Current and future galaxy surveys such as HETDEX, DESI, Euclid, PFS, and SPHEREx aim to address fundamental questions in cosmology. For all these galaxy surveys, galaxy clustering is the primary observable, so modeling the nonlinearities in galaxy clustering is essential to decipher the survey data. Perturbative modeling offers an efficient, flexible, and robust method to achieve this.
In this talk, I will first introduce how we can use perturbation theory to forward model the non-linear matter density field in $N$-body simulations efficiently and accurately with the Fast-Fourier-Transform (FFT) algorithm. Next, I will present a novel n-th order Eulerian Perturbation Theory (nEPT) scheme to model the matter power spectrum and bispectrum, which demonstrates significantly improved convergence and accuracy compared to traditional perturbation theory calculations, even without free parameters. With nEPT, theoretical modeling can extend to smaller scales, allowing us to extract more cosmological information from survey data. Finally, I will introduce the perturbative bias expansion, which describes the non-linear mapping between galaxy number density and underlying matter distribution. I will introduce how to renormalize these bias operators to make the perturbative expansion convergent and implement the expansion at field level. This is an important step towards a robust forward modeling of biased tracers field, which can be useful for field-level inference in galaxy surveys.
BIO
Zhenyuan Wang is a Ph.D. candidate at the Department of Astrophysics and Astronomy at the Penn State University, where he works with Prof. Donghui Jeong. Before going to Penn State, he received his B.S. degree at Tsinghua University in 2018, where he worked with Prof. Yi Mao on 21cm intensity mapping. Currently, his research focuses on developing novel methods to model the large scale structure observed in galaxy surveys.
