Thursday. At Leiden, KiDS meeting.
1107.5169
Clipping the Cosmos: the bias and bispectrum of LSS
Simpson, James, Heavens, Heymans
Introduce new technique which enables extraction of useful information from the bispectrum of galaxies well beyond the conventional limits of perturbation theory: method shown to improve measurements of the galaxy bispectrum, permitting the applicability of the tree-level perturbation theory at z=0 to be extended from k_max~0.1h/Mpc to >~0.5h/Mpc, with correspondingly large improvements in the determination of galaxy bias.
* intro: bispectrum contains cosmo information, but is strongly influenced by nonlinarities. To obtain robust parameter constraints, it is usual to discard information above a limiting k_max. At low z restricted to working on extremely large scales, typically k_max<0.1h/Mpc, where cosmic variance severely hampers precision. Try studying transformation of real space density field to suppress nonlinear growth. Methods: Gaussianization, log(1+delta), but these penalise highest density regions of the field. Goal: extend the range of the relatively simple tree-level perturbation theory by manipulating the field in real space. "Clip" high density regions (remove NL).
* method: extreme peaks of density field correspond to the regions furthest from the quasi-linear evolution, but they also dominate the contribution to higher-order statistics.
1107.5169
Clipping the Cosmos: the bias and bispectrum of LSS
Simpson, James, Heavens, Heymans
Introduce new technique which enables extraction of useful information from the bispectrum of galaxies well beyond the conventional limits of perturbation theory: method shown to improve measurements of the galaxy bispectrum, permitting the applicability of the tree-level perturbation theory at z=0 to be extended from k_max~0.1h/Mpc to >~0.5h/Mpc, with correspondingly large improvements in the determination of galaxy bias.
* intro: bispectrum contains cosmo information, but is strongly influenced by nonlinarities. To obtain robust parameter constraints, it is usual to discard information above a limiting k_max. At low z restricted to working on extremely large scales, typically k_max<0.1h/Mpc, where cosmic variance severely hampers precision. Try studying transformation of real space density field to suppress nonlinear growth. Methods: Gaussianization, log(1+delta), but these penalise highest density regions of the field. Goal: extend the range of the relatively simple tree-level perturbation theory by manipulating the field in real space. "Clip" high density regions (remove NL).
* method: extreme peaks of density field correspond to the regions furthest from the quasi-linear evolution, but they also dominate the contribution to higher-order statistics.
No comments:
Post a Comment