SKLOF:一种新的超新星候选范围约减算法

超新星是宇宙学中的"标准烛光",其在星系中爆发的概率很低,是一种特殊、稀少的天体,只有在大量观测的星系数据中才有机会遇到,而正处于爆发期的超新星会照亮其整个星系从而在观测获得的星系光谱中具有较明显的特征。但是,目前已发现的超新星数量相对于大量的天体而言又是非常稀少的,搜寻它们所用的计算时间成为能否进行后续观测的关键,因此需要寻找高效率的超新星搜寻方法。对超新星候选范围进行约减的LOF算法的时间复杂度较高,计算量大,不适用于大规模数据集。为此通过对LOF算法进行改进,提出了一种在海量星系光谱中快速约减超新星候范围的新方法(SKLOF)。首先对光谱数据集中离中心点近的数据点进行数据剪枝,剪掉那些肯定不是超新星候选体的光谱数据对象,然后利用改进的LOF算法计算剩余的光谱数据的孤立性因子并降序排列进行离群搜索,最后获得超新星候选体的较小的搜索范围以便进行后续的证认。实验结果表明,该算法十分有效,不仅在精确度上有所提高,而且相比于LOF算法还进一步缩短了算法的运行时间,提高了算法的执行效率。     

SKLOF:一种新的超新星候选范围约减算法

超新星是宇宙学中的“标准烛光”, 其在星系中爆发的概率很低, 是一种特殊、稀少的天体, 只有在大量观测的星系数据中才有机会遇到, 而正处于爆发期的超新星会照亮其整个星系从而在观测获得的星系光谱中具有较明显的特征。但是, 目前已发现的超新星数量相对于大量的天体而言又是非常稀少的, 搜寻它们所用的计算时间成为能否进行后续观测的关键, 因此需要寻找高效率的超新星搜寻方法。对超新星候选范围进行约减的LOF算法的时间复杂度较高, 计算量大, 不适用于大规模数据集。为此通过对LOF算法进行改进, 提出了一种在海量星系光谱中快速约减超新星候范围的新方法(SKLOF)。首先对光谱数据集中离中心点近的数据点进行数据剪枝, 剪掉那些肯定不是超新星候选体的光谱数据对象, 然后利用改进的LOF算法计算剩余的光谱数据的孤立性因子并降序排列进行离群搜索, 最后获得超新星候选体的较小的搜索范围以便进行后续的证认。实验结果表明, 该算法十分有效, 不仅在精确度上有所提高, 而且相比于LOF算法还进一步缩短了算法的运行时间, 提高了算法的执行效率。     

海量光谱中激变变星候选体的数据挖掘

提出一套适用于海量光谱自动快速筛选激变变星的方法.利用已证认的激变变星光谱作为模板,使用主分量分析提取主特征后构造光谱特征矩阵,将海量光谱利用光谱特征矩阵映射到特征空间后,使用支持向量机排除大部分非候选体,最后对较少数量的候选体进行模板匹配并证认,结果作为反馈进一步丰富模板库.实验发现了58个新的激变变星候选体,表明了该方法的可行性,为在LAMOST海量光谱中快速搜索激变变星等稀少天体提供了有效途径.     

基于随机森林的激变变星候选体的数据挖掘

提出一种适用于在郭守敬望远镜海量光谱中自动、快速筛选激变变星的方法。利用已证认的激变变星光谱作为模板,通过随机森林分类训练,得到一个分类模型,该模型给出了各个波长对应流量的重要性排序,可根据该排序进行降维并用于激变变星判别,结果作为反馈进一步丰富模板库。实验中共发现了16个新的激变变星候选体,表明了该方法的可行性。     

矮新星候选体的自动搜索研究

美国斯隆数字巡天望远镜已经发布了第9期数据。这些海量的天文光谱数据除了可以用来进行大样本的研究,如探寻银河系的结构和进行多波段证认外,还蕴藏着稀少和特殊的天体,其中就包括矮新星。矮新星是激变变星中所占比例最高的一个亚型,发现更多的矮新星样本对于研究密近双星的演化和参数有积极的意义。目前针对激变变星这类稀少天体的发现主要使用测光粗筛选结合后期观测证认的方法,不但准确率低,而且需要耗费较多的人工处理时间,无法满足在海量光谱数据中快速发现矮新星候选体的需要。本文提出一种适用于在海量光谱中自动、快速发现矮新星的方法。该方法针对SDSS的DR9数据,先使用支持向量机约束主分量分析进行降维,确定特征空间的维数,然后再使用训练后得到的最优分类器对海量光谱进行自动识别,寻找矮新星候选体。实验共发现了276个矮新星,其中6个是未被收录的新的源,表明了该方法的有效性,为在海量光谱中快速发现稀少和特殊天体提供了有效途径。实验中发现的新结果补充了现有的矮新星模板光谱库,可以构造更准确的特征空间。本方法也可用于在其他的巡天望远镜如郭守敬望远镜的海量光谱中进行特殊天体的自动搜索。     

SDSS-DR8中激变变星候选体的数据挖掘

提出一套适用于在海量光谱中快速发现激变变星的方法。针对SDSS发布的DR8数据,尝试流型学习方法在海量光谱数据挖掘中的应用。首先使用非线性局部线性嵌入方法(LLE)对海量光谱数据进行降维,然后使用人工神经网络对低维数据进行分类,最后对较少数量的候选体进行人工证认。实验共发现了6个新的激变变星候选体,并与传统的PCA方法进行了比较,验证了LLE方法在天文数据挖掘中的可行性。     

SDSS-DR10中WDMS光谱的自动搜索研究

SDSS-DR10是美国SLOAN巡天望远镜发布的最新数据,包含了首批APOGEE光谱。这些海量的天文光谱除了可以用来探寻银河系的结构和进行多波段证认外,还蕴藏着包括白矮主序双星在内的特殊天体。白矮主序双星是一类特殊的双星系统,它由两颗主序星演化而来,包含了中低质量恒星演化的终点—白矮星,以及M矮星。白矮主序双星对于密近双星的演化和参数研究有积极的意义。目前针对这类特殊天体的发现主要使用测光筛选结合后期观测证认的方法,不但准确率低,而且需要耗费较多的人工处理时间,无法满足在海量光谱数据中快速发现目标天体的需要。提出一种适用于在海量天文光谱中自动、快速发现白矮主序双星的方法。该方法针对SDSS的DR10数据,使用改进的遗传算法对海量光谱进行自动识别,寻找白矮主序双星候选体。实验共发现了4, 140个白矮主序双星,通过交叉证认,其中24个是未被收录的新的源。验证了遗传算法在天文数据挖掘和自动搜索方面的有效性,为在海量光谱中快速发现特殊天体提供了另一途径。该方法也可用于在其他巡天望远镜的海量光谱中进行特定天体的自动识别。提供了新发现的白矮主序双星的赤经、赤纬等信息,补充了现有的白矮主序双星光谱库。     

基于二元PCA特征谱的星系光谱成分快速扣除

实际观测天体目标光谱如超新星和活动星系核光谱常常混有寄主星系成分,这对目标天体光谱的类别和性质证认识别会造成困难。文章提出了一种快速有效的称为二元PCA特征谱分解的星系扣除算法。该算法首先计算了星系样本模板库和超新星样本模板库各自的PCA特征光谱,然后对特征光谱组通过正交变换得到混合空间的一组标准正交基,进而利用混合光谱在该标准正交基上的分解系数计算该光谱在原特征光谱组的分解系数,获得星系超新星混合光谱的快速分解,系数计算也可通过SVD矩阵分解得到,但计算效率较低。实验表明,该方法优于常用的直接PCA投影重构分解方法,与另一种χ2模板拟合方法扣除星系成分相比,在保持分解效果基本不变的前提下,时间消耗则大大降低,从而使该方法可应用到大规模光谱数据处理中。     

在SDSS DR8恒星光谱中自动搜寻稀有天体

SDSS DR8海量光谱中包含许多有研究价值的稀有天体,如特殊白矮星(DZ,DQ,DC)、碳星、白矮主序双星、激变变星等,如何在海量光谱中自动搜寻稀有天体有着极其重要的意义。提出一种基于核密度估计和K-近邻(K-nearest neighbor, KNN)相结合的方法在SDSS DR8 信噪比大于5的546 383个恒星光谱中搜寻稀有天体。首先对光谱进行高斯核密度估计,选取概率最小的5 000个光谱作为稀有类,概率最大的300 000个光谱作为普通类,然后进行KNN分类,同时也将5 000个稀有光谱的K个最近邻也作为稀有的天体,结果共有21 193条光谱。为了方便分析,对这些光谱聚类后进行人工检查。这些光谱主要包括由于数据缺失、红化、流量定标不准引起的问题光谱、行星状星云、没有物理联系的光谱双星、类星体、特殊白矮星(DZ,DQ,DC)、碳星、白矮主序双星、激变变星等。通过和SIMBAD,NED,ADS及一些主要的文献交叉验证,我们新发现了3个DZ白矮星、1个白矮主序双星、2个伴星为G型星的激变变星,3个激变变星的候选体、6个DC白矮星,1个DC白矮星候选体和1个 BL Lacertae(BL lac)候选体。还发现了1个有CaⅡ三重发射线和MgⅠ发射线的DA白矮星和1个光谱上表现出发射线的晚M恒星但测光图上像是一个星云或星系。     

基于神经网络的模板匹配方法求正常星系红移

星系通常分为正常星系(NG)与活动星系(AG)两类。文章提出了一种自动获取NG红移的快速有效方法: (1) 由NG模板根据红移范围Ⅰ: 0.0～0.3与Ⅱ: 0.3～0.5模拟得到两类星系样本, 进行PCA变换获得样本特征向量; (2) 利用概率神经网络设计两类样本特征向量的Bayes分类器; (3) 对于实际NG光谱数据, 利用Bayes分类器进行分类确定其红移的范围, 然后在此范围内进行模板匹配得到红移的准确值。与在整个红移范围内的模板匹配方法相比, 此方法不但节省了50%的模板匹配运算量, 而且还大大提高了红移值测量的精度。文章研究结果对于大型光谱巡天所产生的海量数据的自动处理具有重要意义。     

Laboratory simulations of astrophysical blast waves with high energy lasers

We review recent efforts to simulate aspects of supernova remnants in laboratory experiments by creating energetic explosions with high energy lasers. High energy pulsed lasers are uniquely suited for these kinds of studies. By focusing a laser with pulse energy of a few joules to many hundreds of joules onto a solid target or into a dense gas target, explosive shock waves of very high Mach number can be created. With a well chosen set of laser and target parameters it has been shown by a number of groups that radiative blast waves can be produced. Such blast waves have dynamics dominated by radiation transport and exhibit unusual characteristics, the most important of which include hydrodynamic instabilities which may play an important role in the structure of the interstellar medium. As a result there are now prospects for gaining new insights into astronomical observations of supernova remnants by studying laboratory laser driven plasma systems.     

A method of searching for supernova candidates from massive galaxy spectra

This paper presents a novel spectroscopic method for searching for supernova candidates from massive galaxy spectra,which is expected to be applied to the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST).This method includes mainly five steps.The first step is spectral preprocessing,including removing spectral noise using wavelet transform,spectral de-redshift,etc.The second step is decomposition of galactic spectra;we can get the galaxy component and supernova component and calculate the Supernova Statistical Characterization Vector (SNSCV) of each galaxy spectrum.The third step is to decrease samples in all the galaxy spectral datasets according to SNSCV of each spectrum,and to use the LOF (Local Outlier Factor)-based outlier detection algorithm to obtain the preliminary selected spectral data.The fourth step is template matching by cross-correlation,according to the matched results we get the secondary selected spectral data.Finally,we choose the final supernova candidates manually through checking the spectral features characteristic of a supernova.By the spectroscopic method proposed in this paper,thirty-six supernova candidates have been detected in a dataset including 294843 galaxy spectra from the Sloan Digital Sky Survey Data Release 7.Nine of these objects are detected first and the other twenty-seven have been reported in other publications (fifteen of which are detected and reported first by us).The twenty-four new super-nova candidates include twenty Ia type supernova candidates,three Ic type supernova candidates and one II type supernova candidate.     

SNO and supernovae

The Sudbury Neutrino Observatory (SNO) has unique capabilities as a supernova detector. In the event of a galactic supernova there are opportunities, with the data that SNO would collect, to constrain certain intrinsic neutrino properties significantly, to test details of the various models of supernova dynamics, and to provide prompt notification to the astronomical community through the Supernova Early Warning System (SNEWS). This paper consists of a discussion of these opportunities illustrated by some preliminary Monte Carlo results.     

Chemistry in space

Summary Molecules have been observed in widely disparate astronomical objects, from comets to supernova remnants. Interstellar space appear to be a real chemical laboratory, able to produce a large number of molecules, some of them relatively complex. The abundances show a very high sensitivity to local physical properties and dynamical history. This sensitivity renders molecular observations and astrochemical modelling very flexible tools for investigating the properties and evolution of the interstellar medium. A brief outline of morfology of interstellar medium is given, together with a discussion of the basic chemical processes leading to the molecular formation. Since interstellar-dust particles play a crucial role in the chemical evolution of interstellar medium, physical and chemical properties of dust are reviewed. Finally, by way of an example of modelling exercise, a toy model of the chemical evolution of interstellar gas is presented. Paper presented at the V Cosmic Physics National Conference, S. Miniato, November 27–30, 1990.     

Astrophysical constraints on particle properties

I summarize some astrophysical phenomenon like gamma ray bursters, astrophysical proof of the existence of blackholes, Active galactic nuclei — as high energy neutrino sources, and some unsolved issues in supernova. I touch on the aspects where novel particle properties (like neutrino mass and magnetic moment) are invoked to understand the astronomical observations.     

LAMOST-DR1中激变变星光谱的数据挖掘研究

LAMOST-DR1是郭守敬望远镜正式巡天发布的首批数据,其数量超过目前世界上所有已知恒星巡天项目的光谱总数。这为进一步扩大特殊和稀少天体如激变变星的数量提供了样本,同时也对天文数据处理方法和技术提出了更高的要求。针对LAMOST的数据特点,提出一种能够在海量天体光谱中自动、快速发现激变变星的方法。该方法使用拉普拉斯特征映射对天体光谱进行降维和重构。结果表明不同类别的天体光谱在拉普拉斯空间中能够得到较明显的区分。在使用粒子群算法对神经网络的参数进行优化后,对LAMOST-DR1的全部数据进行了自动识别。实验共发现了7个激变变星,经过证认,其中2个是矮新星,2个是类新星,1个是高度极化的武仙座AM型。这些光谱,补充了现有的激变变星光谱库。本文验证了拉普拉斯特征映射对天体光谱进行特征提取的有效性,为高维光谱进行降维提供了另一途径。在郭守敬望远镜正式发布的数据中寻找激变变星的首次尝试,实验结果表明该自动化的方法鲁棒性好,速度快,准确率高。该方法也可用于其他大型巡天望远镜的海量光谱处理。     

数字天顶摄影仪的图像处理

垂线偏差的测量过去一直受天文坐标测量困难的影响而效率低下.为了快速、精确地测得垂线偏差,提出了利用大面阵CCD构成数字天顶摄影仪,对拍摄的恒星影像进行搜索、细分定位处理,计算出测站点的天文坐标的基本原理;再根据GPS测得的大地坐标,解出垂线偏差.并且结合本系统自身的条件,提出了快速在大面阵CCD图象中搜索多个恒星目标的新方法.通过软件模拟,结果表明了该方案的可行性.     

Matter Effects on Neutrino Oscillations in Different Supernova Models

In recent years,with the development of simulations about supernova explosion,we have a better understanding about the density profiles and the shock waves in supernovae than before.There might be a reverse shock wave,another sudden change of density except the forward shock wave,or even no shock wave,emerging in the supernova.Instead of using the expression of the crossing probability at the high resonance,PH,we have studied the matter effects on neutrino oscillations in different supernova models.In detail,we have calculated the survival probability of νe(P_s)and the conversion probability of ν_x(P_c) in the Schrodinger equation within a simplified two-flavor framework for a certain case,in which the neutrino transfers through the supernova matter from an initial flavor eigenstate located at the core of the supernova.Our calculations was based on the data of density in three different supernova models obtained from simulations.In our work,we do not steepen the density gradient around the border of the shock wave,which differs to what was done in most of the other simulations.It is found that the mass and the density distribution of the supernova do make a difference on the behavior of P_s and P_c.With the results of P_s and P_c,we can estimate the number of νe(and ν_x) remained in the beam after they go through the matter in the supernova.     

Silicon molecules in space and in the laboratory

Despite the highly non-volatile nature of Si, silicon bearing molecules are a significant trace constituent of the gas associated with late-type stars and star forming regions, representing nearly 10% of the molecular species that have been identified in space. The early astronomical detection of SiO and its remarkable maser emission stimulated laboratory studies of other small silicon species that were crucial to the astronomical identifications. Although the pace of this laboratory work until recently has been slow, the application of Fourier transform microwave spectroscopy to supersonic molecular beams has overcome many experimental difficulties, yielding the detection of more than 20 new silicon chains and rings of astronomical interest during the past three years. Three of these, the rhomboidal ring Sic3 and the isovalent linear molecules SiCN and SiNC, have now been found in space, and with dedicated searches other recently discovered silicon molecules may soon be found. A number of exotic silicon molecules of astronomical and chemical interest should be amenable to laboratory detection with present techniques.     

Searching for neutrino bursts in the galaxy: 36 years of exposure

The Baksan Underground Scintillation Telescope has operated within the program of searching for neutrino bursts since the mid-1980s. We present the current status of the experiment and some results related to the investigation of background events and the stability of facility operation. Over the period from June 30, 1980, to December 31, 2016, the pure observation time was 31.27 years. No neutrino burst candidate event from the explosion of a core-collapse supernova in the Galaxy was recorded in this time. This sets an upper bound of 0.074 yr^–1 on the mean frequency of gravitational stellar collapses in the Galaxy at a 90% confidence level.