The Variable Sky of Deep Synoptic Surveys

TL;DR

This paper analyzes the expected discovery rates of variable and transient sources in deep synoptic surveys like LSST, highlighting the scale, challenges, and complementarity with other surveys such as Gaia.

Contribution

It provides a quantitative analysis of the number of variable objects LSST will discover, and compares it with Gaia, informing survey design and data management strategies.

Findings

LSST will discover ~10^5 variable stars per night initially.

Discovery rates for AGNs and QSOs will decrease significantly over 4 years.

Asteroid discoveries will start high and decline below 1000 per night within 2 years.

Abstract

The discovery of variable and transient sources is an essential product of synoptic surveys. The alert stream will require filtering for personalized criteria -- a process managed by a functionality commonly described as a Broker. In order to understand quantitatively the magnitude of the alert generation and Broker tasks, we have undertaken an analysis of the most numerous types of variable targets in the sky -- Galactic stars, QSOs, AGNs and asteroids. It is found that LSST will be capable of discovering ~10^5 high latitude |b| > 20 deg) variable stars per night at the beginning of the survey. (The corresponding number for |b| < 20 deg is orders of magnitude larger, but subject to caveats concerning extinction and crowding.) However, the number of new discoveries may well drop below 100/night within 2 years. The same analysis applied to GAIA clarifies the complementarity of the GAIA and LSST surveys. Discovery of variable galactic nuclei (AGNs) and Quasi Stellar Objects (QSOs) are each predicted to begin at ~3000 per night, and decrease by 50X over 4 years. SNe are expected at ~1100/night, and after several survey years will dominate the new variable discovery rate. LSST asteroid discoveries will start at > 10^5 per night, and if orbital determination has a 50% success rate per epoch, will drop below 1000/night within 2 years.