A Young Supernova Selection Pipeline For The LSST Era

TL;DR

The paper tackles the bottleneck of obtaining early-time supernova spectra in the LSST era by formulating Young SNe (YSN) criteria that identify transients before peak (phase < -10 days) to enable earlier spectroscopic follow-up with TiDES-like surveys. It develops criteria using ZTF alerts via Lasair and Sherlock, validates them against archival data, and adapts them to LSST simulations for both WFD and DDF surveys, comparing the resulting YSN samples to the existing TiDES_2025 selection. The authors quantify contamination and potential wasted telescope time, demonstrating that the YSN criteria yield significantly earlier targets with manageable purity and can substantially augment early-time SN science when used alongside current selection. They also study optimal 4MOST-like observing strategies to maximize the yield and quality of early-time spectra, recommending strategy choices that balance prompt follow-up with reliable classifications. Collectively, the work provides actionable guidelines for enabling LSST-era early-time spectroscopic SN science and informs the design of future coordinated LSST–4MOST follow-up campaigns.

Abstract

Early-time spectroscopy of supernovae (SNe), acquired within days of explosion, yields crucial insights into their outermost ejecta layers, facilitating the study of their environments, progenitor systems, and explosion mechanisms. Recent efforts in early discovery and follow-up of SNe have shown the potential insights that can be gained from early-time spectra. Surveys such as the Time-Domain Extragalactic Survey (TiDES), conducted with the 4-meter Multi-Object Spectroscopic Telescope (4MOST), will provide spectroscopic follow-up of transients discovered by the Legacy Survey of Space and Time (LSST). Current simulations indicate that early-time spectroscopic studies conducted with TiDES data will be limited by the current SN selection criteria. To enhance early-time SN spectroscopic studies from TiDES-like surveys, we propose a set of selection criteria focusing on young SNe (YSNe), which we define as SNe prior to -10 days before peak brightness. Utilising the Zwicky Transient Facility transient alerts, we developed criteria to select YSNe while minimising the sample's contamination rate to 23 percent. The developed criteria were applied to LSST simulations, yielding a sample of 694 Deep Drilling Field survey SNe and 56,260 Wide Fast Deep survey SNe for follow-up. We demonstrate that our criteria enables the selection of SNe at early-times, enhancing future early-time spectroscopic SN studies from TiDES-like surveys. Finally, we investigated 4MOST-like observing strategies to increase the sample of spectroscopically observed YSNe. We propose that a 4MOST-like observing strategy that follows LSST with a delay of 3 days is optimal for a TiDES-like SN survey in terms of the number of classifiable spectra obtained, while a 1 day delay is most optimal for enhancing the early-time science in conjunction with our YSN selection criteria.

Figures (13)

• Figure 1: Light curves of unclassified contaminants in our YSN candidate sample selected from the ZTF transient alerts. (Left) $g$- and $r$-band light curves of a contaminant that is likely an AGN or quasar. (Middle and right) $g$- and $r$-band light curves of contaminants that have no clear nature.
• Figure 2: Phase distribution of the TNS classified and visually determined SNe in the sample that was produced by applying our YSN selection criteria (see Table \ref{['tab:ztf_selection_criteria']}), with age and brightening rate thresholds of 7 days and 0.2 mag/day respectively, to 60 nights of ZTF transient alerts. Note that the phase has been truncated at -28 days.
• Figure 3: Comparison between the SN Ia selection phase distributions produced by applying our selection criteria (YSN; see Table \ref{['tab:lsst_selection_criteria']}) and the TiDES_2025 selection criteria to the LSST WFD survey simulation. Note that the distributions are normalised and that the phase has been truncated at 50 days.
• Figure 4: Comparison between the SN Ib selection phase distributions produced by applying our selection criteria (YSN; see Table \ref{['tab:lsst_selection_criteria']}) and the TiDES_2025 selection criteria to the LSST WFD survey simulation. Note that the distributions are normalised and that the phase has been truncated at 50 days.
• Figure 5: Comparison between the SLSN selection phase distributions produced by applying our selection criteria (YSN; see Table \ref{['tab:lsst_selection_criteria']}) and the TiDES_2025 selection criteria to the LSST WFD survey simulation. Note that the distributions are normalised and that the phase has been truncated at -150 days and 50 days.