# A boost in the precision of cluster-mass models: Exploiting the extended surface brightness of the lensed supernova Refsdal host galaxy

**Authors:** S. Schuldt, C. Grillo, A. Acebron, P. Bergamini, A. Mercurio, P. Rosati, S. H. Suyu  
**arXiv:** [2602.12329](https://arxiv.org/abs/2602.12329)  
**Date:** 2026-02-12  
**Summary page:** https://www.summarizepaper.com/en/paper/2602.12329v1/

## Key points

- Schuldt et al. present new mass models for the cluster MACS J1149.5+2223 that lenses the supernova (SN) Refsdal.
- The study combines deep Hubble Space Telescope (HST) images with data from the Multi-Unit Spectroscopic Explorer to exploit the source surface-brightness distribution of the SN host for the first time.
- Analysis includes 77,000 HST pixels and 106 point-like multiple images to explore different model variations.
- Including the extended image of the SN host significantly reduces statistical uncertainties for all 34 free model parameters compared to models based solely on point-like images.
- The enhanced precision enables a more accurate reconstruction of both the cluster's total mass distribution and predicted time delays for multiple image positions of SN Refsdal.
- Schuldt and colleagues showcase the delensed image of the spiral galaxy at redshift zSN = 1.49 in multiple HST bands, providing insights into its physical properties.
- Utilizing extended image models in lensing clusters offers unparalleled precision in measuring total mass distribution, determining cosmological parameters, and characterizing high-redshift sources.

## Summary

In their groundbreaking study, Schuldt et al. present new mass models for the cluster MACS J1149.5+2223 that strongly lenses the supernova (SN) Refsdal. By combining deep Hubble Space Telescope (HST) images with extensive data from the Multi-Unit Spectroscopic Explorer, the researchers fully exploit the source surface-brightness distribution of the SN host for the first time. Their analysis incorporates a remarkable 77,000 HST pixels and 106 point-like multiple images to comprehensively explore different model variations. The team considers four distinct models to investigate how the relative weighting of point-like multiple image positions and flux distribution of the SN host impacts model optimization. Remarkably, when including the extended image of the SN host, statistical uncertainties for all 34 free model parameters are significantly reduced by factors ranging from one to two orders of magnitude compared to models based solely on point-like images. This enhanced precision enables a more accurate reconstruction of both the cluster's total mass distribution and predicted time delays for multiple image positions of SN Refsdal. Furthermore, Schuldt and colleagues showcase the delensed image of the SN host - a spiral galaxy at redshift zSN = 1.49 - in multiple HST bands, providing valuable insights into its physical properties. The results demonstrate that utilizing extended image models in lensing clusters offers unparalleled precision in measuring total mass distribution, determining cosmological parameters, and characterizing high-redshift sources. This study highlights how leveraging extended surface brightness information can significantly reduce statistical uncertainties in mass modeling and parameter estimation, underscoring the importance of addressing systematic uncertainties in future analyses within this field.