MARTA-4327: Benchmark High-Redshift Star-Forming Galaxy

• MARTA-4327 is a star-forming galaxy at cosmic noon (z≈2.224) known for deep JWST/NIRSpec observations and direct Tₑ abundance measurements.
• Detailed spectroscopy reveals multiple auroral lines, robust Wolf–Rayet features, and exceptional chemical diagnostics including elevated gas-phase Fe/O ratios.
• Multi-cloud photoionization models show that classical Tₑ analyses underestimate oxygen abundance by up to 0.3 dex due to the influence of dense nebular clumps.

MARTA-4327, denoted as M~4327 in one of the source papers, is a star-forming galaxy at Cosmic Noon in the MARTA program (“Measuring Abundances at high Redshift with the $T_e$ Approach”; JWST PID 1879, PI: Curti) at $z=2.2$, more precisely $z = 2.22344 \pm 0.00007$. It has emerged as a benchmark object for the study of massive stars, direct-method abundances, dust depletion, feedback, and density-structured nebular emission in the high-redshift Universe because it combines deep JWST/NIRSpec spectroscopy, high-S/N detections of multiple auroral lines, and one of the highest-redshift detections of both Wolf–Rayet blue and red bumps in a non-lensed system (Curti et al., 8 Sep 2025).

1. Identification and nomenclature

The designation MARTA-4327 refers, in the relevant astronomical literature here, to a galaxy in the MARTA survey. One paper uses the shortened form M~4327 and explicitly identifies it as “the galaxy MARTA‑4327 from the MARTA survey” (Moreschini et al., 13 Jan 2026). The program acronym MARTA expands as Measuring Abundances at high Redshift with the $T_e$ Approach.

The identifier is not universally attached to all uses of the string “MARTA” in arXiv materials. In particular, a 2025 review of gaseous-detector activities states that “The designation ‘MARTA‑4327’ is not defined anywhere in the text” for the cosmic-ray MARTA project, where only generic “MARTA”, “MARTA stations/modules”, and “MARTA detector module” appear (Souza et al., 30 Nov 2025). This establishes that the galaxy designation should not be conflated with the Muon Array with RPCs for Tagging Air showers instrumentation program.

A plausible implication is that “MARTA-4327” should be treated as an astronomical source designation internal to the MARTA galaxy survey rather than as a generic MARTA hardware or project label. Within that astronomical usage, the object is consistently presented as a chemically informative, compact, actively star-forming system at $z \sim 2.224$.

2. Observations and global properties

MARTA-4327 was targeted in JWST Cycle 1 GO program MARTA and observed with deep JWST/NIRSpec spectroscopy in both medium- and high-resolution gratings. The observations used G140M/F100LP with 32 hr on-source integration, G235M/F170LP with 7 hr, and G235H/F170LP with 3 hr; the high-resolution setting was used specifically for kinematic analysis around H$\alpha$ and N II. The galaxy was a high-priority target because the very deep G140M integration was designed to reach $\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$ for [O III] $\lambda 4363$, enabling a direct $T_e$ analysis.

Its basic measured properties are:

• $z = 2.22344 \pm 0.00007$
• $z=2.2$0
• $z=2.2$1
• $z=2.2$2 (Curti et al., 8 Sep 2025)

A second paper describes the same source as a star-forming galaxy at Cosmic Noon with $z=2.2$3, $z=2.2$4, and $z=2.2$5, and notes that it hosts a population of Wolf–Rayet (WR) stars (Moreschini et al., 13 Jan 2026).

The spectroscopy is unusually rich. High-S/N detections include multiple auroral lines—[S II] $z=2.2$6, [O III] $z=2.2$7, [S III] $z=2.2$8, and [O II] $z=2.2$9—as well as strong nebular lines from the Balmer series, He I recombination lines, [O II] $z = 2.22344 \pm 0.00007$0, [O III] $z = 2.22344 \pm 0.00007$1, [N II] $z = 2.22344 \pm 0.00007$2, [S II] $z = 2.22344 \pm 0.00007$3, [S III] $z = 2.22344 \pm 0.00007$4, [Ar III] $z = 2.22344 \pm 0.00007$5, and the density-sensitive He I $z = 2.22344 \pm 0.00007$6 multiplet (Moreschini et al., 13 Jan 2026). The object is also compact in NIRCam images and does not extend beyond the central MSA shutter in any band (Curti et al., 8 Sep 2025).

3. Stellar population and Wolf–Rayet signatures

MARTA-4327 is one of the clearest high-redshift examples of a WR galaxy observed in deep rest-frame optical spectroscopy. The analysis reports one of the highest-redshift detections of the Wolf–Rayet blue and red bumps in a non-lensed system (Curti et al., 8 Sep 2025). In the blue bump region, the broad He II $z = 2.22344 \pm 0.00007$7 feature has

$z = 2.22344 \pm 0.00007$8

and

$z = 2.22344 \pm 0.00007$9

while the 4605–4650 Å N+C complex has

$T_e$0

In the red bump region, the narrower C IV-centered window gives

$T_e$1

and the extended 5730–5850 Å region gives

$T_e$2

(Curti et al., 8 Sep 2025).

The broad He II $T_e$3 feature is stated to be consistent with a young ($T_e$4 Myr) burst dominated by WN stars, although both SSP models and empirical templates struggle to reproduce the nitrogen stellar features at $T_e$5 Å (Curti et al., 8 Sep 2025). The same paper reports that BPASS models at $T_e$6 Myr with $T_e$7 predict relative WR subpopulations of WNh $T_e$8, WN $T_e$9, and WC $z \sim 2.224$0, and that these models reproduce the He II bump but systematically under-predict the strength of the nitrogen stellar features around 4620–4640 Å.

The observational and template-based comparisons therefore favor a WR population dominated by WN-type stars, possibly including WNh, with a more modest WC contribution. The authors further state that, based on the relative strength of the available optical stellar features, they disfavor the presence of very massive stars (VMS) in this system (Curti et al., 8 Sep 2025). This does not exclude some VMS contribution, but it excludes a dominant VMS interpretation on the available rest-frame optical evidence.

4. Nebular conditions, direct abundances, and depletion

The nebular analysis combines auroral-line diagnostics, Balmer and Paschen decrements, and Bayesian modeling of $z \sim 2.224$1, $z \sim 2.224$2, and $z \sim 2.224$3. The adopted results are: $z \sim 2.224$4

$z \sim 2.224$5

(Curti et al., 8 Sep 2025).

The direct-method ionic abundances are

$z \sim 2.224$6

which yield the total oxygen abundance

$z \sim 2.224$7

The ionization fraction is

$z \sim 2.224$8

(Curti et al., 8 Sep 2025).

The same work derives the following abundance ratios: $z \sim 2.224$9

$\alpha$0

$\alpha$1

$\alpha$2

and

$\alpha$3

(Curti et al., 8 Sep 2025).

A central result is that Ne/O, N/O, and Ar/O align with observations of local star-forming galaxies (including WR galaxies), which the paper interprets as evidence that any impact of the WR population on the chemical enrichment of the ISM is strongly localized (Curti et al., 8 Sep 2025). By contrast, the gas-phase iron abundance behaves differently. The inferred value is

$\alpha$4

with

$\alpha$5

an estimated dust fraction

$\alpha$6

and total

$\alpha$7

(Curti et al., 8 Sep 2025).

The paper interprets the apparently enhanced gas-phase Fe/O relative to local galaxies of similar metallicity as evidence for reduced Fe depletion onto dust grains, possibly linked to localized destruction in WR-driven wind environments. Within the scope of the cited analysis, this is the principal chemical signature that departs from otherwise local-like abundance patterns.

5. Density structure and bias in classical $\alpha$8 metallicities

A separate multi-cloud photoionization analysis places MARTA-4327 at the center of a methodological problem: the relation between unresolved density structure and “direct” abundance estimates. Using the HOMERUN framework, the integrated spectrum is modeled as a linear combination of many constant-density CLOUDY “single-cloud” models, all illuminated by the same ionizing spectrum and sharing the same global chemical composition. The line flux model is

$\alpha$9

and the loss function is

$\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$0

(Moreschini et al., 13 Jan 2026).

For MARTA-4327, three grids are fit:

1. $\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$1,
2. $\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$2,
3. $\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$3.

The fiducial high-density model with $\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$4 yields

$\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$5

$\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$6

$\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$7

$\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$8

$\sim 10^{-19}\,\mathrm{erg\,s^{-1}\,cm^{-2}}$9

$\lambda 4363$0

$\lambda 4363$1

$\lambda 4363$2

$\lambda 4363$3

and

$\lambda 4363$4

(Moreschini et al., 13 Jan 2026).

The same study computes a classical $\lambda 4363$5-based abundance using PyNeb and finds

$\lambda 4363$6

$\lambda 4363$7

$\lambda 4363$8

$\lambda 4363$9

$T_e$0

with

$T_e$1

from S II.

The key result is that classical $T_e$2 metallicity underestimates the true oxygen abundance by $T_e$3 dex in M~4327 once high-density clumps are taken into account. In the low-density HOMERUN run with $T_e$4, the metallicity is $T_e$5, essentially identical to the classical $T_e$6 value $T_e$7; in the fiducial high-density run it rises to $T_e$8 (Moreschini et al., 13 Jan 2026).

The physical explanation is explicit. In the fiducial model:

• H$T_e$9 receives only $z = 2.22344 \pm 0.00007$0 of its flux from $z = 2.22344 \pm 0.00007$1,
• [O III] $z = 2.22344 \pm 0.00007$2 receives $z = 2.22344 \pm 0.00007$3 of its flux from clouds with $z = 2.22344 \pm 0.00007$4,
• [O II] $z = 2.22344 \pm 0.00007$5 receives $z = 2.22344 \pm 0.00007$6 of its flux from $z = 2.22344 \pm 0.00007$7 (Moreschini et al., 13 Jan 2026).

Because collisional de-excitation suppresses low-$z = 2.22344 \pm 0.00007$8 nebular lines such as [O II] $z = 2.22344 \pm 0.00007$9 and [S II] $z=2.2$00 in dense clumps, while high-$z=2.2$01 auroral lines remain efficient, a single-density, two-zone $z=2.2$02 analysis overestimates $z=2.2$03 and biases O/H low. The paper therefore uses MARTA-4327 as the clearest demonstration that “one cloud is not enough” in high-redshift nebular abundance work (Moreschini et al., 13 Jan 2026).

6. Feedback, O I $z=2.2$04, and broader significance

The high-resolution H$z=2.2$05 analysis detects a broad, blueshifted component that is significantly preferred over a single-Gaussian model and is interpreted as an ionized outflow. The measured properties are: $z=2.2$06 after correction for the instrumental line-spread function (Curti et al., 8 Sep 2025).

With an adopted $z=2.2$07, the characteristic and maximum outflow velocities are

$z=2.2$08

giving

$z=2.2$09

and corresponding mass-loading factors

$z=2.2$10

The paper summarizes this as a mass loading factor

$z=2.2$11

(Curti et al., 8 Sep 2025).

Another distinctive feature is the robust detection of O I $z=2.2$12, reported as an $z=2.2$13 line and described as among the first such detections at high redshift (Curti et al., 8 Sep 2025). The upper limits

$z=2.2$14

rule out recombination as the dominant excitation channel and are consistent with Ly$z=2.2$15 fluorescence and/or collisional excitation in dense clumps (Curti et al., 8 Sep 2025). This links the line naturally to the same dense, structured ISM invoked in the multi-cloud analysis.

Taken together, the source papers present MARTA-4327 as a compact $z=2.2$16 galaxy in which several high-value diagnostics coexist: a young WR-dominated stellar burst, normal global $z=2.2$17-element abundance ratios, anomalously elevated gas-phase Fe/O attributed to reduced depletion, an ionized outflow with $z=2.2$18, dense clumps traced by O I $z=2.2$19, and strong auroral-line emission whose interpretation depends sensitively on unresolved density structure (Curti et al., 8 Sep 2025). This suggests that MARTA-4327 is not merely a chemically measured high-redshift galaxy but a laboratory for the joint study of chemical enrichment, feedback, dust processing, and the limitations of single-zone nebular diagnostics (Moreschini et al., 13 Jan 2026).