The effect of intrinsic alignments on weak-lensing statistics in hydrodynamical simulations

The next generation of weak-gravitational-lensing surveys has the potential to place stringent constraints on cosmological parameters. However, their analysis is limited by systematics such as the intrinsic alignments of galaxies, which alter weak-lensing convergence and can lead to biases in cosmological parameter estimations. For the first time, in this work, we investigate the impact of intrinsic alignments on non-Gaussian statistics of the weak-lensing field using galaxy shapes derived from the IllustrisTNG hydrodynamical simulation. We create two catalogs of ray-traced convergence maps: one that includes the measured intrinsic shape of each galaxy and another where all galaxies are randomly rotated to eliminate intrinsic alignments. We compare a range of weak-lensing statistics between the two catalogs, including the shear–shear correlation function, the map-level angular power spectrum, one-point, peak count, and minimum distribution functions, and Minkowski functionals. For each statistic, we assess the level of statistical distinguishability between catalogs for a set of future survey angular areas. Our results reveal strong small-scale correlation in the alignment of galaxies and statistically significant boosts in weak-lensing convergence in both positive and negative directions for high-significance peaks and minima, respectively. We note that our analysis is at a fixed number density of ˜ 5 arcmin^-2, drawn from a single realization of initial conditions, and does not include observational uncertainties or supersample covariance contributions. Weak-lensing analyses utilizing non-Gaussian statistics must account for intrinsic alignments to avoid significantly compromised cosmological inferences.