The effect of primordial non-Gaussianity on the skeleton of cosmic shear maps

TL;DR

This study investigates how primordial non-Gaussianity affects the cosmic web's skeleton in weak lensing maps, showing that skeleton statistics can retain initial condition signatures but currently provide limited constraints.

Contribution

It demonstrates that the skeleton length in cosmic shear maps is sensitive to primordial non-Gaussianity and assesses the potential for future surveys to constrain non-Gaussianity levels.

Findings

Skeleton length deviations depend on initial non-Gaussianity.

Current constraints on $f_{NL}$ are around 300-500.

Future large-area surveys could constrain $f_{NL}$ to a few tens.

Abstract

(abridged) We explore the imprints of deviations from Gaussian primordial density fluctuations on the skeleton of the large-scale matter distribution as mapped through cosmological weak lensing. We computed the skeleton length of simulated effective convergence maps covering $\sim 35$ sq. deg each, extracted from a suite of cosmological $n-$body runs with different levels of local primordial non-Gaussianity. The latter is expected to alter the structure formation process with respect to the fiducial Gaussian scenario, and thus to leave a signature on the cosmic web. We found that alterations of the initial conditions consistently modify both the cumulative and the differential skeleton length, although the effect is generically smaller than the cosmic variance and depends on the smoothing of the map prior to the skeleton computation. Nevertheless, the qualitative shape of these deviations is rather similar to their primordial counterparts, implying that skeleton statistics retain good memory of the initial conditions. We performed a statistical analysis in order to find out at what Confidence Level primordial non-Gaussianity could be constrained by the skeleton test on cosmic shear maps of the size we adopted. At 68.3% Confidence Level we found an error on the measured level of primordial non-Gaussianity of $\Delta f_\mathrm{NL}\sim 300$, while at 90% Confidence Level it is of $\Delta f_\mathrm{NL}\sim 500$. While these values by themselves are not competitive with the current constraints, weak lensing maps larger than those used here would have a smaller field-to-field variance, and thus would likely lead to tighter constraints. A rough estimate indicates $\Delta f_\mathrm{NL} \sim$ a few tens at 68.3% Confidence Level for an all-sky weak lensing survey.