The rates and time-delay distribution of multiply imaged supernovae behind lensing clusters

Time delays of gravitationally lensed sources can be used to constrain the mass model of a deflector and determine cosmological parameters. We here present an analysis of the time-delay distribution of multiply imaged sources behind 17 strong lensing galaxy clusters with well-calibrated mass models. We find that for time delays less than 1000 days, at z=3.0, their logarithmic probability distribution functions are well represented by P (log Δ t)=5.3 x 10^-4 Δ t^β M_250^-2β, with β=0.77, where M_250 is the projected cluster mass inside 250 kpc (in 10^14 M_sun), and β is the power-law slope of the distribution. The resultant probability distribution function enables us to estimate the time-delay distribution in a lensing cluster of known mass. For a cluster with M_250=2 x 10^14 M_sun, the fraction of time delays less than 1000 days is approximately 3%. Taking Abell 1689 as an example, its dark halo and brightest galaxies, with central velocity dispersions larger than 500 km/s, mainly produce large time delays, while galaxy-scale mass clumps are responsible for generating smaller time delays. We estimate the probability of observing multiple images of a supernova in the known images of Abell 1689. A two-component model of estimating the supernova rate is applied in this work. For a magnitude threshold of m_AB=26.5, the yearly rate of Type Ia (core-collapse) supernovae with time delays less than 1000 days is 0.004 +- 0.002 (0.029 +- 0.001). If the magnitude threshold is lowered to m_AB ~ 27.0, the rate of core-collapse supernovae suitable for time delay observation is 0.044 +- 0.015 per year.