# Unveiling the Rich and Diverse Universe of Subsecond Astrophysics through LSST Star Trails

## TL;DR

This paper introduces a novel LSST observing mode using star trail images and deep learning to detect transient astrophysical events on millisecond timescales, opening new avenues for time-domain astronomy.

## Contribution

It proposes a new observational strategy combining star trail imaging with deep learning for rapid transient detection in wide-area surveys.

## Key findings

- First wide-area optical survey on subsecond timescales
- Enhanced detection of stellar flares and fast transients
- Potential to discover new astrophysical phenomena

## Abstract

We present a unique method that allows the LSST to scan the sky for stellar variability on short timescales. The operational component of the strategy requires LSST to take star trail images. The image processing component uses deep learning to sift for transient events on timescales down to 10 ms. We advocate for enabling this observing mode with LSST, as coupling this capability with the LSST's tremendous 319.5 m$^2$deg$^2$ etendue will produce the first wide area optical survey of the universe on these timescales. We explain how these data will advance both planned lines of investigation and enable new research in the areas of stellar flares, cataclysmic variables, active galactic nuclei, Kuiper Belt objects, gamma-ray bursts, and fast radio bursts.