Non-Gaussianity in CMB lensing from full-sky simulations

and Yuqi Kang · 2024

Read Non-Gaussianity in CMB lensing from full-sky simulations by and Yuqi Kang (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2411-02774.txt.

Description

A 2024 arXiv research paper by and Yuqi Kang. The lensing convergence field describing the weak lensing effect of the Cosmic Microwave Background (CMB) radiation is expected to be subject to mild deviations from Gaussianity. We perform a suite of full-sky lensing simulations using ray tracing through multiple lens planes – generated by combining N N -body simulations on smaller scales and low-to-intermediate redshifts with realisations of Gaussian random fields on large scales and at high redshifts.

Catalog record

CREATORand Yuqi Kang
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE48 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2411-02774

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Non-Gaussianity in CMB lensing from full-sky simulations https://github.com/Kang-Yuqi/FLAReS and Yuqi Kang Abstract The lensing convergence field describing the weak lensing effect of the Cosmic Microwave Background (CMB) radiation is expected to be subject to mild deviations from Gaussianity. We perform a suite of full-sky lensing simulations using ray tracing through multiple lens planes – generated by combining N N -body simulations on smaller scales and low-to-intermediate redshifts with realisations of Gaussian random fields on large scales and at high redshifts. We quantify the non-Gaussianity of the resulting convergence fields in terms of a set of skewness and kurtosis parameters and show that the non-Gaussian information in these maps can be used to constrain cosmological parameters such as the cold dark matter density Ω c ​ h 2 \Omega_{\mathrm{c}}h^{2} or the amplitude of primordial curvature perturbations A s A_{\mathrm{s}} . We forecast that for future CMB lensing observations, combining the non-Gaussian parameters with the Gaussian information can increase constraining power on ( Ω c ​ h 2 , A s ) (\Omega_{\mathrm{c}}h^{2},A_{\mathrm{s}}) by a few per cent compared to constraints from Gaussian observables alone. We make the simulation code for the full-sky lensing simulation available for download from GitHub . 1 Introduction On their way from the surface of last scattering to us, the photons of the Cosmic Microwave Background (CMB) are subject to small deflections due to gravitational lensing caused by the large scale structure of the Universe. This leads to slight distortions in observed CMB temperature and polarisation maps, which carry information about the matter distribution at lower-redshift that can be accessed via various reconstruction methods (e.g., [ 1 , 2 , 3 , 4 , 5 ] ). Weak gravitational lensing of the CMB was detected at high statistical significance in Planck data [ 6 ] and high-resolution ground-based CMB polarisation experiments like AdvACT [ 7 ] , SPT-3G [ 8 ] , and the upcoming Simons Observatory [ 9 ] and CMB-S4 [ 10 ] experiments can provide even more detailed CMB lensing measurements. Due to its ability to probe the total matter distribution over a broad range of redshifts, CMB weak lensing has become a staple tool for the inference of cosmological parameters. At the technical level, this usually involves a likelihood analysis based on the power spectrum of maps of the lensing potential – an approach that is limited to extracting the map’s Gaussian information. On the other hand, CMB lensing maps are expected to contain a non-trivial non-Gaussian contribution due to the inherent non-linearity of gravitational lensing itself [ 11 ] , and the fact that the lens (i.e., matter perturbations at lower redshift) is non-Gaussian. The resulting non-Gaussianity of the lensing map is dependent on cosmological parameters, and therefore becomes a useful probe of cosmology in itself. While for current data a focus on the power spectrum is indeed perfectly appropriate given that the non-Gaussianity of the Planck lensing map is dominated by reconstruction noise [ 12 ] , in the future, the non-Gaussian information will become observable [ 13 , 14 , 15 , 16 ] and can be used to complement the power spectrum in parameter inference. In order to actually access this information, one also requires accurate theoretical predictions of non-Gaussian lensing observables though. The problem’s non-linearity necessitates a combination of N -body simulations to model the matter perturbations with ray-tracing techniques. This approach has been applied to the weak gravitational lensing of galaxies [ 17 , 18 , 19 , 20 ] as well as the CMB [ 21 , 22 , 23 , 24 , 25 , 26 ] . In this work, we present a simulation pipeline for the full-sky CMB lensing convergence maps, with an emphasis on their non-Gaussian properties. We keep the computational requirements manageable by employing multi-scale N -body simulations on small angular scales and low-to-intermediate redshifts, and resorting to realisations of Gaussian random fields on large scales and at high enough redshifts. After validating our method and tuning some of its input parameters, we apply the results of our simulations to estimate to what extent the non-Gaussian information can improve cosmological parameter constraints compared to the Gaussian case. The paper is structured as follows. In Section 2 , we describe our lensing simulation algorithm. In Section 3 we validate the simulation results. We forecast the potential of non-Gaussian information to improve constraints on cosmological parameters in Section 4 and present our conclusions in Section 5 . 2 CMB lensing simulations Let us start this Section with a brief summary of the main results related to weak lensing relevant for the discussion of this paper. For more detailed reviews of the weak lensing formalism, we refer the reader to, e.g., Refs. [ 27 , 28 , 29 , 30 ] . The path of a photon traveling a short distance \delta r through a gravitational potential \Psi is subject to a local deflection by an angle \delta\beta=-\dfrac{2}{c^{2}}\delta r\,\nabla_{\perp}\Psi , where \nabla_{\perp} denotes the gradient in the transverse direction. To an observer located at r=0 , a source at comoving distance r_{s} subjected to a local deflection \delta\beta\ll 1 at distance r will appear to be shifted with respect to its true position by an observed local deflection angle \delta\theta=\dfrac{r_{s}-r}{r_{s}}\delta\beta . The total observed deflection angle \alpha can then be obtained by integrating over \delta\theta along the photon path. In the Born approximation, one instead integrates along the line of sight: \alpha^{\mathrm{B}}=-2\int_{0}^{r_{s}}\mathrm{d}r~\frac{r_{s}-r}{r_{s}c^{2}}\nabla_{\perp}\Psi(r\hat{\mathbf{n}},r). (2.1) The observable we consider here is the lensing convergence \kappa which is related to the deflection angle by \kappa=-\frac{1}{2}\nabla_{\hat{\mathbf{n}}}\alpha , where \nabla_{\hat{\mathbf{n}}} is the covariant derivative on the sphere defined by \hat{\mathbf{n}} . Equation 2.1 , combined with the relation \nabla_{\perp}\Psi=\frac{1}{r}\nabla_{\hat{\mathbf{n}}}\Psi then leads to the following expression for \kappa in the Born approximation: \kappa^{\mathrm{B}}=\int_{0}^{r_{s}}\mathrm{d}r~\frac{(r_{s}-r)r}{r_{s}c^{2}}\nabla^{2}_{\hat{\mathbf{n}}}\Psi(r\hat{\mathbf{n}},r). (2.2) The gravitational potential \Psi is related to the density contrast \delta=\frac{\rho-\bar{\rho}}{\bar{\rho}} via Poisson’s equation, \nabla^{2}\Psi=4\pi G\rho a^{2}=\frac{3H_{0}^{2}\Omega_{\mathrm{m}}}{2ac^{2}}\delta, (2.3) where \Omega_{\mathrm{m}} is the total matter density, H_{0} is the Hubble constant and a is the scale factor. Augmenting the spherical Laplacian \nabla_{\hat{\mathbf{n}}}^{2} to the 3-dimensional Laplacian \nabla^{2} , and noting that the derivatives of \Psi along the light path average to zero [ 28 ] , the lensing convergence can be expressed as \kappa^{\mathrm{B}}(\hat{\mathrm{n}})=\frac{3H_{0}^{2}\Omega_{\mathrm{m}}}{2c^{2}}\int_{0}^{r_{s}}\mathrm{d}r~\delta(r,\hat{\mathrm{n}})\frac{\left(r_{s}-r\right)r}{r_{s}a(r)}. (2.4) While the Born approximation may in practice be sufficiently accurate for simulations of the power spectrum at low source redshifts, even for upcoming large lensing surveys, it has been shown that non-Gaussian observables will require post-Born corrections [ 31 , 32 , 33 , 34 , 35 ] to avoid biases. Since the accuracy of the Born approximation decreases with increasing source redshift and we are interested in weak lensing of the CMB, post-Born corrections of higher order in \Psi will have to be taken into account. However, merely going to second order is not sufficient due to the presence of significant partial cancellations between second- and third-order terms [ 34 ] , so we consider post-Born corrections to the convergence up to third order in \Psi : \kappa=\kappa^{\mathrm{B}}+\kappa^{(2)}+\kappa^{(3)}+\mathcal{O}(\Psi^{4}). (2.5) The corresponding correction terms were derived in Ref. [ 34 ] and, with the short-hand notation \Psi_{,a}\equiv\nabla_{\perp}\Psi for transverse derivatives, read, in our notation: 1. Second Order \kappa^{(2)}=\kappa^{2L}+\kappa^{2D} (2.6) \kappa^{2L}=-2\int_{0}^{r_{s}}\mathrm{d}r\int_{0}^{r}\mathrm{d}r^{\prime}~rr^{\prime}~\frac{r_{s}-r}{r_{s}}\frac{r-r^{\prime}}{r}\left[\Psi_{,ac}(r)\Psi_{,ca}\left(r^{\prime}\right)\right] (2.7) \kappa^{2D}=-\frac{3H_{0}^{2}\Omega_{\mathrm{m}}}{c^{2}}\int_{0}^{r_{s}}\mathrm{d}r\int_{0}^{r}\mathrm{d}r^{\prime}~\frac{r^{2}}{a(r)}~\frac{r_{s}-r}{r_{s}}\frac{r-r^{\prime}}{r}\left[\delta_{,c}(r)~\Psi_{,c}\left(r^{\prime}\right)\right] (2.8)

Read the complete text