HyperGal: hyperspectral scene modeling for supernova typing with the Integral Field Spectrograph SEDmachine

TL;DR

HyperGal is a novel hyperspectral scene modeling technique that improves supernova typing accuracy by better separating transient signals from host galaxy backgrounds in integral field spectrograph data.

Contribution

It introduces a fully chromatic scene modeler using pre-transient images and a physically-motivated spectral interpolator to enhance supernova classification in spectroscopic data.

Findings

HyperGal correctly classifies ~95% of SNe Ia.

It improves classification accuracy by 10% over standard methods.

False positive rate is less than 2%, halving the rate of previous methods.

Abstract

Recent developments in time domain astronomy, like the Zwicky Transient Facility, have made possible a daily scan of the entire visible sky, leading to the discovery of hundreds of new transients every night. Among them, 10 to 15 are supernovae (SNe), which have to be classified prior to cosmological use. The Spectral Energy Distribution machine (SEDm), a low resolution Integral Field Spectrograph, has been designed, built, and operated to spectroscopically classify targets detected by the ZTF main camera. The current Pysedm pipeline is limited by contamination when the transient is too close to its host galaxy core; this can lead to an incorrect typing and ultimately bias the cosmological analyses, and affect the SN sample homogeneity in terms of local environment properties. We present a new scene modeler to extract the transient spectrum from its structured background, aiming at improving the typing efficiency of the SEDm. HyperGal is a fully chromatic scene modeler, which uses pre-transient photometric images to generate a hyperspectral model of the host galaxy; it is based on the CIGALE SED fitter used as a physically-motivated spectral interpolator. The galaxy model, complemented by a point source and a diffuse background component, is projected onto the SEDm spectro-spatial observation space and adjusted to observations. The full procedure is validated on 5000 simulated cubes. We introduce the contrast as the transient-to-total flux ratio at SN location. From estimated contrast distribution of real SEDm observations, we show that HyperGal correctly classifies ~95% of SNe Ia. Compared to the standard extraction method, HyperGal correctly classifies 10% more SNe Ia. The false positive rate is less than 2%, half as much as the standard extraction method. Assuming a similar contrast distribution for core-collapse SNe, HyperGal classifies 14% (11%) additional SNe II (Ibc).