RTN-106
First alerts end-to-end demonstration#
Abstract
A roadmap for the end-to-end science demonstration of the LSST alert distribution and follow-up ecosystem, to be executed with the first alerts as part of the early science program.
Introduction#
This is a living document that will evolve as the plans for the alert end-to-end demo evolves.
Goals#
To use Rubin data to test and demonstrate the functionality and science-readiness of individual components of the automated alert follow-up ecosystem.
The target audience of this demo is both the science community and the general public.
Timeline#
The goal is to execute the alerts end-to-end demo by the end of November 2025, with a contingency of mid-December.
Prerequisites#
The conditions under which the first alerts end-to-end demo can occur are:
agency security review complete (done)
all observatories are on-sky post-shutdown
all components are online
compatible fields are scheduled for observations
SOAR is in AEON mode
draft EPO/Comms announcements are prepared
humans are available
The demo#
The following sequence of events might run in one night for one or more science targets, or over a few nights.
Image acquisition: observations are acquired at Rubin Observatory.
Alert production: Rubin DIA and AP pipelines run, distributing alerts to brokers.
ANTARES: a science filter identifies targets of interest for this end-to-end demo.
GOATS: queries and retrieves the targets of interest from ANTARES, down-selects, and triggers AEON.
AEON: receives follow-up trigger request from GOATS and executes observations with SOAR, Gemini, and/or Las Cumbres Observatories.
GOATS: receives raw data from Gemini and runs DRAGONS; receives processed data from SOAR and/or Las Cumbres; runs analysis automatically, e.g., spectral classification.
GOATS/Hermes: results (e.g., supernova classification) are shared with the science community (e.g., with a Transient Name Server report).
Press: announcement from EPO/Comms teams goes out via the usual channels.
Potential targets#
Potential science use cases and targets for the end-to-end demo, and relevant considerations for each.
The most effective science use case will be one that is common (high probability of finding one) and for which same-night follow-up can yield a classification.
Supernovae#
Any extragalactic field (i.e., high Galactic latitude) should yield several SN candidates.
ANTARES already has filters that can identify SN candidates, but the one submitted for automatic reduction should be well offset from its host.
GOATS/AEON could schedule follow-up spectra for several candidates, and it’s likely at least one could be classified (e.g., Type I, II, and sub-type).
In the SN field, classifications are regularly shared by time-domain surveys, and reported to the Transient Name Server (TNS) to obtain an IAU designation.
Variable star#
Similar case as supernovae, but for Galactic fields.
For example, M-dwarf flares are pretty common. They have a much shorter duration (minutes to hours) which would be good for demonstrating the utility of the automatic responses of the alerts ecosystem.
Though this target might work ok for press, since they are quite common it’s unclear how well this target would work to impress the scientific community.
Tidal Disruption Event (TDE)#
TDEs are nuclear transients, good candidates are in host galaxies with no AGN.
They are new and popular in astronomy, and they are good for EPO because they are black holes.
They are more rare in galaxy cores than supernovae, but if we aim for a TDE but end with a supernova, that would be fine.
Microlensing event#
As with variable stars, could be discovered in a Galactic field, but typically the follow-up is rapid-candence photometry not a single spectrum.
Solar system object#
ANTARES does not currently have filters for moving objects, and instead passes them on to the SNAPS broker. Thus, they are probably too challenging a case for automatic follow-up for this demo.
A known moving object that changes in brightness could potentially be automatically identified, but confirmation of cometary activity (e.g., outgassing) might require too long a follow-up timescale.
A new moving object (such as an NEO) would be unlikely on any given single night.
Identification and follow-up#
ANTARES filter#
In November 2025, most alerts are likely to come from the Deep Drilling Fields (DDFs), which have templates and are in observable regions of sky.
Thus, in identifying the potential follow-up target it is likely we’ll have same-night (or recent nights) photometric confirmation, and can design the filter accordingly.
For automated follow-up observations and data reduction, it is best if a transient is well-offset from its host. This could be included in the ANTARES filter or be part of the GOATS target prioritization.
Either we’ll use one or more existing filters, or create a custom temporary filter for this demo.
Follow-up observations and processing#
The alerts end-to-end demo should only use instruments and follow-up observation types that can be automatically processed. This means, the observations can be automatically reduced to calibrated data products in a way that doesn’t require human intervention).
Las Cumbres: imagers and spectrographs have auto-reduction.
SOAR: The Goodman spectrograph data has a quick-look reduction.
Gemini N or S: GMOS would be processed with DRAGONS via GOATS.
Components#
Brief descriptions and links to documentation for each of the components that are being incorporated into this alerts end-to-end demo.
Rubin EPO and Comms#
EPO: Education and Public Outreach
Comms: Communications team
The Rubin EPO and Communications teams will bring this demo to the public with graphics and press articles. These two teams will work with any relevant NSF NOIRLab and DOE communications teams.
Contacts: Kristen Metzger, Stephanie Deppe
Rubin DIA and AP#
DIA: Difference Image Analysis
AP: Alert Production
LSST Science Pipelines documentation
DIA refers to the process of image subtraction and difference-image source detection. AP refers to the production and distribution of alert packets for difference-image sources. This software is run by Rubin Data Management as part of Prompt Processing.
Contact: Eric Bellm
ANTARES#
ANTARES: Arizona–NOIRLab Temporal Analysis and Response to Events System
ANTARES is a full-service, real-time broker that adds contextual value to ingested alerts from multiwavelength astronomical catalogs. Users can write their own filters to identify specific classes of objects, create watch lists for direct notification, or develop catalogs for large-scale comparisons. Users can interact with the system via a web portal, API, or substreams of alerts from filters.
Contact: Tom Matheson
GOATS#
GOATS: Gemini Observation and Analysis of Targets System
GOATS is a browser-based user interface providing end-to-end automation of the entire time-domain/multi-messenger astronomy follow-up process, including target selection, triggering follow-up observations, and data reduction and analysis. It was built using the TOM Toolkit. The DRAGONS software is already integrated into GOATS.
Contact: Monika Soraisam
AEON#
AEON: Astronomical Event Observatory Network
AEON provides rapid, flexible, programmable access to multi-wavelength telescope facilities such as SOAR and Gemini. Alert brokers and Target Observation Managers (TOMs) can submit follow-up observations in real-time without human intervention with AEON.
Contact: César Briceño
Las Cumbres#
Las Cumbres is comprised of many robotic telescopes around the world, functioning as a single facility. Observations can be automatically triggered with AEON.
Contact: Rachel Street
TOM Toolkit#
TOM: Target Observation Manager
Software infrastructure developed by Las Cumbres that enables anyone to build and customize a TOM for their particular science goals. Includes software to interface with ANTARES and AEON.
Contact: Rachel Street
Hermes#
HERMES is a Message Exchange Service for Multi-Messenger Astronomy applications that allow users to both send and review messages related to a variety of events and targets of interest as well as understand opportunities for follow-up.
Contact: Rachel Street
SOAR#
SOAR: Southern Astrophysical Research Telescope
SOAR is a 4m telescope in Chile (Cerro Pachón), offers target-of-opportunity observations for time domain follow-up, and interfaces with AEON.
Contact: César Briceño
Gemini#
Gemini Observatory, 8m twin telescopes in Chile (Cerro Pachón) and Hawaii, offers target-of-opportunity observations for time domain follow-up, and interfaces with AEON. The Gemini Multi-Object Spectrograph (GMOS) is particularly common for transient spectroscopy.
Contact: Bryan Miller
DRAGONS#
DRAGONS: Data Reduction for Astronomy from Gemini Observatory North and South
Software infrastructure for automated processing of imaging and spectroscopic data from Gemini Observatory, including GMOS longslit spectra (but can be extended to other facilities).
Contact: Bryan Miller
GPP#
GPP: Gemini Program Platform
Gemini Observatory’s new web-based platform for proposals and observation preparation, with AEON compatibility.
Contact: Bryan Miller