Probing Galaxy-Formation Physics: A Bayesian Inference Approach

探索星系形成物理学：贝叶斯推理方法

• 批准号: 1109354
• 负责人: Houjun Mo
• 金额: $ 33.9万
• 依托单位: University of Massachusetts Amherst
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2011
• 资助国家: 美国
• 起止时间: 2011-09-01 至 2015-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1109354&HistoricalAwards=false
• 关键词: Probing; Galaxy; Formation; Physics; Bayesian

Semi-analytic models of galaxy formation and evolution (SAMs) use relatively simple, but physically motivated parameterized prescriptions for the effects of complex processes such as merging, gas accretion, star formation, and feedback to follow a simulated population of galaxies through cosmic time. Though inexact, SAMs are necessary since direct simulation of these processes is still beyond the reach of computational hardware and techniques. The disadvantage of SAMs is that the degeneracies among the adjustable parameters can be large and poorly understood. In response, the proposing team has developed a Bayesian inference based SAM. The Bayesian approach yields the posterior probability distribution of the SAM parameters for a given set of data, and an assessment of how well a particular model is supported by the data. In this project, the team will apply their analysis to archival data on high-redshift galaxies, and address the following issues: (1) how galaxies assemble their masses, how star formation proceeds in different galaxies, and whether a universal stellar initial mass function is consistent with the data; (2) how feedback affects the gas component in galaxies, and why most of the baryons today are not in galaxies; (3) the growth of supermassive black holes, and how feedback from active galactic nuclei (AGN) affects galaxy formation and evolution; (4) the properties of ionizing sources at high redshift and how the Universe was re-ionized; (5) whether the Cold Dark Matter (CDM) paradigm can accommodate all observational properties of galaxies. These models will be used to create mock galaxy catalogs tailored to the properties of specific galaxy surveys; the team will make these mock catalogs available to the community. The project will support the work and professional training of a graduate student, and is intended to provide research opportunities for undergraduates.

星系形成和演化的半解析模型（SAMs）使用相对简单，但物理动机的参数化处方的复杂过程的影响，如合并，气体吸积，星星的形成，并通过宇宙时间的反馈遵循模拟的星系人口。虽然不精确，但SAM是必要的，因为这些过程的直接模拟仍然超出了计算硬件和技术的范围。SAM的缺点是可调参数之间的简并可能很大，并且很难理解。作为回应，提议团队开发了基于贝叶斯推理的SAM。贝叶斯方法为给定的一组数据产生SAM参数的后验概率分布，并评估数据对特定模型的支持程度。在这个项目中，研究小组将把他们的分析应用于高红移星系的档案数据，并解决以下问题：（1）星系如何聚集它们的质量，不同星系中星星的形成如何进行，以及一个普遍的恒星初始质量函数是否与数据一致;（2）反馈如何影响星系中的气体成分，以及为什么今天的大多数重子不在星系中;（3）超大质量黑洞的生长，以及活动星系核（AGN）的反馈如何影响星系的形成和演化;（4）高红移下电离源的性质以及宇宙如何被再电离;（5）冷暗物质（CDM）范式是否能容纳星系的所有观测性质。这些模型将用于创建模拟星系目录，以适应特定星系调查的性质;该团队将向社区提供这些模拟目录。该项目将支持研究生的工作和专业培训，并打算为本科生提供研究机会。