Weak-lensing tunnel voids in simulated light cones: A new pipeline to investigate modified gravity and massive neutrinos signatures

Abstract

Cosmic voids offer a unique opportunity to explore modified gravity (MG) models. Their low-density nature and vast extent make them especially sensitive to cosmological scenarios of the class $f(R)$, which incorporate screening mechanisms in dense, compact regions. Weak lensing (WL) by voids, in particular, provides a direct probe for testing MG scenarios. While traditional voids are identified from 3D galaxy positions, 2D voids detected in WL maps trace underdense regions along the line of sight and are sensitive to unbiased matter distribution. To investigate this, we developed an efficient pipeline for identifying and analyzing tunnel voids, i.e., underdensities detected in WL maps, specifically in the signal-to-noise ratio (SNR) of the convergence. In this work, we used this pipeline to generate realistic SNR maps from cosmological simulations featuring different $f(R)$ scenarios and massive neutrinos, comparing their effects against the standard $Λ$CDM model. Using the convergence maps and the 2D void catalogs, we analyzed various statistics, including the probability density function, angular power spectrum, and void size function. We then focused on the tangential shear profile around 2D voids, demonstrating how the proposed void-finding algorithm maximizes the signal. We showed that MG leads to deeper void shear profiles due to the enhanced evolution of cosmic structures, while massive neutrinos have the opposite effect. Furthermore, we found that parametric functions typically applied to 3D void density profiles are not suitable for deriving the shear profiles of tunnel voids. We, therefore, proposed a new parametric formula that provides an excellent fit to the void shear profiles across different void sizes and cosmological models. Finally, we tested the sensitivity of the free parameters of this new formula to the cosmological model.

Figures (16)

• Figure 1: Example of light-cone construction with MapSim from $z=0$ to $z=4$ for a $\Lambda$CDM simulation with a box size of $750 \, h^{-1} \, \mathrm{Mpc}$, tiled to encompass the $5 \times 5 \, \mathrm{deg}^2$ light cone. The light cone geometry is depicted by the red solid line, expanding until its transverse comoving size matches that of the simulation box. Different colors represent different boxes, while the total of $21$ different snapshots are defined by the dashed lines. The source plane (blue solid line) for ray tracing is placed at $z_s = 1$. The snapshots within $z_{\text{obs}} < z < z_s$ are the $13$ utilized ones, while the gray region indicates the snapshots with $z > z_s$ that are not considered in this study. The yellow solid lines represent the $27$ lens planes used, onto which the particles are projected. Among these, only those within the considered snapshots are utilized for constructing the convergence maps used in this work, positioned from $z=0.05$ to $0.98$, respectively.
• Figure 2: Example of a realization of the $\Lambda$CDM light cone, having a FOV of $5\times5$ deg$^2$ aperture and $z_s=1$. In the upper panel, we show the WL convergence map, and in the lower panel, the S/N of the same map with smoothing scale at $\theta_{\mathrm{G}}=2.5 \, \text{arcmin}$. The color scale in both panels represents the convergence value for each pixel in the top panel and the S/N value in the bottom panel, as indicated by the corresponding color bars.
• Figure 3: Top panel: Average PDFs of the noised and smoothed convergence field from $256$ light-cone random realizations with redshift $z_s=1$, for all the analyzed models: $\Lambda$CDM (black), $fR4$ (green), $fR5$ (blue), $fR6$ (red) and $\Lambda$CDM$_{0.15 \, \mathrm{eV}}$ (purple). The black dashed line represents the average PDF without the addiction of GSN for the $\mathrm{\Lambda CDM}$ model. Bottom panel: PDF noised and smoothed corresponding residuals computed with respect to the $\Lambda$CDM model. For the purposes of visualization, only the uncertainties of the noiseless and the reference $\Lambda$CDM measurements are reported in this plot as hatched gray and shaded gray areas, respectively.
• Figure 4: Average angular power spectra of the noised and smoothed convergence field from $256$ light-cone random realizations with redshift $z_s=1$, for all the analyzed models (top panel) and the correspondent residuals with respect to the reference $\Lambda$CDM model (bottom panel). Line styles and color schemes are the same as used in Fig. \ref{['fig:pdf_k']}. The dotted vertical line marks the angular mode of the characteristic smoothing scale in the Fourier transform of the $\kappa$ field.
• Figure 5: Diagram showing the four main steps of the 2D void finder algorithm that allows the identification of voids from a WL convergence map.