Ultra-diffuse Galaxy Analogues in the Subaru Hyper-Suprime Cam Wide-field Clusters

TL;DR

We study ultra-diffuse galaxy analogues (NUDGEs) in a large sample of clusters using Subaru HSC-SSP imaging to quantify their abundances and properties as a function of environment. A uniform detection, modeling, and completeness-corrected pipeline (SExtractor/Galfit with background annuli) yields 5057 NUDGEs in 51 clusters (0.08<z<0.15; M200 ≈ 0.95–8.34×10^14 M⊙). The NUDGE population shares properties with known UDGs (mean re ≈ 2.7 kpc; ⟨μe(g)⟩ ≈ 25.1 mag/arcsec^2; ⟨(g−r)0⟩ ≈ 0.60) and shows a shallow radial density profile, with red NUDGEs more centrally concentrated. The abundance scales with cluster mass as a power law, $N ∝ M_{200}^{0.78±0.28}$, and the size/density distributions are consistent with both internal (high-spin dwarfs) and external (tidal stripping in clusters) formation channels, including cored dark matter halos, underscoring the environmental role in UDG formation while leaving the dominant mechanism open to multiple pathways.

Abstract

We perform a systematic statistical study of ultra-diffuse galaxy analogues (NUDGEs) in a large sample of galaxy clusters to investigate their properties with respect to the host clusters. We used data from the Hyper Suprime-Cam Subaru Strategic Program wide field survey and find a total of 5057 NUDGEs exceeding the background counts in 51 out of 66 galaxy clusters. The clusters span the redshift range 0.08$\,<\,$z$\,<\,$0.15 and they have a mass range of $0.95\times10^{14}\,\text{M}_\odot - 8.34\times10^{14}\,\text{M}_\odot$. The properties of these NUDGEs are found to be similar to UDGs studied in previous works and reaffirm that they are an extension of a continuous galaxy distribution. The number of NUDGEs as a function of cluster halo mass for our sample follows the power law: $N\propto M_{200}^{0.78\pm\,0.28} $. This fit is consistent with previous UDG studies and, together with our NUDGE sizes distributions, matches well with the simulations of UDGs in cored dark matter haloes formed by tidal stripping. The NUDGE density distribution with respect to clustercentric radius of our sample is flatter than previous UDG studies, although the red NUDGEs in this sample show a statistically significant decrease in density with respect to clustercentric radius, indicating that red UDGs may be more affected by their environment than blue UDGs.

Figures (21)

• Figure 1: The spatial positions of the cluster sample are indicated with blue dots as well as the overall coverage of the two wide field survey regions used in our study (black outlined regions). The Galactic plane is highlighted in faint orange shading.
• Figure 2: Subaru image of cluster 9723, both panels show the same image. The red circle indicates the $R_{200}$ cluster radius. In the left panel the positions redMaPPer galaxies are identified by the yellow circles. In the right panel the white crosses identify the positions of NUDGE candidates and the yellow contours indicate the density of the redMaPPer galaxies. The black lines show the individual patches of the Subaru data with respect to the cluster. Each patch was investigated independently.
• Figure 3: Measuring the model properties during completeness testing. Top-left: Subaru image cutout with our model added. Top-right: Model recovered by galfit. Bottom-left: Initial galfit input model that was inserted into the Subaru image. Bottom-right: Residual image created by galfit by subtracting the recovered model from the image. "MAG", "SER", "RAD" and "AXR" refers to the apparent magnitude, Sérsic index, effective radius in arcsecs and axis ratio respectively.
• Figure 4: Plots showing the effective radius vs Mean effective surface brightness in $g$-band for the sample of recovered models from the completeness test. Top-left panel: The input distribution of of our recovered models (each point constitutes up-to 10 models). Bottom-left panel: The distribution of the output properties of the recovered models. The output properties of some models are recovered outside of the completeness grid, particularly toward brighter mean effective surface brightnesses. Top-right panel: Detection plot for the sample of recovered models (input properties). The completeness fraction shown by the colourbar is measured against the total number of models inserted in each bin with respect to their input model properties. Each cluster has a completeness correction with respect to its redshift, which is between the maximum and minimum regions shown (the red and blue rectangles, lowest and highest redshift respectively). Bottom-right panel: Completeness measured with respect to the output properties. The number of models in these bins increase beyond 1, as typically the models are recovered to smaller sizes and brighter surface brightnesses, implying lower recovery fractions in the top-right of this plot and higher fractions toward the bottom left.
• Figure 5: The difference in radius between the recovered galfit radius and the input model radius as a function of the input model radius. The black markers indicate the median value in each bin and the red curve indicates the best-fit second-order polynomial to the median values. The black error-bars indicate the binning of the data, which is identical to the detection and completeness plots (Figure \ref{['fig:completeness_det_char']}), and the red shading indicates the interquartile range. The difference between the measured and input model radii increases as the radius increases, as does the scatter.