Toward Precision Cosmology on Small Scales: A Data-Intensive Approach

关于宇宙学和小尺度星系形成的信息非常丰富，但相关的物理学知识却知之甚少，而且计算能力也很强。 基于晕的模型通过参数化星系和暗物质晕之间的关系来绕过这个问题，但包含系统误差，并且对于可靠的统计测试来说不够准确。 该团队已经提出了一种新的数据密集型方法，该方法在计算上易于处理，可以从模拟中构建真实的目录，并测量聚类统计数据，就像对真实的数据一样。 这项工作将通过概念验证，在小尺度上对Lambda冷暗物质（LCDM）宇宙学模型进行严格的测试。 其结果将是一个软件模拟器，使用模拟网格来预测星系聚类统计数据作为宇宙学和星系晕参数的函数。 所有模拟目录和分析代码将公开发布，这将使许多额外的项目。 该项目还增加了在统计和数据分析的关键领域的培训，以目前的菲斯克-范德比尔特桥梁计划，帮助学生从代表性不足的群体进入博士课程。目前的系统已经被用来测试固定的宇宙学模型对测量斯隆数字巡天，揭示了显着的张力与最简单的形式的LCDM加晕模型。 接下来的研究将通过确定最佳测量值，扩展星系-晕关系的参数化以考虑细微特征（包括非泊松统计，空间和速度偏差以及星系组装偏差）以及加强建模管道来进行。 这将使得对固定宇宙学模型的最严格的测试成为可能，并探测那些已经添加的星系-晕连接的微妙特征。 该项目的最后一个雄心勃勃的部分是允许宇宙学参数在参数搜索范围内变化，方法是利用近似方法使用单一宇宙学模拟来预测其他宇宙学的晕分布。该奖项反映了NSF的法定使命，并且通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

There is a wealth of information about both cosmology and galaxy formation encoded in small scales, but the relevant physics is poorly understood and computationally prohibitive. Halo-based models bypass this problem by parameterizing the relation between galaxies and dark-matter halos, but contain systematic errors and are not sufficiently accurate for reliable statistical tests. A new data-intensive approach that this team has made computationally tractable constructs realistic catalogs from simulations and measures clustering statistics exactly as is done for real data. This work will move past proof-of-concept to perform stringent tests of the Lambda-Cold-Dark-Matter (LCDM) cosmological model on small scales. The result will be a software emulator that uses a grid of simulations to predict galaxy clustering statistics as a function of both cosmology and galaxy-halo parameters. All mock catalogs and analysis codes will be publicly released, which will enable many additional projects. The project also adds training in the critical area of statistics and data analysis to the current Fisk-Vanderbilt bridge program that helps students from underrepresented groups move into a PhD program.The current system has already been used to test fixed cosmological models against measurements in the Sloan Digital Sky Survey, revealing significant tension with the simplest form of the LCDM plus halo model. This next study will be carried out by identifying optimal measurements, extending the parameterization of the galaxy-halo relation to allow for subtle features (including non-Poisson statistics, spatial and velocity bias, and galaxy assembly bias), and strengthening the modeling pipeline. This will enable the most stringent tests of fixed cosmological models, and probe those subtle features of the galaxy-halo connection that have been added. The final ambitious part of the project is to allow cosmological parameters to vary within the parameter search, by exploiting approximate methods to use a single cosmology simulation to predict the halo distributions of other cosmologies.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.