| 研究実績の概要 |
One of the main goals of this project is to develop an analysis pipeline that could look for the effects of massive neutrinos in cosmological observables such as weak lensing and galaxy clustering. In particular, I proposed to develop a fast prediction scheme based on Gaussian Process regression, called emulators, to make theoretical predictions for these observational quantities based on the halo bias and halo mass function. Another goal was to analyze the redshift space clustering statistics from the Hyper Suprime Cam Survey using the emulators.
1.1. Analysis of N-body simulations. I developed tools to analyze the outputs of N-body simulations in redshift space. I measured the monopole and quadrupole correlation functions from N-body simulations using a halo occupation distribution model to model a mock galaxy population. I studied the effect of changing the galaxy properties as well as the effect of adding velocity bias.
I further investigated the halo bias by complementing my existing analysis with bias measurements from clustering statistics.
1.2. Development of Prediction Tools. I continued to develop prediction tools for the analysis of galaxy clustering data, especially those that can be used in redshift space. I performed a detailed study of the parameter space of galaxy properties needed to be covered by the emulator in order for it to be applied to a survey.
I also investigated the possibility of using other statistics, e.g. correlation function, but I found that the sizes of the simulations weren't suffcient.
|
