Speaker
Description
Baryonic physics introduces significant uncertainties in cosmological observables such as weak lensing statistics, while accurate predictions typically require computationally expensive hydrodynamical simulations. This limits the exploration of the parameter space in modern cosmological analyses. Baryonification methods provide a fast alternative by incorporating baryonic effects into dark matter-only simulations, but their accuracy and range of validity remain to be fully established across different implementations and observables.
In this talk, I compare two baryonification approaches: a standard snapshot-based method and a shell-based method applied directly to HEALPix maps, which enables particularly efficient map-level analyses. Their performance is assessed against hydrodynamical simulations from TNG and FLAMINGO, which provide matched dark matter-only counterparts for controlled comparisons, as well as against ensembles of dark matter-only simulations from CosmoGrid.
The comparison focuses on complementary weak lensing statistics, including power spectra and peak counts, allowing for a systematic assessment of how well baryonification captures baryonic effects in both Gaussian and non-Gaussian observables. This setup enables a clear evaluation of the regimes in which baryonification can serve as a reliable substitute for full hydrodynamical simulations.
| Other topic / keywords: | Baryonification, weak lensing, large scale structure, simulation |
|---|