Parametric strong lensing model of the galaxy cluster Abell 2390 from Euclid and MUSE observations

TL;DR

This work delivers a high-precision parametric strong lensing mass model of Abell 2390 by integrating Euclid ERO imaging with reanalyzed MUSE spectroscopy to constrain 35 multiple images from 13 background sources, of which 25 are spectroscopically confirmed. Using Gravity.jl within a Bayesian framework, the authors test 11 mass parametrisations, finding that a single cluster-scale NIE halo plus an external shear term (model M9) best describes the data, with a RMS of $\Delta_{\mathrm{RMS}} = 0.32''$. The total mass profiles from all models agree within uncertainties, with the M9 model showing an isothermal fall-off at large radii; within $40$ kpc the mass is $(1.40 \pm 0.01) \times 10^{13} M_\odot$, and the Einstein radius at $z_s = 4.048$ is $\theta_E = 24.6 \pm 5.4''$. Systematic uncertainties from modelling choices are quantified across models and reduced when focusing on the best-performing set, illustrating Gravity.jl’s efficiency and suitability for large cluster samples from upcoming surveys.

Abstract

We present a new high precision parametric strong lensing total mass reconstruction of the Euclid Early Release Observations (ERO) galaxy cluster Abell 2390, at redshift z = 0.231. We include in this analysis 35 multiple images from 13 background sources, of which 25 are spectroscopically confirmed thanks to observations from the MUSE, spanning a redshift range from z = 0.535 to z = 4.877. After fully reanalysing the MUSE spectroscopy, we combine it with archival spectroscopic catalogues allowing us to select 65 secure cluster members. This sample is further complemented with 114 photometric member galaxies, identified within the Euclid VIS and NISP imaging down to magnitude H = 23. We also measure the stellar velocity dispersions for 22 cluster members, in order to calibrate the Faber-Jackson relation and hence the scaling relations for the subhalo mass components. We test and compare 11 total mass parametrisations of the galaxy cluster with increasing complexity. To do so, we employ the new parametric strong lensing modelling code Gravity.jl. Our best fit total mass parametrisation is characterised by a single large scale halo, 179 subhalo components, and an external shear term. The reference model yields a mean scatter between the model predicted and observed positions of the multiple images of 0.32 arcseconds. We are able to quantify the systematics arising from our modelling choices by taking advantage of all the different explored total mass parametrisations. When our results are compared with those from other lensing studies, we notice an overall agreement in the reconstructed cluster total mass profile in the outermost strong lensing regime: the discrepancy in the innermost region of the cluster (few kiloparsec from the brightest cluster galaxy, where few or no strong lensing features are observed) could possibly be ascribed to the different data and modelling choices.

Figures (11)

• Figure 1: Colour-composite image (red: $\HE$, green: $\JE+\HE$, blue: $\IE$) of the galaxy cluster A2390. The MUSE footprint is shown in red. The spectroscopic (photometric) cluster members are represented with cyan circles (boxes). The 22 cluster members for which we measured the stellar velocity dispersion are further marked with cyan crosses. The spectroscopically confirmed (photometric) multiple images included in our analysis are also shown in magenta (yellow). The multiple images are also labelled with their ID (see Table \ref{['tab:mi']}). The positions of the BCG and galaxy G29 (see Sect. \ref{['results']}) are also marked.
• Figure 2: Spectroscopic redshift distribution of the objects in our final spectroscopic catalogue. Cluster members (i.e. lying in the redshift range from $z = 0.211$ to $z = 0.251$) are in blue, whereas foreground and background objects are in orange and green, respectively. Multiple images are depicted in red. The vertical black line locates the redshift of the galaxy cluster. The insert shows the cluster members selection illustrated in Sect. \ref{['cluster-mems']}. The dashed red line identifies the best-fit Gaussian distribution, whereas the vertical black dotted lines define an interval of $\pm 3\sigma_z$ around the median cluster redshift.
• Figure 3: $\HE$ magnitude distribution of the 179 cluster members included in our models (blue). The 65 spectroscopically confirmed cluster members are highlighted in orange, whereas the sub-sample of the 22 galaxies for which we measured the central stellar velocity dispersion are in red.
• Figure 4: Colour-magnitude $\IE - \HE$ vs $\HE$ diagram. The red dots represent the 60 spectroscopically confirmed cluster members that we employed to fit the RCS, whereas the four cluster members we discarded due to the clipping are in blue. The green dots are the photometric galaxies added to our cluster member sample. The solid red line is the best-fit RCS, while the dotted ones define a range of $\pm \, \sigma_\mathrm{RCS}$ (used for the selection of the cluster members) and $\pm \, 3\sigma_\mathrm{RCS}$ (used for the clipping) around the line, respectively.
• Figure 5: Measured stellar velocity dispersions of 22 MUSE spectroscopically confirmed cluster members as a function of their $\HE$ magnitudes. Their colours encode the mean signal-to-noise ratio of galaxy spectra. The black solid line is the best-fit (maximum likelihood) of the scaling relation in Eq. \ref{['sc1']}. The light orange band corresponds to the best-fit mean scatter, $\Delta \sigma_\mathrm{ap}$, around the best-fit relation. The red solid curve corresponds to the relation in Eq. \ref{['sc1']} as obtained with the best-fit parameters of our reference model (see Sect. \ref{['cluster-mems']}). The light red area is estimated from 300 random values of $\sigma_\mathrm{ref}$ extracted from the Bayesian MCMC realisations of the reference model of this work.