Contents
Images
Upload your image
DSS Images Other Images
Related articles
Random forest automated supervised classification of Hipparcos periodic variable stars
We present an evaluation of the performance of an automatedclassification of the Hipparcos periodic variable stars into 26 types.The sub-sample with the most reliable variability types available in theliterature is used to train supervised algorithms to characterize thetype dependencies on a number of attributes. The most useful attributesevaluated with the random forest methodology include, in decreasingorder of importance, the period, the amplitude, the V-I colour index,the absolute magnitude, the residual around the folded light-curvemodel, the magnitude distribution skewness and the amplitude of thesecond harmonic of the Fourier series model relative to that of thefundamental frequency. Random forests and a multi-stage scheme involvingBayesian network and Gaussian mixture methods lead to statisticallyequivalent results. In standard 10-fold cross-validation (CV)experiments, the rate of correct classification is between 90 and 100per cent, depending on the variability type. The main mis-classificationcases, up to a rate of about 10 per cent, arise due to confusion betweenSPB and ACV blue variables and between eclipsing binaries, ellipsoidalvariables and other variability types. Our training set and thepredicted types for the other Hipparcos periodic stars are availableonline.
| Bayesian inference of stellar parameters and interstellar extinction using parallaxes and multiband photometry
Astrometric surveys provide the opportunity to measure the absolutemagnitudes of large numbers of stars, but only if the individualline-of-sight extinctions are known. Unfortunately, extinction is highlydegenerate with stellar effective temperature when estimated frombroad-band optical/infrared photometry. To address this problem, Iintroduce a Bayesian method for estimating the intrinsic parameters of astar and its line-of-sight extinction. It uses both photometry andparallaxes in a self-consistent manner in order to provide anon-parametric posterior probability distribution over the parameters.The method makes explicit use of domain knowledge by employing theHertzsprung-Russell Diagram (HRD) to constrain solutions and to ensurethat they respect stellar physics. I first demonstrate this method byusing it to estimate effective temperature and extinction from BVJHKdata for a set of artificially reddened Hipparcos stars, for whichaccurate effective temperatures have been estimated from high-resolutionspectroscopy. Using just the four colours, we see the expected strongdegeneracy (positive correlation) between the temperature andextinction. Introducing the parallax, apparent magnitude and the HRDreduces this degeneracy and improves both the precision (reduces theerror bars) and the accuracy of the parameter estimates, the latter byabout 35 per cent. The resulting accuracy is about 200 K in temperatureand 0.2 mag in extinction. I then apply the method to estimate theseparameters and absolute magnitudes for some 47 000 F, G, K Hipparcosstars which have been cross-matched with Two-Micron All-Sky Survey(2MASS). The method can easily be extended to incorporate the estimationof other parameters, in particular metallicity and surface gravity,making it particularly suitable for the analysis of the 109stars from Gaia.
| New GCVS Versions for Three Southern Constellations
We are currently working on a version of the General Catalogue ofVariable Stars (GCVS) revised taking into account the new dataaccumulated since the 4th GCVS edition. A draft new version will bereleased for each constellation as soon as the work for theconstellation is finished. It will contain all stars of the 4th GCVSedition plus a complete catalogue of the stars added to the GCVS in theName Lists of Variable Stars Nos. 67 - 78. Now we are ready for thefirst release, containing more than 1300 variable stars in theconstellations of Antlia, Ara, and Telescopium. When preparing therelease, we actively used modern data-mining possibilities to improvevariability types and light elements. This paper introduces the firstrelease of the new GCVS version and presents new results (types, lightelements), based mainly on data mining, for 213 stars.
| Reprocessing the Hipparcos data of evolved stars. III. Revised Hipparcos period-luminosity relationship for galactic long-period variable stars
We analyze the K band luminosities of a sample of galactic long-periodvariables using parallaxes measured by the Hipparcos mission. Theparallaxes are in most cases re-computed from the Hipparcos IntermediateAstrometric Data using improved astrometric fits and chromaticitycorrections. The K band magnitudes are taken from the literature andfrom measurements by COBE, and are corrected for interstellar andcircumstellar extinction. The sample contains stars of several spectraltypes: M, S and C, and of several variability classes: Mira, semiregularSRa, and SRb. We find that the distribution of stars in theperiod-luminosity plane is independent of circumstellar chemistry, butthat the different variability types have different P-L distributions.Both the Mira variables and the SRb variables have reasonablywell-defined period-luminosity relationships, but with very differentslopes. The SRa variables are distributed between the two classes,suggesting that they are a mixture of Miras and SRb, rather than aseparate class of stars. New period-luminosity relationships are derivedbased on our revised Hipparcos parallaxes. The Miras show a similarperiod-luminosity relationship to that found for Large Magellanic CloudMiras by Feast et al. (\cite{Feast-1989:a}). The maximum absolute Kmagnitude of the sample is about -8.2 for both Miras and semi-regularstars, only slightly fainter than the expected AGB limit. We show thatthe stars with the longest periods (P>400 d) have high mass lossrates and are almost all Mira variables.Based on observations from the Hipparcos astrometric satellite operatedby the European Space Agency (ESA \cite{Hipparcos}).Table \ref{Tab:data1} is only available in electronic form at the CDSvia anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or viahttp://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/403/993
| How many Hipparcos Variability-Induced Movers are genuine binaries?
Hipparcos observations of some variable stars, and especially oflong-period (e.g. Mira) variables, reveal a motion of the photocentercorrelated with the brightness variation (variability-induced mover -VIM), suggesting the presence of a binary companion. A re-analysis ofthe Hipparcos photometric and astrometric data does not confirm the VIMsolution for 62 among the 288 VIM objects (21%) in the Hipparcoscatalogue. Most of these 288 VIMs are long-period (e.g. Mira) variables(LPV). The effect of a revised chromaticity correction, which accountsfor the color variations along the light cycle, was then investigated.It is based on ``instantaneous'' V-I color indices derived fromHipparcos and Tycho-2 epoch photometry. Among the 188 LPVs flagged asVIM in the Hipparcos catalogue, 89 (47%) are not confirmed as VIM afterthis improved chromaticity correction is applied. This dramatic decreasein the number of VIM solutions is not surprising, since the chromaticitycorrection applied by the Hipparcos reduction consortia was based on afixed V-I color. Astrophysical considerations lead us to adopt a morestringent criterion for accepting a VIM solution (first-kind risk of0.27% instead of 10% as in the Hipparcos catalogue). With this moresevere criterion, only 27 LPV stars remain VIM, thus rejecting 161 ofthe 188 (86%) of the LPVs defined as VIMs in the Hipparcos catalogue.Based on observations from the Hipparcos astrometric satellite operatedby the European Space Agency (ESA 1997).Table 1 is also available in electronic form at the CDS, via anonymousftp to cdsarc.u-strasbg.fr (130.79.128.5) or viahttp://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/399/1167
| Hipparcos red stars in the HpV_T2 and V I_C systems
For Hipparcos M, S, and C spectral type stars, we provide calibratedinstantaneous (epoch) Cousins V - I color indices using newly derivedHpV_T2 photometry. Three new sets of ground-based Cousins V I data havebeen obtained for more than 170 carbon and red M giants. These datasetsin combination with the published sources of V I photometry served toobtain the calibration curves linking Hipparcos/Tycho Hp-V_T2 with theCousins V - I index. In total, 321 carbon stars and 4464 M- and S-typestars have new V - I indices. The standard error of the mean V - I isabout 0.1 mag or better down to Hp~9 although it deteriorates rapidly atfainter magnitudes. These V - I indices can be used to verify thepublished Hipparcos V - I color indices. Thus, we have identified ahandful of new cases where, instead of the real target, a random fieldstar has been observed. A considerable fraction of the DMSA/C and DMSA/Vsolutions for red stars appear not to be warranted. Most likely suchspurious solutions may originate from usage of a heavily biased color inthe astrometric processing.Based on observations from the Hipparcos astrometric satellite operatedby the European Space Agency (ESA 1997).}\fnmsep\thanks{Table 7 is onlyavailable in electronic form at the CDS via anonymous ftp tocdsarc.u-strasbg.fr (130.79.128.5) or viahttp://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/397/997
| Long period variable stars: galactic populations and infrared luminosity calibrations
In this paper HIPPARCOS astrometric and kinematic data are used tocalibrate both infrared luminosities and kinematical parameters of LongPeriod Variable stars (LPVs). Individual absolute K and IRAS 12 and 25luminosities of 800 LPVs are determined and made available in electronicform. The estimated mean kinematics is analyzed in terms of galacticpopulations. LPVs are found to belong to galactic populations rangingfrom the thin disk to the extended disk. An age range and a lower limitof the initial mass is given for stars of each population. A differenceof 1.3 mag in K for the upper limit of the Asymptotic Giant Branch isfound between the disk and old disk galactic populations, confirming itsdependence on the mass in the main sequence. LPVs with a thin envelopeare distinguished using the estimated mean IRAS luminosities. The levelof attraction (in the classification sense) of each group for the usualclassifying parameters of LPVs (variability and spectral types) isexamined. Table only available in electronic form at the CDS viaanonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or viahttp://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/374/968 or via ASTRIDdatabase (http://astrid.graal.univ-montp2.fr).
| Stars with the Largest Hipparcos Photometric Amplitudes
A list of the 2027 stars that have the largest photometric amplitudes inHipparcos Photometry shows that most variable stars are all Miras. Thepercentage of variable types change as a function of amplitude. Thiscompilation should also be of value to photometrists looking forrelatively unstudied, but large amplitude stars.
| Infrared colours for Mira-like long-period variables found in the (Mȯ<~10-7 Msolar yr-1) Hipparcos Catalogue
Near-infrared, JHKL, photometry is presented for 193 Mira andsemi-regular variables that were observed by Hipparcos; periods,bolometric magnitudes and amplitudes are derived for 92 of them. Becauseof the way in which the Hipparcos targets were selected, this group ofstars provides a useful data base of Miras with low mass-loss rates(Mȯ<~10-7Msolaryr-1).Various period-colour relationships are discussed in detail. The colour,particularly BCK = 10.86 - 38.10 K (J - K)0 +64.16(J - K)20 - 50.72(J -K)30 + 19, K-L, at a given period is found todepend on the pulsation amplitude of the star. A comparison with modelssuggests that this is a consequence of atmospheric extension, in thesense that large-amplitude pulsators have very extended atmospheres andredder Mȯ<10-7Msolaryr-1, K-L and H-K but bluerJ-H than their lower amplitude counterparts. The stars with veryextended atmospheres also have higher values of K-[12] and hence highermass-loss rates. This finding provides further evidence for the causalconnection between pulsation and mass loss. Two sequences are identifiedin the Hp-K versus logP diagram (where Hp is the Hipparcos broad-bandmagnitude) at short periods (logP<2.35). At a given period these twogroups have, on average, the same pulsation amplitude, but differentJHKL colours and spectral types. The short-period stars in the bluersequence have similar near-infrared colours to the Miras found inglobular clusters. Long-term trends in the infrared light curves arediscussed for stars that have sufficient data.
| Catalogue and Bibliography of Emission-Line Stars of Types Later than B.
Abstract image available at:http://adsabs.harvard.edu/cgi-bin/nph-bib_query?1954ApJS....1..175B&db_key=AST
|
Submit a new article
Related links
Submit a new link
Member of following groups:
|
Observation and Astrometry data
| Constellation: | Télescope |
| Right ascension: | 19h27m39.60s |
| Declination: | -45°52'31.4" |
| Apparent magnitude: | 10.155 |
| Proper motion RA: | 1.9 |
| Proper motion Dec: | -17.9 |
| B-T magnitude: | 12.104 |
| V-T magnitude: | 10.316 |
Catalogs and designations:
|
