恒星大气物理参量的非参数估计方法

恒星大气物理参量（有效温度、表面重力、化学丰度）是导致恒星光谱差异的主要因素．恒星大气物理参量的自动测量是LAMOST等大规模巡天望远镜所产生的海量天体光谱数据自动处理中一个重要研究内容．针对测量大样本的恒星光谱数据估计每个恒星的大气物理参量，提出了一种基于变窗宽核函数的估计算法：变窗宽算法是对固定窗宽算法的改进，分为3个步骤：（1）将历史恒星光谱数据进行PCA处理，得到光谱的低维特征数据；（2）利用特征数据与其物理参数的对应关系，建立一种变窗宽的非参数估计模型；（3）利用该估计模型，直接计算待测恒星光谱的3个物理参量（有效温度、表面重力、金属丰度）．实验结果表明：该方法与固定窗宽估计模型以及在其他文献中报道的方法相比，具有较高的估计精度和鲁棒性．     

The Cannon: A Data Driving Approach to Stellar Label Determination

The new generation of large sky area spectroscopic survey project has produced nearly 10 million low-resolution stellar spectra. Based on these spectroscopic data, this paper introduces a machine learning algorithm named The Cannon. This algorithm is completely based on the known spectroscopic data of stellar atmospheric parameters (effective temperature, surface gravity, and metal abundance, etc.), this algorithm builds the characteristic vector by means of data driving, and establishes the functional relation between spectral flux characteristics and stellar parameters. Then it is applied to the observed spectral data to calculate the atmospheric parameters. The main advantage of The Cannon is that it is not directly based on any stellar physical models, it has an even higher applicability. Moreover, because of the use of full-spectrum information, even for the spectra with a low signal-to-noise ratio (SNR), it still can obtain the parameter solutions of high reliability. This algorithm has significant advantages in the data processing and parameter determination of large-scale stellar spectra. In addition, this paper presents two examples of using The Cannon to obtain the stellar parameters of K and M giants from the LAMOST spectral data.     

The Cannon:一种基于光谱数据的恒星参数测量方法

新一代大规模光谱巡天项目产生了近千万条低分辨率恒星光谱,基于这些光谱数据,介绍一种名为The Cannon的机器学习方法。该方法完全基于已知恒星大气参数(有效温度、表面重力加速度和金属丰度等)的光谱数据,通过数据驱动来构建特征向量,建立光谱流量特征和恒星参数的函数对应关系,进而应用到观测光谱数据中,实现对恒星光谱的大气参数求解。The Cannon的主要优势为不直接基于任何恒星物理模型,适用性更广;由于使用了全谱信息,即便对于低信噪比光谱也能得到较高可信度的参数结果,该算法在大规模恒星光谱的数据处理和参数求解方面具有明显的优势。此外,还利用The Cannon得到LAMOST光谱数据中K巨星和M巨星的恒星参数。     

Recovering physical parameters from galaxy spectra using MOPED

We derive physical parameters of galaxies from their observed spectra using MOPED, the optimized data compression algorithm of Heavens, Jimenez & Lahav. Here we concentrate on parametrizing galaxy properties, and apply the method to the NGC galaxies in Kennicutt's spectral atlas. We focus on deriving the star formation history, metallicity and dust content of galaxies. The method is very fast, taking a few seconds of CPU time to estimate ∼17 parameters, and is therefore specially suited to studying large data sets, such as the Anglo-Australian two-degree-field (2dF) galaxy survey and the Sloan Digital Sky Survey (SDSS). Without the power of MOPED, the recovery of star formation histories in these surveys would be impractical. In Kennicutt's atlas, we find that for the spheroidals a small recent burst of star formation is required to provide the best fit to the spectrum. There is clearly a need for theoretical stellar atmospheric models with spectral resolution better than 1 Å if we are to extract all the rich information that large redshift surveys contain in their galaxy spectra.     

Spectral Synthesis via Mean Field approach to Independent Component Analysis

We apply a new statistical analysis technique, the Mean Field approach to Independent Component Analysis(MF-ICA) in a Bayseian framework, to galaxy spectral analysis. This algorithm can compress a stellar spectral library into a few Independent Components(ICs), and the galaxy spectrum can be reconstructed by these ICs. Compared to other algorithms which decompose a galaxy spectrum into a combination of several simple stellar populations, the MF-ICA approach offers a large improvement in efficiency. To check the reliability of this spectral analysis method, three different methods are used:(1) parameter recovery for simulated galaxies,(2) comparison with parameters estimated by other methods, and(3) consistency test of parameters derived with galaxies from the Sloan Digital Sky Survey. We find that our MF-ICA method can not only fit the observed galaxy spectra efficiently, but can also accurately recover the physical parameters of galaxies. We also apply our spectral analysis method to the DEEP2 spectroscopic data, and find it can provide excellent fitting results for low signal-to-noise spectra.     

History of Star Formation of the Galaxy Cluster Abell 85

Abell 85 is a cD galaxy cluster in the southern hemisphere and has a redshift of 0.055. Based on the spectra of 242 member galaxies provided by the Sloan spectral survey data, using the stellar population constituents and star formation history of these member galaxies obtained from the population synthesis software STARLIGHT, we study the regularities of the variations of star formation properties of galaxies (such as the ages, metal abundances and star formation rates of the characteristic stellar populations) with the local surface density of galaxies. As revealed by the results, the galaxies situated in the highdensity environments of the central region of the cluster possess higher population ages and metal abundances, and their rates of star formation are rather low, the recent activities of star formation are obviously suppressed. Besides, the correlations of the galaxy metal abundance and speci?c star formation rate with the stellar mass are asserted.     

A discussion on metallicity effects on the period-luminosity relations of Cepheids and Miras

We present an automatic, fast, accurate and robust method of classifying astronomical objects. The Self Organizing Map (SOM) as an unsupervised Artificial Neural Network (ANN) algorithm is used for classification of stellar spectra of stars. The SOM is used to make clusters of different spectral classes of Jacoby, Hunter and Christian (JHC) library. This ANN technique needs no training examples and the stellar spectral data sets are directly fed to the network for the classification. The JHC library contains 161 spectra out of which, 158 spectra are selected for the classification. These 158 spectra are input vectors to the network and mapped into a two dimensional output grid. The input vectors close to each other are mapped into the same or neighboring neurons in the output space. So, the similar objects are making clusters in the output map and making it easy to analyze high dimensional data.     

A Method of Estimating the [α/Fe] Ratios from Low-resolution Stellar Spectratwo

A method for the determination of [α/Fe] from low-resolution stellar spectra is presented. The proposed scheme includes the following three steps: firstly, the spectrum is decomposed by the multi-scale Haar wavelet, and the high-frequency components are removed to suppress the high-frequency noise; then, based on the correlation of the spectral data component with [α/Fe], the spectral features are selected by the LASSO (Least Absolute Shrinkage and Selection Operator) algorithm; finally, [α/Fe] is measured by the multiple linear regression method based on the MARCS stellar spectrum library. The effectiveness of the method is verified with the low-resolution stellar spectra of ELODIE, SDSS (Sloan Digital Sky Survey), LAMOST (Large Sky Area Multi-Object Fibre Spectroscopic Telescope), and four star clusters. The systematic deviations and accuracies are as follows: (0.04 dex, 0.064 dex) for the 317 ELODIE spectra; (0.16 dex, 0.065 dex) for the 412 SDSS spectra; (0.05 dex, 0.062 dex) for the 1276 LAMOST spectra (with the signal-noise ratio in the g band (SNRG) greater than 20). The averages of [α/Fe] obtained for the likely members of the globular star clusters (M13, M15) and open star clusters (NGC2420, M67) are in agreement with the literature values.     

Estimating stellar atmospheric parameters based on Lasso features

With the rapid development of large scale sky surveys like the Sloan Digital Sky Survey （SDSS）, GAIA and LAMOST （Guoshoujing telescope）, stellar spectra can be obtained on an ever-increasing scale. Therefore, it is necessary to estimate stel- lar atmospheric parameters such as Teff, log g and [Fe/H] automatically to achieve the scientific goals and make full use of the potential value of these observations. Feature selection plays a key role in the automatic measurement of atmospheric parameters. We propose to use the least absolute shrinkage selection operator （Lasso） algorithm to select features from stellar spectra. Feature selection can reduce redundancy in spectra, alleviate the influence of noise, improve calculation speed and enhance the robustness of the estimation system. Based on the extracted features, stellar atmospheric param- eters are estimated by the support vector regression model. Three typical schemes are evaluated on spectral data from both the ELODIE library and SDSS. Experimental results show the potential performance to a certain degree. In addition, results show that our method is stable when applied to different spectra.     

The sun‐as‐a‐star solar spectrum

The Sun is the only star for which individual surface features can be observed directly. For other stars, the properties of starspots, stellar rotation, stellar flares, etc, are derived indirectly via variation of star‐integrated spectral line profiles or their luminosity measurements. Solar disk‐integrated and disk‐resolved observations allow for investigations of the contribution of individual solar disk features to sun‐as‐a‐star spectra. Here, we provide a brief overview of three sun‐as‐a‐star programs, currently in operation, and describe recent improvements in observations and data reduction for the Integrated Sunlight Spectrometer (ISS), one of three instruments comprising the Synoptic Optical Long‐term Investigations of the Sun (SOLIS) system. Next, we discuss studies employing sun‐as‐a‐star observations (including Ca II K line as proxy for total unsigned magnetic flux and 2800 MHz radio flux) as well as the effects of flares on solar disk‐integrated spectra. (© 2014 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Spectral disentangling of the triple system DG Leo: orbits and chemical composition

DG Leo is a spectroscopic triple system composed of three stars of late-A spectral type, one of which was suggested to be a δ Scuti star. Seven nights of observations at high spectral and high time-resolution at the Observatoire de Haute-Provence with the ELODIE spectrograph were used to obtain the component spectra by applying a Fourier transform spectral disentangling technique. Comparing these with synthetic spectra, the stellar fundamental parameters (effective temperature, surface gravity, projected rotation velocity and chemical composition) are derived. The inner binary consists of two Am components, at least one of which is not yet rotating synchronously at the orbital period though the orbit is a circular one. The distant third component is confirmed to be a δ Scuti star with normal chemical composition.     

Population synthesis of H ii galaxies

We study the stellar population of galaxies with active star formation, determining ages of the stellar components by means of spectral population synthesis of their absorption spectra. The data consist of optical spectra of 185 nearby ( z 0.075) emission-line galaxies . They are mostly H  ii galaxies, but we also include some starbursts and Seyfert 2s, for comparison purposes. They were grouped into 19 high signal-to-noise ratio template spectra, according to their continuum distribution, absorption- and emission-line characteristics. The templates were then synthesized with a star cluster spectral base.
The synthesis results indicate that H  ii galaxies are typically age-composite stellar systems, presenting important contributions from generations up to as old as 500 Myr. We detect a significant contribution of populations with ages older than 1 Gyr in two groups of H  ii galaxies. The age distributions of stellar populations among starbursts can vary considerably despite similarities in the emission-line spectra. In the case of Seyfert 2 groups we obtain important contributions from the old population, consistent with a bulge.
From the diversity of star formation histories, we conclude that typical H  ii galaxies in the local Universe are not systems presently forming their first stellar generation.     

Spectral classification of stars based on LAMOST spectra

In this work, we select spectra of stars with high signal-to-noise ratio from LAMOST data and map their MK classes to the spectral features. The equivalent widths of prominent spectral lines, which play a similar role as multi-color photometry, form a clean stellar locus well ordered by MK classes. The advantage of the stellar locus in line indices is that it gives a natural and continuous classification of stars consistent with either broadly used MK classes or stellar astrophysical parameters. We also employ an SVM-based classification algorithm to assign MK classes to LAMOST stellar spectra. We find that the completenesses of the classifications are up to 90% for A and G type stars, but they are down to about 50% for OB and K type stars. About 40% of the OB and K type stars are mis-classified as A and G type stars,respectively. This is likely due to the difference in the spectral features between late B type and early A type stars or between late G and early K type stars being very weak. The relatively poor performance of the automatic MK classification with SVM suggests that the direct use of line indices to classify stars is likely a more preferable choice.     

基于胶囊网络的恒星光谱分类研究

大型巡天项目的快速发展,产生大量的恒星光谱数据,也使得实现恒星光谱数据的自动分类成为一项具有挑战性的工作.提出一种新的基于胶囊网络的恒星光谱分类方法,首先利用1维卷积网络和短时傅里叶变换将来源于LAMOST(Large Sky Area Multi-Object Fiber Spectroscopy Telescope)Data Release 5(DR5)的F5、G5、K5型1维恒星光谱转化成2维傅里叶谱图像,再通过胶囊网络对2维谱图像进行自动分类.由于胶囊网络具有保留图像中实体之间的分层位姿关系和无需池化层的优点,实验结果表明:胶囊网络具有较好的分类性能,对于F5、G5、K5型恒星光谱的分类,准确率优于其他分类方法.     

Stellar parameters of young brown dwarfs

We present X‐shooter observations of two brown dwarf candidates. We focus on the determination of stellar parameters and their errors. The targets, an accreting class II and a non‐accreting class III objects, are members of the σ Orionis star‐forming region. We derive the spectroscopic spectral types from the VIS spectrum and the stellar parameters. We find that the uncertainties on the stellar parameters have a minor effect on the determination of the mass accretion rate for the accreting star, thus confirming that the discrepancies between the mass accretion rate estimates found with different (simultaneous) tracers are probably due to different physical conditions where the accretion/wind indicators are produced (© 2011 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Stellar Spectral Classification with Locality Preserving Projections and Support Vector Machine

With the help of computer tools and algorithms, automatic stellar spectral classification has become an area of current interest. The process of stellar spectral classification mainly includes two steps: dimension reduction and classification. As a popular dimensionality reduction technique, Principal Component Analysis (PCA) is widely used in stellar spectra classification. Another dimensionality reduction technique, Locality Preserving Projections (LPP) has not been widely used in astronomy. The advantage of LPP is that it can preserve the local structure of the data after dimensionality reduction. In view of this, we investigate how to apply LPP+SVM in classifying the stellar spectral subclasses. In the comparative experiment, the performance of LPP is compared with PCA. The stellar spectral classification process is composed of the following steps. Firstly, PCA and LPP are respectively applied to reduce the dimension of spectra data. Then, Support Vector Machine (SVM) is used to classify the 4 subclasses of K-type and 3 subclasses of F-type spectra from Sloan Digital Sky Survey (SDSS). Lastly, the performance of LPP+SVM is compared with that of PCA+SVM in stellar spectral classification, and we found that LPP does better than PCA.     

EChO spectra and stellar activity II. The case of dM stars

EChO is a dedicated mission to investigate exoplanetary atmospheres. When extracting the planetary signal, one has to take care of the variability of the hosting star, which introduces spectral distortion that can be mistaken as planetary signal. Magneticvariability has to be taken into account in particular for M stars. To this purpose, assuming a one spot dominant model for the stellar photosphere, we develop a mixed observational-theoretical tool to extract the spot’s parameters from the observed optical spectrum. This method relies on a robust library of spectral M templates, which we derive using the observed spectra of quiet M dwarfs in the SDSS database. Our procedure allows to correct the observed spectra for photospheric activity in most of the analyzed cases, reducing the spectral distortion down to the noise levels. Ongoing refinements of the template library and the algorithm will improve the efficiency of our algorithm.     

High-resolution spectroscopic observations of post-asymptotic giant branch candidates from the Edinburgh–Cape Survey

High spectral resolution  ( R ∼ 40 000)  and signal-to-noise optical spectra, obtained at the Very Large Telescope (VLT), are presented for three post–asymptotic giant branch (AGB) candidates selected from the Edinburgh–Cape (EC) Faint Blue Object Survey. The stellar atmospheric parameters and chemical compositions, derived using sophisticated non-local thermodynamic equilibrium calculations, reveal that EC 14102-1337 and EC 20068-7324 are both in an evolved post–horizontal branch (HB) evolutionary state. However, EC 11507-2253 is most likely a post-AGB star.     

Parameterizing Stellar Spectra Using Deep Neural Networks

Large-scale sky surveys are observing massive amounts of stellar spectra.The large number of stellar spectra makes it necessary to automatically parameterize spectral data,which in turn helps in statistically exploring properties related to the atmospheric parameters.This work focuses on designing an automatic scheme to estimate effective temperature(T_(eff)),surface gravity(log g) and metallicity[Fe/H] from stellar spectra.A scheme based on three deep neural networks(DNNs) is proposed.This scheme consists of the following three procedures:first,the configuration of a DNN is initialized using a series of autoencoder neural networks;second,the DNN is fine-tuned using a gradient descent scheme;third,three atmospheric parameters T_(eff),log g and [Fe/H] are estimated using the computed DNNs.The constructed DNN is a neural network with six layers(one input layer,one output layer and four hidden layers),for which the number of nodes in the six layers are 3821,1000,500,100,30 and 1,respectively.This proposed scheme was tested on both real spectra and theoretical spectra from Kurucz's new opacity distribution function models.Test errors are measured with mean absolute errors(MAEs).The errors on real spectra from the Sloan Digital Sky Survey(SDSS) are 0.1477,0.0048 and 0.1129 dex for log g,log T_(eff) and [Fe/H](64.85 K for T_(eff)),respectively.Regarding theoretical spectra from Kurucz's new opacity distribution function models,the MAE of the test errors are 0.0182,0.0011 and 0.0112 dex for log g,log T_(eff) and [Fe/H](14.90 K for T_(eff)),respectively.     

Stellar Spectral Classification Based on Capsule Network

The rapid development of large-scale sky survey project has produced a large amount of stellar spectral data, which make the automatic classification of stellar spectral data a challenging task. In this paper, we have proposed a stellar spectral classification method based on a capsule network. At first, by using the one-dimensional convolutional network and short-time Fourier transform (STFT), the one-dimensional spectra of the F5, G5, and K5 types selected from the LAMOST Data Release 5 (DR5) are converted into the two-dimensional Fourier spectrum images. Then, the two-dimensional Fourier spectrum images are classified automatically by the capsule network. Because the capsule network can preserve the hierarchical pose relationships among the entities in the image, and it does not need any pooling layers, the experimental results show that the capsule network has a better classification performance, for the classifications of the F5, G5, and K5-type stellar spectra, its classification accuracy is superior to other classification methods.