Aim: In this work package we answered the question: what is the best use of the cluster member spectroscopic data to enhance the Euclid cluster cosmology? How can we use spectroscopic data to inform the scaling relation and thus improve the constraining power of the cluster sample? A promising venue to overcome systematics in the cluster mass calibration is by means of multi-wavelength data. Mass proxies derived from data at different wavelengths are subject to different sources of systematics, and thus can be used to enhance the scaling relation calibration. Euclid will provide ~10,000 clusters with enough spectroscopic data for member galaxies (~10 or more) that can be used as an (additional) mass-proxy to calibrate the cluster masses, in the redshift range 0.9≲z≲1.8, where other mass proxies become ineffective (weak lensing and X-ray-based) . 

We found that the traditional techniques were not effective in producing reliable estimates of the cluster velocity dispersion. This was caused by the sparsity of the data available from Euclid spectroscopy, and by the preferential exclusion of passively evolving galaxies (which dominate the cluster population) from the Euclid spectroscopic detection. New techniques have been developed and preliminary results are encouraging albeit not conclusive yet.

Moreover, we worked on the improvement of the algorithm for cluster redshift determination from photometric and spectroscopic data, using the mocks and data sets that were analysed for the mass calibration. The use of spectroscopic data proved to be extremely successful in significantly improving the precision and accuracy of the cluster mean redshift estimates. 

Related Publications:

Girardi et al. (2024) – CLASH-VLT: Galaxy cluster MACS J0329–0211 and its surroundings using galaxies as kinematic tracers.

This work validates the use of galaxy velocity dispersions and spectroscopic data as mass tracers in galaxy clusters, providing an important methodological benchmark for the use of sparse spectroscopic samples such as those expected from the Euclid survey.

Ho et al. (in preparation) – Euclid preparation: Spectroscopy constrains the cluster mass-richness relation

This forthcoming work evaluates the performance of different techniques for estimating galaxy cluster masses from spectroscopic data expected from the Euclid mission. The study compares classical estimators based on velocity dispersion and richness with modern machine-learning approaches. The analysis shows that graph-based neural networks provide significantly improved mass estimates, successfully capturing the complex dynamical information encoded in galaxy phase-space distributions and maintaining good performance across a wide range of cluster richness and redshift (up to z∼1.8z \sim 1.8z∼1.8).

Ghaffari et al. (in preparation) - Euclid preparation. The determination of cluster mean redshifts

This forthcoming work presents the code for the refinement of the cluster mean redshift, both using photometric and spectroscopic data. The scientific performance is benchmarked using a set of simulations. The code is applied to a set of detections from literature using Euclid spectroscopy to evaluate the performance of the code with the actual Euclid redshifts. This work serves as a preparatory step for computing the velocity dispersions of spectroscopic members.