AAS2RTO: Taming the Transient Data Deluge from LSST

* **Introduction:**

* The Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) will soon produce an unprecedented amount of transient astronomical data, with around 10 million alerts every night.

* This data deluge requires intelligent tools to prioritize the most scientifically valuable events for follow-up, especially spectroscopic observations.

* This podcast discusses AAS2RTO, a new tool designed to address this challenge.

* Reference: Sedgewick et al. (2025) "AAS2RTO: Automated Alert Streams to Real-Time Observations"

* **What is AAS2RTO?**

* AAS2RTO is a Python-based tool for prioritizing transient candidates for follow-up observations.

* It uses a **greedy algorithm** to rank candidates based on a user-defined "score".

* The score is calculated from various factors that consider observed properties of the transients, their visibility from a given location, and user specified criteria.

* AAS2RTO is not a broker but rather works with data streams that have already been pre-filtered by brokers.

* AAS2RTO is designed to be flexible and adaptable to different telescopes and scientific goals.

* **How AAS2RTO Works:**

* AAS2RTO ingests transient data from alert brokers like fink, ALeRCE and Lasair, which process data from surveys like the Zwicky Transient Facility (ZTF).

* It can also incorporate data from other sources like the Asteroid Terrestrial-impact Last Alert System (ATLAS) and the Transient Name Server (TNS).

* AAS2RTO filters candidates based on preliminary scores before fitting models to the lightcurves, such as SALT2 models for Type Ia Supernovae (SNe Ia), in order to estimate the time of peak brightness.

* It calculates a final score for each candidate, considering factors like brightness, age, proximity to peak brightness, and visibility from the observing site.

* AAS2RTO generates a ranked list of candidates that is continuously updated.

* **Example Science Case: Type Ia Supernovae**

* The tool was tested on the prioritization of Type Ia Supernovae (SNe Ia) close to their peak brightness.

* AAS2RTO uses lightcurve fitting to predict the time of peak brightness.

* It uses factors such as the latest magnitude, the number of detections, the proximity to peak brightness, the rising light curve, and the time since the first observation to prioritize candidates.

* It also takes into account the visibility of the candidate from the observing site.

* The tests with archival ZTF data show the tool can estimate the time of peak brightness with a precision of ±1.3 - 2.1 days.

* **AAS2RTO in the Context of Other Schedulers**

* AAS2RTO uses a greedy algorithm similar to other telescope schedulers, and is very flexible, quickly responding to new data and conditions.

* AAS2RTO is different from schedulers for facilities with competing programs, as it is designed for a single scientific goal, and does not normalize priority scores.

* Other schedulers use different approaches, such as integer linear programming solutions to provide globally optimal solutions.

* AAS2RTO’s visibility factor accounts for how long a candidate will be visible during the night.

* **Conclusion:**

* AAS2RTO is a valuable tool for prioritizing transient candidates from LSST for spectroscopic follow-up.

* Its flexible design allows for adaptation to different telescopes and scientific objectives.

* It will be used to help optimize observations with the Danish 1.54m telescope, among others.

Acknowledements: Podcast prepared with Google/NotebookLM. Illustration credits: Rubin Obs/NSF/Aura