Lens model of MACSJ1149, version 2.1 Keren Sharon & Traci Johnson (University of Michigan) kerens@umich.edu, tljohn@umich.edu We have modified the version 2 lens model of MACSJ1149 to better reproduce the lensing signal in the vicinity of SN Refsdal. (More information regarding the SN can be found in the discovery paper by Kelly et al. (2014).) We have added the locations of the multiply-imaged SN as a constraint with 13 additional constraints on the host galaxy based on the pre-HFF data. All the other constraints are identical to those in version 2. A full description of the version 2 model can be found in Johnson et al. (2014). The lens products we provide for the version 2.1 model are in the identical format to the version 2 models, with a few additions. We include the magnification maps for source z=1,2,4,9 and additionally a map for the redshift of SN Refsdal (z=1.491). We also include the deflection matrices, kappa, gamma, gamma1, gamma2 computed for z=9 for the best fit lens model. (Note: The new gamma1 and gamma2 maps are identical in format to the gamma maps.) We also include kappa and gamma maps for a simulation of 100 models for estimating the uncertainties on magnifications. For more information regarding the format and use of these FITS files, please see the README file for our version 2 models on MAST. More information regarding the modeling process for this lens model, model components and best fit parameters, source plane reconstructions of the SN host galaxy, and magnification and time delay estimates for SN Refsdal, please refer to Sharon & Johnson (2014, submitted to ApJL). Since heavy constraint is placed on the host galaxy of SN Refsdal, we recommend that this model (version 2.1) only be used for analysis of the SN, the host galaxy, or for lensed galaxies located close to these additional constraints in the image plane. The other multiply-imaged galaxies used as constraints have lower weight in comparison to the host galaxy, and thus this model will not be good at estimating magnifications across the entire image plane as models where all the multiple images have equal weight, as in the version 2 model. For this reason, we still recommend the use of the version 2 models for the analysis of all lensed galaxies in the image plane unrelated to SN Refsdal and its host galaxy. References Johnson, T.L., Sharon, K., Bayliss, M.B., et al. 2014, ApJ, 797, 48 Kelly, P., Rodney, S., Treu T., et al. 2014, arXiv:1405.6009 Sharon, K. & T.L. Johnson, 2014, arXiv:1411.6933