arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2409.14265astro-ph.CO
收藏

大尺度结构系统误差对宇宙学参数估计的影响

Impact of Large-Scale Structure Systematics on Cosmological Parameter Estimation

University of Michigan(密歇根大学);Rutgers University(罗格斯大学);Space Telescope Science Institute(太空望远镜科学研究所)

Humna Awan, Eric Gawiser, Javier Sanchez, Ignacio Sevilla-Noarbe, the LSST Dark Energy Science Collaboration

AI总结:

本文通过端到端模拟推断,系统研究了photo-z、尘埃、观测策略及协方差矩阵等系统误差对LSST宇宙学参数估计的影响,发现photo-z系统误差尤其是光谱校准偏差影响最大,而观测系统误差和不完整协方差矩阵的影响相对次要。

摘要:

近期的大型星系巡天提供了足够的统计能力,使我们的宇宙学分析能够以数据为驱动,其主要限制来自系统误差。因此,理解系统误差的影响至关重要。我们进行了一项端到端分析,通过使用包含各种系统误差的模拟密度图进行推断分析,来研究影响大尺度结构研究的部分系统误差;这些系统误差包括由测光红移(photo-z)引起的误差、银河系尘埃、望远镜观测策略和观测条件所引入的结构,以及不完整的协方差矩阵。具体而言,我们考虑了以下因素的影响:不正确的photo-z分布(测光偏差、弥散、离群值;光谱校准偏差)、尘埃图分辨率、不正确的尘埃消光规律、在去投影时选择零个或仅部分污染模板,以及使用对角协方差矩阵而非完整协方差矩阵。我们利用层析星系角功率谱,量化了这些系统误差对宇宙学参数估计所产生的偏差,重点在于确定在Rubin天文台时空遗产巡天(LSST)的星系聚类分析中,每种系统误差的最大合理水平是否会对关键宇宙学参数的估计产生不利影响。我们发现,在所研究的系统误差中,photo-z系统误差最为紧迫,其中光谱校准偏差造成的不利影响最大,但所幸其最佳拟合模型会以较高的χ²值给出警示。另一方面,大于预期的photo-z弥散会产生显著影响,却不一定表现为拟合不佳。相比之下,在本工作所使用的分析框架中,来自观测系统误差和不完整协方差矩阵的偏差则明显处于次要地位。

英文摘要:

Large near-future galaxy surveys offer sufficient statistical power to make our cosmology analyses data-driven, limited primarily by systematic errors. Understanding the impact of systematics is therefore critical. We perform an end-to-end analysis to investigate the impact of some of the systematics that affect large-scale structure studies by doing an inference analysis using simulated density maps with various systematics; these include systematics caused by photometric redshifts (photo-$z$s), Galactic dust, structure induced by the telescope observing strategy and observing conditions, and incomplete covariance matrices. Specifically, we consider the impacts of incorrect photo-$z$ distributions (photometric biases, scatter, outliers; spectroscopic calibration biases), dust map resolution, incorrect dust law, selecting none or only some contaminant templates for deprojection, and using a diagonal covariance matrix instead of a full one. We quantify the biases induced by these systematics on cosmological parameter estimation using tomographic galaxy angular power spectra, with a focus on identifying whether the maximum plausible level of each systematic has an adverse impact on the estimation of key cosmological parameters from a galaxy clustering analysis with Rubin Observatory Legacy Survey of Space and Time (LSST). We find photo-$z$ systematics to be the most pressing out of the systematics investigated, with spectroscopic calibration biases leading to the greatest adverse impact while helpfully being flagged by a high $χ^2$ value for the best fit model. Larger-than-expected photo-$z$ scatter, on the other hand, has a significant impact without necessarily indicating a poor fit. In contrast, in the analysis framework used in this work, biases from observational systematics and incomplete covariance matrices are comfortably subdominant.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑