Statistical Methodology in Astronomy

赖斯教授建议对两个大型天文项目中出现的统计问题进行研究：MACHO项目的变星数据库(寻找银河系光晕中的暗物质)和台湾美国掩星调查(TAOS)，即搜索柯伊伯带中的彗星。第一个涉及数万条来自变星的光曲线的统计特征和建模。他建议从功能数据分析的角度来处理这些问题，进一步发展和扩展这一快速增长的统计研究领域的方法论。陶斯项目将使用台湾内陆山区的专用望远镜，监测柯伊伯带天体的星场掩星。主要的统计问题围绕着设计图像处理和信号检测程序，这些程序将在高采样率下实时运行，以检测罕见和微弱的信号。这是一项涉及数理统计和天文学的跨学科建议，其核心是两个天文学项目：(1)分析一个大型变星数据库。这些恒星的光线不是恒定的，而是周期性变化的。更好地了解这类恒星对于恒星演化模型以及确定到宇宙中遥远物体的距离都很重要。(2)TAOS项目将在海王星轨道之外的偏远地区探测我们的太阳系。据认为，那里可能有数以亿计的物体，如彗星，但由于它们相对较小和遥远，它们很难被探测到。计算密集的统计方法将在探测这些对象方面发挥关键作用。该项目的部分资金由恒星天文学和天体物理学(SAA)计划提供。

ABSTRACTProfessor Rice proposes research on statistical problems arising in two large astronomical projects: the variable star database of the MACHO project (a search for dark matter in the halo of our galaxy) and Taiwanese American Occultation Survey (TAOS), a search for comets in the Kuiper belt. The first involves statistical characterization and modeling of tens of thousands of light curves from variable stars. He proposes to approach these problems from the viewpoint of functional data analysis, further developing and extending the methodology of this rapidly growing area of statistical research. The TAOS project will monitor star fields for occultations by objects in the Kuiper belt, using dedicated telescopes in the interior mountains of Taiwan. The primary statistical problems center around designing image processing and signal detection procedures that will operate in real time at high sampling rates to detect rare and faint signals. Both projects are anticipated to make valuable contributions to both advancement of knowledge in astronomy and to the development of statistical methodology.This is an interdisciplinary proposal involving mathematical statistics and astronomy, centering around two projects in astronomy: (1) The analysis of a large database of variable stars. These are stars whose light is not constant, but changes in a periodic fashion. Better understanding of this population of stars is important for models of stellar evolution and also for determining distances to remote objects in the universe. (2) The TAOS project will probe our solar system in the remote region beyond the orbit of Neptune. It is thought that there may well be hundreds of millions of objects such as comets there, but because of their relatively small size and remoteness, they are very difficult to detect. Computationally intensive statistical methodology will play a key role in detecting these objects. Partial funding for this project was provided by the Stellar Astronomy and Astrophysics (SAA) Program.