Euclid Quick Data Release (Q1). The metallicity of the intracluster light

Introduction

Galaxy clusters are massive galactic megacities that contain hundreds to thousands of galaxies located within an underlying structure of dark matter, referred to as the dark matter halo. Near the centre of each galaxy cluster is the brightest central galaxy (BCG), which is generally a massive, radially extended, elliptical galaxy, typically with little star formation, except for the case of cooling flow clusters (e.g. [101, 67, 107, 65]).

Unlike similarly massive galaxies, BCGs are particularly radially extended (e.g. [52, 69]) and surrounded by a faint diffuse halo of light commonly referred to as intracluster light (ICL; [115, 116]). Both stacked observations ([114, 112, 15, 3]) and targeted deeper individual cluster observations with long exposure times ([27, 49]) have found that the ICL extends to megaparsec scales. As ICL observations become more common, we have been able to study the ICL as a statistical population (e.g. [62, 47, 48]) and constrain its assembly.

Understanding how the ICL assembles is crucial because it encodes both the recent and long-term growth of galaxy clusters in ways not captured by galaxy light alone or using observations linked to the intracluster medium. The ICL preserves spatial, kinematic, and chemical signatures that record when mass was accreted, and from which progenitors, and along which directions, while also providing information about the overall gravitational potential of the cluster ([81]). Among these signatures, the metallicity and age of the ICL are especially informative, constraining both formation channels and progenitor populations.

Metallicity encodes the chemical enrichment history of a stellar population and is linked to the stellar mass and star-formation history of the galaxies from which the stars were formed ([46, 60]). Since galaxies have radial metallicity gradients (e.g. [24, 63, 7, 66, 113]), stars stripped from the outskirts of massive, metal-rich satellite galaxies will impart a low metallicity on the ICL, as will stars from low-mass, metal-poor dwarf galaxies. In contrast, if the metal-rich cores of massive satellite galaxies merge with the ICL, they will impart a high metallicity on the ICL (e.g. [85, 18, 19]). Examining how the metallicity of the BCG and ICL varies with radius therefore clarifies the progenitor populations and provides a key observable for testing competing models of ICL formation. Models dominated by the tidal stripping of satellite galaxies result in a metallicity gradient ([19, 77, 12]), whereas those driven by mergers with the BCG mix the stellar population and flatten the metallicity gradient ([19, 82]).

In the last two decades, cosmological hydrodynamical simulations (e.g. IllustrisTNG, EAGLE) and semi-analytic models have predicted that the ICL should be systematically more metal-poor than the BCG at all epochs and detect a negative metallicity gradient, such that when measured radially outwards from the BCG, the metallicity decreases (e.g. [18, 19]) Cui et al. 2014; Ahvazi et al. 2024; Mayes et al. 2026). Moreover, these simulations and models predict ICL assembling earlier from low-mass satellites and later from more intermediate-mass systems as clusters grow [84, 94, 18]. Specifically, recent semi-analytic work finds that the typical progenitors of the ICL shift from $\log_{10}(M_\star/M_\odot) \approx9$–10 at early times to $\log_{10}(M_\star/M_\odot) \approx10$–11 at $z \lesssim1$, naturally producing negative colour and metallicity gradients in the combined BCG and ICL (BCG+ICL) systems due to dynamical friction, such that the more massive galaxies fall farther into the core of the cluster before being stripped [19]. As a result of these simulation studies, there are several predictions about the metallicity of the ICL that can be addressed using observational data: (i) the ICL (at $\approx100$ kpc) should be $\approx0.2$–0.4 dex more metal-poor than the BCG [19, 2, 73]; (ii) the BCG+ICL metallicity gradient should be negative and persist to $\gtrsim200$ kpc [2, 4, 73]; (iii) the characteristic ICL metallicity should evolve mildly with redshift as the dominant progenitor mass scale increases after $z \approx1$ [85, 18]; and (iv) clusters with a larger contribution from stripped low-mass satellites should show steeper (more negative) metallicity gradients than systems where the ICL is built predominantly via major mergers with the BCG, a trend testable via joint gradients and satellite galaxy mass functions [18, 19]. Confronting these predictions with spatially resolved metallicities from deep imaging provides a stringent check on ICL assembly pathways and their redshift evolution. Since ICL can account for between 10% and 35% of the stellar mass within a cluster (e.g. [94, 57, 91, 55, 48]), incorporating its metallicity is also essential for estimating the chemical-enrichment budget of the intracluster medium.

The simulation predictions for the metallicity of the ICL have become increasingly testable over the last decade, in part because photometric colours can be used to infer the metallicity of the ICL. Both stacked measurements [114, 112, 110] and individual observations (e.g. [79, 80, 78, 25, 26, 83, 93, 47]) have found that the core of the BCG is redder than the outskirts, which has led some prior observational studies to suggest that the ICL (or more simply the BCG outskirts) is likely metal-poor compared to the BCG core [68, 74, 89, 56, 92]. However, in optical photometry a degeneracy exists between the metallicity and the age of the stellar population, which makes it unclear whether this colour gradient results from a metallicity gradient, an age gradient, or a combination of the two in the stellar population of the ICL. Hubble Space Telescope (HST) studies of the Frontier Fields clusters find average ICL metallicities of $\log_{10}(Z/Z_\odot) \approx-0.3$ to $-0.6$ and clear negative gradients from super-solar values in the BCG-dominated regions to sub-solar values at larger radii [79, 83, 26, 80], while spectroscopic work in nearby clusters (e.g. Coma and $z \approx0.05$ systems) report predominantly older, metal-poor ICL consistent with stripping from lower-mass progenitors [75, 29, 31, 30, 56] (Gu et al. 2020). More recently, deep optical/near-infrared (NIR) observations extend these trends to hundreds of kiloparsecs, with colour-to-metallicity conversions indicating gradients from roughly solar inside $\approx10$ kpc to $\log_{10}(Z/Z_\odot) \approx-0.7$ by $\approx200$–300 kpc in the ICL-dominated regime [32] (Kluger et al. 2025).

Beyond tracing ICL assembly, a well-constrained ICL metallicity, along with the estimated age of the stellar population, can reduce the uncertainty in the stellar mass estimates of the BCG and ICL. The stellar mass of the ICL strongly correlates with the cluster halo mass [114, 53, 47, 48, 20] and simulations have shown that the radial density profile of dark matter in the halo scales with that of the ICL [5, 20, 108, 14]. Thus, constraining ICL metallicity and the age of the stellar population will also optimise the stellar mass measurement and the utility of the ICL as a tracer of the dark matter halo.

Here, we present an analysis using the Euclid Quick Data Release (Q1; Euclid Quick Release Q1 2025) observations of ICL measurements in 168 galaxy clusters with $\log_{10}(M_{\rm halo}/M_\odot) > 14.05$ over the redshift range $0.1 < z < 0.75$ identified using the redMaPPer cluster finding algorithm as part of the Dark Energy Survey (DES; [96, 1]). We chose the DES-redMaPPer cluster catalogue because it provides BCG identification, which is vital to our ICL measurements. However, we note that our results are not dependent on our choice of catalogue and future analyses could be done using Euclid cluster catalogues, which will extend out to much higher redshifts ([36]). In this analysis, we use the combined NIR and optical data from Euclid to measure the colours of the ICL and examine how these colours vary as a function of redshift and radial extent. The remaining sections are as follows. In Sect. 2, we discuss the Q1 data and the DES-redMaPPer cluster catalogue used in this analysis. In Sect. 3, we discuss how we measure the flux and colour of the ICL. In Sect. 4, we present the colours of the ICL. In Sect. 5, we discuss what the NIR colours can tell us about the formation redshift and metallicity of a stellar population. In Sect. 6, we discuss how the colours of the ICL relate to the metallicity and age of the stellar population. We conclude in Sect. 7.

In this analysis, we assume a flat $\Lambda$ cold dark matter Universe, with $\Omega_{\rm m} = 0.30$, $\Omega_\Lambda= 0.70$, and $H_0 = 100 h$ km s$^{-1}$ Mpc$^{-1}$, with $h = 0.7$. Additionally, throughout this analysis we assume a solar metallicity of 0.019 [106] and note that all of the radial extent measurements for the annuli of the ICL are in physical distances.

Data

Euclid Q1 data

Euclid is a 1.2 m telescope with a 0.54 deg$^2$ field of view [41]. The faint detection limit, design of the wide-field telescope, and tight control on scattered light make Euclid ideal for studying the low-surface-brightness (LSB) Universe [43, 37, 23, 71, 76, 98, 105], in particular the ICL [35, 32, 61]. The Q1 data release [33] consists of $\approx63.1$ deg$^2$ in three deep fields: Euclid Deep Field North (EDF-N), Euclid Deep Field South (EDF-S), and Euclid Deep Field Fornax (EDF-F). Each field was observed in a single observation taken to the depth of the Euclid Wide Survey [43]. The Q1 observations were taken in four photometric wavebands ($I_{\rm E}$, $Y_{\rm E}$, $J_{\rm E}$, and $H_{\rm E}$) using the VIS ([38]) and NISP ([39]) instruments. The data used in this analysis were made publicly available as part of the Q1 data release ([33]).¹

In this analysis, we focus only on the data from EDF-S and EDF-F, which are 28.1 deg$^2$ and 12.1 deg$^2$, respectively, and are covered entirely by DES. To obtain images that allow us to recover ICL signal, we followed the process described in Euclid Collaboration: Bamford et al. [34]. We began with the Level 2 (LE2) single-exposure products generated by the Science Ground Segment (SGS) VIS and NIR pipelines ([40, 42]). The calibrated exposures were then post-processed including the removal of the residual atemporal background signal, correction of time-dependent instrumental artefacts (e.g. ‘tartan’ patterns), masking of persistence features (only for NIR images), resampling of individual exposures to a common pixel scale (0.3 arcsec pixel$^{-1}$), background matching across detector chips, and coaddition of the resampled exposures into a 30′ × 30′ image for each cluster in the $I_{\mathrm{E}}$, $Y_{\mathrm{E}}$, $J_{\mathrm{E}}$, and $H_{\mathrm{E}}$ bands. In addition, we produce a coadded $YJH_{\mathrm{E}}$ image by combining the $Y_{\mathrm{E}}$, $J_{\mathrm{E}}$, and $H_{\mathrm{E}}$ images weighted by their relative bandwidths. This additional processing aims to suppress large-scale instrumental background structures while preserving faint extended emission. We note that reprocessing data for LSB observations is a common practice when using large cosmological surveys, such as SDSS and DES. These surveys have different scientific goals than LSB science, so they generally apply a local background subtraction that can remove the diffuse light surrounding bright galaxies (e.g. [9, 107, 111]).

DES-redMaPPer cluster sample

In this analysis, we use the DES Y3 redMaPPer cluster catalogue version 6.4.22+2 ([1]),² based on DES Year 3 Gold photometric data ([102]). This version of the redMaPPer cluster catalogue contains over 21 000 galaxy clusters with a redMaPPer membership probability-weighted total richness, $\lambda$, greater than 20 ([95, 96]), which corresponds to a halo mass lower limit of ≈ $10^{14.05}\,M_{\odot}/h$ ([1]).

redMaPPer is a probabilistic cluster finding algorithm that identifies galaxy clusters and their member galaxies as overdensities in the colour–magnitude phase space. The photometric redshift provided by redMaPPer for each cluster is determined based on a fit to the red sequence. redMaPPer aims to select the luminous cluster galaxy located nearest to the centre of the gravitational potential well of the cluster as the central galaxy and provides the information for the five most probable central candidates per cluster ([96, 97]). We assume that the most likely central galaxy candidate is the BCG, although we visually confirm each of our BCGs. Based on multi-wavelength measurements, redMaPPer identifies the correct BCG in ≈ 80% of all clusters ([99, 111, 10, 58]). We note that Golden-Marx et al. [48] showed that BCG+ICL statistics are not biased by the inclusion of low-probability central galaxies.

Additionally, we note that Zhang et al. [110, 109] and Golden-Marx et al. [48] use the same cluster catalogue for their analysis of the ICL. However, because we are using the Q1 photometry, here we are limited to the overlap region between the Q1 (EDF-S and EDF-F) and DES fields, which leaves us with a subset of 168 clusters covering the redshift range 0.1 < $z$ < 0.75. In Figure 1, we plot the halo mass distribution, where $M_{\mathrm{halo}}$ refers to $M_{200m}$, estimated using the mass-richness relation given in appendix D from Abbott et al. [1], as a function of redshift. We note that the halo mass distribution of our sample is relatively uniform across the redshift range of interest. However, the redshift distribution of our sample has a deficit of clusters at $z \approx0.5$, which results from limitations in the redMaPPer cluster catalogue. Upcoming cluster catalogues from Euclid Data Release 1 (DR1) data will not have these limitations and will therefore be free of the possible bias that the cluster sample selection might have on our results.

ICL measurement pipeline

Background subtraction and masking

The first step of our analysis was to measure and subtract the background flux within our BCG+ICL images. For this, we constructed a background mask, which identifies and masks all sources in the image, which includes bright and faint satellite galaxies, bright foreground stars, artefacts, and a conservative mask of the region where the BCG and ICL fluxes are significant. In this mask, high-surface-brightness sources located outside of the BCG+ICL region were grown around each detected source by 50% to minimise the contamination of ICL from their extended halos. Within the BCG+ICL masked region, the masks are more conservatively grown by 25%, to retain as much diffuse light as possible while excluding bright compact sources. In all NIR-band images, we used the source masks generated from the coadded $YJH_E$ image, while a separate source mask was used for the $I_E$-band image.

After applying this mask, we used photutils.Background2D to detect any large-scale gradients in our background. To measure the 2D background substructure within the image, we used a variable box size of $1\ \mathrm{Mpc} \times1\ \mathrm{Mpc}$ at the redshift of the cluster. The resulting background map was first median-filtered using a $3 \times3$ grid to mitigate overestimation caused by any residual ICL contamination, which can extend beyond 1 Mpc from the BCG, and then subtracted from our image to create our measurement image.

We unmasked the region of BCG+ICL to construct masks that allow us to measure the BCG+ICL photometry. Since we combined NIR and optical photometry for the Euclid colours, we combined the $YJH_E$ mask with the $I_E$ mask to create a more conservative $IYJH_E$ measurement mask, which removes bright sources and potential artefacts from our BCG+ICL signal.

Construction of the surface brightness profiles

The key measurement in this analysis is the colour of the ICL. For this colour to be physically meaningful, we required each photometric measurement to cover the same spatial region and have the same measurement mask applied (the final measurement mask). To construct the surface brightness (SB) profiles, we used the coadded $YJH_E$ image to extract the AutoProf isophotal shape profiles of the BCG+ICL (Stone et al. [104]).

The isophotes were fixed to have semi-major axes starting at 1 pixel and increase in radius by 0.15 dex for subsequent ellipses.

To run AutoProf, we first provided a fixed centre, given by the DES-redMaPPer cluster catalogue. AutoProf applies the measurement mask ($IYJH_E$ mask) to the image and then estimates the position angle (PA) and ellipticity of each isophote using a Fourier analysis of the flux within circular apertures. Starting from the initial apertures, for each isophote, the PA and ellipticity are iteratively refined using a regularisation approach, which continues until convergence is achieved (Stone et al. [104]). Thus, the final output is a series of isophote profiles with ellipticity and PA, computed out to where the signal-to-noise ratio (S/N) drops to 1. Beyond this radius the ellipticity and PA of the ellipses are fixed to the last robustly measured values. The typical accuracy of the PA and ellipticity is $10^\circ$ and 0.1, respectively (Euclid Collaboration: Bamford et al. [34]). To get our final photometric measurements, we used the elliptical isophotes defined by AutoProf to extract consistent 1D SB profiles across all four wavebands by measuring the median sigma-clipped flux within each annulus.

ICL flux measurements and uncertainties

Using the SB profiles, we then integrated the 1D profiles (assuming an elliptical shape based on the AutoProf isophotal annuli) to measure the flux contained within each annulus. We then interpolated over those fluxes to estimate the total flux contained in elliptical annuli measured in fixed physical sizes as opposed to fixed angular sizes to compare clusters across redshift space. Based on the red sequence estimated photometric redshifts from redMaPPer, we measured the flux in the following physical annuli: 0–10 kpc, 10–30 kpc, 30–50 kpc, 50–100 kpc, 100–150 kpc, 150–200 kpc, 200–300 kpc, 300–400 kpc, 400–500 kpc, and 500–600 kpc. This process was then repeated for each cluster. We did not apply a Galactic extinction correction to our flux (or colour) measurements because, as shown in Euclid Collaboration: Tucci et al. [44], the extinction corrections are less than 0.05 in magnitude for EDF-S and EDF-F, which is smaller than the uncertainty in our measurements.

To estimate the uncertainty of our flux measurement within each of our selected apertures, we performed the previously described analysis on two $2.5^\circ\times2.5^\circ$ sky patches, one for each deep field. We used the same background box size for the sky patch as for each cluster, so each cluster has a bespoke uncertainty in each annulus. After subtracting the background, for each field we measured the photometry centred on 1000 random points within each sky patch, with elliptical apertures (with an ellipticity of 0.5) extending out to 1000″ with the same spacing in arcseconds as our ICL measurements as described in Euclid Collaboration: Bamford et al. [34]. As for the ICL measurements, we interpolated these measurements to measure the light within the same physical annuli as described in the previous section. Using the 1000 SB profiles, we then measured the standard deviation within each radial bin to determine the background noise levels for each physical annulus. We note that the noise properties on large scales are dominated by systematic background variation, not Poisson noise, and therefore the additional masked area due to the satellite galaxies in each cluster, relative to the control field, does not impact our uncertainty estimate.

Colour of the BCG+ICL

BCG+ICL colour in individual clusters

As previously described, for each of the DES-redMaPPer clusters, we measured the flux as well as the associated uncertainty in the four wavebands and three colours ($Y_E - H_E$, $J_E - H_E$, and $I_E - H_E$) in the ten non-overlapping physical apertures. Quantitatively, these colours are the ratios of the two fluxes, and the uncertainty in each colour was estimated using standard error propagation based on the measured flux ratio and the uncertainty in each respective flux.

In Figure 2 we present the colour as a function of redshift for the first six radial apertures. To be included in this figure, we required that the S/N of the flux in both bands used to measure the colour be $>10$, which is why we excluded the outer four bins as there are very few clusters that meet this criterion. We provide the number of clusters that meet these criteria in the lower left corner of each sub-plot of Figure 2.

We can robustly measure the colour within 50 kpc for $>90\%$ of all the clusters. However, the number of clusters for which we can measure robust colours decreases significantly out to larger radii and is less than 10% by 200 kpc. We note that at such large radii the ICL is preferentially detectable among both higher-halo-mass and lower-redshift clusters, in agreement with some of the predictions from Euclid Collaboration: Bellhouse et al. [35]. To enhance our understanding of whether there is any trend between the colours, we take a weighted mean of the photometric colour within four redshift bins (with median redshifts of 0.30, 0.39, 0.56, 0.69), shown as unfilled circles. The black error bars represent the 16th and 84th percentiles of the colour measurements within the redshift range at that aperture, and therefore highlight the scatter in the BCG+ICL colour between individual clusters at a certain redshift. We see a strong consistency in the colours of the BCGs within each redshift bin within the inner 30 kpc. This implies that the stellar populations of the BCGs within these different clusters are likely to be similar. However, at larger extents, we see a larger scatter in the colours of the ICL in different clusters within the same redshift bin. The scatter in colours is significantly larger than the individual error bars on the colour for each cluster. This spread in colour implies that the stellar population of the ICL differs from cluster to cluster, which suggests that individual clusters, or small cluster samples, should not be used to characterise the average properties of the ICL population.

Average BCG+ICL colour

To avoid the bias that is caused by looking at individual clusters, and to extend our ICL colour measurements out to larger radii, we stacked our 1D light profiles in redshift bins. The flux in each stack was constructed from the mean of 42 clusters in the following redshift intervals: $0.1 < z < 0.348$, $0.348 \le z < 0.432$, $0.432 \le z < 0.63$, $0.63 \le z < 0.75$. We did not apply any sigma clipping in our measurements. To measure the uncertainty in the average ICL colour, we performed 1000 bootstrap resamplings of our clusters and used the 16th and 84th percentile values as our uncertainties. To ensure that we were not biased by any particular cluster, we only retained radial colour data when more than 80% of the bootstrap resampled measurements had fluxes in both bands that had a $S/N > 10$. For the inner 150 kpc where the weighted means and the stacks overlap, the two values provide consistent results. In Figure 3, we present the 1D stacked measurements and uncertainties on the average colour of the BCG+ICL for each redshift bin.

The colour of the average BCG+ICL profiles varies as a function of radius for each of the different redshift intervals. The location of each point on the $x$ axis represents the midpoint of the aperture. We can measure the colour of the ICL out to 400 kpc at $z = 0.3$, 300 kpc at $z = 0.39$ in the stacked profiles, but only out to 200 kpc in the two higher redshift bins.

Figure 3 presents a clear gradient in the average BCG+ICL colour between 10 kpc and $\approx150$ kpc in each of the three colours. At larger radii, the uncertainties are large enough that the data are consistent with the ICL colour either decreasing further or flattening out. In general, we note that the BCG (0–10 kpc, 10–30 kpc) is significantly redder than the outskirts, which are associated with the ICL. The presence of a radial gradient in the BCG+ICL colour is in strong agreement with prior analyses of both individual ICL studies and ICL populations, including stacked analyses (e.g. [114, 80, 112, 26, 47]).

Since dust can redden the stellar colours, we also include a red arrow in Figure 3, which is representative of the impact of $A_V = 0.2$ mag of intrinsic extinction on each colour at each median redshift. As shown in [51], dust only contributes a small amount to the light within the ICL, as any dust in the ICM is expected to be sputtered by the X-rays from the ICM. In Figure 3, we can see that the impact of dust is quite small in the $Y_{\mathrm{E}}-H_{\mathrm{E}}$ and $J_{\mathrm{E}}-H_{\mathrm{E}}$. However, the $I_{\mathrm{E}}-H_{\mathrm{E}}$ is much more susceptible to the impact of dust and it is possible that the colours are reddened as a result of dust.

Relationship between NIR colours, metallicity, and age

Assuming that the dust content is very low, a colour gradient can be caused by either an underlying metallicity gradient, such that the stars that are bluer have a lower metallicity than the BCG, or the stars in the outskirts are younger than the old ‘red and dead’ stellar population in the core of the BCG, or a combination of the two, as shown in Montes & Trujillo [80]. Using the E-MILES stellar population synthesis (SPS) models (Vazdekis et al. [106]), which assume a Kroupa universal initial mass function (Kroupa [64]), we estimated how the colour of the stellar population depends on both formation redshift, $z_{\mathrm{form}}$, and the metallicity. In this model we assumed a single burst of star formation that occurs at $z_{\mathrm{form}}$. We note that it is possible that a more complex star-formation model is required to fully model the colour of the ICL. However, the inclusion of additional parameters, such as bursts of star formation, would likely lead to degeneracies between potential models. We calculated each model for a specific observed redshift. We note that these models do not account for specific abundance element ratios and assume that [M/H] = [Fe/H] (Vazdekis et al. [106]). Thus, we only constrained the total metallicity, not the amount of Fe.

In this analysis, using the four Euclid bands we chose to measure the $Y_{\mathrm{E}} - H_{\mathrm{E}}$, $J_{\mathrm{E}} - H_{\mathrm{E}}$, and $I_{\mathrm{E}} - H_{\mathrm{E}}$ colours. We constructed SEDs at observed redshifts of $z = 0.3$, $0.39$, $0.56$, and $0.69$ using the E-MILES SPS models (Vazdekis et al. [106]). The models have fixed stellar metallicities of $\log_{10}(Z/Z_{\odot}) = 0.2$, $0.0$, $-0.38$, $-0.68$, $-1.28$, $-1.68$, and $-2.28$, and the stellar ages were fixed such that the stars formed at a redshift of $0.5$, $1$, $2$, $5$, or $10$. We determined the three colours by convolving the SEDs with the filter transmission curves of the Euclid filters. Finally, we performed a smooth interpolation across the metallicity-formation redshift space to determine how the stellar colours depend on the parameter space at the median redshift of the four bins used in this analysis. The results are shown in Figure 4. While the colour bars in Figure 4 show the photometric colour of a given stellar population model, we also overlay colour isocontours (in white), to show how a fixed photometric colour depends on both $z_{\mathrm{form}}$ and metallicity.

For the three Euclid colours we explore, we find that there is a strong relationship between metallicity and colour, with the higher-metallicity stars having redder colours. However, there is no coherent relationship between the colours and stellar formation redshift. For $Y_{\mathrm{E}} - H_{\mathrm{E}}$ and $J_{\mathrm{E}} - H_{\mathrm{E}}$, the colour does not depend on $z_{\mathrm{form}}$ if the stellar population formed at $z_{\mathrm{form}} > 1$ (e.g. Anders et al. [6]). This can be seen as the colour contours are almost vertical lines in the metallicity-$z_{\mathrm{form}}$ space. When the stellar population formed at $z_{\mathrm{form}} < 1$, there is a relationship between colour and stellar age: the higher-metallicity stars tend to be bluer at younger stellar ages, whereas for the lowest-metallicity stars, the younger stars tend to be slightly redder in $J_{\mathrm{E}} - H_{\mathrm{E}}$, but have no dependency on age for the $Y_{\mathrm{E}} - H_{\mathrm{E}}$ colour. At $z < 0.39$, the $I_{\mathrm{E}} - H_{\mathrm{E}}$ colour is mostly independent of stellar age when the stars formed at $z_{\mathrm{form}} > 1$, but at $z_{\mathrm{form}} < 1$ the colour depends strongly on stellar formation redshift, with younger stars having bluer colours for all metallicities. At $z \ge0.56$ the $I_{\mathrm{E}} - H_{\mathrm{E}}$ colour depends both on stellar age and metallicity unless the stars have a formation redshift of $z_{\mathrm{form}} \gtrsim3$, which can be seen as the colour contours begin to tilt away from a vertical line.

In summary, if the formation redshift of the stellar population is greater than 1, then the $Y_{\mathrm{E}} - H_{\mathrm{E}}$ and $J_{\mathrm{E}} - H_{\mathrm{E}}$ colours can constrain the metallicity of the ICL. On the other hand, the $I_{\mathrm{E}} - H_{\mathrm{E}}$ colour may be able to provide some weak constraints on the stellar age.

Age and metallicity of the ICL

The $I_{\mathrm{E}} - H_{\mathrm{E}}$ colours presented in Figure 3 are very red, even for the ICL between 200 kpc and 400 kpc. By comparing these measured colours to the isocontours in the bottom panel of Figure 4, we see that these observed colours imply that the ICL stars were formed at a redshift of $z_{\mathrm{form}} > 1$. Dust extinction would impact this colour (see red arrow in Figure 3), but we do not expect dust to be significant in the ICL as X-rays from the ICM would sputter the dust grains. Therefore, the $I_{\mathrm{E}} - H_{\mathrm{E}}$ colour implies that the stellar population of the ICL is relatively old, with a formation epoch of $z_{\mathrm{form}} > 1$. This finding is supported by simulations and observations, which find that the median age of the ICL is of the order of 10 Gyr (e.g. Edwards et al. [31]; Brown et al. [12]). Given that we have constrained the stellar population to have a formation epoch of $z_{\mathrm{form}} > 1$, we are now able to use the $Y_{\mathrm{E}} - H_{\mathrm{E}}$ and $J_{\mathrm{E}} - H_{\mathrm{E}}$ colours to constrain the metallicity of the ICL (Figure 4). In Figure 3 we mark the expected colour corresponding to a range of metallicities, assuming the median redshift of each bin and that all the stars formed at $z_{\mathrm{form}} = 5$. Our conclusions do not change if we assume $z_{\mathrm{form}}$ to be in the range of $2 < z_{\mathrm{form}} < 10$.

We use the values of $Y_{\mathrm{E}} - H_{\mathrm{E}}$, $J_{\mathrm{E}} - H_{\mathrm{E}}$ and $I_{\mathrm{E}} - H_{\mathrm{E}}$ measured in each physical aperture and read off the metallicity that each colour corresponds to within the colour-metallicity space shown in Figure 4, along the line of $z_{\mathrm{form}} = 5$. We note that as a result of our choice of SPS models, we have an upper limit on the metallicity of $\log_{10}(Z/Z_{\odot}) = 0.20$. Thus, for some of the inner regions, we are unable to adequately constrain the metallicity and can only place a lower limit on the value (as shown by the red arrows in Figure 5).

In Figure 5, we present the metallicity of the BCG+ICL based on the colour measurements of the 1D stacked light profiles. The error bars on the metallicity are based on the bootstrap uncertainty shown in Figure 3. We measure a metallicity profile out to

$\approx300$ kpc for the first time using a statistical sample and see a metallicity gradient in all three colours. We note that there is some variation between the three colours and across the different redshift ranges. Generally, the $I_{\mathrm{E}} - H_{\mathrm{E}}$ and $J_{\mathrm{E}} - H_{\mathrm{E}}$ colours yield slightly higher metallicities. However, across all three colours, we detect a metallicity gradient between 10 kpc and $\approx150$ kpc such that the core of the BCG is super-solar and the ICL (beyond 100 kpc) is sub-solar. At larger radii, we are unable to fully determine the trends either a flattening in metallicity gradient or continuously decreasing metallicity is consistent within the large uncertainties. Despite this uncertainty, we conclude that using Euclid colours has allowed us to put a lower limit on the stellar age of the ICL, and detect a radial gradient in metallicity.

While the detection of a metallicity gradient is independent of our choice of SPS model, the metallicity values estimated from these models and presented here are dependent on our choice of SPS model and colour. In Appendix § 9, we show the metallicity measurements when assuming a Bruzual & Charlot [13] model, which would yield a similar gradient, but differing metallicities of the order of 0.2 dex. Thus, we emphasise that while this metallicity gradient is a real feature of the BCG+ICL, including more parameters in our SPS models (such as multiple bursts of star formation) and additional optical photometry bands is needed to improve estimation of the metallicity measurements. Furthermore, it is possible that spectroscopic measurements of the ICL may be needed to break the degeneracies that exist between more complex star-formation histories.

In Figure 5, we overlay the 2D projected luminosity-weighted mean metallicity profiles for the BCG+ICL in the TNG100 simulation of the IllustrisTNG Project (in blue; Nelson et al. [88]; Pillepich et al. [90]; Springel et al. [103]; Nelson et al. [87]; Naiman et al. [86]; Marinacci et al. [70]) and the Hydrangea simulation (in purple; Bahé et al. [8]). Hydrangea is a zoom-in simulation of galaxy clusters that have been evolved within the EAGLE galaxy formation model (Schaye et al. [100]). For both simulations, the measurements are based on a single 2D projection for each cluster. To separate ICL particles from those of the BCG and galaxies, for TNG, we followed Brown et al. [11] and applied a threshold in instantaneous local stellar density to identify ICL emulating AdaptahOP (Dubois et al. [28]), while for Hydrangea, we used an unbinding threshold based on the instantaneous kinetic and potential energies. Though these are differing ICL definitions, both simulations show a clear gradient between the metallicity of the BCG core and that of the ICL, in agreement with our observations. For the TNG100 measurement, we measure the mean luminosity weighted metallicity based on the 15 most massive halos [at $z = 0$ the halo mass range is $13.95 < \log_{10}(M_{\mathrm{halo}}/M_{\odot}) < 14.6$]. For Hydrangea, we measure the mean luminosity weighted metallicity based on 54 clusters with $\log_{10}(M_{\mathrm{halo}}/M_{\odot}) > 14.0$. We see that the BCG (inner 10 kpc) metallicity is slightly higher in Hydrangea than TNG; however, we note that the metallicities from Hydrangea can be biased high (Schaye et al. [100]), which may result in this difference.

In Sect. § 1, we noted a few predictions about the metallicity of the ICL from simulations and models. (i) We similarly find that the ICL (the region beyond 100 kpc) is $\approx0.4$ dex more metal-poor than the BCG (inner 30 kpc). (ii) We detect a persistent metallicity gradient out to $\approx150$ kpc, in agreement with what has been observed in TNG and Hydrangea, as shown in Figure 5. However, we note that beyond 150 kpc, we are unable to conclude that a gradient is present. (iii) In Figure 6, we compare the metallicity gradient for the simulations and the $J_{\mathrm{E}} - H_{\mathrm{E}}$ measurements from Euclid for the lowest and second-highest redshift bins³. There is a slight evolution in the metallicity gradient in the simulations, such that the ICL is more metal-poor at higher redshifts. The observations agree with this trend, but there is not sufficient signal to confirm it exists in the Universe. Since this comparison suggests the possibility of evolution, we note that dividing the radius by $R_{200\mathrm{m}}$ removes much of this offset in the simulations and decreases it in the observations. As the offset does not entirely disappear in observations, this suggests the evolution of the metallicity gradient is not entirely self-similar. However, our metallicity measurements within each redshift bin at large radii are within $1\sigma$ of one another, so it remains unclear whether these results are statistically robust. Thus, using the upcoming DR1 data (1900 deg$^2$) and cluster catalogues will significantly increase the sample size, allowing us to investigate whether this change in the metallicity gradient is real. (iv) Our results show a steep (non-zero) metallicity gradient, indicating that the ICL is likely built predominantly via tidal stripping.

Comparison to previous observational results

As noted in Sect. 1, a colour gradient within the ICL has been commonly detected in both individual observations and large statistical studies, such that the ICL is bluer than the core of the BCG (e.g. [114, 80, 26, 111, 93, 47]). As a result of our stacking and tight constraints on the uncertainty of the flux measurements at large radii, we measure a colour profile out to $\approx300\,\mathrm{kpc}$ and detect a colour gradient out to $\approx150\,\mathrm{kpc}$. Beyond these radii, we are unable to adequately determine whether the gradient in colour and metallicity persists or flattens.

Before Euclid it was challenging to characterise the metallicity within the ICL because of the degeneracy that exists between the age and metallicity of the ICL when estimated using optical colours (e.g. Contini [17]). Therefore, there are only a few single target or small sample studies, most of which only extend out to $\approx200\,\mathrm{kpc}$, with which we can directly compare our results. Specifically, Montes & Trujillo [79] and Montes Montes & Trujillo [80] detect both a metallicity and an age gradient in individual ICL measurements using the HST Frontier Fields clusters, which are generally more massive than those observed in this analysis, while DeMaio et al. [25] assume that the age is fixed when detecting a metallicity gradient in their small ICL sample. Moreover, while ICL metallicity measurements are not commonplace, there is a strong consensus among spectroscopic studies that the BCG is more metal-rich (super-solar) compared to the outskirts of the ICL, which is sub-solar (e.g. [16, 29, 31, 30, 56] (Gu et al. 2020)).

In Figure 7, we compare the metallicities derived from our $J_{\mathrm{E}}-H_{\mathrm{E}}$ colours to the measurements from the literature that cover a similar redshift range ([79, 80, 25, 32]). This figure shows that our measured BCG and ICL metallicities, as well as the presence of a metallicity gradient, are in good agreement with these previous observational studies for the region of overlap. Moreover, our results are similar to those from Klug et al. (2025), who use the $J_{\mathrm{E}}-H_{\mathrm{E}}$ and $Y_{\mathrm{E}}-J_{\mathrm{E}}$ colours and estimate the metallicity of the ICL in Perseus as $\log_{10}(Z/Z_{\odot}) \approx-0.8$ at $\approx150\,\mathrm{kpc}$. Additionally, while our results agree with previous measurements, they do highlight that in order to fully characterise the metallicity profile of the ICL, large samples of clusters must be used.

Building on this analysis, there are a number of questions we plan to explore to improve our understanding of ICL formation with the upcoming data from Euclid DR1. In particular, the larger and more statistically robust dataset from DR1 will extend these stacked ICL measurements out to much larger radii. These added data should also allow us to address whether the ICL colour and metallicity are influenced by either the cluster mass as suggested by some models, simulations, and observations ([18, 19, 22, 91]) or the dynamical state of the cluster, estimated using parameters including the magnitude gap or ICL fraction (e.g. [48, 59]), further enhancing our understanding of how the underlying dark matter halo correlates with and influences the formation of the ICL.

Formation history of the ICL

The current consensus, based on observations and simulations, is that the ICL grows as a result of tidal stripping (e.g. [19, 12, 21]). Moreover, simulations suggest that the majority of ICL stars come from massive [$\log_{10}(M_{\star}/M_{\odot}) > 10.5$] satellite galaxies, though we note that the fractional mass loss from lower-mass galaxies is higher (e.g. [18, 19, 12, 11, 21]). Based on the colour gradient and sub-solar metallicity, our results strongly support the idea that the tidal stripping of satellite galaxies is the primary growth mechanism for the ICL out to $\approx150\,\mathrm{kpc}$. Using our ICL metallicities in the 200–300 kpc region, we estimated the progenitor galaxy population using the mass-metallicity relation from Gallazzi et al. [46], which is based on observations of the local Universe, to be $9.7 < \log_{10}(M_{\star}/M_{\odot}) < 10.5$. However, since the progenitor satellite galaxies are unlikely to be entirely stripped, just the metal-poor outskirts, these stellar mass estimates are lower limits, which thus align with simulations.

In this analysis, we have placed constraints on the metallicity of the ICL out to large radii. However, the NIR and limited available optical data do not allow us to strongly constrain the age of the stellar population. Thus, we present an explanation for the physical formation of the metallicity gradient that is not dependent on age. If, based on the presence of the colour and metallicity gradient between 10 kpc and $\approx150$ kpc, we assume that at least part of the ICL forms predominantly via tidal stripping, then the outskirts of these satellite galaxies will be stripped upon infall into the cluster, because they lie in a shallower part of the gravitational potential well of the cluster [72], and those stars will populate the ICL. Meanwhile, the more metal-rich satellite galaxy cores lose energy through dynamical friction (e.g. Contini [17]) and either merge with the BCG [12] or, depending on their orbits and infall velocities, remain as satellites. Following this model, there are two unique physical processes that are responsible for growing the BCG+ICL: the inner region assembles as a result of the hierarchical merging of the cores of satellite galaxies with the BCG (yielding a metal-enhanced core), while the outer region (the ICL) grows via the tidal stripping of the outskirts of satellite galaxies, in agreement with previous results [48]. However, we note that deeper data from Euclid DR1 will inform whether the colour and metallicity measurements beyond $\approx150$ kpc are consistent with this formation picture.

Conclusions

In this work, we use the Euclid Q1 observations of 168 DES-redMaPPer clusters over the redshift range $0.1 < z < 0.75$ with $\log_{10}(M_{\mathrm{halo}}/M_{\odot}/h) > 14.05$ to determine how the stellar properties (colour, age, and metallicity) of the ICL vary with radius and redshift to inform the progenitor population that forms the ICL. We summarise the primary results of this analysis as follows.

• Using individual light profiles, robust ICL colours can be measured for 75% of the clusters out to 100 kpc, but this rapidly decreases to $< 10\%$ at 200 kpc. We find that the ICL signal is predominantly detectable among both more massive and lower-redshift clusters at larger radii.

• Using the individual profiles, we see that the colours of the inner 30 kpc, which is representative of the BCG, are remarkably consistent. Beyond this extent, the colours have a large scatter. This suggests that the stellar population of the ICL has large cluster-to-cluster variation, and therefore care should be taken not to infer average ICL properties from small samples or individual clusters.

• Using stacked 1D light profiles allows us to robustly measure the ICL colours out to $\approx300$ kpc (Fig. 3) and detect a colour gradient in each of the $Y_{\mathrm{E}}-H_{\mathrm{E}}$, $J_{\mathrm{E}}-H_{\mathrm{E}}$, and $I_{\mathrm{E}}-H_{\mathrm{E}}$ colours between 10 kpc and $\approx150$ kpc. At larger radii, we are unable to consistently characterise the colour trend as the uncertainties mean the profiles are consistent with either a flattening in the trend or a continually decreasing colour with radius.

• The red $I_{\mathrm{E}}-H_{\mathrm{E}}$ colour of the outer ICL (Fig. 3) implies the ICL stellar population must have formed at a redshift earlier than $z=1$, based on the E-MILES SPS models [106].

• Using the E-MILES SPS models [106], we show that at $z_{\mathrm{form}}>1$ a NIR colour gradient can be explained by a metallicity gradient, which suggests that this trend may be independent of the age of the stellar population, potentially breaking the degeneracy that exists between colour, age, and metallicity.

• Using the metallicity–$z_{\mathrm{form}}$–colour parameter space (Fig. 4), we determined how the observed colour profiles (see Fig. 3), assuming the E-MILES SPS models [106], transform into a metallicity profile (see Fig. 5). We find that the core of the BCG has a super-solar metallicity, while the outskirts of the ICL are metal-poor. Using this stacked data, we measured how the metallicity of the BCG+ICL changes as a function of radius within four redshift bins. In each of the three colours, we detect a clear metallicity gradient out to $\approx150$ kpc. We place the first statistical constraints on the metallicity of the ICL population out to $\approx300$ kpc from the BCG. These metallicity constraints can be used to constrain galaxy formation and merger history models of the ICL and also improve ICL stellar mass estimates in future observations.

• Based on the metallicity of the ICL and the colour and metallicity gradients, our results support the idea that the ICL (at least out to $\approx150$ kpc) grows through the tidal stripping of satellite galaxies while the BCG (as supported by its enhanced metallicity) grows via mergers with the dense metal-rich cores of satellite galaxies.

We find that the NIR colours from Euclid statistically constrain the metallicity of the ICL. To constrain the age along with the metallicity, we need to sample more of the spectral energy distribution of the ICL. By combining the Euclid NIR colours with the deeper optical colours – using narrower filters – from the Vera C. Rubin observatory [54], we expect to be able to accurately determine the ages of the stellar population within the ICL, which will give us all the information needed to fully characterise the stellar population within the ICL.

Acknowledgements. The authors would like to thank the anonymous referee for their feedback on this paper, which helped to greatly strengthen the presented analysis. J.B.G.M., N.A.H., T.K., H. Gao, H. Gully, M. U., and M.B. gratefully acknowledge financial support from The Leverhulme Trust. M.M. acknowledges support from grant RYC2022-036949-I financed by the MICIU/AEI/10.13039/501100011033 and by ESF+, grant CNS2024-154592 financed by MICIU/AEI/10.13039/501100011033, and program Unidad de Excelencia María de Maeztu CEX2020-001058-M. Y.M.B. acknowledges support from UK Research and Innovation through a Future Leaders Fellowship (grant agreement MR/X035166/1) and financial support from the Swiss National Science Foundation (SNSF) under project 200021_213076. This work used the DiRAC@Durham facility managed by the Institute for Computational Cosmology on behalf of the STFC DiRAC HPC Facility (www.dirac.ac.uk). The equipment was funded by BEIS capital funding via STFC capital grants ST/K00042X/1, ST/P002293/1, ST/R002371/1 and ST/S002502/1, Durham University and STFC operations grant ST/R000832/1. DiRAC is part of the National e-Infrastructure. W.C. gratefully acknowledges Comunidad de Madrid for the Atracción de Talento fellowship no. 2020-T1/TIC19882 and Agencia Estatal de Investigación (AEI) for the Consolidación Investigadora Grant CNS2024-154838. He further acknowledges the Project PID2024-156100NBC21 financed by MICIU/AEI/10.13039/501100011033/FEDER, EU and ERC: HORIZON-TMA-MSCA-SE for supporting the LACEGAL-III (Latin American Chinese European Galaxy Formation Network) project with grant number 101086388 and the science research grants from the China Manned Space Project. F.D. acknowledges support from CNES. A.H.G. acknowledges the support of NASA ROSES Grant 12-EUCLID11-0004. L.M. acknowledges the financial contribution from the PRIN-MUR 2022 20227RNLY3 grant “The concordance cosmological model: stress-tests with galaxy clusters” supported by Next Generation EU and from the grant ASI n. 2024-10-HH.0 “Attività scientifiche per la missione Euclid- fase E”. R.R. acknowledges financial support through grants PRIN-MIUR 2020SKSTHZ and through INAF WEAVE StePS funds. This work has made use of the Euclid Quick Release Q1 data from the Euclid mission of the European Space Agency (ESA), 2025, https://doi.org/10.57780/esa-2853f3b. The Euclid Consortium acknowledges the European Space Agency and a number of agencies and institutes that have supported the development of Euclid, in particular the Agenzia Spaziale Italiana, the Austrian Forschungsförderungsgesellschaft funded through BMIMI, the Belgian Science Policy, the Canadian Euclid Consortium, the Deutsches Zentrum für Luft- und Raumfahrt, the DTU Space and the Niels Bohr Institute in Denmark, the French Centre National d’Etudes Spatiales, the Fundação para a Ciência e a Tecnologia, the Hungarian Academy of Sciences, the Ministerie van Economische Zaken, the Innovation y Universidades, the National Aeronautics and Space Administration, the National Astronomical Observatory of Japan, the Netherlands Onderzoekschool voor Astronomie, the Norwegian Space Agency, the Research Council of Finland, the Romanian Space Agency, the Swiss Space Office (SSO) at the State Secretariat for Education, Research, and Innovation (SERI), and the United Kingdom Space Agency. A complete and detailed list is available on the Euclid web site (www.euclid-ec.org/consortium/community/). The IllustrisTNG simulations were undertaken with compute time awarded by the Gauss Centre for Supercomputing (GCS) under GCS Large-Scale Projects GCS-ILLU and GCS-DWAR on the GCS share of the supercomputer Hazel Hen at the High Performance Computing Center Stuttgart (HLRS), as well as on the machines of the Max Planck Computing and Data Facility (MPCDF) in Garching, Germany.

Authors and affiliations

Euclid Collaboration: J. B. Golden-Marx⋆¹, N. A. Hatch¹, S. P. Bamford¹, H. Gao¹, T. Kolcu¹, M. Montes², H. J. Brown¹, Y. M. Bahé¹, ³, H. Gully¹, G. Martin¹, M. Urbano⁴, S. L. Ahad⁵, ⁶, C. Bellhouse¹, M. Bolzonella⁷, M. Butler¹, Weiguang Cui⁸, ⁹, ¹⁰, F. Durret¹¹, A. Ellien¹², A. H. Gonzalez¹³, M. Kluge¹⁴, L. Moscardini¹⁵, ⁷, ¹⁶, F. R. Pearce¹, R. Ragusa¹⁷, B. Altieri¹⁸, S. Andreon¹⁹, N. Auricchio⁷, C. Baccigalupi²⁰, ²¹, ²², ²³, M. Baldi²⁴, ⁷, ¹⁶, S. Bardelli⁷, P. Battaglia⁷, A. Biviano²¹, ²⁰, E. Branchini²⁵, ²⁶, ¹⁹, M. Brescia²⁷, ¹⁷, J. Brinchmann²⁸, ²⁹, ³⁰, S. Camera³¹, ³², ³³, G. Cañas-Herrera³⁴, ³⁵, V. Capobianco³³, C. Carbone³⁶, V. F. Cardone³⁷, ³⁸, J. Carretero³⁹, ⁴⁰, S. Casas⁴¹, ⁴², M. Castellano³⁷, G. Castignani⁷, S. Cavuoti¹⁷, ⁴³, K. C. Chambers⁴⁴, A. Cimatti⁴⁵, C. Colodro-Conde⁴⁶, G. Congedo⁹, C. J. Conselice⁴⁷, L. Conversi⁴⁸, ¹⁸, Y. Copin⁴⁹, A. Costille⁵⁰, F. Courbin⁵¹, ⁵², ⁵³, H. M. Courtois⁵⁴, M. Cropper⁵⁵, J.-C. Cuillandre⁵⁶, A. Da Silva⁵⁷, ⁵⁸, H. Degaudenzi⁵⁹, G. De Lucia²¹, H. Dole⁶⁰, F. Dubath⁵⁹, C. A. J. Duncan⁹, X. Dupac¹⁸, S. Dusini⁶¹, S. Escoffier⁶², M. Farina⁶³, R. Farinelli⁷, S. Ferriol⁴⁹, F. Finelli⁷, ⁶⁴, S. Fotopoulou⁶⁵, M. Frailis²¹, E. Franceschi⁷, L. Gabarra⁶⁶, S. Galeotta²¹, K. George⁶⁷, B. Gillis⁹, C. Giocoli⁷, ¹⁶, J. Gracia-Carpio¹⁴, A. Grazian⁶⁸, F. Grupp¹⁴, ⁶⁹, W. G. Hartley⁵⁹, S. V. H. Haugan⁷⁰, W. Holmes⁷¹, I. M. Hook⁷², F. Hormuth⁷³, A. Hornstrup⁷⁴, ⁷⁵, K. Jahnke⁷⁶, M. Jhabvala⁷⁷, B. Joachimi⁷⁸, E. Keihänen⁷⁹, S. Kermiche⁶², A. Kiessling⁷¹, B. Kubik⁴⁹, M. Kümmel⁶⁹, M. Kunz⁸⁰, H. Kurki-Suonio⁸¹, ⁸², A. M. C. Le Brun⁸³, S. Ligori³³, P. B. Lilje⁷⁰, V. Lindholm⁸¹, ⁸², I. Lloro⁸⁴, M. Magliocchetti⁶³, G. Mainetti⁸⁵, D. Maino⁸⁶, ³⁶, ⁸⁷, E. Maiorano⁷, O. Mansutti²¹, S. Marcin⁸⁸, O. Marggraf⁸⁹, M. Martinelli³⁷, ³⁸, N. Martinet⁵⁰, F. Marulli¹⁵, ⁷, ¹⁶, R. J. Massey⁹⁰, N. Mauri⁴⁵, ¹⁶, S. Maurogordato⁹¹, C. J. R. McPartland⁷⁵, ⁹², E. Medinaceli⁷, S. Mei⁹³, ⁹⁴, M. Melchior⁹⁵, Y. Mellier⋆⋆¹¹, ⁹⁶, M. Meneghetti⁷, ¹⁶, E. Merlin³⁷, G. Meylan³, P. Monaco⁹⁷, ²¹, ²², ²⁰, A. Mora⁹⁸, M. Moresco¹⁵, ⁷, C. Moretti²¹, ²⁰, ²², ²³, R. Nakajima⁸⁹, C. Neissner⁹⁹, ⁴⁰, R. C. Nichol¹⁰⁰, S.-M. Niemi³⁴, C. Padilla⁹⁹, S. Paltani⁵⁹, F. Pasian²¹, K. Pedersen¹⁰¹, W. J. Percival⁵, ⁶, ¹⁰², V. Pettorino³⁴, A. Pezzotta¹⁹, S. Pires⁵⁶, G. Polenta¹⁰³, M. Poncet¹⁰⁴, L. A. Popa¹⁰⁵, C. Porciani⁸⁹, L. Pozzetti⁷, F. Raison¹⁴, R. Rebolo⁴⁶, ¹⁰⁶, ¹⁰⁷, A. Renzi¹⁰⁸, ⁶¹, J. Rhodes⁷¹, G. Riccio¹⁷, I. Risso¹⁹, ²⁶, E. Romelli²¹, M. Roncarelli⁷, R. Saglia⁶⁹, Z. Sakr¹⁰⁹, ¹¹⁰, ¹¹¹, D. Sapone¹¹², B. Sartoris⁶⁹, ²¹, J. A. Schewtschenko⁹, M. Schirmer⁷⁶, P. Schneider⁸⁹, T. Schrาบback¹¹³, A. Secroun⁶², G. Seidel⁷⁶, S. Serrano¹¹⁴, ¹¹⁵, ², P. Simon⁸⁹, C. Sirignano¹⁰⁸, ⁶¹, G. Sirri¹⁶, L. Stanco⁶¹, J. Steinwagner¹⁴, P. Tallada-Crespí³⁹, ⁴⁰, A. N. Taylor⁹, H. I. Teplitz¹¹⁶, J. Tereno⁵⁷, N. Tessore⁵⁵, S. Toft¹¹⁵, ⁵⁹, R. Toledo-Moreo⁷, F. Torradeflot⁴⁰, ³⁹, I. Tutusaus¹⁴, J. Valiviita⁸¹, ⁸², T. Valenzuela²¹, A. Veropalumbo¹⁹, ²⁶, ²⁵, Y. Wang¹¹⁶, J. Weller⁶⁹, ¹⁴, C. Zacchei²¹, ²⁰, G. Zamorani⁷, F. M. Zerbi¹⁹, I. A. Zinchenko¹²⁰, E. Zucca⁷, M. Ballardini¹²¹, ¹²², ⁷, E. Bozzo⁵⁹, C. Burigana¹²³, ⁶⁴, R. Cabanac¹¹⁰, M. Calabrese¹²⁴, ³⁶, A. Cappi⁹¹, ⁷, T. Castro²¹, ²², ²⁰, ¹²⁵, J. A. Escartin Vigo¹⁴, J. García-Bellido¹²⁶, R. Maoli¹²⁷, ³⁷, J. Martín-Fleitas¹²⁸, R. B. Metcalf¹⁵, ⁷, M. Pöntinen⁸¹, V. Scotten¹¹, ¹²⁹, M. Sereno⁷, ¹⁶, M. Tenti¹⁶, M. Viel²⁰, ²¹, ²³, ²², ¹²⁵, M. Wiesmann⁷⁰, Y. Akrami¹²⁶, ¹³⁰, I. T. Andika⁶⁹, G. Angora¹⁷, ¹²¹, S. Anselmi⁶¹, ¹⁰⁸, ¹³¹, M. Archidiacono⁸⁶, ⁸⁷, F. Atrio-Barandela¹³², L. Bazzanini¹²¹, ⁷, D. Bertacca¹⁰⁸, ⁶¹, M. Bethermin⁴, A. Blanchard¹³³, ⁸³, H. Böhringer¹⁴, ⁶⁷, ¹³⁴, M. Bonici³⁶, ³⁵, S. Borgani⁹⁷, ²⁰, ²¹, ²², ¹²⁵, M. L. Brown⁴⁷, S. Bruton³⁷, B. Camacho Quevedo²⁰, ²³, ²¹, F. Caro³⁷, C. S. Carvalho¹¹⁷, Y. Charles⁵⁰, F. Cogato¹⁵, ⁷, S. Conseil⁴⁹, A. Coraudio¹³⁶, M. Coraud¹²⁷, ²¹, ²⁰, O. Cucciati⁷, S. Davini²⁶, G. Desprez¹³⁷, A. Díaz-Sánchez¹³⁸, S. Di Domizio²⁵, ²⁶, J. M. Diego¹³⁹, P. Dimauro¹⁴⁰, ³⁷, P.-A. Duc⁴, M. Y. Elkhassab²¹, ²², ⁹⁷, ²⁰, A. Enia⁷, Y. Fang⁶⁹, F. Fontanot²¹, ²⁰, A. Franco¹⁴¹, ¹⁴², ¹⁴³, K. Ganga⁹³, T. Gasparetto³⁷, V. Gautard¹⁴⁴, R. Gavazzi⁵⁰, ⁹⁶, E. Gaztanaga², ¹¹⁴, ¹⁴⁵, F. Giacomini¹⁶, F. Gianotti⁷, G. Gozalias¹⁴⁶, ⁸¹, M. Guidi²⁴, ⁷, C. M. Gutierrez¹⁴⁷, A. Hall⁹, H. Hildebrandt¹⁴⁸, J. Hjorth¹⁰¹, L. K. Hunt¹⁴⁹, J. J. E. Kajava¹⁵⁰, ¹⁵¹, Y. Kang⁵⁹, V. Kansal¹⁵³, ¹⁵⁴, D. Karagiannis¹²¹, ¹⁵⁵, K. Kivinen⁷⁹, J. Kimm⁷⁸, C. C. Kirkpatrick⁷⁹, S. Kruk¹⁸, M. Lattanzio¹²², L. Legrand¹⁵⁶, ¹⁵⁷, G. Leroy¹⁵⁸, ⁵⁹, G. F. Lesci¹⁵⁵, J. Lesgourgues⁴¹, L. Leuzzi⁷, T. I. Liaudat¹⁶⁰, S. J. Liu⁶³, A. Loureiro¹⁶¹, ¹⁶², J. Macias-Perez¹⁶³, F. Mannucci¹⁴⁹, C. J. A. P. Martins¹⁶⁴, ²⁸, L. Maurin⁶⁰, M. Miluzio¹⁸, ¹⁶⁵, A. Montoro², ¹¹⁴, G. Morgante⁷, S. Nadathur¹⁴⁵, K. Naidoo¹⁴⁵, P. Natoli¹²¹, ¹²², A. Navarro-Alsina⁸⁹, S. Nesseris¹²⁶, D. Paoletti⁷, ⁶⁴, F. Passalacqua¹⁰⁸, ⁶¹, K. Paterson⁷⁶, A. Pisani⁶², D. Potter¹⁵⁸, G. W. Pratt⁵⁶, S. Quaiⁱ¹⁵, ⁷, M. Radovich⁶⁸, G. Rodighiero¹⁰⁸, ⁶⁸, W. Roster¹⁴, S. Sacquegna¹⁶⁶, M. Sahlén¹⁶⁷, D. B. Sanders⁴⁴, S. Sarpa¹²³, ¹²⁵, ²², J. Schaye³⁵, A. Schneider¹⁵⁸, M. Schultheis⁹¹, D. Sciotti⁷⁷, E. Sellentin¹⁶⁸, ³⁵, F. Shankar¹⁶⁹, L. C. Smith¹⁷⁰, J. G. Sorce¹⁷¹, ⁶⁰, S. A. Stanford¹⁷², K. Tanidis⁷, F. Tarsitano⁹, G. Testera²⁶, R. Teyssier¹⁷⁴, S. Tosi²⁵, ²⁶, ¹⁹, A. Troja¹⁰⁸, ⁶¹, M. Tucci⁵⁹, A. Venhola¹⁷⁵, D. Vergani⁷, G. Verza⋆⋆¹⁷⁶, ¹⁷⁷, P. Vielzeuf⁶², S. Vinciguerra⁵⁰, and N. A. Walton¹⁷⁰ ⁴ Université de Strasbourg, CNRS, Observatoire astronomique de Strasbourg, UMR 7550, 67000 Strasbourg, France

⁵ Waterloo Centre for Astrophysics, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada

⁶ Department of Physics and Astronomy, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada

⁷ INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy

⁸ Departamento de Física Teórica, Facultad de Ciencias, Universidad Autónoma de Madrid, 28049 Cantoblanco, Madrid, Spain

⁹ Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK

¹⁰ Centro de Investigación Avanzada en Física Fundamental (CIAFF), Facultad de Ciencias, Universidad Autónoma de Madrid, 28049 Madrid, Spain

¹¹ Institut d’Astrophysique de Paris, 98bis Boulevard Arago, 75014, Paris, France

¹² OCA, P.H.C Boulevard de l’Observatoire CS 34229, 06304 Nice Cedex 4, France

¹³ Department of Astronomy, University of Florida, Bryant Space Science Center, Gainesville, FL 32611, USA

¹⁴ Max Planck Institute for Extraterrestrial Physics, Giessenbachstr. 1, 85748 Garching, Germany

¹⁵ Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Università di Bologna, via Piero Gobetti 93/2, 40129 Bologna, Italy

¹⁶ INFN-Sezione di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy

¹⁷ INAF-Osservatorio Astronomico di Capodimonte, Via Moiariello 16, 80131 Napoli, Italy

¹⁸ ESAC/ESA, Camino Bajo del Castillo, s/n., Urb. Villafranca del Castillo, 28692 Villanueva de la Cañada, Madrid, Spain

¹⁹ INAF-Osservatorio Astronomico di Brera, Via Brera 28, 20122 Milano, Italy

²⁰ IFPU, Institute for Fundamental Physics of the Universe, via Beirut 2, 34151 Trieste, Italy

²¹ INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy

²² INFN, Sezione di Trieste, Via Valerio 2, 34127 Trieste TS, Italy

²³ SISSA, International School for Advanced Studies, Via Bonomea 265, 34136 Trieste TS, Italy

²⁴ Dipartimento di Fisica e Astronomia, Università di Bologna, Via Gobetti 93/2, 40129 Bologna, Italy

²⁵ Dipartimento di Fisica, Università di Genova, Via Dodecaneso 33, 16146, Genova, Italy

²⁶ INFN-Sezione di Genova, Via Dodecaneso 33, 16146, Genova, Italy

²⁷ Department of Physics "E. Pancini", University Federico II, Via Cinthia 6, 80126, Napoli, Italy

²⁸ Instituto de Astrofísica e Ciências do Espaço, Universidade do Porto, CAUP, Rua das Estrelas, PT4150-762 Porto, Portugal

²⁹ Faculdade de Ciências da Universidade do Porto, Rua do Campo Alegre, 4150-007 Porto, Portugal

³⁰ European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching, Germany

³¹ Dipartimento di Fisica, Università degli Studi di Torino, Via P. Giuria 1, 10125 Torino, Italy

³² INFN-Sezione di Torino, Via P. Giuria 1, 10125 Torino, Italy

³³ INAF-Osservatorio Astrofisico di Torino, Via Osservatorio 20, 10025 Pino Torinese (TO), Italy

³⁴ European Space Agency/ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands

³⁵ Leiden Observatory, Leiden University, Einsteinweg 55, 2333 CC Leiden, The Netherlands

³⁶ INAF-IASF Milano, Via Alfonso Corti 12, 20133 Milano, Italy

³⁷ INAF-Osservatorio Astronomico di Roma, Via Frascati 33, 00078 Monteporzio Catone, Italy

³⁸ INFN-Sezione di Roma, Piazzale Aldo Moro, 2 - c/o Dipartimento di Fisica, Edificio G. Marconi, 00185 Roma, Italy

³⁹ Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Avenida Complutense 40, 28040 Madrid, Spain

⁴⁰ Port d’Informació Científica, Campus UAB, C. Albareda s/n, 08193 Bellaterra (Barcelona), Spain

⁴¹ Institute for Theoretical Particle Physics and Cosmology (TTK), RWTH Aachen University, 52056 Aachen, Germany

⁴² Deutsches Zentrum für Luft- und Raumfahrt e. V. (DLR), Linder Höhe, 51147 Köln, Germany

⁴³ INFN section of Naples, Via Cinthia 6, 80126, Napoli, Italy

⁴⁴ Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822, USA

⁴⁵ Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Università di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy

⁴⁶ Instituto de Astrofísica de Canarias, E-38205 La Laguna, Tenerife, Spain

⁴⁷ Jodrell Bank Centre for Astrophysics, Department of Physics and Astronomy, University of Manchester, Oxford Road, Manchester M13 9PL, UK

⁴⁸ European Space Agency/ESRIN, Largo Galileo Galilei 1, 00044 Frascati, Roma, Italy

⁴⁹ Université Claude Bernard Lyon 1, CNRS/IN2P3, IP2I Lyon, UMR 5822, Villeurbanne, F-69100, France

⁵⁰ Aix-Marseille Université, CNRS, CNRS, Marseille, France

⁵¹ Institut de Ciències del Cosmos (ICCUB), Universitat de Barcelona (IEEC-UB), Martí i Franquès 1, 08028 Barcelona, Spain

⁵² Institució Catalana de Recerca i Estudis Avançats (ICREA), Passeig de Lluís Companys 23, 08010 Barcelona, Spain

⁵³ Institut de Ciencies de l’Espai (IEEC-CSIC), Campus UAB, Carrer de Can Magrans, s/n Cerdanyola del Vallés, 08193 Barcelona, Spain

⁵⁴ UCB Lyon 1, CNRS/IN2P3, IUF, IP2I Lyon, 4 rue Enrico Fermi, 69622 Villeurbanne, France

⁵⁵ Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey RH5 6NT, UK

⁵⁶ Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France

⁵⁷ Departamento de Física, Faculdade de Ciências, Universidade de Lisboa, Edifício C8, Campo Grande, PT1749-016 Lisboa, Portugal

⁵⁸ Instituto de Astrofísica e Ciências do Espaço, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal

⁵⁹ Department of Astronomy, University of Geneva, ch. d’Ecogia 16, 1290 Versoix, Switzerland

⁶⁰ Université Paris-Saclay, CNRS, Institut d’astrophysique spatiale, 91405, Orsay, France

⁶¹ INFN-Padova, Via Marzolo 8, 35131 Padova, Italy

⁶² Aix-Marseille Université, CNRS/IN2P3, CPPM, Marseille, France

⁶³ INAF-Istituto di Astrofisica e Planetologia Spaziali, via del Fosso del Cavaliere, 100, 00100 Roma, Italy

⁶⁴ INFN-Bologna, Via Irnerio 46, 40126 Bologna, Italy

⁶⁵ School of Physics, HH Wills Physics Laboratory, University of Bristol, Tyndall Avenue, Bristol, BS8 1TL, UK

⁶⁶ Department of Physics, University of Oxford, Keble Road, Oxford OX1 3RH, UK

⁶⁷ University Observatory, LMU Faculty of Physics, Scheinerstr. 1, 81679 Munich, Germany

⁶⁸ INAF-Osservatorio Astronomico di Padova, Via dell’Osservatorio 5, 35122 Padova, Italy

⁶⁹ Universitäts-Sternwarte München, Fakultät für Physik, Ludwig-Maximilians-Universität München, Scheinerstr. 1, 81679 München, Germany

⁷⁰ Institute of Theoretical Astrophysics, University of Oslo, P.O. Box 1029 Blindern, 0315 Oslo, Norway

⁷¹ Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA

⁷² Department of Physics, Lancaster University, Lancaster, LA1 4YB, UK

⁷³ Felix Hormuth Engineering, Goethestr. 17, 69181 Leimen, Germany

⁷⁴ Technical University of Denmark, Elektrovej 327, 2800 Kgs. Lyngby, Denmark

⁷⁵ Cosmic Dawn Center (DAWN), Denmark

⁷⁶ Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany

⁷⁷ NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA

⁷⁸ Department of Physics and Astronomy, University College London, Gower Street, London WC1E 6BT, UK

⁷⁹ Department of Physics and Helsinki Institute of Physics, Gustaf Hällströmin katu 2, University of Helsinki, 00014 Helsinki, Finland

⁸⁰ Université de Genève, Département de Physique Théorique and Centre for Astroparticle Physics, 24 quai Ernest-Ansermet, CH-1211 Genève 4, Switzerland

⁸¹ Department of Physics, P.O. Box 64, University of Helsinki, 00014 Helsinki, Finland

⁸² Helsinki Institute of Physics, Gustaf Hällströmin katu 2, University of Helsinki, 00014 Helsinki, Finland

⁸³ Laboratoire d’etude de l’Univers et des phenomenes eXtremes, Observatoire de Paris, Université PSL, Sorbonne Université, CNRS, 92190 Meudon, France

⁸⁴ SKAO, Jodrell Bank, Lower Withington, Macclesfield SK11 9FT, UK

⁸⁵ Centre de Calcul de l’IN2P3/CNRS, 21 avenue Pierre de Coubertin 69627 Villeurbanne Cedex, France

⁸⁶ Dipartimento di Fisica "Aldo Pontremoli", Università degli Studi di Milano, Via Celoria 16, 20133 Milano, Italy

⁸⁷ INFN-Sezione di Milano, Via Celoria 16, 20133 Milano, Italy

⁸⁸ University of Applied Sciences and Arts of Northwestern Switzerland, School of Computer Science, 5210 Windisch, Switzerland

⁸⁹ Universität Bonn, Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany

⁹⁰ Department of Physics, Institute for Computational Cosmology, Durham University, South Road, Durham, DH1 3LE, UK

⁹¹ Université Côte d’Azur, Observatoire de la Côte d’Azur, CNRS, Laboratoire Lagrange, Bd de l’Observatoire, CS 34229, 06304 Nice cedex 4, France

⁹² Niels Bohr Institute, University of Copenhagen, Jagtvej 128, 2200 Copenhagen, Denmark

⁹³ Université Paris Cité, CNRS, Astroparticule et Cosmologie, 75013 Paris, France

⁹⁴ CNRS-UCB International Research Laboratory, Centre Pierre Binétruy, IRL2007, CPB-IN2P3, Berkeley, USA

⁹⁵ University of Applied Sciences and Arts of Northwestern Switzerland, School of Engineering, 5210 Windisch, Switzerland

⁹⁶ Institut d’Astrophysique de Paris, UMR 7095, CNRS, and Sorbonne Université, 98 bis boulevard Arago, 75014 Paris, France

⁹⁷ Dipartimento di Fisica - Sezione di Astronomia, Università di Trieste, Via Tiepolo 11, 34131 Trieste, Italy

⁹⁸ Telespazio UK S.L. for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

⁹⁹ Institut de Física d’Altes Energies (IFAE), The Barcelona Institute of Science and Technology, Campus UAB, 08193 Bellaterra (Barcelona), Spain

¹⁰⁰ School of Mathematics and Physics, University of Surrey, Guildford, Surrey, GU2 7XH, UK

¹⁰¹ DARK, Niels Bohr Institute, University of Copenhagen, Jagtvej 155, 2200 Copenhagen, Denmark

¹⁰² Perimeter Institute for Theoretical Physics, Waterloo, Ontario N2L 2Y5, Canada

¹⁰³ Space Science Data Center, Italian Space Agency, via del Politecnico snc, 00133 Roma, Italy

¹⁰⁴ Centre National d’Etudes Spatiales – Centre spatial de Toulouse, 18 avenue Edouard Belin, 31401 Toulouse Cedex 9, France

¹⁰⁵ Institute of Space Science, Str. Atomistilor, nr. 409 Măgurele, Ilfov, 077125, Romania

¹⁰⁶ Consejo Superior de Investigaciones Cientificas, Calle Serrano 117, 28006 Madrid, Spain

¹⁰⁷ Universidad de La Laguna, Dpto. Astrofísica, E-38206 La Laguna, Tenerife, Spain

¹⁰⁸ Dipartimento di Fisica e Astronomia "G. Galilei", Università di Padova, Via Marzolo 8, 35131 Padova, Italy

¹⁰⁹ Institut für Theoretische Physik, Universität Heidelberg, Philosophenweg 16, 69120 Heidelberg, Germany

¹¹⁰ Institut de Recherche en Astrophysique et Planétologie (IRAP), Université de Toulouse, CNRS, UPS, CNES, 14 Av. Edouard Belin, 31400 Toulouse, France

¹¹¹ Université St Joseph; Faculty of Sciences, Beirut, Lebanon

¹¹² Departamento de Física, FCFM, Universidad de Chile, Blanco Encalada 2008, Santiago, Chile

¹¹³ Universität Innsbruck, Institut für Astro- und Teilchenphysik, Technikerstr. 25/8, 6020 Innsbruck, Austria

¹¹⁴ Institut d’Estudis Espacials de Catalunya (IEEC), Edifici RDIT, Campus UPC, 08860 Castelldefels, Barcelona, Spain

¹¹⁵ Satlantis, University Science Park, Sede Bld 48940, Leioa-Bilbao, Spain

¹¹⁶ Infrared Processing and Analysis Center, California Institute of Technology, Pasadena, CA 91125, USA

¹¹⁷ Instituto de Astrofísica e Ciências do Espaço, Faculdade de Ciências, Universidade de Lisboa, Tapada da Ajuda, 1349-018 Lisboa, Portugal

¹¹⁸ Cosmic Dawn Center (DAWN)

¹¹⁹ Universidad Politécnica de Cartagena, Departamento de Electrónica y Tecnología de Computadoras, Plaza del Hospital 1, 30202 Cartagena, Spain

¹²⁰ Astronomisches Rechen-Institut, Zentrum für Astronomie der Universität Heidelberg, Mönchhofstr. 12-14, 69120 Heidelberg, Germany

¹²¹ Dipartimento di Fisica e Scienze della Terra, Università degli Studi di Ferrara, Via Giuseppe Saragat 1, 44122 Ferrara, Italy

¹²² Istituto Nazionale di Fisica Nucleare, Sezione di Ferrara, Via Giuseppe Saragat 1, 44122 Ferrara, Italy

¹²³ INAF, Istituto di Radioastronomia, Via Piero Gobetti 101, 40129 Bologna, Italy

¹²⁴ Astronomical Observatory of the Autonomous Region of the Aosta Valley (OAVdA), Loc. Lignan 39, I-11020, Nus (Aosta Valley), Italy

¹²⁵ ICSC - Centro Nazionale di Ricerca in High Performance Computing, Big Data e Quantum Computing, Via Magnanelli 2, Bologna, Italy

¹²⁶ Instituto de Física Teórica UAM-CSIC, Campus de Cantoblanco, 28049 Madrid, Spain

¹²⁷ Dipartimento di Fisica, Sapienza Università di Roma, Piazzale Aldo Moro 2, 00185 Roma, Italy

¹²⁸ Aurora Technology for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

¹²⁹ ICL, Jilin, Université Catholique de Lille, LITL, 59000 Lille, France

¹³⁰ CERCA/ISO, Department of Physics, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106, USA

¹³¹ Laboratoire Univers et Théorie, Observatoire de Paris, Université PSL, Université Paris Cité, CNRS, 92190 Meudon, France 132 Departamento de Física Fundamental. Universidad de Salamanca. Plaza de la Merced s/n. 37008 Salamanca, Spain

133 Center for Data-Driven Discovery, Kavli IPMU (WPI), UTIAS, The University of Tokyo, Kashiwa, Chiba 277-8583, Japan

134 Max-Planck-Institut für Physik, Boltzmannstr. 8, 85748 Garching, Germany

135 California Institute of Technology, 1200 E California Blvd, Pasadena, CA 91125, USA

136 Department of Physics & Astronomy, University of California Irvine, Irvine CA 92697, USA

137 Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands

138 Departamento Física Aplicada, Universidad Politécnica de Cartagena, Campus Muralla del Mar, 30202 Cartagena, Murcia, Spain

139 Instituto de Física de Cantabria, Edificio Juan Jordá, Avenida de los Castros, 39005 Santander, Spain

140 Observatorio Nacional, Rua General Jose Cristino, 77-Bairro Imperial de Sao Cristovao, Rio de Janeiro, 20921-400, Brazil

141 INFN, Sezione di Lecce, Via per Arnesano, CP-193, 73100, Lecce, Italy

142 Department of Mathematics and Physics E. De Giorgi, University of Salento, Via per Arnesano, CP-I93, 73100, Lecce, Italy

143 INAF-Sezione di Lecce, c/o Dipartimento Matematica e Fisica, Via per Arnesano, 73100, Lecce, Italy

144 CEA Saclay, DFR/IRFU, Service d’Astrophysique, Bat. 709, 91191 Gif-sur-Yvette, France

145 Institute of Cosmology and Gravitation, University of Portsmouth, Portsmouth PO1 3FX, UK

146 Department of Computer Science, Aalto University, PO Box 15400, Espoo, FI-00 076, Finland

147 Instituto de Astrofísica de Canarias, E-38205 La Laguna; Universidad de La Laguna, Dpto. Astrofísica, E-38206 La Laguna, Tenerife, Spain

148 Ruhr University Bochum, Faculty of Physics and Astronomy, Astronomical Institute (AIRUB), German Centre for Cosmological Lensing (GCCL), 44780 Bochum, Germany

149 INAF-Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125, Firenze, Italy

150 Department of Physics and Astronomy, Vesilinnantie 5, University of Turku, 20014 Turku, Finland

151 Finnish Centre for Astronomy with ESO (FINCA), Quantum, Vesilinnantie 5, University of Turku, 20014 Turku, Finland

152 Serco for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

153 ARC Centre of Excellence for Dark Matter Particle Physics, Melbourne, Australia

154 Centre for Astrophysics & Supercomputing, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia

155 Department of Physics and Astronomy, University of the Western Cape, Bellville, Cape Town, 7535, South Africa

156 DAMTP, Centre for Mathematical Sciences, Wilberforce Road, Cambridge CB3 0WA, UK

157 Kavli Institute for Cosmology Cambridge, Madingley Road, Cambridge, CB3 0HA, UK

158 Department of Astrophysics, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland

159 Department of Physics, Centre for Extragalactic Astronomy, Durham University, South Road, Durham, DH1 3LE, UK

160 IRFU, CEA, Université Paris-Saclay 91191 Gif-sur-Yvette Cedex, France

161 Oskar Klein Centre for Cosmoparticle Physics, Department of Physics, Stockholm University, Stockholm, SE-106 91, Sweden

162 Astrophysics Group, Blackett Laboratory, Imperial College London, London SW7 2AZ, UK

163 Univ. Grenoble Alpes, CNRS, Grenoble INP, LPSC-IN2P3, 53, Avenue des Martyrs, 38000, Grenoble, France

164 Centro de Astrofísica da Universidade do Porto, Rua das Estrelas, 4150-762 Porto, Portugal

165 HE Space for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

166 INAF - Osservatorio Astronomico d’Abruzzo, Via Maggini, 64100, Teramo, Italy

167 Theoretical astrophysics, Department of Physics and Astronomy, Uppsala University, Box 516, 751 37 Uppsala, Sweden

168 Mathematical Institute, University of Leiden, Einsteinweg 55, 2333 CA Leiden, The Netherlands

169 School of Physics & Astronomy, University of Southampton, Highfield Campus, Southampton SO17 1BJ, UK

170 Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK

171 Univ. Lille, CNRS, Centrale Lille, UMR 9189 CrIStAL, 59000 Lille, France

172 Department of Physics and Astronomy, University of California, Davis, CA 95616, USA

173 Institute for Particle Physics and Astrophysics, Dept. of Physics, ETH Zurich, Wolfgang-Pauli-Strasse 27, 8093 Zurich, Switzerland

174 Department of Astrophysical Sciences, Peyton Hall, Princeton University, Princeton, NJ 08544, USA

175 Space physics and astronomy research unit, University of Oulu, Pentti Kaiteran katu 1, FI-90014 Oulu, Finland

176 Center for Computational Physics, Flatiron Institute, 162 5th Avenue, 10010, New York, NY, USA

177 International Centre for Theoretical Physics (ICTP), Strada Costiera 11, 34151 Trieste, Italy

Impact of SPS model selection

We investigate the impact of alternative choices of SPS models on our estimate of the metallicity of the stellar population of the BCG and ICL. In the main text of this analysis, we used the E-MILES SPS models (Vazdekis et al. [106]). Here, we use a Bruzual & Charlot [13] SPS model. We note that we follow the same methodology as discussed in Sects. 5 and 6. In Fig. A.1, we present the metallicity distributions for the same three colours as in Figure 5, but estimated using a Bruzual & Charlot [13] SPS model. We note that the maximum metallicity for this model is $\approx0.4$, which is greater than that of the E-MILES SPS models (Vazdekis et al. [106]). As shown, we see that the metallicities based on the $I_{\mathrm{E}}-H_{\mathrm{E}}$ and $Y_{\mathrm{E}}-H_{\mathrm{E}}$ are more consistent at lower redshifts than the $J_{\mathrm{E}}-H_{\mathrm{E}}$ values. This is a result of degeneracies that exist within the Bruzual & Charlot [13] models for NIR wavelengths since that portion of the SED is based on extensions between the STELIB and Pickles spectral libraries using the BaSeL library and the higher wavelength Pickles library is of only one metallicity. These same degeneracies do not exist within the E-MILES models, which are derived from the Infrared Telescope Facility Spectral Library (Vazdekis et al. [106]). We note that at higher redshifts, the NIR colours are redshifted to lower wavelengths that are not in the regions covered by these extensions. However, regardless of the difficulty of using the $J_{\mathrm{E}}-H_{\mathrm{E}}$ colours at lower redshifts, we find nearly identical trends, in terms of a metallicity gradient, its extent, and trends beyond $\approx150$ kpc to what we detect when using the E-MILES SPS models (Vazdekis et al. [106]). As such we find that our detection of a metallicity gradient is independent of our choice of SPS model. However, we note that the actual metallicity values are dependent on the choice of SPS model and differ by $\approx0.2$ dex (when not considering the $J_{\mathrm{E}}-H_{\mathrm{E}}$ colours).