海量天文星表数据的交叉与融合

发布时间：2018-09-07 11:28

【摘要】：天文学作为一门古老的学科,一直在人类历史上扮演着十分重要的角色。交叉证认在天体研究中起着重要的作用,是多波段天文数据融合的关键技术和方法。通过对某一天体的多波段证认,可以获得该天体不同波段的信息,进而可以揭露更多的物理本质信息。随着观测仪器和观测技术的大幅提高,使得搜集数据的能力大大加强。2012年10月开始,郭守敬望远镜(大天区面积多目标光纤光谱望远镜,以下简称LAMOST)开始正式巡天,巡天过程中将会产生海量的光谱数据。与此同时,一些新发布的巡天星表数据也在近期发布。大视角近红外巡天WISE (Wide-field Infrared Survey Explorer)星表数据于2012年发布,星表中包含563,921,584个目标；FIRST (Faint Images of the Radio Sky at Twenty-Centimeters)星表数据于2012年2月16日发布新版本,星表中包含946,464条记录；SDSS DR9星表数据于2012年7月发布,星表中包含1,231,051,050条记录。除此之外,还有一些巡天星表数据,2MASS星表中包含472,640,569个目标,GALEX星表中包含225,565,338条记录。对于如此海量的星表数据,多波段天文星表数据的交叉和融合就显得极为重要。然而,不同的星表数据,光学波段不同,分辨率不同,位置精度(误差半径)也不尽相同。这就为星表数据的交叉证认增加了难度。本研究的主要工作主要分为两部分： (1)研究HTM和HEALPix这两种不同的伪二维球面索引方法,设计一种算法在一定程度上解决边界漏源问题,使得交叉认证的精度有所提高,减少漏源情况的发生 (2)对于算法,我们通过一系列的实验进行验证;同时对星表之间的系统差进行检验；并使用ThreadPool技术加速交叉认证过程,减少花费时间。并且对于星表交义证认中出现的“一对多”和“多对一”的情况,我们进行了讨论,并提出了一种判断光谱数据同源的方法。
[Abstract]:Astronomy, as an ancient subject, has been playing a very important role in human history. Cross-identification plays an important role in celestial research and is the key technology and method of multi-band astronomical data fusion. Through the multi-band identification of a certain celestial body, the information of different bands of the object can be obtained, and more physical essential information can be revealed. With the rapid improvement of observational instruments and observation techniques, the ability to collect data has been greatly enhanced. Since October 2012, the Guo Shoujing Telescope (the large Sky area Multitarget Optical Fiber Spectral Telescope, hereinafter referred to as LAMOST) has begun its official sky survey. A great deal of spectral data will be produced during the survey. At the same time, a number of newly released star catalog data are also released in the near future. The large angle near infrared survey WISE (Wide-field Infrared Survey Explorer) catalog data was released in 2012, the catalog contains 563921584 target FIRST (Faint Images of the Radio Sky at Twenty-Centimeters) the new version was released on February 16, 2012. The catalog contains 946464 records. The SDSS DR9 catalog, released in July 2012, contains 1231051050 records. In addition, there are a number of surveys in the 2MASS catalog containing 4726,40569 targets and the Galex catalog contains 225565338 records. For such a large number of star catalog data, the cross and fusion of multi-band astronomical catalog data is extremely important. However, the position accuracy (error radius) varies with different optical band and resolution. This increases the difficulty of cross-identification of catalog data. The main work of this study is divided into two parts: (1) two different pseudo-two-dimensional spherical indexing methods, HTM and HEALPix, are studied, and an algorithm is designed to solve the boundary leakage source problem to a certain extent. The accuracy of cross-certification is improved, and the occurrence of leakage source is reduced. (2) for the algorithm, we verify the algorithm through a series of experiments, at the same time, we test the system difference between the star catalog. And the use of ThreadPool technology to speed up the cross-certification process, reducing the cost of time. We also discuss the cases of "one-to-many" and "many-to-one" in the synonym of the catalog, and propose a method to judge the homology of spectral data.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P114;TP202

【共引文献】