legacy Survey of Space and Time and the Roman Space Telescope will provide many more supernovae than we can follow up on spectroscopically. Therefore, we do our analysis entirely photometrically. We correct for the biases introduced by core collapse contamination using our simulations and incorporate the error in photometric redshift measurement by calculating efficiency in a two dimensional matrix of simulated versus recovered redshifts. Finally, this method will allow us to combine multiple surveys for additional constraining power on the volumetric rate. We discuss preliminary results on fitting rate models to the Dark Energy Survey 5-year survey dataset as well as discuss rate measurement forecasts for the Roman Space Telescope.

Co-Authors: Masao Sako, Richard Kessler, James Lasker

Gemini Next Generation Operations Software: Gemini Program Platform and Scheduler

By: Bryan Whitfield Miller

Type: Poster

Abstract: Gemini Observatory will be an important facility for following up time-domain discoveries in the Rubin and multi-messenger eras. Gemini will work within the Astronomical Event Observatory Network (AEON) along with SOAR, NOIRLab, Las Cumbres Observatory, and likely Rubin in-kind telescope time contributions. Queue observing allows Gemini to easily execute target-of-opportunity (ToO) observations and this capability is very important for transient follow-up. However, updates are needed to support AEON and the potential increase in ToO requests. This will involve major updates to the observatory control system, archive, and data reduction software. Most of the presentation will provide updates on the Gemini Program Platform, a major software project that is rewriting and re-envisioning Gemini's proposal and observation preparation software. It will also provide APIs for querying status information about the observatory and for submitting and managing observations programmatically, e.g. via the TOM Toolkit. Observing plans will be updated dynamically with scheduling software that is also under construction. First use is expected for the 2027A call for proposals so the testing and deployment plans will be reviewed. Gemini's new tools will be an important part of the MMA ecosystem and may inspire or impact future software planning.

Co-Authors: Andy Stephens, Aleksandar Cikota, Arturo Nunez, Andy Adamson, Kristin Chiboucas, German Gimeno, Mark Everett

Metallicity dependence of pair-instability supernova properties

By: Takashi Moriya

Type: Poster

Abstract: The next generation time-domain surveys such as those conducted by Roman Space Telescope are expected to discover many high-redshift transients. Especially, pair-instability supernovae (PISNe) predicted to appear preferentially at high redshifts are important objectives for high-redshift transient surveys. While PISNe can appear in various redshifts with different metallicities, most PISN models constructed so far assume zero or extremely low metallicities. In this presentation, we investigate the metallicity dependence of PISN properties with a focus on hydrogen-free PISNe. We evolved helium stars from the helium zero-age main sequence to the pair-instability explosion phase with the initial metallicity range of 0-10 Zsun and calculate the light curves of the subsequent PISN explosions. We find that the explosion energies and synthesized 56Ni masses from PISNe with the helium core mass above 80 Msun do not strongly depend on metallicity. When the helium core mass is below 80 Msun, the explosion properties above 2 Zsun are found to be significantly different from those below 2 Zsun. PISN light-curve properties are found to have metallicity dependence. The rise times of the PISN bolometric light curves tend to be longer for higher metallicity PISNe because of higher opacity. The peak bolometric luminosity is not affected by metallicity much. The metallicity dependence is more significant in shorter wavelengths. We conclude this presentation by discussing the effect of metallicity on the detectability of hydroge-free PISNe in the future high-redshift transient surveys.

Fink Broker: Live Demo and Data Access

By: Anais Möller

Type: Talk

Abstract: The Vera C. Rubin Observatory will revolutionize time-domain astronomy with millions of nightly alerts. Fink is a community-driven broker currently processing ZTF alerts and a selected Rubin community broker.

In this demo, I will showcase Fink's capabilities and how to use it for your science. I will demonstrate our machine learning classification and filtering pipelines across diverse science cases—supernovae, kilonovae, solar system objects, and stellar variability—and show how alerts are prioritized in real-time. I will walk through accessing Fink data via our web portal and API, and demonstrate how to design custom filters for your specific science needs. For science cases ranging from extragalactic transients to galactic variables, I will demonstrate efficient alert extraction tailored to specific research programs.

SCIMMA: Adding Value to Astrophysical Surveys with AI as a Service

By: Gautham Narayan

Type: Talk

Abstract: With the start of the Legacy Survey of Space and Time, and soon the launch of the Nancy Grace Roman Space Telescope and the Lazuli Space Telescope, astrophysicists will have to contend with a fire-hose of data. The rate of discovery in the transient and variable sky will not only increase by an order of magnitude over existing surveys, Moreover, many of the sources these surveys will discover are likely to have counterparts in the gravitational wave and astroparticle sky. Identifying these and other rare events within the flood is a challenge in itself, but the complexity of the multi-experiment, multi-modal data also renders many of our traditional analysis tools moot. SCIMMA is working to deploy AI models as a service, enabling the community to make informed decisions by automating several key inference tasks. Our Blast service is designed to provide information about the host galaxies of transients, DASH is designed to process spectra at scale, and SELDON is designed to generate forecasts from partial light curves, as well as other tasks. Together these form the basis of a foundation AI model for the time-domain sky, and together with the interfaces that SCIMMA is building to connect astronomers and observatories, they are infrastructure to enable multi-mission science. I will demo these services live and highlight how quickly they can be integrated into your analysis pipelines.

Co-Authors: On behalf of the Scalable Infrastructure for Multi-Messenger Astrophysics (SCIMMA) Team

Training Is Not Enough: Making ML Models Work in the LSST Alert Stream Era

By: Jakob Nordin

Type: Talk

Abstract: Machine-learning models are now ubiquitous in astronomy and beyond. As new architectural variants are continuously introduced and the available training power continues to scale, the central challenges for ML applications in science are shifting toward issues of provenance, validation, and application. Here I outline recent developments in these areas based on the AMPEL platform: the noiztf approach to constructing unbiased training samples; built-in guarantees for traceability and reproducibility; and how pipelines are designed and evaluated for specific science goals, ranging from multi-messenger science and exotic transients to cosmology. Finally, I will explore how AMPEL models that successfully completed the ELAsTiCC challenge perform when confronted with the LSST alert stream.

Exploring Transient Host Galaxies in the Big Data Era with FrankenBlast

By: Anya Elizabeth Nugent

Type: Talk

Abstract: The next few years will mark golden era of astrophysical transient discovery, with high-cadence observing missions like the Rubin Observatory and NASA's Roman mission promising to detect up to tens of thousands of events nightly. While these missions will greatly expand known transient populations and reveal new phenomena, most transients will solely have scarce photometric coverage, making their classification and progenitor inference extremely difficult. Importantly, transient host galaxies provide a powerful and complementary avenue for distinguishing transients and understanding their origins, independent of transient emission. Here, I discuss FrankenBlast - a novel tool that associates transients to their host galaxies and runs a machine-learning based SED-fitting algorithm to robustly constrain host stellar population properties, all within minutes per object. I present FrankenBlast-determined stellar population property results for ~10,000 host galaxies of YSE and ZTF-discovered transients. I highlight our key findings on how host properties can be used to identify events misclassified from traditional photometric classifiers and how host stellar population properties provide vital clues into transient progenitors. I conclude by discussing how FrankenBlast can be adapted for Rubin and Roman and by outlining the fundamental questions about transient origins that host galaxy studies will enable us to address in the upcoming Big Data era.

SCORPIO - a new facility instrument for the Gemini South telescope geared for RAPID response

By: Swayamtrupta Panda

Type: Poster

Abstract: As time-domain astronomy transitions from discovery-limited to response-limited, the ability to rapidly characterize transients across wavelengths has become a defining challenge of the field. Meeting this challenge requires not only powerful surveys and alert streams, but also agile, multi-band instrumentation capable of immediate, information-rich follow-up. The Spectrograph and Camera for Observations of Rapid Phenomena in the Infrared and Optical (SCORPIO) is the next facility instrument for the Gemini South telescope, designed explicitly to address these needs in the era of large-scale, low-latency time-domain surveys. SCORPIO will provide simultaneous imaging and spectroscopy in eight wavelength bands spanning 0.4 to 2.2 microns. The instrument can be configured in two fundamental operation modes: imaging and spectroscopy. In imaging mode, SCORPIO’s eight detectors observe the astronomical scene simultaneously through broadband filters. Two main observing modes are available: an imaging mode that samples the full 180″ × 180″ field of view delivered by the instrument optics at a spatial scale of 0.18″ per pixel; and a spectroscopic mode that performs medium-resolution long-slit spectroscopy using volume phase holographic (VPH) grisms and the choice between six slit widths ranging from 0.36″ to 4.32″. SCORPIO is currently in its Assembly, Integration, and Verification phase, with commissioning at Cerro Pachón (Chile) expected in 2027 and availability to the Gemini user community in 2028. The instrument will be highly versatile, addressing forefront science cases from Solar System studies to extragalactic astrophysics, and will play a leading role in Gemini follow-up of time-domain events triggered by large-scale surveys. In doing so, SCORPIO will help fulfill key strategic goals of NOIRLAB / Gemini Observatory and will play a leading role as a workhorse instrument for Rubin follow-up of diverse, dynamic, and yet-unknown cosmic objects.

Co-Authors: Rodolfo Angeloni - Project Scientist (SCORPIO), International Gemini Observatory, NSF/NOIRLab, La Serena, Chile

Unraveling the Nature of the Nuclear Transient AT2020adpi

By: Paarmita Pandey

Type: Poster

Abstract: Transient events associated with supermassive black holes (SMBHs) offer rare glimpses into accretion physics and the dynamic environments of galactic nuclei. In this talk, I will present a comprehensive multiwavelength study of AT2020adpi (ZTF20acvfraq), a luminous optical/UV transient at z = 0.26 that exhibits characteristics of an ambiguous nuclear transient (ANT), which are a new and emerging class of events with properties that are different from both tidal disruption events (TDEs) and standard active galactic nuclei (AGN) variability. I will discuss the temporal and spectral evolution of AT2020adpi, its host galaxy properties, and how it compares to other nuclear transients in the broader luminosity–timescale phase space. The transient’s peculiar light curve, strong MIR flare, and evolving emission-line features point toward an accretion episode possibly triggered by a stellar disruption within an existing accretion disk, or by instabilities in an active nucleus. AT2020adpi highlights the challenges of classifying nuclear transients and the need for coordinated, multiwavelength monitoring to disentangle AGN variability from tidal disruption signatures.

Co-Authors: Jason Hinkle, Christopher Kochanek

Hourglass 2: Updated Simulations for the Roman Space Telescope's High-Latitude Time Domain Survey

By: Jillian Marie Paulin

Type: Talk

Abstract: The Roman Space Telescope’s High-Latitude Time Domain Survey (HLTDS) is designed to optimize the number of Type Ia supernovae and other time-domain sources detected over the course of 5 years after its upcoming launch in September. We present Hourglass 2, the most up-to-date set of HLTDS simulations based on current observing plans as outlined by the Roman Time Allocation Committee (ROTAC). These simulations are a follow-up to the original Hourglass simulations published in 2025, which only considered the two Core years of HLTDS. Hourglass 2 contains all components of HLTDS (Pilot, Core, Extended). Furthermore, the observation strategy, particularly for the initial Pilot observations, has changed since the original simulations, which is reflected in Hourglass 2. These simulations contain Type Ia and Core Collapse supernovae, as well as rare transients (including tidal disruption events, kilonovae, pair-instability supernovae, and superluminous supernovae). They also include simulated spectra from Roman’s Prism. We also present light curve fitting using the SALT3 model, and SCONE supernova classification.

Co-Authors: Masao Sako

Rapid-response follow-of transients with the Canary Islands telescopes: present, future, and some science highlights

By: Ismael Perez-Fournon

Type: Poster

Abstract: Two of the best observatories in the northern hemisphere are located in the Canary Islands: Teide Observatory in Tenerife and

Roque de los Muchachos Observatory in La Palma. In this talk I will review the main telescopes that are available for rapid follow up of astronomical transients, the different schemes of observing time allocation, in particular the IAC-Barcelona-Madrid in-kind contributions to LSST (a fraction of the 10.4-m Gran Telescopio de Canarias observing time will be offered to the LSST community), and summarize some highlights of observations of different types of transients, including nearby supernovae, GRBs and Einstein Probe fast X-ray transients. I will also discuss our experience with ZTF/LSST brokers for different projects and the planned extension to LSST.

Co-Authors: Main projects: LensWatch, SGLF, IAC-China collaboration on supernovae and Einstein Probe FXTs, and IAC LSST science collaboration.

ELEPHANT Tracks: Identifying Hostless Transients in Real Time

By: Priscila Jael Pessi

Type: Talk

Abstract: The next decade of time-domain astronomy will be defined by the ability to rapidly identify rare and physically informative transients within overwhelming alert streams from multi-wavelength surveys. Among the most enigmatic of these are hostless transients, events with no obvious association to a host galaxy, which can probe extreme stellar evolution, explosion physics, and galaxy environments, but are challenging to systematically identify at scale.

In this talk, I will present ELEPHANT (ExtragaLactic alErt Pipeline for Hostless AstroNomical Transients), a low-latency, alert-level framework designed to filter and study hostless extragalactic transients in real time. ELEPHANT operates on alerts from the Zwicky Transient Facility (ZTF) distributed through the Fink alert broker. It combines machine-learning classifications provided by the broker with targeted image-analysis methods applied to alert cutouts, allowing efficient selection of hostless candidates directly from the alert stream.

I will describe the performance of ELEPHANT since its deployment, including the contamination rate of the selected sample and basic statistics of the flagged events that have been classified by the community. Finally, I will discuss ongoing efforts to adapt the pipeline to future surveys, with a focus on preparing ELEPHANT for the Rubin Observatory LSST alert stream.

Co-Authors: R. Durgesh, E. E. O. Ishida, L. Nakazono, E. E. Hayes, R. A. P. Oliveira, A. Moitinho, A. Krone-Martins, B. Moews, R. S. de Souza , R. Beck, M. A. Kuhn, K. Nowak, and S. Vaughan

EuroFlash: a commensal search for fast transients < 300 MHz with LOFAR 2.0

By: Ziggy Pleunis

Type: Poster

Abstract: Low-frequency (~< 300 MHz) observations of transients allow precise measurements of propagation effects (dispersion, scatter-broadening and Faraday rotation) that probe the local environments of the sources that produce these transients. The detection of bursts from repeating fast radio burst (FRB) sources 20180916B and 20181030A with the LOFAR telescope indeed has put the tightest constraints on free-free absorption local to these two sources. It has also revealed a chromaticity in the periodic activity of FRB 20180916B. LOFAR has also recently detected two long-period transients: ILT J1101+5521, originating in an M dwarf-white dwarf binary and ILT J163430+445010, with pulses that are almost 100% circularly polarized. The LOFAR interferometer is spread through Europe and sensitive from 10 to 240 MHz. It has been undergoing a major upgrade recently: LOFAR 2.0 will have a much improved correlator network bandwidth that allows for commensal observations. We have been developing the EuroFlash survey, that will piggyback on all LOFAR observations and independently beamform and image to search for transients on a dedicated compute cluster. A combination of targeted and untargeted searches will run 24/7 and year-round and will greatly increase the yield of fast transients at low frequencies. It will reveal any frequency-evolution in the population of FRBs through comparison with the CHIME/FRB survey and surveys at L-band, and it may well uncover a completely new class of fast transients that only manifests at low frequencies. Inclusion of the international stations with baselines up to 1000 km will allow for ~arcsecond localization of bursts. In this talk, I will review low-frequency observations of transients, give an update of the commissioning of EuroFlash on LOFAR 2.0 and discuss synergies with other transient surveys.

Co-Authors: The EuroFlash team.

The rapidly evolving and luminous supernova SN 2021lwz

By: Frédérick Poidevin

Type: Poster

Abstract: SN 2021lwz is a rapidly evolving and luminous supernova. It is studied using optical spectroscopy, photometry and imaging linear polarimetry obtained during several follow-up campaigns. These data are used to analyse and model the evolution of this explosion. Before peak SN 2021lwz spectral features do not match the H-poor or He-Rich H-poor superluminous supernova type, or Ic-BL type. On the other hand, after maximum light, the spectral evolution looks similar to the Ic/Ic-BL type. It also has an intermediate luminosity between SNe Ic/Ic-BL and typical SLSNe, and a rise-time similar to iPTF16asu. Therefore, it has aspects of many related classes. We will present these results and discuss the possibility that SN 2021lwz is a nice example of how differences in ejecta mass and engine parameters could produce a wide range of engine-driven SESNe.

Co-Authors: S. West, C. M. Omand, R. Könyves-Tóth, S. Schulze, L. Yan, T. Kangas, I. Pérez-Fournon, S. Geier, J. Sollerman, P. J. Pessi, C. M. Gutiérrez, T.-W. Chen, K-Ryan Hinds, R. Marques-Chaves, R. Shirley, C. Jimenez Angel, R. Lunnan, D. A. Perley, N. Sarin, Y. Yao, R. Dekany, J. Purdum, A. Wold, R. R. Laher, M. J. Graham, M. M. Kasliwal, and T. Jegou Du Laz.

YSN-Class (Young SuperNova Classifier)

By: Oleksandra Pyshna

Type: Poster

Abstract: We present YSN-Class (Young SuperNova Classifier), a multi-modal supernova classifier with emphasis on early-time observations (e.g., <1 week post-explosion). YSN-Class combines photometry and early spectra with varying representations and features in a fusion architecture, which is then processed by the gradient boosting algorithm to obtain the classification. The classifier was trained and tested on 1050 supernovae from the Census of the Local Universe (CLU) survey—a volume-limited, spectroscopically complete Zwicky Transient Facility (ZTF) sample—and on 5107 supernovae from the Bright Transient Survey (BTS). In the photometry-only, no-redshift, mode, YSN-Class achieves a macro-averaged F-1 score of 51% and 76% accuracy for BTS early-time classification (using only data obtained at least 10 days before peak) across 5 classes (Ia/II/Ib/c/IIn/SLSN). For the CLU sample (using data within 5 days of explosion), the model reaches a macro-averaged F-1 score of 61% and 64% accuracy for 3 broader classes (Ia/II/Ib/c). When including CLU early spectra for the classification (spectra <1 week since explosion) to the photometric data within 10 days since explosion, YSN-Class attains a macro-averaged F-1 score of 76% across the three classes (Ia/II/Ib/c). In addition, the framework can sub-classify Type II supernovae into “flash” and “no-flash” categories, which trace interaction between the ejecta and confined circumstellar material, achieving a 73% average F-1 score using photometry covering maximum brightness, and 80% average F-1 score by including early spectra as well. Our multi-modal approach demonstrates a promising way for early-time classification of supernovae with ZTF, and can be combined with the upcoming Vera C. Rubin Observatory (VRO) transient alert stream and existing real-time classification pipelines to provide robust classifications within the first few days after explosion, which plays a crucial role for studying supernova events.

Co-Authors: Wynn Jacobson-Galan, Ashish Mahabal, Mansi Kasliwal

General Coordinates Network (GCN): NASA’s Next Generation Time-Domain and Multimessenger Astronomy Alert System

By: Judith Racusin

Type: Talk

Abstract: The Gamma-ray Coordinates Network (GCN, https://gcn.nasa.gov) is a public collaboration platform run by NASA for the astronomy research community to share alerts and rapid communications about high-energy, multimessenger, and transient phenomena. Over the past 34 years, GCN has helped enable many seminal advances by disseminating observations, quantitative near-term predictions, requests for follow-up observations, and observing plans. GCN distributes alerts between space- and ground-based observatories, physics experiments, and thousands of astronomers around the world. With new transient discovery and follow-up instruments, this coordination effort is more important and complex than ever. Maximizing the science return from this era requires interoperability between alert brokers, new alert standards, and increased machine readability. The General Coordinates Network, the modern evolution of GCN is built on modern, open-source, reliable, and secure alert distribution technologies, and deployed in the cloud. We present a status of the new GCN, features and new alert types enabling discovery in time-domain, multimessenger, and high-energy astrophysics, and our plans for the future.lorem

Catching Kilonovae and Other Fast Transients in the Era of Rubin and Roman

By: Jillian Rastinejad

Type: Talk

Abstract: The landmark discovery of the first binary neutron star (NS) merger through gravitational waves (GWs) and its accompanying kilonova opened a new era of multi-messenger astronomy. Despite the treasure trove of data provided by this discovery, numerous unknowns remain about these elusive events, including the median and span of their heavy element yields and their light curves' viewing angle dependency. In this talk, I will review how Rubin and Roman offer unprecedented opportunities to probe these unknowns through detections kilonovae following GW events and "untriggered" in their surveys. I will discuss lessons learned from previous searches with wide-field optical surveys and follow-up to GRBs. Finally, I will highlight emerging tools to facilitate fast transient candidate vetting, combining data streams to populate fast-evolving light curves, and rapid follow-up with multi-wavelength and space-based facilities.

Fast radio transients

By: Vikram Ravi

Type: Talk

Abstract: Time-domain studies on approximately sub-second timescales have been part of radio astronomy for around six decades. Observations of radio pulsars reveal endpoints of massive stellar evolution, the physics of ultra-dense matter and strong magnetic fields beyond the Schwinger limit, and the properties of strong-field gravity. Radio pulsars are indeed themselves now used as gravitational-wave detectors of binary supermassive black holes via precision, long-term timing of their rotations. Over the past two decades, fast radio bursts (FRBs) at extragalactic distances have also been discovered in high-time-resolution radio data. Based on one Galactic detection, and several arguments based on their emission properties and host galaxies / environments, FRBs are likely commonly associated with magnetars like those observed in the Milky Way. Some FRBs, however, appear to have even more exotic origins, in globular clusters and in dense star-forming regions of dwarf galaxies. I will review open problems in these fields, and what is expected in the coming decade. I will present prospects for the Deep Synoptic Array, which when complete in 2029 will be the most powerful radio telescope yet built.

Automating real-time workflows for transient discovery and rapid response follow-up

By: Nabeel Rehemtulla

Type: Talk

Abstract: Significant filtering of data streams from current and future time domain surveys is required due to the breadth and volume of astrophysical information they contain. However, numerous aspects of transient discovery and follow-up workflows using these streams have stagnated and pose considerable, oft under-appreciated, barriers to advancement. For example, transient discovery and follow-up still relies predominantly on manual action by experts, requiring substantial time investments and, critically, slowing responses to time-sensitive events. In this talk, I introduce a suite of “bots” I have deployed into the real-time production environments of ZTF and the La Silla Schmidt Southern Survey (LS4). Real/bogus classification is essential to image-differencing time domain surveys, but current state-of-the-art classifiers require substantial labeled data sets for training. I introduce RBbot, the real/bogus classifier for LS4 trained using unsupervised domain adaptation techniques, enabling RBbot to learn from the large quantities of unlabeled data which are traditionally discarded. After filtering bogus alerts, transients of interest must be selected and followed-up spectroscopically, processes traditionally done entirely manually. I also introduce BTSbot, a mutli-modal CNN now capable of fully autonomously discovering new, nearby (d<60 Mpc) supernovae and requesting target-of-opportunity follow-up to SEDM, MMT, Gemini, LCO facilities, SOAR, WINTER, and Swift. This one-of-a-kind workflow has reduced the typical latency between a transient’s first detections and its first follow-up by more than an order of magnitude. In the case of SN 2025qtt, Swift/UVOT began observing just 6 minutes after its automated discovery by BTSbot. Advanced infrastructure like these bots and their tight integration into ZTF, LS4 and follow-up facilities are mandatory to maximize the output of our time-domain surveys.

Co-Authors: Adam A. Miller, Wynn V. Jacobson-Galán

ULTRASAT from the NASA Perspective

By: James E Rhoads

Type: Poster

Abstract: The Ultraviolet Transient Astronomy Satellite, ULTRASAT, is an ambitious near-ultraviolet wide field imaging mission combining high sensitivity and an unprecedented field of view. These give ULTRASAT an unprecedented survey efficiency for hot astrophysical sources, and ULTRASAT will thereby open new vistas in time domain astronomy. Development of ULTRASAT is led by the Weizmann Institute of Science and Israel Space Agency, with partner organizations in Germany and the United States. This presentation will give a brief overview of ULTRASAT, including a summary of the driving science objectives, and the resulting hardware design and capabilities. It will also the structure of the partnership, including NASA's role in the mission, and the mission's role in NASA's portfolio. Finally, it will describe current science planning activities, including the field selection and cadence planning for ULTRASAT’s high cadence, wide area, and all-sky surveys.

Co-Authors: The ULTRASAT collaboration

Time domain astronomy with REX

By: James E Rhoads

Type: Poster

Abstract: We discuss the potential time-domain applications of the REX mission concept. REX, the Reionization Explorer, is an explorer concept that will combine a ~0.7m telescope, a square degree field, and a unique complement of near-infrared narrowband filters. REX will cover ~100 square degrees (spread over many fields), in tens of epochs spread over a two year prime mission, as it builds up the sensitivity required to identify z>7 galaxies in deep image stacks. A typical object in the survey area will have light curves with a per-epoch sensitivity >~ 23.5 AB magnitudes and sampling intervals of <~ 10 days in each of ~eight narrowband wavelengths between 0.95 and 1.45 microns. We will present initial estimates for the expected samples of variable objects that might be found in REX data. We will also discuss potential ToO applications of the REX hardware. Finally, we consider synergies with other missions that would be active during the likely REX prime mission. We welcome discussion and collaboration with other community members as we continue to develop REX science plans.

Co-Authors: Sangeeta Malhotra and the REX team

Stellar Alerts: lessons from ZTF

By: Jan van Roestel

Type: Talk

Abstract: Stars can exhibit a wide range of periodic and irregular optical/IR variability, and many of these variable stars will trigger alerts. However, because the location of a star is typically well-known and its variability is typically persistent, methods other than alerts are used to study variable stars. But as always, there are exceptions. The Zwicky Transient Facility (ZTF) has been observing the entire northern sky for 8 years now, and I will give a short overview of the kinds of variable stars detected through alerts and how the ZTF alerts have been used. This includes positive alerts triggered by Microlensing events, Cataclysmic variables, YSOs, and other outbursts and flares, as well as negative alerts triggered by long-duration eclipses, disk transits, or planetesimal disruption events. I will briefly discuss the potential science we can do by quickly following up on alerts associated with stars, and what the prospects are with current and future time-domain surveys.

Probing early signatures of dust formation in core-collapse supernovae with ground based optical/NIR observations

By: Sam Rose

Type: Poster

Abstract: Despite the ubiquitous presence of dust across cosmic time and in a variety of environments the formation of dust remains an open question. It has been observed that large amounts of dust can form in the ejecta of supernovae, but supernovae may also destroy large quantities of dust when they explode. Existing measurements of supernovae dust mass contributions are also complicated by the difficulty in disentangling newly formed dust in the ejecta from pre-existing dust in the circumstellar medium (CSM) of the progenitor star. Therefore, the overall contribution of supernovae dust to the cosmic dust budget is debated. In this talk I will present results from the Red Astronomical Transient (RAT) survey which combines data from the Zwicky Transient Facility (ZTF) alert stream with ground-based optical spectroscopic and NIR photometric and spectroscopic follow-up observations to probe the early signatures of new dust formation as well as the presence of pre-existing CSM dust. With the large sample of supernovae discovered and characterized by the ZTF collaboration it is possible to obtain a larger and more complete sample of dust and CSM diagnostics in core-collapse supernovae including narrow line ejecta-CSM interaction features, NIR excesses, and the spectroscopic detection of molecular precursors to dust formation. I will also discuss the upcoming roles of Rubin, which will detect faint precursor and post-explosion emission which probes CSM formation and evolution, as well as Roman which will provide sensitive NIR observations.

The core collapse supernova landscape: infrastructure and science opportunities

By: David Sand

Type: Talk

Abstract: What are the explosion mechanisms and progenitor star systems of supernovae? In the early hours to days after explosion, core collapse supernovae provide clues to how they explode, and what their progenitor star systems were. Here we will discuss the science that will open up at the start of the LSST era, as well as the time domain infrastructure that is necessary to make the most of this discovery engine. Opportunities that were possible only once or twice per year will increase by an order of magnitude, and will benefit from direct ties between discovery and follow-up facilities. We will discuss specific examples of ephemeral, but critical, signatures in early supernova data (e.g. early light curve ‘bumps’, flash spectroscopy, shock breakout cooling) and what they are teaching us about these cosmic explosions, with an eye towards what even deeper/fainter observations might provide. Finally, we will look ahead to the next few years and what Roman, ULTRASAT, UVEX and Lazuli will add to the picture.

AGN Variability-Informed Photometric Redshifts for Large Time-Domain Surveys

By: Sarath Satheesh Sheeba

Type: Poster

Abstract: Accurate redshift estimates for Active Galactic Nuclei (AGNs) are essential for placing them in a cosmological framework and for deriving intrinsic properties such as luminosities, black hole masses, and accretion rates. However, photometric redshifts for AGNs remain challenging due to their mostly featureless spectral energy distributions and intrinsic variability, particularly in the absence of large-scale spectroscopic follow-up for upcoming surveys such as the Rubin LSST.

In recently submitted work (Satheesh-Sheeba et al., submitted after revision to A&A; arXiv:2509.13308), we introduced VAR-PZ, a framework that incorporates AGN variability as a physically motivated prior for photo-z estimation. VAR-PZ models AGN variability as a damped random walk process and leverages cosmological time dilation as well as the luminosity and wavelength dependence of variability to construct redshift priors. By combining these variability-based priors with traditional SED fitting, we demonstrated a reduction in photometric-redshift outliers by more than 10 percentage points using SDSS data. LSST-like simulations further indicate that this approach can reduce the outlier fraction from approximately 32 percent to below 8 percent for Deep Drilling Fields by the end of the survey.

Building on this framework, we are currently developing a complementary, data-driven approach using machine learning. In this work in progress, we employ Fully Connected Neural Networks trained on optical and near-infrared colors together with variability features extracted by the ALeRCE light-curve classifier. Using a sample of 72,000 ZTF AGNs with known spectroscopic redshifts, we achieve an outlier fraction of approximately 9 percent on a held-out test set of 20,000 sources. These advanced results demonstrate the strong potential of AGN variability in machine-learning models for scalable AGN redshift estimation.

Together, these two approaches highlight AGN variability as a powerful and complementary tool for improving photometric redshifts and distance estimates in the time-domain era.

Co-Authors: Prof. Roberto J. Assef Dr. Paula Sánchez Sáez Prof. Timo Anguita

Where Are All of the Off-Axis GRBs? A Late-Time Radio Campaign of Ic Broad Line Supernovae

By: Genevieve Schroeder

Type: Talk

Abstract: Long-duration gamma-ray bursts (GRBs) are the most luminous explosions in the Universe, visible out to cosmological redshifts. Almost every spectroscopically confirmed supernova (SN) associated with a long GRB has been hydrogen and helium poor with broad lines (Ic-bl), indicating a stripped massive star progenitor with fast winds that produced a highly-collimated jet viewed on-axis. A critical prediction of this progenitor model is that there should be a large number of off-axis jets accompanying SNe Ic-bl - however, this has yet to be observationally confirmed. Years after the explosion, the off-axis jet will have decelerated and become spherical, resulting in a late-rising radio light curve. Here, I will present my recent paper on late time radio observations of a sample of 14 SNe Ic-bl that have been shown to have exploded in the same host environments as long GRBs, making them some of the strongest candidates for off-axis GRBs to-date. I will use these observations to place constraints on the existence of off-axis jets within this sample. Additionally, I will present preliminary results from my 26A VLA campaign of 80 SNe Ic-bl, and lay the path forward to detecting off-axis GRBs in the era of large transient surveys at X-ray (Einstein Probe/SVOM), optical (LSST/Roman), and radio (DSA-2000) wavelengths.

General Coordinates Network (GCN) Circulars: Large Language Model Driven Analysis

By: Vidushi Sharma

Type: Talk

Abstract: The General Coordinates Network (GCN) is NASA's time-domain and multi-messenger alert system. GCN distributes two data products - automated ``Notices,'' and human-generated ``Circulars,'' that report the observations of high-energy and multi-messenger astronomical transients. The flexible and non-structured format of GCN Circulars, comprising of more than 40500 Circulars accumulated over three decades, makes it challenging to manually extract observational information, such as redshift or observed wavebands. In this work, we employ large language models (LLMs) to facilitate the automated parsing of transient reports. We develop a neural topic modeling pipeline with open-source tools for the automatic clustering and summarization of astrophysical topics in the Circulars database. Using neural topic modeling and contrastive fine-tuning, we classify Circulars based on their observation wavebands and messengers. Additionally, we separate gravitational wave (GW) event clusters and their electromagnetic (EM) counterparts from the Circulars database. Finally, using the open-source Mistral model, we implement a system to automatically extract gamma-ray burst (GRB) redshift information from the Circulars archive, without the need for any training. Evaluation against the manually curated Neil Gehrels Swift Observatory GRB table shows that our simple system, with the help of prompt-tuning, output parsing, and retrieval augmented generation (RAG), can achieve an accuracy of 97.2 % for redshift-containing Circulars. Our neural search enhanced RAG pipeline accurately retrieved 96.8 % of redshift circulars from the manually curated database. Our study demonstrates the potential of LLMs, to automate and enhance astronomical text mining, and provides a foundation work for future advances in transient alert analysis.

Co-Authors: Vidushi Sharma, Ronit Agarwala, Judith L. Racusin, Leo P. Singer, Tyler Barna, Eric Burns, Michael W. Coughlin, Dakota Dutko, Courey Elliott, Rahul Gupta, Ashish Mahabal, Nikhil Mukund

Autonomous Surveys and Community ToOs on the UltraViolet EXplorer (UVEX): Towards a Science Operations Center without Humans in the Loop

By: Leo Pound Singer

Type: Talk

Abstract: The UltraViolet EXplorer (UVEX) is a wide-field ultraviolet space telescope selected as a NASA Medium-Class Explorer (MIDEX) mission for launch in 2030. UVEX will undertake deep, cadenced surveys of the entire sky to probe low mass galaxies and explore the ultraviolet (UV) time-domain sky, and it will carry the first rapidly deployable UV spectroscopic capability for a broad range of science applications. In this talk, I will give an overview of the strategies and algorithms under development to optimally plan and execute UVEX’s science mission in real time and without any need for humans in the loop for rapid response to time-sensitive targets of opportunity.

Hiding in Plain Sight: Wandering AGNs in the LSST Era

By: Kendall Sippy

Type: Poster

Abstract: In the complex kinematics of galaxy mergers, the end result will not always be a successful merger of the galaxies’ central massive black holes (MBHs). As a galaxy experiences major and minor mergers over its lifetime, it should accumulate a large number of non-nuclear, wandering MBHs which may emerge as active galactic nuclei (AGNs), providing a way to constrain the rate of these unsuccessful BH mergers. Wandering AGNs, if they are as numerous as some recent work predicts, have the potential to dramatically influence galaxy evolution via feedback and quenching star formation, yet we have virtually no constraints on this population. The time-domain capabilities of the Legacy Survey of Space and Time (LSST) provide a uniquely powerful tool to search for wandering AGNs. AGNs can be readily detected via their variability, and the difference imaging provided by LSST gives us a position estimate for these sources, independent of the contaminating host galaxy light, putting us in a position to identify wandering AGNs directly from the publicly-available data and observationally characterize them for the first time. We can then perform multi-resolution forward modeling of the LSST and Euclid images with Scarlet2, leveraging Euclid’s higher resolution and LSST’s deeper imaging, with machine learning priors to enable the best possible characterization of the galaxy structure and confirm strong candidates. I will present some preliminary results with the currently available LSST data.

Co-Authors: Charlotte Ward

Brokering alerts in real time in the Rubin era with ANTARES

By: Monika Soraisam

Type: Talk

Abstract: As Rubin/LSST begins its science operations, it is poised to fundamentally transform the way we study the dynamic Universe by going deeper, faster, and wider in panchromatic bands. It will generate up to 10 million alerts per night, which is unprecedented for time-domain astronomy. Hidden in this stream of alerts will be rare, time-critical events requiring immediate follow-up, such as FBOTs, shock breakout flashes from SNe, kilonovae, etc. Brokers are software systems that will enable users to sift through this barrage of data and select high-priority events. ANTARES is one such broker developed by NOIRLab. In this talk, I will review the features of ANTARES and demonstrate how users can interact with the system. Additionally, I will highlight how ANTARES is integrated with other NOIRLab time-domain services in an end-to-end system, called GOATS, enabling users to go from alert to science via a single, seamless interface.

Old dog, new tricks - Introducing the Rapid Response Mode at the Nordic Optical Telescope

By: Samuel Grund Sorensen

Type: Poster

Abstract: Time-domain astronomy is undergoing rapid growth, driven by new ground- and space-based facilities discovering extreme and short-lived transients at an unprecedented rate. These developments place increasing demands on follow-up facilities to respond quickly to capture rapidly evolving phenomena. To address this, a Rapid Response Mode (RRM) has been implemented and is now operational at the 2.56-meter Nordic Optical Telescope, located at the Observatorio del Roque de los Muchachos on La Palma, Spain. The system allows authorized users to submit observation requests through a machine-to-machine web-API, bypassing manual communication, preparation, and scheduling. Requests are automatically validated, compiled, and executed immediately, interrupting ongoing observations if necessary. Typical response times are approximately 2 minutes for photometric observations and 3 minutes for spectroscopic observations. We present the system architecture and design choices underlying this implementation, demonstrating how rapid response capabilities can be retrofitted to existing facilities while providing a practical framework for developing similar systems.

The Diversity of Massive Stellar Deaths and Relativistic Jets: Unraveling a Cosmic Tandem

By: Gokul Prem Srinivasaragavan

Type: Poster

Abstract: Some of the most powerful probes of black holes and neutron stars in the Universe are stellar engine-driven transients, which are rare explosions triggered by a massive star undergoing a core-collapse supernova (SN), collapsing into a compact object and launching a relativistic jet. In addition to probing extreme physics, the study of these explosions sheds light on broader astrophysical questions such as the complexity of massive stellar evolution and the origin of systems detected in gravitational waves, making their characterization incredibly important. Historically, most telescopes focused on discovering long gamma-ray bursts (LGRBs) to study these transients. Thanks to new observational capabilities, we are beginning to discover engine-driven transients that do not produce observable gamma-rays, illustrating the diversity of massive stellar deaths and relativistic jets. During this talk, I will highlight this diversity, presenting multiwavelength studies on stellar engine-driven transients that do not produce observable gamma-rays, specifically orphan afterglows and X-ray Flashes. I will conclude by discussing how the next generation of time-domain surveys such as the Einstein Probe and Rubin Telescope will revolutionize the study of stellar engine-driven transients, transforming our understanding of how compact objects launch the fastest outflows in the Universe.

AEON+: Collaboration of Multiwavelength Observatories for Time Domain Science

By: Rachel Street

Type: Talk

Abstract: AEON+ is a collaboration between astronomical observatories designed to foster access to responsive,

flexible modes of observations that are essential for the characterization of time domain

phenomena. Modern and upcoming astronomical surveys are producing discovery alerts on an unprecedented

scale. This requires a paradigm shift in the operations of most astronomical telescope

facilities towards two key capabilities: requests for observations submitted programmatically

and flexible modes of executing the requested observations. Building on the success of the

original AEON Collaboration of optical observatories, the AEON+ program is expanding these

capabilities to a broader range of facilities, and wavelengths including Keck Observatories,

the Canada-France-Hawai'i Telescope, the Skynet Network and the National Radio Astronomical

Observatory. This talk will give an overview of the program, summarize new and planned

capabilities at each observatory, and demonstrate the new AEONlib software which will provide

more flexible programmable access for observers. It will also present related

upgrades to the TOM Toolkit to support science with the Vera C. Rubin Observatory.

Co-Authors: J. Chatelain, W. Lindstrom, A. Riba

An Extensible TOM for the Time-Domain Era: SAGUARO as a Scalable Discovery Platform

By: Bhagya Madimugar Subrayan

Type: Talk

Abstract: Searches After Gravitational waves Using ARizona Observatories (SAGUARO) is a mature, end-to-end framework for rapid transient discovery, vetting, and coordinated follow-up, originally developed to identify any optical counterparts to gravitational wave (GW) events. Since its deployment during the LIGO/Virgo’s O3 observing run, SAGUARO has demonstrated the ability to tile large localization regions with wide-field optical imagers, perform real-time image subtraction, and rapidly assess candidate viability through coordinated follow-up, establishing a proven foundation for low-latency time-domain science. Recent infrastructure upgrades implemented for the LVK O4 era have significantly expanded SAGUARO’s capabilities. These include deeper reference imaging, machine learning (ML) classification, and a fully integrated, web based Target and Observation Manager (TOM). The SAGUARO TOM aggregates photometric, contextual, and host galaxy information from public data streams, enabling rapid and informed prioritization of viable transients from a single, centralised portal. Together, these improvements have broadened SAGUARO’s scope beyond a mission specific follow-up program to a flexible, survey-independent system. In this talk, we focus on SAGUARO’s core capabilities for GW follow up and show how the same TOM infrastructure can support additional science programs without significant re-engineering. We present the scope of integration of the new blue-sensitive 90Prime instrument on the 2.1m Bok telescope as an additional wide-field discovery resource, and demonstrate the use of SAGUARO infrastructure within the SHADOW program, which ingests and vets DECam candidates to identify extremely young supernovae by shadowing LSST observations. SAGUARO offers a scalable and adaptable model for time-domain infrastructure in the LSST and Roman era, enabling rapid response and discovery across diverse transient science cases.

Co-Authors: Griffin Hosseinzadeh, Noah Franz, Phillip Noel, David Sand, Wen-fai Fong, K. Azalee Bostroem, Charlie Kilpatrick, Nick Viera, Conor Ransome, Manisha Shrestha, Jillian C. Rastinejad, Jennifer Andrews, Kerry Paterson, M. J. Lundquist

Every Deep Field Lives: Discovering the Transients through AI without Image Subtraction

By: Nao Suzuki

Type: Poster

Abstract: The advent of JWST, together with the upcoming Rubin LSST and Roman Space Telescope, marks the beginning of a transformative era in time-domain astrophysics. For the first time, we can systematically probe a transient universe that remains largely hidden from ground-based observatories. Every JWST deep field contains transient sources, including supernovae and other short-lived phenomena, quietly embedded in the data yet often overlooked because time variability was not the original intent of most observations. As we enter the era of multiple space- and ground-based facilities, a major opportunity lies in integrating their discovery pipelines and follow-up capabilities. Advances in AI and machine learning now enable the identification and characterization of transients across platforms without traditional image subtraction, reshaping how we detect and interpret change in the cosmos. In this talk, I will present early examples of cross-mission AI discoveries and discuss how coordinated, intelligent surveys may revolutionize our understanding of the dynamic universe in the coming decade.

Machine learning methods and data products for kilonova observations

By: Andrew Toivonen

Type: Poster

Abstract: Multi-messenger searches for electromagnetic transients such as kilonovae or gamma-ray bursts associated with neutron star mergers continue to be of great interest in time-domain astronomy due to the wealth of scientific knowledge they provide. Future observing runs of the International Gravitational-Wave Network combined with all-sky surveys carried out by the Zwicky Transient Facility and Vera C. Rubin Observatory promise numerous exciting transients, and will necessitate data products and machine learning algorithms to sort through the abundance of triggers. We present the performance of a new gravitational-wave data product that provides estimates of the binary chirp mass, which can be used by astronomers to determine whether a gravitational-wave candidate event may contain a neutron star, a new pipeline for realistic simulations of optical observations of kilonovae, and introduce a machine learning architecture being designed for the identification, parameter estimation, and forecasting of kilonovae light curves from early-time observations. These efforts are intended to inform target-of-opportunity searches and maximize our chances of kilonova detection.

Co-Authors: Soumi De Felipe Fontinele-Nunes Michael Coughlin Ingo Tews

Update on SNAPS, the Solar System Notification Alert Processing System -- A ZTF and LSST broker

By: David Trilling

Type: Talk

Abstract: We are operating the Solar System Notification Alert Processing System (SNAPS), a downstream broker that presently ingests alerts from the Zwicky Transient Facility (ZTF), and will soon from the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST). SNAPS serves several purposes that together enable a wide range of science cases. (1) SNAPS is a clearinghouse for data from these surveys. A community user can query object(s) of interest and receive data records, derived properties (such as color and rotational period), and ancillary data (for example, orbital elements). (2) We have several tools that detect unusual asteroids in the SNAPS database. These can be population outliers (objects whose intrinsic properties are unusual compared to other small bodies) or individual outliers (objects whose properties change with time). An example of the former case is an object with a very large lightcurve amplitude, and an example of the latter case is an active asteroid. To date we have published both science and techniques papers, and a number of additional papers are in progress. In this presentation we will briefly summarize our infrastructure capabilities and highlight some of our most recent science highlights, including, hopefully, first results from LSST data. Finally, we will present tools that the community can use to interact with SNAPS.

Co-Authors: Gowanlock, Michael ; Munugala, Revanth Reddy ; Kramer, Daniel ; Chernyavskaya, Maria ; Clark, Erin ; Chappus, Savannah ; Johnson, Gunnar ; Metcalf, Connor ; Mule, Graceson

Application of time-series analysis methods to a multiple-sector TESS observations

By: Ashutosh Tripathi

Type: Poster

Abstract: Due to its unprecedented high cadence and uniform sampling compared to ground-based optical telescopes, the Transiting Exoplanet Survey Satellite (TESS) has proven an ideal instrument for studying bright AGNs. However, some issues should be resolved in the time-domain study of these high-quality data, such as gaps, noise, and red noise modeling. In this talk, I will discuss various time series analysis methods that could be used to analyze observations of bright AGN and examine whether issues such as gaps and noise in these data can be mitigated. I will present an analysis of the ~ 300-day TESS observation of a blazar using power spectrum density, structure-function, and weighted wavelet Z-transform approaches. Continuous auto-regressive moving averages, Bartlett periodogram, and wavelet decomposition methods are used to reduce the effects of gas and uncorrelated noise in the data. I will also discuss the potential of the presence of non-linearity in these observations by employing recurrence analysis, where the variability or periodicity can be quantified as the recurrent state.

Co-Authors: Paul J. Wiita (The College of New Jersey) Ryne Dingler (Texas A\&M University) Krista Lynne Smith (Texas A\&M University) R. A. Phillipson (Villanova University) Matthew J.\ Graham (California Institute of Technology) and Lang Cui (Xinjiang Astronomical Observatory)

Better Together: A multi-survey modeling framework to generate light curves and characterize host galaxies across Rubin, Euclid, Roman and LS4.

By: Charlotte Ward

Type: Talk

Abstract: We have entered an exciting decade for the detection and characterization of transients and their host galaxies across complementary surveys: Rubin will provide deep optical photometry, while Euclid and Roman will provide high resolution imaging and IR coverage. I will present our scalable, GPU-compatible, and neural network-stabilized approach to modeling transients and hosts in multi-resolution ground and space-based datasets. Our software enables efficient extraction of multi-survey light curves, host galaxy SEDs and morphology models, and transient-host separations. This joint survey analysis approach is designed to overcome challenges such as galaxy detection and characterization in deep, highly blended fields, as well as the risk of reference image mismatch when producing multi-survey light curves. I will demonstrate early applications to non-nuclear tidal disruption events, supernovae, and lensed transients.

ACROSS: Enabling Time Domain and Multi-Messenger Astrophysics with NASA Missions

By: Samuel Wyatt

Type: Talk

Abstract: The Astro2020 Decadal Survey recommended an investment in Time Domain and Multi-Messenger Astrophysics (TDAMM) as the top-priority sustaining activity in space for the coming decade. One aspect of NASA’s response to this recommendation is a pilot project, the Astrophysics Cross-Observatory Science Support (ACROSS) initiative, designed to provide support to both missions and observers as they pursue TDAMM science. Here, we present our observations of needs in the community and initial plans for ACROSS activities, including services to facilitate and improve cross-observatory follow-up planning and execution; a multi-messenger web portal with links to existing observatory resources, community tools, and information targeted for TDAMM General Observers.

Co-Authors: Jamie Kennea (PSU), Craig Pellegrino (NASA-GSFC), Chris Roberts (NASA-GSFC), Brian Humensky (NASA-GSFC), Tyler Pritchard (UMD, NASA-GSFC), Michelle Hui (NASA-MFSC), Dan Kocevski (NASA-MSFC), Nitzan Frock (NASA-GSFC), Kirill Vorobyev (NASA-GSFC)

BHTOM: Standardized, automated time-domain follow-up from a global volunteer telescope network

By: Lukasz Wyrzykowski

Type: Talk

Abstract: The rapid growth of time-domain alert streams demands a scalable and interoperable follow-up infrastructure that can operate across heterogeneous facilities and observer communities. We present BHTOM (Black Hole Target Observation Manager) — a global, voluntary network of nearly 200 telescopes ranging from professional observatories to advanced amateur systems, coordinated through the bhtom.space platform. Built originally for the Gaia Science Alerts program on the TOM Toolkit paradigm, BHTOM provides centralized target lists, automated access to time-domain data and almost real-time monitoring of follow-up observations. A key unique capability of BHTOM is automated processing and standardization of data from any telescope, including reduction, quality checks, metadata homogenization, uniform photometric extraction, and pipeline-derived light-curve products, enabling consistent multi-site time-series analysis.

BHTOM explicitly integrates citizen-science participation, providing tools and workflows that allow non-professional observers to contribute calibrated data products to professional campaigns with minimal friction. This effort is coordinated in particular by the European Astronomical Society of Small Telescopes (EASST.eu), which manages the community of small-telescope observers and supports sustained engagement in time-domain follow-up.

The system supports a broad range of science cases, including supernovae, microlensing events, stellar variability and flares, long-term quasar monitoring, and Solar System targets (comets, asteroids), as well as exoplanet transit monitoring.

In addition to new observations, BHTOM incorporates multi-wavelength time-domain archives (e.g., ZTF, Gaia, CRTS, LINEAR, and space-based UV/IR time series such as Swift/UVOT, GALEX, NEOWISE) and is being extended toward high-energy transient archives. We present the network, its capabilities and the tools built for follow-up operations and future readiness for Rubin and Roman-era alert-driven astronomy.

Co-Authors: P. Mikolajczyk (Uni.Wroclaw/NCBJ, Poland) K. Kotysz (Uni.Wroclaw/Uni.Warsaw, Poland) J. Majumdar (Uni.Warsaw, Poland) A.F.Gillan (NCBJ, Poland) P.Pessi (NCBJ,Poland)

Roman hIgh-redshift transient SciencE (RISE): Enabling a Non-Cosmology Time-Domain Program

By: Lin Yan

Type: Talk

AT2025ulz and S250818k: zooming in with the Hubble Space Telescope

By: Yuhan Yang

Type: Poster

Abstract: AT2025ulz is an optical/near-infrared transient discovered during follow-up of the candidate gravitational wave (GW) event S250818k. Its young age ($\lesssim$1 d), rapid decline and strong color evolution over the first 48 hr classify it as a potential kilonova candidate.

In this work, we present the results of our observing campaign, carried out with the Gran Telescopio Canarias (GTC) and the Hubble Space Telescope (HST). Although the early time evolution of AT2025ulz resembles some aspects of a kilonova, its rapid onset (~3 hr after the GW trigger) and luminosity (a factor of ~5 brighter than AT2017gfo in g-band) are difficult to reproduce. Only a small subset of our kilonova models matches its multi-color light curve, and the inferred ejecta mass is uncomfortably large given the low chirp mass ($\lesssim 0.87 M_{\odot}$) of the GW candidate. HST observations place the transient within a nearby (z=0.08489) spiral galaxy with on-going star-formation and measure a color (F336W-F160W ≈1.4 mag) that is too blue to match with a kilonova.

Our data support the classification of AT2025ulz as a supernova, initially undergoing a shock-cooling phase and later entering its photospheric phase, and spectroscopically identified via its broad absorption features.

SN 2024iss: A Multi-Wavelength Exposé of a Type IIb Supernova with an Early-Time UV Spectrum

By: Rujula Yete

Type: Poster

Abstract: Type IIb supernovae (SN IIb) result from the core-collapse of massive stars whose hydrogen envelopes have been stripped via binary interaction and/or enhanced mass-loss episodes. Ultraviolet spectroscopy of these supernovae within their first week is rare, but provides key insights into the characteristics and environment of the star before explosion, allowing for direct constraints on red supergiant atmospheres, the structure of supernova ejecta, and characterization of the chemical elements ejected by the progenitor star preceding explosion. We present multi-wavelength observations and a comprehensive analysis of the SN IIb 2024iss, one of the closest type IIb supernovae ever detected (14.04 Mpc). Observations of SN2024iss include a ZTF detection ~40 minutes after explosion (one of the earliest detections of a SN IIb to date) and an ultraviolet spectrum obtained from the Hubble Space Telescope at 7 days after explosion (the earliest HST UV spectrum obtained of a SN IIb to date). Photometric data in optical and ultraviolet bands show a double-peaked light curve typical of type IIb supernovae. We fit several shock cooling emission models to the first peak in the light curve (lasting up to 7 days), and obtained a progenitor radius of 101 - 320 R⊙ and a hydrogen envelope mass of 0.07 - 0.46 M⊙. The +7 day ultraviolet spectrum displays features from several iron-group elements present in the outer SN ejecta during the end of shock cooling emission. Furthermore, we compare SN2024iss to other similar type II and type IIb supernovae and find unique features in SN2024iss that stem from its CSM interaction. This study demonstrates the utility of a rich, multiwavelength dataset for constraining the progenitor systems and explosion dynamics of SNe IIb, with an emphasis on rapid detection and early-time ultraviolet observations.

Co-Authors: Wynn Jacobson-Galán, Mansi Kasliwal

Real-time detection of changing-state AGN in LSST data

By: Miranda Zak

Type: Poster

Abstract: Typical active galactic nuclei (AGN) exhibit some amount of stochastic but statistically predictable variability. A rare subset of AGN, called changing-state AGN (CSAGN), demonstrate more drastic changes, namely a change in the presence of optical broad emission lines. This change is thought to be a result of an instability in the accretion disk that affects the accretion rate of gas onto the supermassive black hole. Most CSAGN are selected through spectroscopic surveys, directly measuring the changes in the broad lines, but optical light curves by themselves have proven to be sufficient to detect changing-state events. The Legacy Survey of Space and Time (LSST) of the newly commissioned Rubin Observatory will produce millions of AGN light curves, from which we expect around 10,000 changing-state events will be detectable in the first year of the survey. We outline our work in developing a classifier for selecting changing-state events in LSST light curves using simulation-based inference (SBI) based on physical models of CSAGN variability (such as the disk-tearing model). First steps include building the SBI framework using light curves generated from general relativistic magnetohydrodynamic simulations and using this framework to estimate physical parameters describing the AGN. Ultimately, we plan to expand this work into a real-time detection system for CSAGN in LSST data, which will allow for rapid multi-wavelength and/or spectroscopic follow-up of these unusual events.

Co-Authors: Charlotte Ward