Magnified or multiply imaged? – Search strategies for gravitationally lensed supernovae in wide-field surveys

Abstract

Strongly lensed supernovae can be detected as multiply imaged or highly magnified transients. In order to compare the performances of these two observational strategies, we calculate expected discovery rates as a function of survey depth in five grizy filters and for different classes of supernovae (Ia, IIP, IIL, Ibc and IIn). We find that detections via magnification is the only effective strategy for relatively shallow pre-LSST surveys. For survey depths about the LSST capacity, both strategies yield comparable numbers of lensed supernovae. Supernova samples from the two methods are to a large extent independent and combining them increases detection rates by about 50 per cent. While the number of lensed supernovae detectable via magnification saturates at the limiting magnitudes of LSST, detection rates of multiply imaged supernova still go up drastically at increasing survey depth. Comparing potential discovery spaces, we find that lensed supernovae found via image multiplicity exhibit longer time delays and larger image separations making them more suitable for cosmological constraints than their counterparts found via magnification. We find that the ZTF will find about 2 type Ia and 4 core-collapse lensed supernovae per year at a limiting magnitude of 20.6 in the r band. Applying a hybrid method which combines searching for highly magnified or multiply imaged transients, we find that LSST will detect 89 type Ia and 254 core-collapse lensed supernovae per year. In all cases, lensed core-collapsed supernovae will be dominated by type IIn supernovae contributing to 80 per cent of the total counts, although this prediction relies quite strongly on the adopted spectral templates for this class of supernovae. Revisiting the case of the lensed supernova iPTF16geu, we find that it is consistent within the 2\sigma contours of predicted redshifts and magnifications for the iPTF survey.