English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Fast and accurate simulations of the non-linear evolution of the cosmicdensity field are a major component of many cosmological analyses, but thecomputational time and storage required to run them can be exceedingly large.For this reason, we use generative adversarial networks (GANs) to learn acompressed representation of the 3D matter density field that is fast and easyto sample, and for the first time show that GANs are capable of generatingsamples at the level of accuracy of other conventional methods. Usingsub-volumes from a suite of GADGET-2 N-body simulations, we demonstrate that adeep-convolutional GAN can generate samples that capture both large- andsmall-scale features of the matter density field, as validated through avariety of n-point statistics. The use of a data scaling that preserveshigh-density features and a heavy-tailed latent space prior allow us to obtainstate of the art results for fast 3D cosmic web generation. In particular, themean power spectra from generated samples agree to within 5% up to k=3 andwithin 10% for k<5 when compared with N-body simulations, and similar accuracyis obtained for a variety of bispectra. By modeling the latent space with aheavy-tailed prior rather than a standard Gaussian, we better capture samplevariance in the high-density voxel PDF and reduce errors in power spectrum andbispectrum covariance on all scales. Furthermore, we show that a conditionalGAN can smoothly interpolate between samples conditioned on redshift. Deepgenerative models, such as the ones described in this work, provide greatpromise as fast, low-memory, high-fidelity forward models of large-scalestructure.

Nonlinear 3D Cosmic Web Simulation with Heavy-Tailed Generative Adversarial Networks