标题： 大规模结构系统效应对宇宙学参数估计的影响 显示英文标题

标题： Impact of Large-Scale Structure Systematics on Cosmological Parameter Estimation

摘要： 大型近未来星系巡天提供了足够的统计能力，使我们的宇宙学分析成为数据驱动的，主要受限于系统误差。 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的分析，通过使用带有各种系统误差的模拟密度图进行推理分析，来研究影响大尺度结构研究的一些系统误差的影响；这些系统误差包括由 photometric redshifts（photo-$z$s）、银河尘埃、望远镜观测策略和观测条件引起的结构以及不完整的协方差矩阵所导致的系统误差。 具体而言，我们考虑了错误的 photo-$z$分布（光度偏倚、散射、异常值；光谱校准偏倚）、尘埃图分辨率、错误的尘埃定律、选择无或仅部分污染模板用于反投影，以及使用对角协方差矩阵而不是完整协方差矩阵的影响。 我们使用断层分析星系角功率谱量化了这些系统误差对宇宙学参数估计产生的偏差，重点关注确定每个系统误差的最大可能水平是否会对从鲁宾天文台空间与时间遗留巡天（LSST）的星系聚类分析中估算关键宇宙学参数产生不利影响。 我们发现，被调查的系统误差中，photo-$z$系统误差最为紧迫，光谱校准偏倚导致了最大的不利影响，但幸运的是，最佳拟合模型的高$\chi^2$值对此进行了标记。 另一方面，超出预期的 photo-$z$散射显著影响，而无需一定表明拟合效果差。 相比之下，在本研究采用的分析框架中，观测系统误差和不完整协方差矩阵的偏差可以舒适地处于次主导地位。 显示英文摘要

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 $\chi^2$ 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.