MACS J0138-2155: Lensing Cluster Insights

• MACS J0138-2155 is a massive, relaxed galaxy cluster with a strong-lensing core, notable for producing two multiply imaged Type Ia supernovae.
• Analyses combine multiwavelength data from HST, JWST, VLT/MUSE, and Chandra to constrain its mass distribution and time-delay cosmography.
• Detailed lens modeling and spectroscopy provide insights into dark matter self-interaction limits and galaxy evolution via ram-pressure stripping.

Searching arXiv for papers on MACS J0138-2155 to ground the article in the latest literature. MACS J0138-2155, also written MACS J0138.0-2155 or MACSJ0138, is a massive, relaxed, strong-lensing galaxy cluster at redshift $z=0.336$ that lenses a background galaxy at $z=1.949$ and is notable as the first known system to produce two multiply imaged Type Ia supernovae in the same host, SN Requiem and SN Encore (Acebron et al., 12 Mar 2025). In current work, the cluster functions simultaneously as a laboratory for strong-lensing reconstruction, time-delay cosmography, cluster-galaxy environmental processing, and dark-matter phenomenology, with analyses combining Hubble Space Telescope, James Webb Space Telescope, VLT/MUSE, and Chandra data (Gibson et al., 27 Aug 2025).

1. Cluster identity and observational setting

MACS J0138-2155 is part of the MAssive Cluster Survey and is repeatedly characterized as a massive, relaxed cluster with a prominent strong-lensing core (Flowers et al., 2024). The system has a cluster redshift measured either as $z_{\rm L}=0.336$ or, from MARZ spectroscopy, a biweight location $z=0.3367$, and the bright lensed source galaxy MRG-M0138 lies at $z_{\rm S}=1.949$ (Flowers et al., 2024). The field contains two giant tangential arcs and one prominent radial arc, with giant arcs extending to $r \approx 19''$, and this central geometry underlies both lens-modeling and cosmographic applications (Ertl et al., 12 Mar 2025).

The observational foundation is unusually broad. HST imaging includes broad-band optical and near-infrared data such as F390W, F555W, F814W, F105W, F125W, F140W, and F160W, while JWST/NIRCam imaging adds F115W, F150W, F200W, F277W, F356W, and F444W (Acebron et al., 12 Mar 2025). VLT/MUSE observations cover the central $\approx 1$ arcmin$^2$ with a total depth of $\sim 3.7$ hr, combining a 2019 program and a 2023 Target of Opportunity campaign, and provide secure redshifts, emission-line diagnostics, and stellar kinematics (Granata et al., 2024). Chandra ACIS-I observations supply X-ray morphology and hydrostatic mass constraints, including a cool core with $kT=(4.00\pm0.25)$ keV within 25 kpc and $z=1.949$0–7 keV beyond in one analysis, and an $z=1.949$1 temperature $z=1.949$2 keV in another (Acebron et al., 12 Mar 2025).

This multiwavelength coverage has made MACS J0138-2155 a benchmark cluster for inner-halo mass reconstruction. A plausible implication is that the system’s value derives not only from its rare supernova configuration, but also from the availability of mutually constraining probes on scales from the brightest cluster galaxy to the arc-defined lensing region.

2. Strong lensing, arcs, and the multiply imaged supernovae

The defining feature of MACS J0138-2155 in the lensing literature is that it is the first known cluster to show two multiply imaged supernova siblings, SN Requiem and SN Encore, in the same host galaxy at $z=1.949$3 (Acebron et al., 12 Mar 2025). SN Requiem was discovered in archival HST data, while SN Encore was discovered by JWST and spectroscopically confirmed with JWST/NIRSpec IFU and MUSE (Suyu et al., 15 Sep 2025). This configuration has motivated blind, multi-team modeling campaigns aimed at predicting image positions, magnifications, and time delays for eventual $z=1.949$4 inference (Suyu et al., 15 Sep 2025).

Current lensing analyses use a vetted catalog of eight “gold” systems comprising 23 multiple images with secure spectroscopic redshifts spanning $z=1.949$5 (Suyu et al., 15 Sep 2025). In the host system itself, the compact core of MRG-M0138 is lensed into at least five images, while the supernova images occupy the southern and western tangential arcs and the radial arc near the cluster center (Ertl et al., 12 Mar 2025). Some analyses also note a likely “silver” system near $z=1.949$6, but it is excluded from the most conservative model sets because its redshift is not secure (Suyu et al., 15 Sep 2025).

The lensing formalism follows standard notation. The critical surface density is

$z=1.949$7

with convergence $z=1.949$8, and magnification

$z=1.949$9

Time delays are written in terms of the Fermat potential,

$z_{\rm L}=0.336$0

so that for fixed mass models $z_{\rm L}=0.336$1 scales approximately as $z_{\rm L}=0.336$2 (Suyu et al., 15 Sep 2025).

Predictions for the supernova images are now comparatively mature. For SN Encore, all models predict at least four images, often with a fifth central image in some fraction of posterior samples; for SN Requiem, all models predict at least one more image and often two (Suyu et al., 15 Sep 2025). In one multi-plane modeling analysis, the ultimate model predicts for Encore $z_{\rm L}=0.336$3, $z_{\rm L}=0.336$4, $z_{\rm L}=0.336$5, and the future radial image $z_{\rm L}=0.336$6, with a possible future central image $z_{\rm L}=0.336$7 predicted by $z_{\rm L}=0.336$8 of MCMC samples; for Requiem it predicts $z_{\rm L}=0.336$9, $z=0.3367$0, $z=0.3367$1, the future radial image $z=0.3367$2, and a possible future central image $z=0.3367$3 predicted by $z=0.3367$4 of samples (Ertl et al., 12 Mar 2025). The existence fraction of faint central images differs substantially across models, reflecting differences in the inferred inner mass-profile slope and local substructure treatment rather than a settled observational ambiguity (Suyu et al., 15 Sep 2025).

A major methodological issue is the sensitivity of magnification predictions to model assumptions. One enhanced lens model finds systematically larger magnifications for SN Requiem images than earlier work, by factors $z=0.3367$5–3.0 versus Newman et al. (2018), and $z=0.3367$6–5.5 versus Rodney et al. (2021), and attributes the discrepancy primarily to lens-model choices such as anchored Faber–Jackson scaling and flexible subhalo truncations (Acebron et al., 12 Mar 2025). This has direct implications for intrinsic supernova luminosities and host-galaxy property inference.

3. Mass distribution, relaxation, and lens-model architectures

MACS J0138-2155 is consistently described as relaxed in X-ray, spectroscopic, and lensing analyses, although different methods emphasize different structural aspects (Flowers et al., 2024). Chandra imaging in the 0.5–2 keV band shows round morphology, and within $z=0.3367$7 the X-ray surface brightness is fit by a 2D elliptical $z=0.3367$8-model with ellipticity $z=0.3367$9, core radius $z_{\rm S}=1.949$0 kpc, and $z_{\rm S}=1.949$1 (Flowers et al., 2024). In the strong-lensing regime, one parametric model measures a projected total mass

$z_{\rm S}=1.949$2

while the corresponding Chandra hydrostatic mass is $z_{\rm S}=1.949$3, consistent within $z_{\rm S}=1.949$4 (Acebron et al., 12 Mar 2025).

Several lens-model families have been tested. One study adopts Lenstool with Bayesian MCMC, single-plane lensing, and image-plane $z_{\rm S}=1.949$5 optimization, with a cluster-scale non-truncated elliptical dPIE halo, external shear, 84 cluster members, and two line-of-sight perturbers (Acebron et al., 12 Mar 2025). Another constructs seven models using different dark-matter halo parameterizations and explores both multi-plane and approximate single-plane treatments, with mass components consisting of cluster-scale dark-matter halos and galaxy-scale dPIE halos (Ertl et al., 12 Mar 2025). In that framework, two elliptical cluster-scale halos are required: a primary halo centered within $z_{\rm S}=1.949$6 of the BCG and a secondary, cored, extended halo $z_{\rm S}=1.949$7 southwest of the BCG (Ertl et al., 12 Mar 2025).

The choice of halo profile matters. In the seven-model study, six of seven models fit the image positions well, but the model with an elliptical NFW primary cluster halo is statistically disfavored, with an image-position $z_{\rm S}=1.949$8 about four times higher than the cored-isothermal models (Ertl et al., 12 Mar 2025). In the enhanced strong-lensing analysis, the reference best-fit model reproduces all 23 observed image positions with a root-mean-square offset of $z_{\rm S}=1.949$9, and the supernovae plus host-galaxy images with a mean precision of only $r \approx 19''$0 (Acebron et al., 12 Mar 2025). This indicates that, within the inner arc-bracketed region, the cluster mass distribution is tightly constrained despite remaining model-family dependence.

One persistent point of comparison is the apparent tension between X-ray roundness and earlier highly elliptical lens reconstructions. Flowers et al. report that Rodney et al. (2021) inferred a dark-matter halo ellipticity $r \approx 19''$1, substantially higher than the X-ray ellipticity $r \approx 19''$2 (Flowers et al., 2024). The papers do not resolve this discrepancy into a single interpretation. A plausible implication is that the difference reflects differing sensitivities of X-ray isophotes, parametric lens components, and substructure prescriptions rather than a straightforward contradiction about the cluster’s dynamical state.

4. Spectroscopy, cluster membership, and galaxy-scale substructure

The cluster now has a comparatively mature spectroscopic census. MUSE spectroscopy has delivered 107 reliable extragalactic redshifts in the field, including 50 secure cluster members and multiple background sources (Granata et al., 2024). Membership in one analysis is defined by a rest-frame line-of-sight velocity cut of $r \approx 19''$3 km s$r \approx 19''$4 around the mean, corresponding to $r \approx 19''$5, and the member population is dominated by passively evolving early-type spectra, with a few active or jellyfish galaxies and a BCG that shows broad emission lines from an AGN (Granata et al., 2024).

Photometric augmentation increases the modeled membership to 84 galaxies. Using HST and JWST color–magnitude and color–color selections calibrated on the spectroscopic sample, one study supplements the 50 spectroscopic members with 34 photometric members, producing the 84-galaxy catalog used in lens modeling (Ertl et al., 12 Mar 2025). In parallel, multiple analyses emphasize two important line-of-sight perturbers: a foreground galaxy near the southern arc at $r \approx 19''$6 and a background galaxy northwest of the BCG at $r \approx 19''$7 (Ertl et al., 12 Mar 2025). Explicit inclusion of these perturbers improves the fidelity of magnification predictions near the arc system, and omitting the foreground galaxy biases magnifications for nearby systems (Acebron et al., 12 Mar 2025).

A central technical ingredient is the cluster-specific Faber–Jackson calibration used to regularize galaxy-scale subhalos. With deeper MUSE data and HST/F160W photometry, Granata et al. measure a slope $r \approx 19''$8, a reference dispersion $r \approx 19''$9 km s$\approx 1$0, and intrinsic scatter $\approx 1$1 km s$\approx 1$2 for 13 early-type members excluding the BCG (Granata et al., 2024). Flowers et al., using 18 bright, non-blended, quiescent galaxies and 81 systematic variations, find $\approx 1$3 (stat.) $\approx 1$4 (sys.) and $\approx 1$5 (stat.) $\approx 1$6 (sys.) km s$\approx 1$7 at a reference velocity dispersion of $\approx 1$8 km s$\approx 1$9 (Flowers et al., 2024). The agreement between these calibrations is explicitly noted.

In one lensing implementation, the adopted scaling is

$^2$0

with $^2$1 when $^2$2 is assumed (Acebron et al., 12 Mar 2025). In another, the equivalent dPIE scaling is written in terms of Einstein radius and truncation radius relative to a reference galaxy (Ertl et al., 12 Mar 2025). In both cases, the purpose is the same: to reduce the degeneracy between velocity scale and truncation scale for member halos, thereby stabilizing the substructure model that governs local image positions and magnifications.

5. Intracluster medium, galaxy transformations, and jellyfish systems

The cluster core also hosts a distinct environmental-physics program centered on ram-pressure stripping. In MACS J0138.0-2155, three cluster members, J1–J3, are classified as jellyfish galaxies, and a fourth object, J4, lies in the foreground at $^2$3 but close in projection to a lensed arc (Gibson et al., 27 Aug 2025). The physical mechanism is ram-pressure stripping, for which the paper states the relation

$^2$4

where $^2$5 is the galaxy’s velocity through the ICM and $^2$6 is the local ICM density (Gibson et al., 27 Aug 2025).

HST imaging reveals elongated, disturbed tails in J1–J3, while MUSE maps show extraplanar ionized gas, tail-confined H$^2$7 emission, and velocity gradients of a few hundred km s$^2$8 across the tails (Gibson et al., 27 Aug 2025). J1 and J2 are identified as late-stage stripping systems: they show strong Balmer absorption and little or no emission in their heads, with H$^2$9 emission confined to the tails, indicating post-starburst disk populations and recent quenching within $\sim 3.7$0 Gyr (Gibson et al., 27 Aug 2025). J3 differs morphologically and spectroscopically, with both absorption and emission present and a main star-forming knot aligned with the stripping direction, suggesting an earlier stripping phase (Gibson et al., 27 Aug 2025).

The resolved kinematics are specific. J1 lies 42 kpc from the BCG, has a tail length of $\sim 3.7$1 kpc, and shows a tail mean gas velocity offset by $\sim 3.7$2 km s$\sim 3.7$3 relative to the head, with a gradient $\sim 3.7$4 km s$\sim 3.7$5 across the tail; in its strongest star-forming knot, $\sim 3.7$6 km s$\sim 3.7$7 (Gibson et al., 27 Aug 2025). J2 lies 15 kpc from the BCG, has a tail length of $\sim 3.7$8 kpc, and shows a clear tail gradient, with $\sim 3.7$9 km s$kT=(4.00\pm0.25)$0 at the tip where star formation occurs (Gibson et al., 27 Aug 2025). J3, at 111 kpc, shows a gas velocity gradient consistent with outgoing motion post-pericenter, and at its main star-forming knot both $kT=(4.00\pm0.25)$1 and $kT=(4.00\pm0.25)$2 are $kT=(4.00\pm0.25)$3 km s$kT=(4.00\pm0.25)$4 (Gibson et al., 27 Aug 2025).

The H$kT=(4.00\pm0.25)$5-based star-formation analysis uses an intrinsic Balmer ratio $kT=(4.00\pm0.25)$6, the Cardelli extinction law, and a Chabrier-IMF calibration

$kT=(4.00\pm0.25)$7

with internal extinction from the Balmer decrement (Gibson et al., 27 Aug 2025). Peak resolved star-formation surface densities are $kT=(4.00\pm0.25)$8 $kT=(4.00\pm0.25)$9 yr$z=1.949$00 kpc$z=1.949$01 in a J1 tail knot, $z=1.949$02 at the J2 tail tip, $z=1.949$03 in J3, and $z=1.949$04 on the arc-facing side of J4 (Gibson et al., 27 Aug 2025). BPT diagnostics place J2 and J4 in purely star-forming loci; J1 includes some composite bins; J3 is predominantly composite with some spatially extended LINER-like bins, interpreted in the paper as consistent with shock excitation in stripped gas rather than AGN (Gibson et al., 27 Aug 2025).

These results make MACS J0138-2155 unusual among strong-lensing clusters in that the lens-modeling literature and the environmental-processing literature intersect directly. J4, although not a stripped cluster member, lies close enough in projection to a lensed arc that its mild rotation and low star formation are relevant for accurate lens modeling (Gibson et al., 27 Aug 2025).

6. Cosmography, blind model comparisons, and dark-matter constraints

MACS J0138-2155 is now a cluster-scale platform for time-delay cosmography. Seven independent mass models built using six software packages—glafic, GLEE, Lenstool I, Lenstool II, Zitrin-analytic, MrMARTIAN, and WSLAP+—have been compared under a blind protocol to quantify software- and modeling-choice systematics (Suyu et al., 15 Sep 2025). Using the gold images, the better-fitting models are broadly consistent in predicted positions, magnifications, and time delays, especially when $z=1.949$05 (Suyu et al., 15 Sep 2025).

The key short-delay measurement comes from SN Encore. The measured delay between images $z=1.949$06 and $z=1.949$07 is

$z=1.949$08

with the negative sign indicating that $z=1.949$09 arrives earlier than $z=1.949$10 by $z=1.949$11 days (Pierel et al., 15 Sep 2025). Combining this delay with the seven blind lens models, weighted by image-position likelihoods, yields

$z=1.949$12

for a flat $z=1.949$13CDM cosmology with $z=1.949$14 (Pierel et al., 15 Sep 2025). The uncertainty is currently dominated by the time-delay measurement rather than lens-model dispersion (Pierel et al., 15 Sep 2025).

The long delays predicted for future supernova images are central to the cluster’s cosmographic significance. For a fiducial cosmology with $z=1.949$15 km s$z=1.949$16 Mpc$z=1.949$17 and $z=1.949$18, the next Encore image is expected after $z=1.949$19 days and the next Requiem image after $z=1.949$20 to 4000 days, with the four lowest-$z=1.949$21 models placing the Requiem $z=1.949$22 reappearance in $z=1.949$23April–December 2026 if $z=1.949$24 km s$z=1.949$25 Mpc$z=1.949$26, and $z=1.949$27March–November 2027 if $z=1.949$28 km s$z=1.949$29 Mpc$z=1.949$30 (Suyu et al., 15 Sep 2025). A separate joint strong-lensing plus stellar-kinematics SIDM analysis predicts Requiem “d” around November 2027 and “e” around April 2028, with a $z=1.949$31 window between January 2027 and November 2028 at $z=1.949$32 km s$z=1.949$33 Mpc$z=1.949$34 (O'Donnell et al., 27 Aug 2025). The difference reflects different mass models rather than disagreement about the qualitative expectation that the next Requiem image should appear on a multi-year timescale.

The same cluster also yields a dark-matter constraint. By combining strong lensing with spatially resolved stellar kinematics of the BCG and a self-consistent SIDM halo profile, one study reports a 95% confidence upper limit on the self-interaction cross section of

$z=1.949$35

at an interaction velocity of $z=1.949$36 km s$z=1.949$37 (O'Donnell et al., 27 Aug 2025). In that analysis, lensing-only posteriors favor a higher $z=1.949$38 mode near $z=1.949$39 cm$z=1.949$40 g$z=1.949$41, but the IFU kinematics suppress this mode and tighten the upper limit (O'Donnell et al., 27 Aug 2025). This suggests that MACS J0138-2155 is valuable not only for cosmography, but also for velocity-aware tests of SIDM in a regime where baryonic gravity from the BCG must be modeled jointly with the dark halo.

Taken together, the recent literature defines MACS J0138-2155 as a convergent system: a relaxed strong-lensing cluster with a high-value arc configuration, a uniquely informative pair of lensed Type Ia supernovae, measurable galaxy-scale substructure, active ram-pressure stripping in the core, and sufficient data quality to support both percent-level future cosmography and detailed constraints on inner-halo dark-matter physics.