Spider-Webb: JWST Near Infrared Camera resolved galaxy star formation and nuclear activities in the Spiderweb protocluster at z = 2.16

ABSTRACT

Near-infrared (NIR) emission is less affected by dust than ultraviolet and optical emission and is therefore useful for studying the properties of dust-obscured galaxies. Although rest-frame NIR observations of high-redshift galaxies have long been made using space telescopes, their structures were unresolved due to the lack of angular resolution. This letter reports the early results from the analysis of high-resolution Pa|$\beta$| imaging of the Spiderweb protocluster at |$z=2.16$| with the JWST Near Infrared Camera. We investigate radial profiles of Pa|$\beta$| lines and rest-frame NIR continua from luminous H |$\alpha$|-emitting galaxies (HAEs) in the protocluster. Particularly, we compare those of 11 HAEs (N-HAEs) on the star-forming main sequence with those of 8 HAEs (X-HAEs) with X-ray active galactic nuclei (AGNs). Resultant composite Pa|$\beta$| line images of N-HAEs indicate significant star formation in galactic discs. In contrast, X-HAEs are dominated by point source components rather than outer star formation, as inferred from our earlier work based on multiwavelength spectral energy distribution fitting. Given their higher stellar potentials suggested from their rest-frame NIR images, the different characteristics may be driven by the impact of AGN feedback.

galaxies: clusters: individual: PKS 1138−262, galaxies: evolution, galaxies: formation, galaxies: high-redshift

1 INTRODUCTION

The redshift interval of |$z=$| 2–4 is known as the peak epoch of cosmic star formation rate (SFR) density, providing a key to understanding galaxy formation and evolution (e.g. Lilly et al. 1996; Madau et al. 1996; Hopkins & Beacom 2006; Madau & Dickinson 2014, and references therein). While significant progress has been made over the past decades, a framework of regulation mechanisms of star formation in massive galaxies remains controversial (e.g. Shapley 2011; Silk & Mamon 2012; Schawinski et al. 2014; Somerville & Davé 2015; Naab & Ostriker 2017). Active galactic nucleus (AGN) feedback (e.g. Silk & Rees 1998; Sijacki et al. 2007; Fabian 2012; Förster Schreiber & Wuyts 2020) is currently the most successful solution to explain observations, such as stellar mass to halo mass relation in the bulk of modern hydrodynamic simulations (e.g. Dubois et al. 2014; Davé et al. 2019; Terrazas et al. 2020; Wellons et al. 2023).

Thus, understanding the intricate interplay (co-evolution) between galaxies and supermassive black holes (SMBHs) has become a central theme in the latest research on galaxy formation and evolution. Modern hydrodynamic simulations commonly implement at least two channels (so-called radiative and kinetic modes) in AGN feedback (e.g. Schaye et al. 2015; Weinberger et al. 2017), and some model suites also include other mechanisms, such as X-ray heating (Choi et al. 2012; Davé et al. 2019) and cosmic ray feedback (Wellons et al. 2023; Hopkins et al. 2024). These simulations suggest that AGN feedback at low accretion rates in SMBHs is considered important for massive galaxies, as it dominates the time-integrated feedback energy. In addition, unravelling AGN feedback mechanisms is tightly linked to understanding black hole (BH) accretion mechanisms following a well-known M_BH–|$\sigma$| relation (Magorrian et al. 1998; Ferrarese & Merritt 2000; King 2003; Kormendy & Ho 2013; Heckman & Best 2014). Otherwise, over(under)massive BHs result in over(under)quenching in the stellar mass to halo mass relation (Wellons et al. 2023), requiring further adjustments to feedback models.

However, investigating solely luminous AGNs may not contribute significantly to understanding AGN feedback because it only traces a short period of AGN activities, and because energy contributions from radiative AGN feedback at high accretion states may be fairly low compared with the total budget of energy injections (Terrazas et al. 2020; Piotrowska et al. 2022; Ward et al. 2022; Bluck, Piotrowska & Maiolino 2023; Hartley et al. 2023; Bluck et al. 2024). Weak signs of AGNs have been frequently reported in quiescent and post-star-forming galaxies through deep observations or stacking analyses (e.g. Kriek et al. 2009; Olsen et al. 2013; Marsan et al. 2015; Man et al. 2016; Gobat et al. 2017; Ito et al. 2022; Kubo et al. 2022; Carnall et al. 2023; Belli et al. 2024; Bugiani et al. 2024; López et al. 2024; Park et al. 2024; Shimakawa et al. 2024b). However, a systematic study of post-star-forming galaxies that host SMBHs at moderate-to-low accretion rates is quite expensive, since deep and wide-field observations with various facilities in multiwavelength are required.

An efficient strategy for addressing such an observational challenge is to target maturing protoclusters at |$z=$| 2–4, where a number of massive galaxies under star formation quenching can be observed at a lower cost owing to a smaller number of pointings. One of the best-studied protoclusters, the Spiderweb protocluster (Carilli et al. 1997; Pentericci et al. 1997, 2000; Kurk et al. 2000 ), would be a good test bed for examining AGN feedback on massive galaxies. Based on the X-ray to the sub-mm data, Shimakawa et al. (2024b, S24 hereafter) reported that a third of massive H |$\alpha$| emitters (HAEs) are undergoing a transition from star-forming to quiescent, and significant fractions of their H |$\alpha$| + [N ii] lines would originate from low-luminosity AGNs such as low-ionization nuclear emission-line regions (Heckman 1980) rather than star formation. However, the results were inconclusive as they heavily relied on complex spectral energy distribution (SED) modelling; therefore, more direct verification is required to resolve star formation and AGN activities in these massive HAEs.

Given such a backdrop, this letter examines radial profiles of Pa|$\beta$| lines of 19 massive HAEs with |$M_\star > 2\times 10^{10}$| M|$_\odot$| in the Spiderweb protocluster at |$z=2.16$|⁠, based on newly obtained JWST/Near Infrared Camera (NIRCam) F405N and F410M images (Pérez-Martínez et al. 2024; Shimakawa et al. 2024a). Pa|$\beta$| gives us with a less dust-biased view compared to H |$\alpha$| (Kennicutt 1998; Calzetti et al. 2007) and its wavelength redshifted to the protocluster is just covered by the F405N filter. We select the targets from S24 (table A), which have measured stellar masses and SFRs for 84 HAEs based on the multiwavelength photometry from X-ray to sub-mm data, including 14 HAEs with X-ray counterparts (Tozzi et al. 2022). This work is motivated to resolve star formation and AGN activities for these massive HAEs at |$\sim$|1 kpc resolution with NIRCam, enabling us to validate the previous arguments. Throughout the letter, we assume a flat lambda cold dark matter model with |$h=0.693$| and |$\Omega _M=0.286$|⁠, which are consistent with those obtained from the Wilkinson Microwave Anisotropy Probe 9 yr data (Hinshaw et al. 2013). We use the Chabrier (2003) stellar initial mass function and the AB magnitude system (Oke & Gunn 1983). When we refer to figures or tables in this letter, we designate their initials by capital letters (e.g. Fig. 1 or Table 1) to avoid confusion with those in the literature (e.g. fig. 1 or table 1).

2 TARGETS AND DATA

Our primary goal is to study radial profiles of Pa|$\beta$| lines of massive HAEs with X-ray counterparts in S24, where X-ray sources are observed using a deep survey with the Chandra X-ray observatory (Tozzi et al. 2022). Fig. 1 shows the redshift distributions of our targets, if their spectroscopic redshifts are available, and known protocluster members with spectroscopic confirmations by previous studies on H |$\alpha$| and CO lines (Shimakawa et al. 2014; Jin et al. 2021; Pérez-Martínez et al. 2023). It indicates that the target HAEs are adequately covered by the both narrow-band filters, NB2071 on Subaru/MOIRCS in H |$\alpha$| line (Shimakawa et al. 2018) and F405N on the JWST/NIRCam (Rieke, Kelly & Horner 2005) in Pa|$\beta$| line, enabling us to analyse their Pa|$\beta$| line properties with the F405N image.

Figure 1.

Filter response curves and redshift distribution of spec-|$z$| members in the Spiderweb protocluster. The solid, dash–dotted, and dotted lines indicate the filter throughput of the NIRCam/F405N, F410M, and MOIRCS/NB2071, respectively. Spec-|$z$| confirmations with H |$\alpha$| and CO(⁠|$1\!\!-\!\!0$|⁠) lines are shown separately by pink and grey (Shimakawa et al. 2014; Jin et al. 2021; Pérez-Martínez et al. 2023). Moreover, if available, spec-|$z$| distribution of our targets (N-HAEs and X-HAEs) is shown by the magenta histogram.

Fig. 2 shows SED-inferred SFRs (SFR|$_\mathrm{SED}$|⁠) averaged over the last 100 Myr and stellar masses of HAEs in S24 (fig. 4), which are derived from the x-cigale SED fitting code (Boquien et al. 2019; Yang et al. 2020, 2022) based on the multiband photometry from various pieces of literature (Kurk et al. 2000; Miley et al. 2006; Kodama et al. 2007; Seymour et al. 2007; Koyama et al. 2013; Valtchanov et al. 2013; Dannerbauer et al. 2014; Shimakawa et al. 2018; Tozzi et al. 2022).S24 have suggested that a large fraction of X-ray-detected HAEs consists of post-star-forming galaxies well below the star-forming main sequence (Brinchmann et al. 2004; Daddi et al. 2007; Elbaz et al. 2007; Noeske et al. 2007; Salim et al. 2007) despite them having H |$\alpha$| (+[N ii]) line emissions similar to those of HAEs without X-ray counterparts at a similar stellar mass range. They also report that most of them host low-luminosity X-ray AGNs with |$L_\mathrm{X,2-10 \ keV}\lesssim 4\times 10^{43}$| erg s|$^{-1}$| that cannot be explained by star formation (Tozzi et al. 2022). Based on these previous studies, this letter focuses on 19 massive HAEs with stellar masses of |$>\!\! 2\times 10^{10}$| solar mass (M|$_\odot$|⁠), except for the Spiderweb radio galaxy, among 58 HAEs covered by our JWST/NIRCam images (S24, fig. 2). They are divided into 8 HAEs with X-ray counterparts and 11 HAEs without X-ray detection for comparison (Fig. 2). The identification numbers in the respective samples are [40, 46, 48, 55, 58, 71, 77, 83] with X-ray detection and [9, 15, 22, 28, 30, 35, 39, 54, 80, 82, 93] without X-ray detection (S24, table A).¹ For convenience, these two samples are referred to as ‘X-HAEs’ and ‘N-HAEs’ hereafter. Each of them exhibits similar properties in SFR|$_\mathrm{SED}$| and stellar mass within each sample, except an X-HAE (ID = 58) showing significantly higher SFR than the other X-HAEs (Fig. 2). However, it should be noted that the conclusions of this letter remain the same even if this X-HAE is not considered. In that case, the central Pa|$\beta$| component of the composite Pa|$\beta$| radial profile in X-HAEs (Section 3) becomes slightly weaker.

Figure 2.

SED-based SFRs versus stellar masses for HAE members of the Spiderweb protocluster (coloured symbols) and reference samples from the GOODS-S field at |$z=$| 1.5–2.5 (grey symbols) in S24. Galaxies with and without X-ray counterparts are represented by circles and squares, respectively. We adopt 11 N-HAEs (yellow) and 8 X-HAEs (magenta) with |$M_\star > 2\times 10^{10}$| M|$_\odot$| covered by the NIRCam imaging. The star symbols are based on the stacking analysis, where their stellar masses are reproduced from median flux densities and their SFRs are based on dust-corrected Pa|$\beta$| fluxes (see Section 3).

The Pa|$\beta$| line image (F405N) and corresponding continuum image (F410M) are obtained by the JWST/NIRCam at the long wavelength channel through the Cycle-1 GO 1572 programme, in parallel with F115W and F182M imaging at the short wavelength channel (Pérez-Martínez et al. 2024; Shimakawa et al. 2024a). The total integrations are 63 and 21 min, respectively. The typical |$3\sigma$| flux limit in the narrow-band selection is |$\sim\!\! 2\times 10^{-18}$| erg s|$^{-1}$| cm|$^{-2}$|⁠, which is comparable with the H |$\alpha$| + [N ii] flux limit of |$\sim\!\! 3\times 10^{-17}$| erg s|$^{-1}$| cm|$^{-2}$| in the Subaru/MOIRCS H |$\alpha$| imaging (Shimakawa et al. 2018), under the dust-free line ratio of H |$\alpha$|/Pa|$\beta =17.5$| in the case B recombination (Luridiana, Morisset & Shaw 2015). Details of the JWST/NIRCam observation and data are described in Shimakawa et al. (2024a). Fig. 2 also plots the 3D-HST sources at |$z=$| 1.5–2.5 in the GOODS-S field (Skelton et al. 2014) for reference, where their stellar masses and SFR|$_\mathrm{SED}$| are obtained in the same manner by S24 (section 4.1), based on compilation data from Giavalisco et al. (2004), Erben et al. (2005), Hildebrandt et al. (2006), Wuyts et al. (2008), Nonino et al. (2009), Cardamone et al. (2010), Retzlaff et al. (2010), Grogin et al. (2011), Koekemoer et al. (2011), Brammer et al. (2012), Hsieh et al. (2012), Ashby et al. (2013), Liu et al. (2017), and Luo et al. (2017).

3 RESOLVING GALAXY STAR FORMATION AND NUCLEAR ACTIVITIES

The remainder of this letter focuses on 19 massive HAEs with |$M_\star > 2\times 10^{10}$| M|$_\odot$| in the Spiderweb protocluster covered by the NIRCam imaging. Their RGB cut-outs from F410M/F182M/F115W images are shown in Fig. 3. This section attempts to resolve star formation and nuclear activities in these N-HAEs and X-HAEs at an |$\sim$|1 kpc angular resolution using their composite Pa|$\beta$| line images.

Figure 3.

RGB colour postage stamps of (a) N-HAEs and (b) X-HAEs based on JWST/NIRCam F410M/F182M/F115W band images (Pérez-Martínez et al. 2024; Shimakawa et al. 2024a).

The resultant composite Pa|$\beta$| and continuum images and their radial profiles are shown in Fig. 4. We here focus on the radial profiles owing to the lack of signal-to-noise ratios per pixel in the composite Pa|$\beta$| line images. They are based on the median stacking procedure described in Shimakawa et al. (2022) and |$1\sigma$| errors of radial profiles are determined from standard deviations of the corresponding aperture or annulus on 500 random composite backgrounds in each band. For the noise estimation, we randomly cut out images from background areas with similar weight values to individual HAEs based on the weight map. Matching weight values in the random background selection is critical because the imaging depth is not homogeneous across the survey area (S24, section 2.2). We do not rotate images in the stacking analysis to keep a consistent point spread function (PSF) within the samples. The Pa|$\beta$| surface brightness (SB|$_\mathrm{Pa\beta }$|⁠) and stellar continuum at |$\lambda _\mathrm{rest}=1.3$||$\mu$|m (SB|$_\mathrm{cont}$|⁠) within each annulus are estimated as follows:

$$\begin{eqnarray} \mathrm{SB_{Pa\beta }} = \frac{1}{A}\frac{\Delta _\mathrm{NB}(f_\mathrm{NB}-f_\mathrm{BB})}{1-\Delta _\mathrm{NB}/\Delta _\mathrm{BB}}, \end{eqnarray}$$

(1)

$$\begin{eqnarray} \mathrm{SB_{cont}} = \frac{1}{A}\frac{f_\mathrm{BB}-f_\mathrm{NB}\Delta _\mathrm{NB}/\Delta _\mathrm{BB}}{1-\Delta _\mathrm{NB}/\Delta _\mathrm{BB}}, \end{eqnarray}$$

(2)

where A is an area of each annulus in arcsec|$^2$|⁠, and |$f_\mathrm{NB}$| and |$f_\mathrm{BB}$| are flux densities within the corresponding annulus in the F405N and F410M bands, respectively. We adopt filter bandwidths of |$\Delta _\mathrm{NB}$| = 460 Å and |$\Delta _\mathrm{BB}$| = 4360 Å. We also confirm that their PSFs are consistent with each other based on composite images of unresolved point sources (Fig. 4).

Figure 4.

Radial profiles of Pa|$\beta$| lines (top), rest-frame 1.3 |$\mu$|m continua (middle), and the rest-frame EWs (bottom). The radial profiles are based on the composite Pa|$\beta$| and continuum images embedded in each panel, which are produced using the arcsinh stretch for the sake of visibility (the white circles indicate |$r=$| 10 kpc). The yellow and magenta symbols indicate median radial profiles of N-HAEs and X-HAEs, respectively. The colour regions depict |$1\sigma$| uncertainties and inverse triangles indicate |$2\sigma$| upper limits. The grey dashed and dotted lines show composite PSFs in F405N and F410M, respectively. The dash–dotted lines are the possible PSF components decomposed from the continuum images (see Section 4).

Fig. 4 indicates that radial profiles of surface brightness are clearly different between N-HAEs and X-HAEs, particularly at a radius of |$r=$| 2–5 kpc. The composite N-HAEs show disc star formation traced by Pa|$\beta$| from |$\sim\!\! 2$| to |$\sim\!\! 8$| kpc, and a moderate increase in Pa|$\beta$| equivalent width (EW) is consistent with the inside-out growth observed in typical star-forming galaxies (e.g. Nelson et al. 2012, 2016; Shen et al. 2024). On the other hand, we do not observe any clear sign of extended Pa|$\beta$| component above |$2\sigma$| levels in X-HAEs. Rather, there is an unresolved Pa|$\beta$| emission in the centre, while their composite stellar radial profile is resolved (more extended) relative to the PSF. The central point source dominates the entire Pa|$\beta$| flux in X-HAEs, and it is reasonable to consider that it originates from X-ray AGNs rather than star formation (Tozzi et al. 2022). Thus, the comparison results of radial profiles in Pa|$\beta$| and stellar continuum exhibit direct evidence of significantly lower star formation in X-HAEs than those in N-HAEs, which is claimed by S24 based on the multiwavelength SED fitting. However, it is currently unclear whether an offset Pa|$\beta$| peak at |$r=0.6$| kpc seen in N-HAEs is intrinsic or not: it could be due to stronger stellar absorption in the centre, the natures of the individuals, and measurement errors in object positions. Because some N-HAEs may host Type 2 AGNs according to line diagnostics by Shimakawa et al. (2015) and Pérez-Martínez et al. (2023), there may be partial point source contributions to their stacked Pa|$\beta$| image.

Then, total Pa|$\beta$| fluxes are estimated to be |$(11.4\pm 0.2)\times 10^{-18}$| erg s|$^{-1}$| cm|$^{-2}$| for N-HAEs and |$(5.0\pm 0.2)\times 10^{-18}$| erg s|$^{-1}$| cm|$^{-2}$| for X-HAEs. We calculate their SFR|$_\mathrm{Pa\beta }$| through the Kennicutt (1998) prescription with a conversion factor of H |$\alpha$|/Pa|$\beta =17.5$| and assuming Pa|$\beta$| extinctions of |$A_\mathrm{Pa\beta }=$| 0.59 and 0.40 mag for N-HAEs and X-HAEs, respectively, based on the SED fitting for their median flux densities (S24). A possible impact of Pa|$\beta$| stellar absorption is ignored in this study (cf. EW = 1–2 Å according to Seillé et al. 2024, corresponding to 5–14 per cent of the total flux in our sample). As a result, their dust-corrected SFR|$_\mathrm{Pa\beta }$| are estimated to be |$59\pm 9$| and |$<\!\! 19$| M|$_\odot$| yr|$^{-1}$| in N-HAEs and X-HAEs, respectively, which are consistent with those inferred from the SED fitting (Fig. 2). We here adopt |$2\sigma$| upper limit for X-HAEs as we do not detect a residual Pa|$\beta$| component other than the point source in their composite Pa|$\beta$| radial profile (Fig. 4). The total Pa|$\beta$| flux is tentatively adopted for N-HAEs, although they could include AGN contributions. Multiple line information is required for individual sources to resolve their possible AGN impacts and better constrain their SFRs individually.

4 IMPACT OF AGN FEEDBACK ON STAR FORMATION

We find that star formation is significantly suppressed in massive HAEs hosting X-ray AGNs by comparing their spatially resolved Pa|$\beta$| radial profile with those of the control HAE sample without X-ray counterparts. Thus, this section discusses the obtained results in light of AGN feedback. Today, the BH mass is considered to be a proxy for the total amount of AGN feedback energy, which is roughly proportional to the integral of BH accretion rates. The cumulative feedback energy budget is thought to be critical in regulating star formation quenching for massive galaxies in cosmological hydrodynamic simulations (e.g. Terrazas et al. 2020; Wellons et al. 2023). It should be borne in mind that this scenario is still debated (Garza et al. 2024; Wang et al. 2024). However, as we cannot derive BH masses for our sample from the current data, we discuss AGN feedback instead by adopting their stellar potential values, |$\phi _\star \equiv M_\star /r_\mathrm{eff}$|⁠, as suggested by Terrazas et al. (2020), Piotrowska et al. (2022), and Bluck et al. (2023, 2024). This is based on the assumptions that the BH mass is tightly correlated with the central velocity dispersion, that is, the gravitational potential (e.g. Ferrarese & Merritt 2000; Gebhardt et al. 2000; Gültekin et al. 2009; Kormendy & Ho 2013), and the stellar potential approximates the latter quantity (Bluck et al. 2023, section 4.2). We estimate circularized effective radii (⁠|$r_\mathrm{eff}$|⁠) for our samples with galfit (version 3.0.5; Peng et al. 2010) based on the F410M imaging data with the stacked PSF image (⁠|$\lambda _\mathrm{rest}=1.3$||$\mu$|m). In addition to a Sérsic profile (Sérsic 1963, 1968), a PSF model is included in the parameter fitting to consider a possible AGN contribution (Fig. 5), and confirm larger effects of additional point source on the size measurements in X-HAEs compared to N-HAEs. For reference, point source contributions are estimated to be |$f_\mathrm{PSF}=4\pm 1{{\ \rm per\, cent}}$| in the stacked N-HAEs and |$12\pm 2{{\ \rm per\, cent}}$| in the stacked X-HAEs (see also Figs 4 and 5).

Figure 5.

Model fitting to the continuum images (shown using the arcsinh stretch). From above, the fitting results on the composite images of N-HAEs and X-HAEs are shown. From left to right, each column represents the composite images, best-fitting models, residual images, and decomposed PSFs, respectively, with a linear stretch in the same flux range (see also Fig. 4).

Fig. 6 shows the specific SFR (SSFR) against the stellar potential for our targets, suggesting that X-HAEs with low SSFRs tend to have higher stellar potentials compared to N-HAEs. The anticorrelation is also supported by Spearman’s rank correlation test (⁠|$\rho _\mathrm{ s}=-0.62$| with |$p$|-values |$=0.005$|⁠). For the composite samples, we measure SSFRs from dust-corrected SFR|$_\mathrm{Pa\beta }$| obtained in Section 3. Advanced mass concentrations in X-HAEs suggest that they may host more massive BHs and thus be injected by larger amounts of AGN feedback energies in their formation histories. Altogether, we conclude that X-HAEs may be at the stage of AGN quenching in the galaxy–BH co-evolution while emitting weak signs of AGNs in H |$\alpha$| + [N ii], Pa|$\beta$|⁠, and X-ray before permanent quenching. At the moment it is restricted to Pa|$\beta$| and continuum profiles, but resolving ionization mechanisms, e.g. by observing multiple lines with an integral field unit, may lead to greater understanding of these AGN host galaxies.

Figure 6.

SSFRs versus stellar potentials (⁠|$\phi _\star$|⁠) for the selected HAEs with |$M_\star > 2\times 10^{10}$| M|$_\odot$|⁠. The symbols in the figure are the same as those in Fig. 2 but grey colours are based on size measurements using only Sérsic components without additional point source contributions. SSFRs of the composite N-HAEs and X-HAEs are based on dust-corrected SFR|$_\mathrm{Pa\beta }$|⁠. Spearman’s rank correlation coefficient (⁠|$\rho _\mathrm{ s}=-0.62$|⁠) and |$p$|-values (⁠|$=\!\!0.005$|⁠) are appended in the top right corner.

5 CONCLUSIONS

We study radial profiles of Pa|$\beta$| and rest-frame near-infrared stellar continua for massive HAEs that host X-ray AGNs (X-HAEs) and not (N-HAEs) in the Spiderweb protocluster at |$z=2.16$|⁠, based on the NIRCam images and the source catalogue from our previous work (S24). The highlights of this letter are summarized as follows:

• We find that Pa|$\beta$| radial profiles are clearly different between N-HAEs and X-HAEs in the sense that AGNs would dominate Pa|$\beta$| line emissions in X-HAEs. This provides direct evidence of low star formation in X-HAEs claimed by our previous study via multiwavelength SED fitting (S24).

• We also investigate the structural parameter, |$\phi _\star$|⁠, which can be used as an observable proxy of the gravitational potential. We then confirm that X-HAEs tend to have higher stellar potentials with lower SSFRs, further supporting the AGN quenching scenario to explain their low star formation in the outskirts of massive galaxies.

The Spiderweb protocluster is an ideal test bed for understanding the formation of the cluster red sequence at the peak epoch of cosmic star formation and BH growth. High-quality NIRCam images enable various detailed analyses that would address transitions from star-forming to red sequence galaxies and their environmental dependence by combining valuable archive data. Although we leave many of them to future research, this letter successfully illustrates some of such remarkable capabilities as an early report.

ACKNOWLEDGEMENTS

We thank the anonymous referee for useful comments. This work is based on observations made with the NASA/ESA/CSA JWST. The data were obtained from the Mikulski Archive for Space Telescopes at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-03127 for JWST. The observation is associated with program #1572 in cycle 1. This work is in part based on observations taken by the 3D-HST Treasury Program (GO 12177 and 12328) with the NASA/ESA HST, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS5-26555.

This work was supported by a Waseda University Grant for Special Research Projects (2023C-590 and 2024R-057) and MEXT/JSPS KAKENHI Grant Numbers 23H01219 and 18H03717. We would like to thank Editage (www.editage.com) for English language editing. TK acknowledges financial support from JSPS KAKENHI Grant Number 24H00002 (Specially Promoted Research by T. Kodama et al.). HD, JMP-M, and YZ acknowledge financial support from the Agencia Estatal de Investigación del Ministerio de Ciencia e Innovación (AEI-MCINN) under grant (La evolución de los cúmulos de galaxias desde el amanecer hasta el mediodía cósmico) with reference (PID2019-105776GB-I00/doi:10.13039/501100011033) and Agencia Estatal de Investigación del Ministerio de Ciencia, Innovación y Universidades (MCIU/AEI) under grant (Construcción de cúmulos de galaxias en formación a través de la formación estelar oscurecida por el polvo) and the European Regional Development Fund (ERDF) with reference (PID2022-143243NB-I00/doi:10.13039/501100011033). JMP-M acknowledges funding from the European Union’s Horizon-Europe research and innovation programme under the Marie Skłodowska-Curie grant agreement no. 101106626. CDE acknowledges funding from the MCIN/AEI (Spain) and the NextGenerationEU/PRTR (European Union) through the Juan de la Cierva-Formacion program (FJC2021-047307-I). YZ acknowledges the support from the China Scholarship Council (202206340048), and the National Science Foundation of Jiangsu Province (BK20231106). This work made use of the following tools: numpy (Harris et al. 2020), matplotlib (Hunter 2007), topcat (Taylor 2005), astropy (Astropy Collaboration 2013, 2018, 2022), and pandas (Reback et al. 2022).

DATA AVAILABILITY

The targets are selected from the publicly available source catalogues in Shimakawa et al. (2018, 2024b). The original science frames are stored in and can be retrieved from the archive system termed the Barbara A. Mikulski Archive for Space Telescopes (MAST).²

Footnotes

1

ID = 28 was regarded as an X-HAE in Tozzi et al. (2022) and S24, whereas Naufal et al. (2024) noted that its neighbouring source is the actual X-ray counterpart.Another N-HAE (ID = 89) is also covered, but was excluded from the sample as it is likely to be a background emitter (S24).

2

doi:10.17909/vx25-q902