Periodicity Recovery with LSST Alert Production Light Curves

We report a preliminary investigation of the periodicity recovery
using synthetic Legacy Survey of Space and Time (LSST) survey light
curves during the real-time Alert Production (AP) latency of 60
seconds. Our light curves emerge from the Photometric LSST
Astronomical Classification Challenge (PLAsTiCC) and an in-house
light curve generation tool using LSST OpSim V.1.7.1 baseline. We
find that simple periodic phenomena such as RR Lyrae will have a
higher completeness in periodicity compared to more complex light
curves such as eclipsing binaries.

Transcript

1. Periodicity Recovery with LSST Alert Production Light Curves Summary We report a preliminary investigation of the periodicity recovery using synthetic Legacy Survey of Space and Time (LSST) survey light curves during the realtime Alert Production (AP) latency of 60 seconds. Our light curves emerge from the Photometric LSST Astronomical Classification Challenge (PLAsTiCC) and an in-house light curve generation tool using LSST OpSim V.1.7.1 baseline. We find that simple periodic phenomena such as RR Lyrae will have a higher completeness in periodicity compared to more complex light curves such as eclipsing binaries. Floating Mean Multi-Band Periodogram Example observed light curve after ~365 days from PLAsTiCC Example phase-folded PLAsTiCC light curve after ~365 days Period grid, model complexity, number of detections, false positive rate, algorithm run-time, outliers Parameters to Consider RR Lyrae Eclipsing Binaries P alias = P true 1 + MPtrue P harmonic = MP true Common Period Failure Modes Conclusions Based on our preliminary analysis using open-source Lomb-Scargle Periodogram implementation GatsPy, reporting at least the top three highest periods will lead to periodicity completeness of ~80% for RR Lyrae and up to 50% for eclipsing binaries.Future work will examine more carefully the effects of completeness trade-off for more complex models and coarser period grids, and false alarm probability assessment. Authors Anastasios Tzanidakis & Professor Eric Bellm Department of Astronomy, University of Washington, Seattle, WA atzanida@uw.edu \\ atzanida.github.io Citations [1]: VanderPlas & Ivezić 2015 - Periodograms for Multiband Astronomical Time Series [2]: Bellm 2021 - Review of Timeseries Features DMTN-118 [3]: Hložek & Ponder et al. 2020 - Results of the Photometric LSST Astronomical Time- series Classification Challenge [4]: Gatspy (software) - General tools for Astronomical Time Series in Python. GitHub.com/astroML/gatspy The results of this study will be reported on the DMTN-221. dmtn-221.lsst.io Ptop = Ptrue Ptop = 2 × Ptrue Ptop = Ptrue Completeness Performance