A real-time search for Type Ia Supernovae with late-time CSM interaction in ZTF

TL;DR

This study implements a real-time, binning-based monitoring program using ZTF to search for late-time circumstellar interaction in 6,914 Type Ia supernovae. By updating light curves every four weeks and pursuing rapid photometric and spectroscopic follow-up, the team identifies several late-time signals, most of which are either artefacts or nuclear transients, with SN 2020qxz emerging as a robust SN Ia-CSM case exhibiting four transient emission lines at highly blueshifted velocities. The findings demonstrate both the potential of large sky surveys to catch late-time CSM phenomena and the challenges posed by host-galaxy contamination and follow-up logistics, emphasizing the need for rapid, multi-instrument observations. The work also suggests a two-region CSM structure around SN 2020qxz and discusses implications for progenitor systems, while pointing to LSST as a transformative resource for future such studies.

Abstract

The nature of Type Ia supernova (SN Ia) progenitor systems and the mechanisms that lead up to their explosions are still widely debated. In rare cases the SN ejecta interact with circumstellar material (CSM) that was ejected from the progenitor system prior to the SN. The unknown distance between the CSM and SN explosion site makes it impossible to predict when the interaction will start. If the time between the SN and start of CSM interaction is of the order of months to years the SN has generally faded and is not actively followed up anymore, making it even more difficult to detect the interaction while it happens. Here we report on a real-time monitoring program which ran between 13-11-2023 and 09-07-2024, monitoring 6914 SNe Ia for signs of late-time rebrightening using the Zwicky Transient Facility (ZTF). Flagged candidates were rapidly followed up with photometry and spectroscopy to confirm the late-time excess and its position. We report the discovery of a $\sim50$ day rebrightening event in SN 2020qxz around 1200 days after the peak of its light curve. SN 2020qxz had signs of early CSM interaction but faded from view over 2 years before its reappearance. Follow-up spectroscopy revealed 4 emission lines that faded shortly after the end of the ZTF detected rebrightening. Our best match for these emission lines are H$β$ (blue shifted by $\sim5900$ km s$^{-1}$) and CaII$_{\lambda8542}$, NI$_{\lambda8567}$, and KI$_{λλ8763, 8767}$, all blue shifted by 5100 km s$^{-1}$ (although we note that these identifications are uncertain). This shows that catching and following up on late-time interactions as they occur can give new clues about the nature of the progenitor systems that produce these SNe by putting constraints on the possible type of donor star, and the only way to do this systematically is to use large sky surveys such as ZTF to monitor a large sample of objects.

Figures (9)

• Figure 1: Left: Binned light curves in flux space in the SN rest frame of the three objects with a late time signal that we could not follow up on. The three ZTF bands $g$, $r$, and $i$ are shown in green, orange, and red, respectively. Individual observations are shown before the binning starts. Bins are shown as coloured blocks showing the size of the bin, mean value, and $1\sigma$ uncertainty of the binned observations, with the point showing the mean observation date of the bin. Bins that are $>\ 5\sigma$ above zero flux are filled, and bins $\leq5\sigma$ from zero flux are dashed. Middle: Same plot as on the left but in absolute magnitude, corrected for MW extinction. $5\sigma$ detections in individual observations are shown before the binning starts, and the $5\sigma$ binned detections are shown in the same way as in the left plots. $5\sigma$ upper limits are shown with a downward arrow for the binned non-detections. The gray vertical line shows when the excess was first discovered. Right: PS1 cutouts centred on the SN location, which is marked in red.
• Figure 2: Same as Fig. \ref{['non_followup_lcs']}, but for the objects for which follow-up observations were made. The grey vertical line shows when the excess was first discovered, and the grey dashed and dotted vertical lines show the photometric and spectroscopic follow-up observations, respectively. The coloured points at the follow-up dates show the follow-up detections and the crosses show the follow-up $5\sigma$ upper limits. Note that for SN 2020qxz the time period between the SN tail and late-time signal has been cut out to better show the detections and follow-up campaign.
• Figure 3: Left:$r$-band image of the location and host galaxy of SN 2019zbq, taken on 23 November 2023 with GTC+OSIRIS. The SN location is marked in red and the purple dot is the galaxy nucleus location. Right: Difference image of the left region after subtracting a Pan-STARRS template image. There is a residual visible at the host galaxy location. Note that the colour scaling is different for the two images to highlight the important sources.
• Figure 4: Pan-STARRS DR1 colour image of the location and host galaxy of SN 2020qxz. The SN location is marked in red, and the two host galaxy lobes are marked in white.
• Figure 5: Combined late-time spectra of SN 2020qxz and its host in the SN rest frame, zoomed in on the regions containing the transient emission lines and the H$\alpha$ region. The transient emission lines are marked with grey vertical lines, and the best-fit Gaussians are overlaid on top of the combined March spectrum with the $1\sigma$ uncertainties of the Gaussian fits shown in grey. The dashed vertical lines show the location of host galaxy lines (as well as H$\beta$ to compare its location to the transient emission line) at the host redshift of $z=0.0975$. We also show the classification spectrum of SN 2020qzx taken around peak magnitude in light blue and the late-time spectrum obtained for SN 2015cp in magenta, scaled down by a factor of 2 for readability.