The large-scale structure of the observable Universe is one of the most important probes for testing cosmological models. Sample variance affects both observations and numeric simulations of cosmological structure formation, since these are only sampling a finite volume of a single realization of the cosmic density field. In this seminar, Gabor Racz will present his new method for estimating and reducing this source of error in cosmological N-body simulations. In his novel technique, simulations are run in matching pairs. The average statistical properties of a pair of such simulations have greatly suppressed variance. Using this new method can significantly reduce the computational costs of predicting the properties of cosmic structures for a given cosmological model by enabling use of smaller simulated volumes without compromising the precision. This technique also can be used for estimating the effects of the sample variance in existing simulations by comparing the original simulation and its complementary pair.
About the Speaker:
Gabor Racz is an NPP postdoctoral fellow at JPL working with Alina Kiessling in the Structure of the Universe Group (3266). He received his Ph.D. in Eotvos Lorand University, Budapest. His research focuses on cosmological structure formation, the effects of the large-scale structures on the measurements, and numerical methods in cosmology.