Abstract
The first resolved, multiply imaged supernova Type Ia, iPTF16geu, was observed 4 years ago, five decades after such systems were first envisioned. Because of the unique properties of the source, these systems hold a lot of promise for the study of galaxy structure and cosmological parameters. However, this very first example presented modelers with a few puzzles. It was expected that to explain image fluxes a contribution from microlensing by stars would be required, but to accommodate the magnitude of microlensing, the density slope of the elliptical power law lens model had to be quite shallow, $\rho_{2D} \propto r^{-0.7}$. Furthermore, the center of mass had to be displaced from that of observed light by ~0.1 kpc, and the position angle of light distribution was misaligned with that of mass by ~40 degrees. In this paper we present mass models that resolve the first two problems, and suggest a resolution of the third. Motivated by observations of local ellipticals, and some recent analysis of galaxy-scale lenses, our mass models consist of two offset (baryonic) mass components. The resulting mass distributions have a single centroid, but are lopsided, and have isodensity contours that are not purely elliptical and not self-similar with radius. For many of our models the microlensing requirements are modest, and the ring formed by the extended supernova host galaxy resembles the observed one.
Highlights
The pioneering work of Refsdal (1964) described a new method to measure the Hubble constant, using a phenomenon—gravitational lensing—that would not be observed for another 15 years (Walsh et al 1979)
To motivate our mass model we assume that the lens galaxy at z ≈ 0.2 must be analogous to nearby ellipticals of similar dynamical mass and effective radius of the light distribution, Re
The two steps of our search method resemble those of the particle swarm optimization, though the latter looks for a single solution, while we look for many
Summary
The pioneering work of Refsdal (1964) described a new method to measure the Hubble constant, using a phenomenon—gravitational lensing—that would not be observed for another 15 years (Walsh et al 1979). Varying the fraction of dark matter compared to total mass, or including external shear does not help (Yahalomi et al 2017) This conclusion suggests that an isothermal power law lens is unlikely. Williams & Zegeye orientation by ∼ 40◦ Taken together, these properties of the lens model lead us to conclude that the actual mass distribution is more complicated than an elliptical density power law. Addition of dark matter substructure, as predicted by ΛCDM cosmological model (Klypin et al 1999; Moore et al 1999) is unlikely to reconcile the simple mass model with observations because substructure cannot change profile slope, or affect image positions significantly enough to reconcile light and mass centroids. We do compare extended rings and time delays generated by our models with observations, in § 4.3 and § 4.4
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
