Lessons from the first multiply imaged supernova: A revised Light-Traces-Mass strong lensing model for the galaxy cluster MACS J1149.5+2223

Abstract: Our light-traces-mass (LTM) strong-lensing model for MACS J1149.5+2223 has played several key roles over the last decade: it aided the identification of multiple images in this cluster and the study of MACS1149-JD1 at redshift $z\simeq9$, it was used to estimate the properties of the first multiply imaged supernova, Refsdal, in its discovery paper, and of the first caustic crossing event by a cluster, Lensed Star 1. Supernova Refsdal supplied an invaluable opportunity to conduct a blind test of the ability of common lens-modeling techniques to accurately describe the properties of SN Refsdal's images and predict the reappearance of one of its counter images that was due about a year post-discovery of the original Einstein cross. Thanks to this practice, in which our submitted model yielded some outlying results, we located a numerical artifact in the time delay (TD) calculation part of the code, which was now fixed. This artifact did not influence the reproduction of multiple images (i.e., the deflection fields -- which are those constrained directly from the observations) or the derived mass model, and so it remained unnoticed prior to supernova Refsdal, emphasizing the importance of blind tests in astronomy. Here we update our model and present revised LTM measurements for Refsdal. These are important not only for completing the LTM view of the Refsdal event, but also because they affect the range of values predicted from different lens-modeling techniques and thus the range of systematic uncertainties for the TD calculation and the resulting Hubble constant.