Automated morphological classification of galaxies based on projection gradient nonnegative matrix factorization algorithm

The development of automated morphological classification schemes can successfully distinguish between morphological types of galaxies and can be used for studies of the formation and subsequent evolution of galaxies in our universe. In this paper, we present a new automated machine supervised learning astronomical classification scheme based on the Nonnegative Matrix Factorization algorithm. This scheme is making distinctions between all types roughly corresponding to Hubble types such as elliptical, lenticulars, spiral, and irregular galaxies. The proposed algorithm is performed on two examples with different number of image (small dataset contains 110 image and large dataset contains 700 images). The experimental results show that galaxy images from EFIGI catalog can be classified automatically with an accuracy of ～93% for small and ～92% for large number. These results are in good agreement when compared with the visual classifications.     

The scatter in the near-infrared colour–magnitude relation in spiral galaxies

We have determined a dust-free colour–magnitude (CM) relation for spiral galaxies, by using I  −  K colours in edge-on galaxies above the plane. We find that the scatter in this relation is small and approximately as large as can be explained by observational uncertainties. The slope of the near-IR CM relation is steeper for spirals than for elliptical galaxies. We suggest two possible explanations. First, the difference could be caused by vertical colour gradients in spiral galaxies. In that case these gradients should be similar for all galaxies, on average ∼0.15 dex in [Fe/H] per scaleheight, and should increase for later galaxy types. The most likely explanation, however, is that spirals and ellipticals have intrinsically different CM relations. This means that the stars in spirals are younger than those in ellipticals. The age, however, or the fraction of young stars in spiral galaxies would be determined solely by the luminosity of the galaxy, and not by its environment.     

Mass decomposition of galaxies using DECA software package

The new DECA software package, which is designed to perform photometric analysis of the images of disk and elliptical galaxies having a regular structure, is presented. DECA is written in Python interpreted language and combines the capabilities of several widely used packages for astronomical data processing such as IRAF, SExtractor, and the GALFIT code used to perform two-dimensional decomposition of galaxy images into several photometric components (bulge+disk). DECA has the advantage that it can be applied to large samples of galaxies with different orientations with respect to the line of sight (including edge-on galaxies) and requires minimum human intervention. Examples of using the package to study a sample of simulated galaxy images and a sample of real objects are shown to demonstrate that DECA can be a reliable tool for the study of the structure of galaxies.     

The shapes of galaxies in the Sloan Digital Sky Survey

We determine the underlying shapes of spiral and elliptical galaxies in the Sloan Digital Sky Survey Data Release 6 (SDSS DR6) from the observed distribution of projected galaxy shapes, taking into account the effects of dust extinction and reddening. We assume that the underlying shapes of spirals and ellipticals are well approximated by triaxial ellipsoids. The elliptical galaxy data are consistent with oblate spheroids, with a correlation between luminosity and ellipticity: the mean values of minor to middle axis ratios are 0.41 ± 0.03 for   M _r ≈−18  ellipticals and 0.76 ± 0.04 for   M _r ≈−22.5  ellipticals. Ellipticals show almost no dependence of axial ratio on galaxy colour, implying a negligible dust optical depth.
There is a strong variation of spiral galaxy shapes with colour indicating the presence of dust. The intrinsic shapes of spiral galaxies in the SDSS DR6 are consistent with flat discs with a mean and dispersion of thickness to diameter ratio of (21 ± 2) per cent, and a face-on ellipticity, e , of  ln( e ) =−2.33 ± 0.79  . Not including the effects of dust in the model leads to discs that are systematically rounder by up to 60 per cent. More luminous spiral galaxies tend to have thicker and rounder discs than lower luminosity spirals. Both elliptical and spiral galaxies tend to be rounder for larger galaxies.
The marginalized value of the edge-on r -band dust extinction E ₀ in spiral galaxies is   E ₀≃ 0.45  mag for galaxies of median colours, increasing to   E ₀= 1  mag for   g − r > 0.9  and   E ₀= 1.9  for the luminous and most compact galaxies, with half-light radii  <2  h ⁻¹ kpc  .     

The Milky Way Galaxy as a strong gravitational lens

We study the gravitational lensing effects of spiral galaxies by taking a model of the Milky Way and computing its lensing properties. The model is composed of a spherical Hernquist bulge, a Miyamoto–Nagai disc and an isothermal halo. As a strong lens, a spiral galaxy like the Milky Way can give rise to four different imaging geometries. They are (i) three images on one side of the galaxy centre ('disc triplets'), (ii) three images with one close to the centre ('core triplets'), (iii) five images and (iv) seven images. Neglecting magnification bias, we show that the core triplets, disc triplets and fivefold imaging are roughly equally likely. Even though our models contain edge-on discs, their image multiplicities are not dominated by disc triplets. The halo is included for completeness, but it has a small effect on the caustic structure, the time delays and brightnesses of the images.
The Milky Way model has a maximum disc (i.e. the halo is not dynamically important in the inner parts). Strong lensing by nearly edge-on disc galaxies breaks the degeneracy between the relative contributions of the disc and halo to the overall rotation curve. If a spiral galaxy has a submaximum disc, then the astroid caustic shrinks dramatically in size, whilst the radial caustic shrinks more modestly. This causes changes in the relative likelihood of the image geometries, specifically (i) core triplets are now ∼9/2 times more likely than disc triplets, (ii) the cross-section for threefold imaging is reduced by a factor of ∼2/3, whilst (iii) the cross-section for fivefold imaging is reduced by ∼1/2. Although multiple imaging is less likely (the cross-sections are smaller), the average total magnification is greater. The time delays are smaller, as the total projected lensing mass is reduced.     

Point spread function tails and the measurements of diffuse stellar halo light around edge-on disc galaxies

Measuring the integrated stellar halo light around galaxies is very challenging. The surface brightness of these haloes is expected to be many magnitudes below dark sky and the central brightness of the galaxy. Here, I show that in some of the recent literature the effect of very extended Point Spread Function (PSF) tails on the measurements of halo light has been underestimated; especially in the case of edge-on disc galaxies. The detection of a halo along the minor axis of an edge-on galaxy in the Hubble Ultra Deep Field can largely be explained by scattered galaxy light. Similarly, depending on filter and the shape one assumes for the uncertain extended PSF, 20–80 per cent of the halo light found along the minor axis of scaled and stacked Sloan Digital Sky Survey (SDSS) edge-on galaxy images can be explained by scattered galaxy light. Scattered light also significantly contributes to the anomalous halo colours of stacked SDSS images. The scattered light fraction decreases when looking in the quadrants away from the minor axis. The remaining excess light is well modelled with a Sérsic profile halo with shape parameters based on star count halo detections of nearby galaxies. Even though, the contribution from PSF scattered light does not fully remove the need for extended components around these edge-on galaxies, it will be very challenging to make accurate halo light shape and colour measurements from integrated light without very careful PSF measurements and scattered light modelling.     

The onset of warps in Spitzer observations of edge-on spiral galaxies

We analyse warps in the nearby edge-on spiral galaxies observed in the Spitzer /Infrared Array Camera (IRAC) 4.5-μm band. In our sample of 24 galaxies, we find evidence of warp in 14 galaxies. We estimate the observed onset radii for the warps in a subsample of 10 galaxies. The dark matter distribution in each of these galaxies are calculated using the mass distribution derived from the observed light distribution and the observed rotation curves. The theoretical predictions of the onset radii for the warps are then derived by applying a self-consistent linear response theory to the obtained mass models for six galaxies with rotation curves in the literature. By comparing the observed onset radii to the theoretical ones, we find that discs with constant thickness can not explain the observations; moderately flaring discs are needed. The required flaring is consistent with the observations. Our analysis shows that the onset of warp is not symmetric in our sample of galaxies. We define a new quantity called the onset-asymmetry index and study its dependence on galaxy properties. The onset asymmetries in warps tend to be larger in galaxies with smaller disc scalelengths. We also define and quantify the global asymmetry in the stellar light distribution, that we call the edge-on asymmetry in edge-on galaxies. It is shown that in most cases the onset asymmetry in warp is actually anticorrelated with the measured edge-on asymmetry in our sample of edge-on galaxies and this could plausibly indicate that the surrounding dark matter distribution is asymmetric.     

Spin vector orientation of galaxies in the region 15h 48m≤α(2000) ≤ 19h 28m, −68°≤δ(2000) ≤−62°

We present an analysis of the orientations of 1433 galaxies found in the region  15^h 48^m≤α(2000) ≤ 19^h 28^m, −68°≤δ(2000) ≤−62°  . In this region we investigated three Abell clusters (S0794, S0797, S0805) of richness Class 0 and the Triangulum Australis cluster. Our aim is to examine non-random effects in galaxy orientations in clusters. In addition, we classified the investigated galaxies into subsamples on the basis of their axial ratio, major diameter and morphology. The spin vector orientations of total galaxies in the investigated region is found to be random. No preferred orientation is found in the clusters. We could not note any morphological dependence of the galaxy orientations in our samples. No preferred orientations can be seen for the spiral galaxies. The morphologically unidentified galaxies, galaxies having major diameters of <47 arcsec, and the nearly edge-on galaxies  ( b / a < 0.5, 0.4 < b / a ≤ 0.5)  show anisotropy: spin vectors of galaxies tend to be oriented perpendicular to the Local Supercluster plane and spin vector projections tend to point radially with respect to the Virgo cluster centre.     

Vertical Scale Parameter Estimates for 48 Non-edge-on Spiral Galaxies

In the first paper of this series, we directly studied the mathematical forms, symmetry of spiral structure, and the projection of galactic discs on the images, and measured the pitch angles of the spiral arms and inclination angles of the galactic discs for 60 spiral galaxies. In this second paper, we estimate the vertical scale parameters of 48 non-edge-on spiral galaxies based on the method proposed by Peng et al. and on the results given in Paper Ⅰ. As we know, for edge-on disc galaxies we can obtain the vertical scale parameter from the photometry, once a mathematical form is specified for the vertical light distribution. For non-edgeon galaxies, some other methods have to be used. The statistical result was that the vertical scale parameter is comparable for edge-on and non-edge-on galaxies, although it is obtained from two very different methods.     

Stellar Population Analysis of Galaxies based on Genetic Algorithms

We present a new method for determining the age and relative contribution of different stellar populations in galaxies based on the genetic algorithm. We apply this method to the barred spiral galaxy NGC 3384, using CCD images in U, B, V, R and I bands. This analysis indicates that the galaxy NGC 3384 is mainly inhabited by old stellar population (age >109yr). Some problems were encountered when numerical simulations are used for determining the contribution of different stellar populations in the integrated color of a galaxy. The results show that the proposed genetic algorithm can search efficiently through the very large space of the possible ages.     

The global structure of galactic discs

A statistical study of global galaxy parameters can help to improve our understanding of galaxy formation processes. In this paper we present the analysis of global galaxy parameters based on optical and near-infrared observations of a large sample of edge-on disc galaxies. We found a correlation between the ratio of the radial to vertical scale parameter and galaxy type: galaxies become systematically thinner when going from type S0 to type Sc, whereas the distribution seems to level off for later types. The observed scalelength ratios (and thus the radial colour gradients) largely represent the dust content of the galaxies. On average, the colour gradients indicated by the scalelength ratios increase from type Sa to at least type Sc. For galaxy types later than Sc, the average colour gradient seems to decrease again. The distribution of K -band (edge-on) disc central surface brightnesses is rather flat, although with a large scatter. However, the latest-type sample galaxies ( T  > 6) show an indication that their average disc central surface brightnesses may be fainter than those of the earlier types. This effect is probably not the result of dust extinction.     

Ellipticity of Nearby Elliptical Galaxies Identified with Extended Extragalactic Radio Sources

The dependence of the ellipticity of nearby optical galaxies associated with extragalactic radio sources on the type of radio galaxy is considered. It is shown that elliptical galaxies identified with radio galaxies of the FRI type and with radio sources having a radio image with a small elongation have a smaller ellipticity, on the average, than those associated with radio galaxies of the FRII type and with radio sources with greater elongation. As one possible explanation for this phenomenon, one can assume that the optical galaxies associated with radio galaxies of type FRI and with radio sources of small elongation are oriented in space so that they are observed edge-on relatively rarely.     

Constraining the star formation histories of spiral bulges

Stellar populations in spiral bulges are investigated using the Lick system of spectral indices. Long-slit spectroscopic observations of line strengths and kinematics made along the minor axes of four spiral bulges are reported. Comparisons are made between central line strengths in spiral bulges and those in other morphological types [elliptical, spheroidal (Sph) and S0]. The bulges investigated are found to have central line strengths comparable to those of single stellar populations of approximately solar abundance or above. Negative radial gradients are observed in line strengths, similar to those exhibited by elliptical galaxies. The bulge data are also consistent with correlations between Mg₂, Mg₂ gradient and central velocity dispersion observed in elliptical galaxies. In contrast to elliptical galaxies, central line strengths lie within the loci defining the range of 〈Fe〉 and Mg₂ achieved by Worthey's solar abundance ratio, single stellar populations (SSPs). The implication of solar abundance ratios indicates significant differences in the star formation histories of spiral bulges and elliptical galaxies. A 'single zone with infall' model of galactic chemical evolution, using Worthey's SSPs, is used to constrain the possible star formation histories of our sample. We show that the 〈Fe〉, Mg₂ and H β line strengths observed in these bulges cannot be reproduced using primordial collapse models of formation but can be reproduced by models with extended infall of gas and star formation (2–17 Gyr) in the region modelled. One galaxy (NGC 5689) shows a central population with a luminosity-weighted average age of ∼5 Gyr, supporting the idea of extended star formation. Kinematic substructure, possibly associated with a central spike in metallicity, is observed at the centre of the Sa galaxy NGC 3623.     

Evolutionary synthesis models for galaxy transformation in clusters

The galaxy population in rich local galaxy clusters shows a ratio of one quarter elliptical galaxies, two quarters S0 galaxies, and one quarter spiral galaxies. Observations of clusters at redshift 0.5 show a perspicuously different ratio, the dominant galaxy type are spiral galaxies with a fraction of two quarters while the number of S0 galaxies decreases to a fraction of one quarter (Dressler et al. 1997). This shows an evolution of the galaxy population in clusters since redshift 0.5 and it has been suspected that galaxy transformation processes during the infall into a cluster are responsible for this change. These could be merging, starburst or ram-pressure stripping. We use our evolutionary synthesis models to describe various possible effects of those interactions on the star formation of spiral galaxies infalling into clusters. We study the effects of starbursts of various strengths as well as of the truncation of star formation at various epochs on the color and luminosity evolution of model galaxies of various spectral types. As a first application we present the comparison of our models with observed properties of the local S0 galaxy population to constrain possible S0 formation mechanisms in clusters. Application to other types of galaxies is planned for the future. This revised version was published online in August 2006 with corrections to the Cover Date.     

Ultraviolet-background-induced bifurcation of galactic morphology

Based upon a novel paradigm of galaxy formation under the influence of an ultraviolet background, the evolutionary bifurcation of pre-galactic clouds is compared with observations of elliptical and spiral galaxies. The theory predicts that the dichotomy between dissipational and dissipationless galaxy formation stems from the degree of self-shielding from the ultraviolet background radiation. This is demonstrated on a bifurcation diagram of collapse epochs versus masses of pre-galactic clouds. Using the observed properties, the collapse epochs and mass are assessed for each type of galaxy . By direct comparison of the theory with observations, it turns out that the theoretical bifurcation branch successfully discriminates between elliptical and spiral galaxies. This suggests that the ultraviolet background radiation could play a profound role in the differentiation of galactic morphology into the Hubble sequence.     

Near-infrared imaging of the host galaxies of BL Lacertae objects

Near-infrared images have been obtained of nine BL Lacertae objects in order to investigate their host galaxy properties. From numerical modelling of the data we find that five of the nine BL Lacs have contributions from extended galaxy emission in their profiles. Tentative morphologies were derived for two of the BL Lacs, namely B2 0752+258, for which a disc morphology is slightly preferred, and S4 0954+65, for which an elliptical morphology is preferred. From our modelling we derive host galaxy absolute magnitudes of M_H =−25.6±0.5 for B2 0752+258, and M_H =−26.3±0.8 for S4 0954+65. We also find an average K -band absolute K -corrected host galaxy magnitude, for the BL Lacs, of 〈 M_K 〉=−26.3±0.6 asssuming an elliptical galaxy model, and 〈 M_K 〉=−26.1±0.9 assuming a disc galaxy model. The derived absolute magnitudes are similar to those found for the putative parent population of FRI radio galaxies, predicted by unified schemes. For those BL Lacs in which host galaxies have been previously detected at optical wavelengths we derive rest frame colours which are generally consistent with those of 'normal' galaxies. However, for TEX 0836+182 we find a rather blue colour which, together with the elongated morphology, may be indicative of interaction or spiral structure. Our results suggest that the host galaxies of BL Lacs are indeed similar to those of FRI radio galaxies, adding further weight to the unification scenario. In our present infrared frames of relatively modest resolution, we do not find that infrared observations are any more or less effective than optical observations at revealing the underlying host galaxy.     

Imaging of the protoelliptical NGC 1700 and its globular cluster system

An excellent candidate for a young elliptical, or 'protoelliptical' galaxy is NGC 1700. Here we present new B -, V - and I -band imaging using the Keck telescope, and reanalyse existing V - and I -band images from the Hubble Space Telescope . After subtracting a model of the galaxy from the Keck images, NGC 1700 reveals two symmetric tidal tail-like structures. If this interpretation is correct, it suggests a past merger event involving two spiral galaxies. These tails are largely responsible for the 'boxiness' of the galaxy isophotes observed at a radius of ∼13 kpc.
We also show that the B − I colour distribution of the globular cluster system is bimodal. The mean colour of the blue population is consistent with that of old Galactic globular clusters. Relative to this old, metal-poor population, we find that the red population is younger and more metal-rich. This young population has an age and metallicity similar to that inferred for the central stars, suggesting that both populations are associated with an episode of star formation triggered by the merger that may have formed the galaxy. We find that, although they have large errors, the majority of the age estimates of NGC 1700 are reasonably consistent and we adopt a 'best estimate' for the age of 3.0±1.0 Gyr. This relatively low age places NGC 1700 within the age range where there is a notable lack of obvious candidates for protoellipticals. The total globular cluster specific frequency is rather low for a typical elliptical, even after taking into account the fading of the galaxy over the next 10 Gyr. We speculate that NGC 1700 will eventually form a relatively 'globular cluster poor' elliptical galaxy.     

基于SE-Inception-v3的星系形态分类模型

随着天文探测技术的快速发展, 海量的星系图像数据不断产生, 能够及时高效地对星系图像进行形态分类对研究星系的形成与演化至关重要. 针对传统的星系形态分类模型特征选择困难、分类速度慢、准确率受限等难题, 提出一种以Inception-v3神经网络为主干结构, 融合压缩激励(Squeeze and Excitation Network, SE)通道注意力机制的星系形态分类模型. 该模型在斯隆数字巡天(Sloan Digital Sky Survey, SDSS)样本的测试集准确率高达99.37%. 旋涡星系、圆形星系、中间星系、雪茄状星系与侧向星系的F1值分别为99.33%、99.58%、99.33%、99.41%与99.16%. 该模型与Inception-v3、MobileNet (Mobile Neural Network)和ResNet (Residual Neural Network)网络模型相比, SE-Inception-v3宽度和深度优势表现出更强的特征提取能力, 可以高效识别不同形态的星系, 为未来大型巡天计划的大规模星系形态分类问题提供了一种新方法.     

EfficientNetV2-S-Triplet7: 一种改进的星系newline 形态学分类算法

星系的形态与星系的形成和演化息息相关, 其形态学分类是星系天文学后续研究的重要一环. 当前海量天文观测数据的出现使得天文数据自动分析方法越来越得到重视, 针对此问题, 利用先进的深度学习骨干网络EfficientNetV2, 分析不同的注意力机制类型和使用节点对网络性能的影响, 构建了一种命名为EfficientNetV2-S-Triplet7 (即在EfficientNetV2-S stage7的$1\times1$卷积层后加入Triplet模块)的改进算法模型来实现星系形态学的自动分类. 使用第二期星系动物园(Galaxy Zoo 2, GZ2)中超过24万张的测光图像作为初始数据进行实验测试. 在对数据进行预处理时采取了尺寸抖动、翻转、色彩畸变等图像增强手段来解决图像数量的不平衡问题. 在同一系列经典和前沿的深度学习算法模型AlexNet、ResNet-34、MobileNetV2、RegNet进行对比实验后, 得出EfficientNetV2-S-Triplet7算法在分类准确率、查全率和F1分数等指标上具有最好的测试结果. 在9375张测试图像中的3项指标值分别可达到89.03%、90.21%、89.93%, 查准率达到89.69%, 在其他模型中排在第3位. 该结果表明将EfficientNetV2-S-Triplet7算法应用于大规模星系数据的形态学分类任务中有很好的效果.     

Nucleated dwarf elliptical galaxies in the Virgo cluster

Using images from a charge-coupled device survey with the Wide Field Camera on the Isaac Newton Telescope, we performed B - and I -band photometry on 156 Virgo cluster dwarf elliptical (dE) galaxies, 25 candidate new cluster dwarfs, and nine candidate field dwarfs. Galaxies were modelled with Sérsic profiles, using both 1D χ² and 2D cross-correlation methods, with nuclei modelled as point sources. The intensity profiles of 50 galaxies previously classified as dE, dE?, or ? are more accurately fitted if a nucleus is included, and this results in the majority of dwarfs now being classified as nucleated dwarf ellipticals (dE,N). Some faint galaxies with B magnitudes of 18–21 have particularly large relative nuclei, while a small number have apparent central dimmings. For cluster dE,N galaxies the nucleus magnitude is correlated with the magnitude of the host galaxy. The profile parameters of dE and dE,N galaxies are not significantly different, and there is no evident discontinuity in relative nucleus size between non-nucleated and nucleated dwarfs, suggesting that they may form a continuum. Nuclei are on average redder than their underlying galaxies, though a spread of relative colours was found, and two-fifths of nuclei are bluer. Formation mechanisms of nuclei are discussed: at least some appear to have formed in an already existing non-nucleated galaxy, though others may have formed simultaneously with their galaxies and subsequently evolved within them.