Time-domain astronomy with Rubin and the Fink broker

The Vera C. Rubin Observatory began releasing its first public data alerts in February 2026. Every night, Rubin can detect millions of changes in the sky, from exploding and varying stars to AGN to new phenomena. Fully harnessing this power requires methods capable of dealing with its enormous data volumes to maximise science and optimise follow-up resources.

In this talk I will present Fink, a Rubin broker developed to face these challenges.  Fink receives and processes Rubin’s time-domain detections in real-time and is community driven: science teams contribute classifiers to identify the most promising events for a wide range of science cases using filtering and state-of-the-art machine learning.  I will present our current classification methods, as well as first results with both simulations and real data. I will then talk about the role of brokers and ML in supernova cosmology using the Dark Energy Survey and its supernova cosmology results as an example. I will conclude by discussing how Rubin and brokers will allow us to study the diversity of transients and test whether Dark Energy is a cosmological constant or evolves with time.

Talk in presence and via Zoom:

https://uni-hamburg.zoom.us/j/66006535328?pwd=aGkrSjJIYmZjK0VpYlpGL0ZrdHg2UT09