A spectroscopic and proper motion search of Sloan digital Sky Survey: red subdwarfs in binary systems
Z. H. Zhang, D. J. Pinfield, B. Burningham, H. R. A. Jones, M. C. Galvez-Ortiz, S. Catalan, R. L. Smart, S. Lepine, J. R. A. Clarke, Ya. V. Pavlenko, D. N. Murray, M. K. Kuznetsov, A. C. Day-Jones, J. Gomes, F. Marocco, B. Sipocz
aa r X i v : . [ a s t r o - ph . GA ] J u l Mon. Not. R. Astron. Soc. , 1–25 (2013) Printed 7 November 2018 (MN L A TEX style file v2.2)
A spectroscopic and proper motion search of Sloan digitalSky Survey: red subdwarfs in binary systems
Z. H. Zhang, , , ⋆ D. J. Pinfield, B. Burningham, H. R. A. Jones, M. C. G´alvez-Ortiz, , S. Catal´an, R. L. Smart, S. L´epine, J. R. A. Clarke, , Ya. V. Pavlenko, , D. N. Murray, M. K. Kuznetsov, A. C. Day-Jones, J. Gomes, F. Marocco and B. Sip˝ocz Centre for Astrophysics Research, Science and Technology Research Institute, University of Hertfordshire, Hatfield AL10 9AB, UK Shanghai Astronomical Observatory, Chinese Academy of Sciences, 80 Nandan Road, Shanghai 200030, China Istituto Nazionale di Astrofisica, Osservatorio Astronomico di Torino, Strada Osservatrio 20, 10025 Pino Torinese, Italy Centro de Astrobiolog´ıa (CSIC-INTA), Ctra. Ajalvir km 4, E-28850 Torrej´on de Ardoz, Madrid, Spain Department of Astrophysics, Division of Physical Sciences, American Museum of Natural History, New York, NY 10024, USA Departamento de F´ısica y Astronom´ıa, Universidad de Valpara´ıso, Av. Gran Breta˜na 1111, Casilla 5030, Valpara´ıso, Chile Main Astronomical Observatory, Academy of Sciences of Ukraine, Golosiiv Woods, Kyiv-127, 03680, Ukraine
Accepted 2013 June 06. Received 2013 May 15; in original form 2012 August 13
ABSTRACT
Red subdwarfs in binary systems are crucial for both model calibration and spec-tral classification. We search for red subdwarfs in binary systems from a sample of highproper motion objects with Sloan digital Sky Survey spectroscopy. We present herediscoveries from this search, as well as highlight several additional objects of interest.We find 30 red subdwarfs in wide binary systems including: two with spectral type ofesdM5.5, 6 companions to white dwarfs and 3 carbon enhanced red subdwarfs withnormal red subdwarf companions. 15 red subdwarfs in our sample are partially resolvedclose binary systems. With this binary sample, we estimate the low limit of the redsubdwarf binary fraction of ∼
10 %. We find that the binary fraction goes down withdecreasing masses and metallicities of red subdwarfs. A spectroscopic esdK7 subdwarf+ white dwarf binary candidate is also reported. 30 new M subdwarfs have spectraltype of > M6 in our sample. We also derive relationships between spectral types andabsolute magnitudes in the optical and near-infrared for M and L subdwarfs, and wepresent an M subdwarf sample with measured
U, V, W space velocities.
Key words: brown dwarfs – stars: carbon – stars: late-type – stars: Population II –subdwarfs – Galaxy: halo.
Dwarf stars with subsolar metallicity are bluer thansolar abundance dwarfs or main-sequence stars of theequivalence mass. They lie below the main sequence inthe Hertzsprung–Russell diagram and appear less lumi-nous than main-sequence stars. These objects were thuscalled “subdwarfs” by Kuiper (1939). Evolving subd-warfs are referred to as cool subdwarfs to provide distinc-tion from hot subdwarfs, a different class of objects (e.g.Han, Podsiadlowski, & Lynas-Gray 2007). Cool dwarfs andcool subdwarfs are observationally and kinematically dis-tinct, and were separated into Population I and II categories, ⋆ E-mail: [email protected] respectively by Baade (1944). Population I and II are as-sociated with the Galactic disc and spheroid respectively.Roman (1950, 1952, 1954) find that the old, high-velocityPopulation II stars were also metal deficient.Red dwarfs are low-mass and relatively cool stars onthe main sequence with spectral types of late-type K andM, masses between ∼ ∼ M ⊙ , and surfaceeffective temperatures between ∼ ∼ c (cid:13) Z. H. Zhang et al. of red dwarfs by chemical abundance. RSDs are significantlyrarer than red dwarfs. The Kapteyn’s Star (sdM1 type) isthe only cool subdwarf among the sample of 8 pc within theSun which contains more than 244 known stars and browndwarfs, including 157 M dwarfs (Kirkpatrick et al. 2012).The known sample of late-type M subdwarfs is significantlysmaller than that of early-type M subdwarfs because theyare fainter and have lower space density according to thehalo mass function (e.g. Chabrier 2003). Late-type M sub-dwarfs have the most complex stellar atmospheres becausethey are ultracool and have large-scale variation of chemicalabundance and gravity.Although RSDs are less luminous than F, G and mid-K subdwarfs, they are numerous and have more notablespectral features caused by chemical abundance and grav-ity, thus they are better targets for observational and the-oretical studies. It is difficult to distinguish F, G and mid-K subdwarfs from normal dwarf stars of the same spectraltype using their optical spectra because they are feature-less. Dwarfs and subdwarfs with spectral types of late-typeK and M, however, have a number of very different spec-tral features. Model atmospheres also suggest that opticalspectra of M subdwarfs are dramatically affected by metal-licity variations. It is thus possible to use low-resolutionspectroscopy for exploring effective temperature, metallic-ity and gravity effects on the spectra of these cool objects(Allard & Hauschildt 1995).Classification and characterization of M subdwarfs is arapidly evolving field (Gizis 1997; L´epine, Rich, & Shara2007; Jao et al. 2008; Dhital et al. 2012). M subdwarfs areclassified into three metal classes: subdwarf (sdM), extremesubdwarf (esdM) and ultra subdwarf (usdM) based on theratio of TiO to CaH indices (L´epine, Rich, & Shara 2007).CaH and TiO indices are easy to measure and sensitiveto temperature and metallicity. However, Jao et al. (2008)found that the CaH and TiO indices are affected in compli-cated ways by combinations of temperatures, metallicitiesand gravities of RSDs. Model spectra show that the TiO5index is more sensitive to metallicity while the CaH2 andCaH3 indices are more sensitive to gravity. This suggeststhat the effect of gravity, which was previously ignored,should be considered in the classification of M subdwarfs.An ideal testbed for the impact of gravity on spectra of Msubdwarfs is binary systems with two M subdwarfs whichshare the same age and metallicity. With the same effect onspectra from metallicity, it is possible to measure the dif-ference of gravity on the broad indices of CaH2, CaH3 andTiO5.M subdwarfs in binary systems are the key for bothmodel calibration, and spectral classification and charac-terization. Discovery of a sample of RSD binary systemsis therefore crucial. In this paper we present the discov-ery of 45 RSD binary systems from the Sloan Digital SkySurvey (SDSS, York et al. 2000) and the UKIRT InfraredDeep Sky Survey (UKIDSS, Lawrence et al. 2007). At leastone companion in each system is confirmed as an RSD withSDSS spectra. The selection and classification processes ofour RSD sample are presented in Section 2. The identifica-tion of RSD binary systems is presented in Section 3. Sec-tion 4 presents further discussion of RSD binary systems ofparticular interests. Summary and conclusions is describedin Section 5. -1 0 1 2 3 4r−z H r dM:sdM:esdM:usdM:WD: Figure 1.
SDSS r band reduced PMs H r and r − z colour ofour PM selected objects. Symbols in the figure are black down-pointing triangles: M dwarfs; green squares: sdMs; red circles:esdMs; blue up-pointing triangles: usdMs; grey left-pointing tri-angles: WDs. The eighth data release (DR8) of SDSS includes 14555 deg of imaging data, and 9274 deg of spectroscopic data. Thereare over 1.84 million spectra in total, including 0.6 mil-lion stars, 0.13 million quasars and 0.95 million galaxies(Aihara et al. 2011). The SDSS DR8 also includes propermotions (PMs) for objects derived by combining SDSSastrometry with USNO-B positions, re-calibrated againstSDSS. The errors of PMs are typically less than 10 mas · yr − (Munn et al. 2004). We selected our candidates using SDSS CasJobs by combin-ing the spectroscopic and PM catalogues. We required PMgreater than 100 mas · yr − . No photometric criteria wereapplied but spectral observations for red dwarfs in SDSS islimited to r ∼ . , i ∼
20 and z ∼ .
5. Most of RSDsin SDSS spectroscopic data base are at distances of 200 ∼
400 pc. Objects with tangential velocity of 200 km s − ata distance of 400 pc (or 100 km s − and 200 pc) will havePMs higher than 100 mas · yr − . A PM cut of 100 mas · yr − allows a good balance in order to select most late-type K,early-type M and almost all mid-late type M subdwarfs inthe combined SDSS PM + spectroscopic catalogues whileminimizing the contamination by dwarf stars. Since not allspectra in DR7 (Abazajian et al. 2009) are reproduced inDR8 we applied our search to both data releases finding7499 and 8445 spectra respectively. There were 9146 spec-tra in total for 8236 objects (some objects had more thanone spectrum). c (cid:13) , 1–25 ed subdwarfs in binary systems r − z dM:sdM:esdM:usdM:WD:sdC:dC:MS:CWD:WD:DQ WD:WD+MS: (a) 0 1 2g−r0123 u − g (b) Figure 2. g − r versus r − z (left) and g − r versus r − z (right) colours of our PM selected sample. Symbols in the figure are blackdown-pointing triangles: dMs; green squares: sdMs; red circles: esdMs; blue up-pointing triangles: usdMs; filled magenta pentagons:carbon dwarfs (dC); magenta pentagons filled with black: carbon subdwarfs (sdC); open magenta pentagons: DQ WDs; blue hexagonfilled with red: WD + MS binaries; dark grey left-pointing triangles: WDs; cyan left-pointing: cool WDs (CWD); light grey dots: 3028point sources (MS) with 17 < r <
18 selected from 10 square degrees of SDSS.
SDSS spectra were reduced and classified with the idlspec2dcode by SDSS (Aihara et al. 2011). For 8236 objects in oursample, 5687 were classified as stars, 1558 as galaxies and348 as QSOs, 643 of them were not classified. Not all objectsin the sample are classified by SDSS. RSDs are generallyclassified as stars or galaxies in some cases.To pick out and classify RSDs in our sample properlywe used a K and M subdwarf classification code developedby L´epine, Rich, & Shara (2007). We ran the code on thespectra of all 8236 objects to identify subdwarfs and assigntheir spectral types. The code classified objects into ninegroups: dM (2455), dK (80), sdM (689), sdK (326), esdM(483), esdK (189), usdM (256), usdK (189), and unclassi-fied (3442). We found many objects, originally classified asgalaxies by SDSS, that were re-classified as sdK, esdK orusdK.We inspected the spectra by eye in each group to ensurethat the correct classification was applied in each case. Wefound 2004 galaxies with false PMs were selected into oursample: 383 of them were classified as stars (mostly late-typeK subdwarfs). We found that 463 late-type K subdwarfsand 1363 M subdwarfs survived the eyeball check. 54 ob-jects were classified as sdM but removed from our subdwarfsample because they do not have typical halo kinematics.Fig. 1 shows reduced PMs and r − z colour of these objects. Three sequences from left to right show the location of whitedwarfs (WDs), M subdwarfs and M dwarfs. Fig. 2 shows twocolour diagrams of g − r versus r − z and g − r versus r − z of the sample. The WD sample will be discussed in a futurepaper. M, L and T dwarfs are known in relatively large numbersin the solar neighbourhood and have recently improved ab-solute magnitude versus spectral type relationships (e.g.Faherty et al. 2012; Dupuy & Liu 2012). M and L subd-warfs are much less numerous in nearby space and their ab-solute magnitude and spectral type relationships have notbeen well constrained. To estimate distances of our M sub-dwarfs we determined relationships between spectral typesand absolute magnitudes ( M r,i,z,J,H,Ks ) based on SDSS and c (cid:13) , 1–25 Z. H. Zhang et al. M r d:sd:esd: (a) 0 5 10 15 2051015 M J (b)0 5 10 15 20101520 M i (c) 0 5 10 15 20510 M H (d)0 5 10 15 20Spectral type1015 M z (e) 0 5 10 15 20Spectral type510 M K (f) Figure 3.
Spectral type and absolute magnitude relationships of dwarfs (black pluses), subdwarfs (red dots) and extreme subdwarfs(blue diamonds). M0 = 0, M5 = 5, L0 = 10, L5 = 15. Green and black lines on the left-hand panels are best fits of dwarfs and allsubdwarfs respectively. Green, cyan and magenta lines on the right-hand panels are best fits of dwarfs, subdwarfs and extreme subdwarfs,respectively. M and L subdwarfs with two independent parallax measurements are plotted twice. and subdwarfs with parallax distances from the literature. References of parallax distances for M and L dwarfs:Gliese & Jahreiß (1991); Perryman et al. (1997); Gizis (1997);Dahn et al. (2002); Vrba et al. (2004); Costa et al. (2005,2006); Jao et al. (2005, 2011); Henry et al. (2006); L´epine et al.(2009); Smart et al. (2010); Riedel et al. (2010); Andrei et al.(2011); Faherty et al. (2012). See table 1 for references of parallax distances of M and Lsubdwarfs.
Fig. 3 shows spectral type - absolute magnitude rela-tionships of M and L types of dwarfs and subdwarfs. Theparallax sample of available M subdwarfs is classified underthe system of Gizis (1997) which has three metal classes:dM, sdM and esdM. M subdwarfs with metal class of usdM(see Fig. 3 of L´epine, Rich, & Shara 2007) are included inthe esdM metal class of Gizis (1997). Table 1 shows par-allax measurements of M and L subdwarfs used in Fig. 3.We fitted M r,i,z of M and L subdwarfs and extreme subd-warfs together with straight lines. We fitted M J,H,Ks of M c (cid:13) , 1–25 ed subdwarfs in binary systems Table 1.
K7, M and L subdwarfs with parallax measurementsName 2MASS SpT Ref1 π (mas) Ref2 π (mas) Ref3LP 406 −
47 J01002474+1711272 sdM5 13 15.7 ± ± −
036 J02025226+0542205 sdM0 8 35.9 ± ± − ± −
001 J03285302+3722579 sdK7 8 35.3 ± ± − ± −
059 J03422933+1231368 sdM1.5 8 45.1 ± ± − ± − ± ± −
033 J05480018+0822142 sdM0 8 18.8 ± ± −
023 J06140146+1509570 sdM2 8 30.6 ± ± −
044 J07432434+7248500 sdK7 8 18.7 ± ± − ± − − ± − − ± − ± −
040 J11324528+4359444 sdM0.5 8 18.3 ± ± ± ± ± −
035 J12023365+0825505 sdM2 8 26.0 ± ± − ± − ± − ± ± ± −
52 J14390030+1839385 sdM7 8 28.4 ± ± −
008 J15281403+1643109 sdK7 8 18.7 ± ± − ± −
27 J15454034 − ± ± −
059 J16420431+1025583 sdM2 8 26.1 ± ± − ± −
023 J18413636+0055145 sdK7 8 36.2 ± ± − ± ± − − ± −
24 J19442199 − ± ± −
21 J19451476 − ± ± −
052 J19464860+1204580 sdM1 8 22.4 ± ± ± ± −
019 J20272905+3559245 sdM1.5 8 21.1 ± ± ± − ± ± ± − ± − ± −
034 J23082608+3140240 sdM0.5 8 22.7 ± ± subdwarfs and M0–L7 extreme subdwarfs with straight linesseparately. The differences in absolute magnitudes betweenesdM and usdM subdwarfs of the same subtype are smallerthan the fitting errors.Tables 2 & 3 show the coefficients of polynomial fitsof the SDSS and 2MASS magnitudes as a function of spec-tral type for the M and L subdwarfs and dwarfs plottedin Fig. 3. Table 4 shows average absolute magnitudes in2MASS J, H, Ks bands for K7 subdwarfs in Tables 3 & 2.Early-type M subdwarfs are fainter than the same subtypedwarfs in optical bands and even more so in near-infraredbands. It is clear that esdMs are fainter than sdMs, andsdMs are fainter than dMs for M5 types. However, late-type M and L subdwarfs appear to be brighter than nor-mal dwarfs with the same subtypes for M r,i,z,J , and similarto that of dwarfs for M H,Ks . The H collision-induced ab-sorption (Saumon et al. 1994) becomes stronger as metal- licity goes down and suppresses the H and Ks band flux,thus the near-infrared spectra become bluer. While thedust cloud delays the suppressing of spectra below 1 µ m(Witte, Helling, & Hauschildt 2009). From Fig. 3 we cansee ultracool subdwarfs (UCSDs) may not be a suitablename for metal-deficient ultracool dwarfs (UCDs) becausethey are not less luminous than the same subtype UCDs.‘Purple dwarfs’ might be a sensible name for such bluishand very-red UCDs with subsolar abundance. We estimated distances of our M dwarf and subdwarf sam-ples based on spectral type – absolute magnitude rela-tionships derived in Section 2.3.1. For objects detected in2MASS, we used the mean value of distances estimated c (cid:13) , 1–25 Z. H. Zhang et al.
Table 1. continued.Name 2MASS SpT Ref1 π (mas) Ref2 π (mas) Ref3G 004 −
029 J02341234+1745527 esdM3 17 27.3 ± −
047 J02524557+0155501 esdM2 17 25.3 ± −
022 J03132412+1849390 esdK7 8 30.9 ± −
059 J03501388+4325407 esdM0 8 23.4 ± −
017 J04013654+1843423 esdM0.5 8 16.7 ± ± −
31 J04305244+2812001 esdM1 8 10.2 ± ± −
42 J05103896+1924078 esdM5.5 8 13.4 ± ± −
44 J05195663+2010545 esdM4.5 8 10.4 ± ± ± ± − − ± ± − ± ± − − ± −
047 J14065553+3836577 esdM1.5 8 37.4 ± −
48 J14313832 − ± −
32 J15202946+1434391 esdM5 8 8.9 ± ± ± −
36 J16595712 − ± ± −
14 J17395137+5127176 esdM3.5 8 10.3 ± −
219 J18215294+7709303 esdM2.5 8 10.4 ± ± − ± ± −
96 J20253705 − ± ± −
13 J21073416 − ± ± −
051 J22284904+0548128 esdK7 8 26.8 ± ± ± Table 3.
Coefficients of polynomial fits to magnitudes versus spectral types of M and L dwarfs M abs c c c c c c rms (mag) M r M i M z M J M H M Ks M abs = X i =0 c i × (SpT) i where SpT = 1 for M1, SpT = 5 for M5, SpT = 10 for L0, SpT = 15 for L5. Optical spectral types are applied to fits of SDSS r, i, z magnitudes. Near-infrared spectral types are applied to fits of 2MASS J, H, Ks magnitudes. The rms errors are indicated in the lastcolumn. The fits are applicable from M1 to L8 for M r,i,z and from M1 to L9 for M J,H,Ks . with spectral type versus M J,H,Ks relationships, since theyhave smaller root-mean-square errors than the optical bands(usdMs are treated as esdMs). For objects not detected in2MASS or which have no errors for
J, H, Ks band magni-tudes, we estimated their distances with spectral type versus M r,i,z relationships, and the mean value of three distancesfor each object is adopted as a final distance.Stellar Doppler shifts are computed using the ELODIElibrary (Prugniel & Soubiran 2001) spectra as templates with the SDSS pipeline (Adelman-McCarthy et al. 2008;Aihara et al. 2011). These Doppler shifts represent the bestestimate of the radial velocity of the star. Fig. 4 shows thenormalized radial velocity and error distribution of dMs,sdMs, esdMs and usdMs. The full width at half-maximum(FWHM) of the best Gaussian fits for dMs, sdMs, esdMsand usdMs are 85.71, 257.54, 296.54 and 303.11 km s − re-spectively. The M subdwarfs velocities are larger than thatof M dwarfs, which is consistent with these M subdwarfs be- c (cid:13) , 1–25 ed subdwarfs in binary systems Table 2.
Coefficients of polynomial fits to magnitudes versusspectral types of M and L subdwarfs M abs c c rms (mag) SpT range M r M i M z M J M J M H M H M Ks M Ks M abs ) as a function of spectraltypes (SpT) for M0–L7 subdwarfs in Fig. 3. The fits are definedas M abs = c + c × SpT where SpT = 0 for M0, SpT = 5 for M5,SpT = 10 for L0, SpT = 15 for L5. The rms errors and applicableranges of spectral types are indicated in the last two columns.
Table 4.
Absolute magnitudes of K7 type subdwarfs M abs sdK7 esdK7 M J ± ± M H ± ± M Ks ± ± ing members of the Galactic halo; while the M dwarfs derivefrom the Galactic disc.With PMs and radial velocities from SDSS and spec-troscopic distances estimated from Section 2.3.1, we calcu-lated the U, V, W space velocities of our M subdwarfs.Fig. 5 shows the space velocities in
V - U and
V - W spaces. Fig. 6 shows distributions of
U, V, W
Galactic ve-locities for dMs, sdMs, esdMs and usdMs. Fig. 7 shows cu-mulative histograms of errors of
U, V, W and total spacevelocities. The lack of objects around U ∼ − re-flects the fact that only objects with PMs higher than 100mas · yr − are selected in our sample, thus some distant early-type M subdwarfs are missed. The V space velocity distri-butions of dMs, sdMs, esdMs and usdMs have their maximaat − . , − . , − .
08 and − .
19 km s − , and wefind FWHM of 67.95, 172.43, 171.65 and 257.59 km s − re-spectively, assuming a Gaussian distribution. The V velocitydistribution of M dwarfs can not be fitted well with a singleGaussian line (Fig. 6). It appears that some of the M dwarfs( ∼ V < −
100 km s − ). A sin-gle Gaussian also can not fit the distribution of W velocity(17% have W >
50 km s − or W < −
80 km s − ). This meansa fraction of M dwarfs have halo like kinematics. There arealso some M subdwarfs which have disc kinematics in theoriginal sample. The study by Spagna et al. (2010) basedon FGK stars shows that the metallicity tail of the thick discpopulation goes down to [ M /H] ∼ − M /H] ∼ − −4 −2 0 2 4Radial velocity (100 km/s)0.00.51.0 N o r m a li s e d c o un t dMsdMesdMusdM Figure 4.
Radial velocity and error distribution of dMs (black),sdMs (green), esdMs (red) and usdMs (blue). All distributions arenormalized so that the area, e.g. the sample for each class, is equalto one. Dotted lines are best Gaussian fits. A yellow dashed lineshows the half-maximum of dMs fit. Additional fit lines of sdMs,esdMs and usdMs are also plotted and shifted to the yellow lineby their half-maximum. Typical errors of radial velocities are 3-10km · s − . Figure 7.
Cumulative histograms of errors of
U, V, W
Galacticvelocities and total space velocity for dM, sdM, esdM and usdMdwarfs (from top left to bottom right).
Cool subdwarfs with spectral types of late-type M and L arereferred to as UCSDs (e.g. Burgasser et al. 2009) followingthe definition of UCDs (e.g. Kirkpatrick, Henry, & Irwin1997). UCSDs are important for our understanding of metal-poor ultracool atmospheres. UCSDs exhibit complex spectradominated by molecular absorption bands and metal lines. c (cid:13) , 1–25 Z. H. Zhang et al. -600 -500 -400 -300 -200 -100 0 100V (km/s)-300-200-1000100200300400 U ( k m / s ) (a) -600 -500 -400 -300 -200 -100 0 100V (km/s)-300-200-1000100200300400 W ( k m / s ) dM:sdM:esdM:usdM:(b) Figure 5.
U, V, W
Galactic velocities of dM (grey down-pointing triangles), sdM (green squares), esdM (red circles) and usdM (blueup-pointing triangles) dwarfs. Note that U is positive towards the Galactic anti centre. (cid:0) (cid:1) (cid:2) (cid:3) (cid:4) (cid:5) (cid:6) N o r m a li s e d c o un t dMsdMesdMusdM Figure 6.
Histograms of
U, V, W
Galactic velocities for dM, sdM, esdM and usdM dwarfs. All distributions are normalized so that thearea, e.g. the sample for each class, is equal to one. Dotted lines are best Gaussian fits. A yellow dashed line show the half-maximum ofdMs fit. Additional fit lines of sdMs, esdMs and usdMs are also plotted and shifted to the yellow line by their half-maximum.
Spectra of UCSDs are affected by their low effective tem-perature, subsolar abundance and gravity in a complicatedway. Current atmospheric models do not reproduce observedspectra of UCSDs (Burgasser, Cruz, & Kirkpatrick 2007).UCSDs with different properties ( T eff , [ M /H], gravity, multi-plicity) are very useful to test and calibrate models of ultra-cool atmospheres (Burrows et al. 2001; Marley et al. 2002;Helling et al. 2008; Witte, Helling, & Hauschildt 2009)and low-mass stellar evolution scenarios (Baraffe et al.1997, 2003; Montalb´an, D’Antona, & Mazzitelli 2000).To date, there have been only about 80 UCSDsdiscovered Gizis (1997); Gizis & Harvin (2006);Schweitzer et al. (1999); L´epine, Shara, & Rich (2003b); L´epine, Rich, & Shara (2003c);L´epine, Shara, & Rich (2004); L´epine, Rich, & Shara(2007); L´epine & Scholz (2008); Burgasser et al. (2003);Burgasser (2004); Burgasser & Kirkpatrick (2006);Burgasser, Cruz, & Kirkpatrick (2007); Scholz et al.(2004); Scholz, Lodieu, & McCaughrean (2004); Marshall(2008); Jao et al. (2008); Sivarani et al. (2009);Cushing et al. (2009); Lodieu et al. (2010, 2012);Kirkpatrick et al. (2010). 42 subdwarfs in our samplehave spectral types of > M6 , 12 of which are known > M7 (L´epine & Scholz 2008), 30 of which are new > M6including 9 M6.5–M7.5. Table 5 shows photometry andPMs of these 30 new > M6 subdwarfs. Fig. 8 shows spectral c (cid:13) , 1–25 ed subdwarfs in binary systems Figure 8.
SDSS optical spectra of 30 RSDs with spectral types of > sdM6, > esdM6 and > usdM6. The SDSS name and spectral typeare indicated above each spectrum. Absorption bands of CaH2, CaH3 and TiO5 are also indicated above top spectra. All spectra arenormalized at 8000 ˚A.c (cid:13) , 1–25 Z. H. Zhang et al.
Figure 9.
SDSS spectra of RSDs with different gravities. Sub-types indicated in the plot are flawed for high gravity subdwarfs.Small shifts between lines of different spectra along wavelengthare due to different high radial velocities. All spectra are normal-ized at 8000 ˚A. sequences of these > sdM6, > esdM6 and > usdM6 subdwarfs.Spectral types are assigned according to a metallicity index ζ TiO / CaH defined by absorption bands of CaH2, CaH3 andTiO5 (L´epine, Rich, & Shara 2007). From these spectrawe can see that both CaH and TiO bands are sensitive toeffective temperature, but that the TiO bands are moresensitive to metallicity compared to CaH bands. Spectraltypes of these late-type M subdwarfs have uncertaintiesof 0.5-1.0 because some spectra do not have a very highsignal-to-noise ratio. The actual uncertainty of the spectraltype classification could be larger because the effects ofgravity are not included in the ζ TiO / CaH index. In someextreme cases, gravity could changes the ζ TiO / CaH index byan equivalent of three subtypes. We will discuss the impactof gravity on the spectra of M subdwarfs in Section 2.3.4.
During our visual inspection of M subdwarf spectra we foundsome M subdwarfs that have very strong CaH bands, andappear up to three subtypes later than normal M subdwarfswith the same overall profile. The CaH and TiO indices areused to assign spectral types and metal class for M sub-dwarfs (Gizis 1997; L´epine, Rich, & Shara 2007). Grav- -400 -200 0 200µ RA -400-2000200 µ D ec Figure 10.
30 subdwarf common PM pairs. Companions in eachbinary are plotted in black and red error bars. A multiplicationsymbol indicates the location of (0, 0). Light grey dots are RSDsin our sample. ity is not considered in the classification of M subdwarfs.The study by Jao et al. (2008) based on Gaia model grids(Brott & Hauschildt 2005) suggests that CaH and TiO ab-sorption bands are both good indicators of effective temper-ature. The TiO is more sensitive to metallicity changes com-pared to CaH. The CaH is very sensitive to gravity changesbut the TiO does not appear to be sensitive to gravity at all.Jao et al. (2008) also suggests that overall spectral profilescould be used as a major indicator of effective temperature.In this system, spectra with similar overall profiles and TiOindices would have similar effective temperatures and metal-licities, and the variation of CaH indices would representsgravity changes.We inspected all subdwarf spectra for unusual, rela-tive CaH strength and found a large variation of the depthof CaH bands among M subdwarfs with the same overallprofile and depth of TiO band. Fig. 9 shows SDSS spec-tra of M subdwarfs with different gravity features. Spec-tra of M subdwarfs with normal gravity are over-plottedfor comparison: SDSS J091559.72+290817.4 (SDSS J0915;sdK5), SDSS J010131.32 − c (cid:13) , 1–25 ed subdwarfs in binary systems Table 5.
30 new subdwarfs with spectral types of M6 and later.SDSS name SDSS g SDSS r SDSS i SDSS z µ RA (mas/yr) µ Dec (mas/yr) SpTJ002552.58+010924.9 21.06 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± and spectral types of these five high gravity M ultra subd-warfs.We only found such large variation of CaH band inour usdM subdwarf sample. The TiO band in spectra ofusdM subdwarfs is very weak and barely visible, thus haslarge measurement errors. If strengthening CaH bands inthese five usdM subdwarfs does not represent high gravity,it probably indicates low metallicity beyond normal usdMsubdwarfs. Dahn et al. (1977) identified the first dwarf carbon starG 77–61 (LHS 1555), and hypothesized that this objectwas in fact a double star. The primary ejected carbonmaterial on to the surface of its lower mass companionduring its giant branch phase, and then evolves to be-come a cool WD and is much fainter than the carbondwarf secondary (e.g. Steinhardt & Sasselov 2005). Ra-dial velocity variations proved this hypothesis of an un-seen component in G 77-61 (Dearborn et al. 1986). The
U, V, W space motion (Dahn et al. 1977) and spectrum fits(Gass, Wehrse, & Liebert 1988; Plez & Cohen 2005) of G77–61 indicate that it is a low metallicity object of the Galac-tic halo.Carbon dwarfs are rare objects compared to normal reddwarfs, with only about 120 published (e.g. Margon et al.2002; Lowrance et al. 2003; Downes et al. 2004). The metal-poor carbon dwarfs have features of both carbondwarfs and RSDs, and have not been distinguished fromcarbon dwarfs as a new population.Five cool carbon dwarfs with strong CaH indices werenoticed by Margon et al. (2002). They argue that CaH in-dices present in these stars may be an effective low-resolutionluminosity indicator. However, it may be more natural toexplain the presence of CaH indices with low metallicity.Strong CaH and weak TiO indices are main features oflate-type K and M subdwarfs. These carbon-enriched metal-deficient objects could be called ’carbon subdwarfs’ becausethey have features of both cool subdwarfs and carbon dwarfs(see Section 4.3 for further discussion). Nine late-type K andM type carbon subdwarfs have been identified in our sam-ple. Five mid-K-type carbon subdwarf candidates are alsoidentified. Table A1 shows photometry and PMs of thesecarbon subdwarfs and 22 cool carbon dwarfs..
We used three different methods to identify subdwarf binarysystems with different separations.
Common PM is one of the most useful indicators of wide bi-nary systems ( >
100 au). Many ultracool dwarf binary sys-tems have been successfully identified by this method (e.g. c (cid:13) , 1–25 Z. H. Zhang et al.
Table 6.
Five M ultra subdwarfs with high gravitySDSS name SDSS g SDSS r SDSS i SDSS z µ RA (mas/yr) µ Dec (mas/yr) SpT a J075526.13+482837.3 19.71 ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± a Spectral types listed in the table are based on the classification system of L´epine, Rich, & Shara (2007) which has no considerationfor gravity effects, and thus are flawed for these high-gravity objects. More practicable subtypes of these high-gravity M subdwarfs areindicated in parentheses after their spectral types. H r (a) 0 1 2r−z182022 H r (b) Figure 11.
Reduced PMs and r − z colour of our 30 RSDs binaries and 9 binaries containing > M6 dwarfs. Panel (b) on the right isa zoom in plot for panel (a) on the left. Our PM selected sample (known galaxies excluded) are also plotted on the left-hand panelfor reference. Black circles represent spectroscopically confirmed RSDs, black pluses represent their companions without spectra. RSD+ RSD binaries are joined with black lines; RSD + WD binaries are joined with cyan lines. Red squares represent spectroscopicallyconfirmed > M6 companions, red crosses x are companions without spectra, they are joined with red lines. Binaries with carbon subdwarfcompanions are joined with magenta lines. Faherty et al. 2010; Zhang et al. 2010; Burningham et al.2010; Day-Jones et al. 2011; Pinfield et al. 2012).
The statistical probability that two objects with commonPM higher than 100 mas · yr − and errors less than ∼ · yr − within a few arcminutes could occur by randomis usually very small (e.g. ≪ · yr − from SDSS DR8 and cross matched thissample with our PM and spectroscopy sample. To includesome possible very wide binaries we used a separation limitof 9 arcmin, and PM difference of 15 mas · yr − during thiscross match. Objects in the SDSS PM catalogue generallyhave errors better than 15 mas · yr − for r <
20 (Figure 4 of Munn et al. 2004) . Objects with errors larger than 15mas · yr − are not reliable.We first estimate the expected number of random com-mon PM pairs within our RSD sample and SDSS PM cat-alogue. The SDSS PM catalogue has 1.81 million objectswith PM >
80 mas · yr − . For each of these 1.81 million ob-jects, we counted the number of common PM pairs withPM differences of less than 15 mas · yr − in the PM sampleof 1.81 million objects, without separation constraints. Wethen divided the total number of common PM pairs (1.642billion) by the total number of objects (1.808 million) andthe total coverage of the PM catalogue (14555 arcmin ) toget the average random common PM density of the wholesample. The possibility of finding common PM companionswithin 15 mas · yr − by random within a small area of ra-dius of 9 arcmin is 4 . × − . Thus, we would expect tofind 1880 × . × − = 8 . c (cid:13) , 1–25 ed subdwarfs in binary systems r − z MS:sdM:esdM:usdM:sdC:dC: (a) 1 2g−r12 r − z (b) Figure 12. g − r versus r − z colours of our 30 RSD binaries. Green squares: sdMs; red circles: esdMs; blue up-pointing triangles:usdMs; filled magenta pentagons: carbon dwarfs; magenta pentagons filled with black: carbon subdwarfs; grey dots: 3028 point sourceswith 17 < r <
18 selected from 10 square degrees of SDSS. Black circles represent spectra confirmed RSDs, black pluses represent theircompanions without spectra. RSD + RSD binaries are joined with black lines; RSD + WD binaries are joined with cyan lines. RSD +carbon subdwarf binaries are joined with magenta lines. between our RSD sample and the PM catalogue with 1.81million objects.50 M subdwarf common PM pairs were found in ourPM pair search. Fig. 10 (a) shows PMs of 30 RSD binaries.Their PMs are listed in Table 7. To confirm the binary sta-tus of our common PM pairs, we did a colour consistencycheck of our common PM pairs according to three rules: (I)fainter companion should have redder colours; (II) compan-ions of a binary should associate and line up on the sameridge in the reduced PM versus r − z colour plot (Fig. 11);(III) companions of a binary should associate and line up onthe same metallicity sequences (Figure 12). Rule (I) is alsoapplied when we use rule (II) or (III) for binarity checking.These three rules do not apply on binaries with WD com-panions. 32 survived rule (II), and 24 of them also survivedrule (III). We thus believe these 24 common PM pairs aregenuine binary systems.Table 7 shows properties of the RSD binaries identifiedwith common PMs. Of all our wide binaries identified withcommon PMs, at least one of the companions is a confirmedRSD with SDSS spectra. Colours and relative brightnessesof these RSD are consistent with their common PM com-panions. Fig. 13 shows SDSS spectra of some of these com-panions. In five systems spectra of both components weretaken by SDSS. Fig. 11 shows the SDSS r band reducedPM and r − z colours of these 30 RSD binaries and 9 > M6dwarf binaries (discovered as a by-product, see Table A2 inthe appendix). Our sample (grey dots) are separated intothree sequences: WD, RSD and M dwarfs from left to right. Fig. 12 shows the g − r and r − z colours of 30 RSD binarysystems. Four metal sequences of dwarfs, M subdwarfs, Mextreme subdwarfs, and M ultra subdwarfs are plotted forcomparison. A number of carbon dwarfs/subdwarfs are alsoover plotted. Three carbon subdwarfs look inconsistent inFigures 11 and 12, suggesting that only one companion ineach system is a carbon subdwarf (see Section 4.3 for furtherdiscussion). We conducted a systematic search for companions to our Msubdwarfs. This search was conducted by visual inspectionof the region of sky around each of our M subdwarfs usingthe SDSS Navigate Tool. We inspected images covering sep-arations out to 1 arcmin on the sky, and looked for objectsthat could be K or M subdwarf companions according to rule(I) (Section 3.1.1). Then we measured PMs of all such se-lected candidates following the method described in section5 of Zhang et al. (2009) to test their companionship. Im-ages available from online data bases of the SDSS, UKIDSSand POSS are used for our PM measurements. Five commonPM pairs were found by this method.We applied this method to a larger SDSS RSD samplewithout PM measurements to search for fainter companions.We select binary candidates by their colours, then measuretheir common PMs to confirm their binary status. SDSSJ150015.11+473937.5 (SDSS J1500; usdM0.5) is found tohave a fainter companion with spectral type of usdM3 ac- c (cid:13) , 1–25 Z. H. Zhang et al.
Figure 13.
SDSS spectra of nine subdwarf companions in eight wide binary systems. All spectra are normalized at 8000 ˚A. Spectra arebinned by 11 pixels, original spectra are plotted in yellow. cording to Fig. 3. We are following up more binary candi-dates from a colour + spectroscopy selected RSD sample.
Objects classified as galaxies by imaging data, but as starsby spectroscopy or PMs, are often in fact partially resolvedbinary systems. A binary system with relative small separa-tion (e.g. 0.5 arcsec – 3 arcsec for SDSS images) and similarluminosity for each companion (e.g. dM+dM, dM+WD orWD+WD) will be classified as an extended source, e.g. agalaxy.Four spectroscopically confirmed M subdwarfs inour sample were classified as galaxies by the SDSSpipelines based on their imaging data. We foundthat they have a peanut-like configuration and dou-ble peaks in their images (SDSS J093517.25+242139.4,esdM1; SDSS J121502.52+271706.7, esdM0.5; SDSSJ131304.72 − r band images, and havesubstantially moved their positions in the SDSS images.They have elliptical shapes in POSS2 ir images (1.0 arcsecpixel − ), and are consistent with their SDSS i and z bandimages (0.4 arcsec pixel − ). SDSS J1215 and SDSS J1313are also detected in the UKIDSS images (0.2 arcsec pixel − ,0.4 arcsec pixel − ). These peaks in each pair generally the same separations and position angles in SDSS and UKIDSSimages. As an example, Fig. 14 (o) and (p) show that SDSSJ1313 is passing by a background object to its westernside from SDSS to UKIDSS epochs. Thus we conclude thatthese four objects are common PM binary systems.Four late-type K subdwarf binary systems (SDSSJ091956.86+324844.2, sdK7; SDSS J111523.79+270216.3,sdK7; SDSS J124951.09+324521.4, sdK6.5 and SDSSJ152733.23+113853.2, sdK7.5) were found in the same way.Companions in these eight binary systems generally havesimilar magnitudes. To find close binary systems which con-tain fainter companions (and may not be classified as galax-ies by SDSS) in our sample, we visually inspected all Msubdwarfs in our sample in SDSS and UKIDSS, and foundanother seven close RSD binary systems. They all have dou-ble flux peaks and common PMs. In total we found 15 par-tially resolved binary systems from our RSD sample. Nearbystars around these binaries do not have double peaks.Five M and two late-type K subdwarfs have faint com-panions detected nearby but there are not good enough sec-ond epoch images to confirm their common PMs. Table 8shows photometry and PMs of 15 partially resolved RSDbinaries and 7 candidate systems. SDSS spectra of theseRSD binaries can not be distinguished visually from thatof single RSDs. A combined spectrum of two equal spectraltype companions in a close RSD binary would looks similarto spectra of each companion. While a combined spectrumof two companions with more than 2-3 subtypes differentwould be dominated by the brighter companions. c (cid:13) , 1–25 ed subdwarfs in binary systems DSS2 ir 2000-01-11SDSS J0935 (esdM1) (a)
SDSS i 2004-12-15SDSS J0935 (esdM1) (b)
SDSS z 2004-12-15SDSS J0935 (esdM1) (c)
DSS2 ir 1997-07-10SDSS J1422 (esdM0.5) (d)
SDSS i 2005-05-12SDSS J1422 (esdM0.5) (e)
SDSS z 2005-05-12SDSS J1422 (esdM0.5) (f)
DSS2 ir 1996-04-23SDSS J1215 (esdM0.5) (g)
SDSS i 2005-01-18SDSS J1215 (esdM0.5) (h)
SDSS z 2005-01-18SDSS J1215 (esdM0.5) (i)
UKIDSS J 2010-06-29SDSS J1215 (esdM0.5) (j)
UKIDSS H 2010-02-25SDSS J1215 (esdM0.5) (k)
UKIDSS K 2010-02-25SDSS J1215 (esdM0.5) (l)
DSS2 ir 1993-04-02SDSS J1313 (usdM1) (m)
SDSS i 2000-03-03SDSS J1313 (usdM1) (n)
SDSS z 2000-03-03SDSS J1313 (usdM1) (o)
UKIDSS H 2010-07-29SDSS J1313 (usdM1) (p)
UKIDSS K 2010-07-29SDSS J1313 (usdM1) (q)
Figure 14.
DSS2, SDSS and UKIDSS images of close binaries. All images have a size of 6 arcsec × (cid:13) , 1–25 Z . H . Z ha n ge t a l . Table 7.
30 common PM confirmed RSD binaries.Comp SDSS name SDSS u SDSS g SDSS r SDSS i SDSS z µ RA (mas/yr) µ Dec (mas/yr) Sep (arcsec) SpT a
01A J000633.94 − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± − ± ± ± ± ± − ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± c (cid:13) R A S , M N R A S , e d s u b d w a r f s i n b i n a r y s y s t e m s Table 7. continued.Comp SDSS name SDSS u SDSS g SDSS r SDSS i SDSS z µ RA (mas/yr) µ Dec (mas/yr) Sep (arcsec) SpT a
22A J145725.85+234125.4 c ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± b ± ± ± ± ± − ± ± b ± ± ± ± ± − ± ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± c ± ± ± ± ± − ± − ± c ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± a Spectral types in brackets are estimated from relationships of spectral types and r − , i − , z − band absolute magnitudes (Fig. 3). b This binary system is not in the original PM search. c Objects have carbon lines and are classified as carbon subdwarfs. c (cid:13) R A S , M N R A S , Z. H. Zhang et al.
Table 9.
Statistics of binaries of late-type K and M subdwarfsGroup Number Binary FractionsdK 204 11 5.39 per centsdM 622 15 2.41 per centesdK 175 4 2.29 per centesdM 486 12 2.47 per centusdK 84 0 —usdM 255 2 0.78 per centK 463 15 3.24 per centM 1363 29 2.13 per centsd 826 26 3.15 per centesd 661 16 2.42 per centusd 339 2 0.59 per centTotal 1826 44 2.41 per cent
The multiplicity fraction of dwarf stars decreases withmass, from 57% for nearby solar-type main-sequencestars (Duquennoy & Mayor 1991) to 42% for M dwarfs(Fischer & Marcy 1992). The M subdwarf multiplicityfraction is still not clear. Recent searches for M sub-dwarf binary systems (Riaz, Gizis, & Samaddar 2008;Lodieu, Zapatero Osorio, & Mart´ın 2009) show a small bi-nary fraction of ∼ ±
6% for K and M type cool subd-warfs.Forty four RSDs from our original PM selected sam-ple are found in binary systems with projected separationof >
100 au. Fainter companions are missed due to the sur-vey depth and the incompleteness of the PM catalogue usedfor the companion search. Although our binary search is notcomplete, this binary sample does however indicate a chang-ing trend of binary fraction by masses and metallicities. Wefind that the binary fraction of RSDs reduces with decreas-ing masses and metallicities. Table 9 shows the statistics ofbinary frequency of RSDs from our sample by spectra andmetallicity classes.Companions of wide binaries in Table 7 without SDSSspectra are not in our original sample. We group binaries byspectral types of companions in our original sample. Bothcompanions of two sdK + sdM systems 18AB and 30AB inTable 7 are in our original sample, we count as one in bothsdK and sdM groups.This binary sample also allows us to put a lower limit onthe binary fraction ( >
100 au) of RSDs. 2.41% of our RSDsare confirmed in binary systems. As our search of binaries isnot complete the binary fraction will be higher than 2.41%.There are another seven binary candidates listed in Table 8to be confirmed with second epoch imaging. The complete-ness of SDSS+USNO PM catalogue is 0.7 for SDSS i = 19,and 0.3 for SDSS i = 20 (Munn et al. 2004). RSDs in oursample are at distances of 100–500 pc. Faint UCSD com-panions of these RSDs would be missed due to the surveydepth. Massive ( M/M ⊙ >
1) companions of RSDs wouldhave evolved (e.g. cool WDs, neutron stars and black holes) and are too faint to be detected by SDSS and UKIDSS. Thusthe RSD binary fraction at >
100 au should be & >
100 au and >
100 au measured by Jao et al. (2009) are 14 and 12 percent, respectively. Assuming that RSD binary fractions arecomparable, the total RSD binary fraction would be & <
100 au). These new searches will allow us to put a tighterconstraint on the binary fraction of RSDs.
In this section we discuss in more detail systems of particularinterest, including those where the secondary is a relativelynearby late-type star, and also those where a companion hassomewhat unusual properties.
The spectrum of SDSS J151650.33+605305.4 (SDSS J1516)is shown in Fig. 13. It has been spectroscopically classifiedas an esdM5.5 subdwarf, and is a wide companion to theesdK2 subdwarf G 224–58 (22A in Table 7). With a sep-aration of 93 arcsec, this system is one of our widest bi-naries. An optical spectrum of G224–58 was observed withthe Intermediate Dispersion Spectrograph mounted on theIsaac Newton Telescope on 2010 December 24. The datawere reduced and the spectrum extracted using standardsoftware packages, and we measured the radial velocity of G224–58 by cross-correlation with a radial velocity standardover several wavelength ranges. We avoided regions contam-inated with telluric lines, and also avoided possible emissionlines, and any lines that appeared to be broadened. When wecross-correlated with the reference star HD3765, which hasa radial velocity of − − (Udry, Mayor, & Queloz1999), we derived a radial velocity of − ± − for G 224–58. We also measured the radial velocity usinga different reference star; HD10780 (+2.70 km s − ), andmeasured a consistent value. The radial velocity of G 224–58 is consistent with that of SDSS J1516 or G 224–58 B, − ± − measured from the SDSS spectrum.This is consistent with expectations for a physically associ-ated system. G 224–58 AB has the highest PM (449.77 ± · yr − ) amongst our binary sample. It is at a distance of137 ±
28 pc, estimated using the relationship between abso-lute magnitudes and spectral type (Fig. 3), and has clearhalo space velocities ( U = 278 km s − , V = −
202 km s − , W = −
39 km s − ).Precise metallicity measurements of M dwarfs forcalibration have become popular in the last few years(e.g. Rojas-Ayala et al. 2010, 2012; Terrien et al. 2012;¨Onehag et al. 2012; Neves et al. 2013). These works arebased on M dwarfs in binary systems with FGK dwarfs pri-maries. Precise metallicities are measured from high resolu-tion spectra of early-type primaries adapted to the M dwarfsecondaries to calibrate metallicity features in their spectra.The lowest M dwarf metallicity calibrations are currently[ M /H] ∼ − . c (cid:13) , 1–25 ed subdwarfs in binary systems Table 8.
15 partially resolved RSD binaries and seven candidate systems.SDSS Name SDSS i µ RA (mas/yr) µ Dec (mas/yr) SpT Dis (pc) Sep (arcsec) Sep (au) BinarityJ012958.44+073745.0 15.48 ± ± − ± ± − ± − ± ± ± − ± ± ± − ± ± ± − ± ± ± − ± ± − ± − ± ± − ± ± − ± − ± ± ± − ± − ± − ± − ± − ± ± − ± − ± ± − ± − ± ± − ± ± ± ± ± ± ± − ± ± − ± − ± ± − ± − ± ± − ± − ± ± ± − ± ± − ± − ± ± ± ± subdwarfs. With an early-type esdK subdwarf and a late-type esdM subdwarf, G 224–58 AB is an ideal benchmarkfor subsolar metallicity calibration of M subdwarfs down to= − . < [ M /H] < − . − M /H] ∼ − . ± .
1) that have been found:SDSS J1416AB (Burningham et al. 2010); HIP 73786B(Murray et al. 2011); BD+01 ◦ − The spectrum of SDSS J210105.44 − − ±
37 parsec, derived using our relationship between ab- solute magnitudes and spectral type (Fig. 3). The systemhas halo space velocities ( U = −
90 km s − , V = −
310 kms − , W = −
49 km s − ). This system is of particular use asa test for the M subdwarf classification methods which isstill in debate (e.g. L´epine, Rich, & Shara 2007; Jao et al.2008). Three of the companion objects have features that are char-acteristic of RSDs and also have features that are char-acteristic of carbon dwarfs. We have examined these ob-jects closely in the SDSS i − and z − band images (0.4 arc-sec pixel − grz colour-space. The companionsto these three objects appear to be normal RSDs occupy-ing typical colour space in the g − r versus r − z diagram.This indicates that the carbon in subdwarfs did not orig-inate in their formation environment, but has presumablycomes from the progenitors of unseen WD companions. SDSS J145725.85+234125.4 (SDSS J1457 A) was clas-sified as an usdK6.5 subdwarf according to the metallicityindex ζ TiO / CaH defined by L´epine, Rich, & Shara (2007).It has a high PM of µ RA = − . ± .
73 mas · yr − , µ Dec = − . ± .
17 mas · yr − . SDSS J1457AB is a ro-bust genuine binary with separation of 3.96 arcsec. The toppanel of Fig. 16 shows the spectrum of SDSS J1457A plottedalong with a best-fitting carbon dwarf spectrum (also fromSDSS) and an esdK7 type subdwarf spectrum. We can seethat the spectrum of the carbon dwarf is very similar to thatof SDSS J1457A, apart from the CaH region around 6700–7000 ˚A. However, the spectrum of the esdK7 subdwarf SDSSJ092302.40+301919.7 (SDSS J0923) compares well with the c (cid:13) , 1–25 Z. H. Zhang et al. r − z (a) Figure 15. g − r versus r − z colours of the three carbon subd-warf binaries. Grey down-pointing triangles: dMs; green squares:sdMs; red circles: esdMs; blue up-pointing triangles: usdMs; filledmagenta pentagons: carbon dwarfs; black pentagons: carbon sub-dwarfs. Blue circles with black pentagons inside are spectroscop-ically confirmed carbon subdwarfs in binaries (SDSS J1535B,SDSS J1457A and SDSS J1614A, from left to right), and blackdiamonds are their companions which do not have measured spec-tra. Companions in confirmed binaries are joined with black lines. spectrum of SDSS J1457A in this 6400–7900 ˚A range. Fea-tures of both RSDs (CaH, TiO bands) as well as carbondwarfs (e.g. the C swan bands) are clear in the spectrumof SDSS J1457A. Thus, we think SDSS J1457A is a K7-typecarbon subdwarf.Although the CaH and TiO indices match well withthat of an esdK7 subdwarf, but SDSS J1457A is actuallyan sdK7 subdwarf. The reason why the SDSS J1457A ap-pears like an esdK7 is because the TiO index is sensitive toboth metallicity and carbon abundance. When the C/O ra-tio is greater than 1, all of the oxygen is bound in CO, andnone is left to bond with titanium to form TiO (chapter2, Reid & Hawley 2005). Thus it is not possible to measurethe correct metallicity of carbon subdwarfs by their CaH andTiO indices without consideration of carbon abundance. Inthe case of SDSS J457A, we can measure the metallicity fromits binary companion. SDSS J145726.02+234122.2 (SDSSJ1457B) is an M6.5-type subdwarf according to the M r,i,z –spectral type relationships shown in Fig. 3. Fig. 15 showsthat SDSS J1457B is an sdM subdwarf, as it is located atthe sdM sequence and very close to the dM sequence. SDSSJ1457A should share the same metallicity as SDSS J1457B.Thus, we conclude that SDSS J1457AB is an sdK7+sdM6.5carbon subdwarf system. SDSS J153554.81+105323.7 (SDSS J1535B) was clas-sified as an usdK6.5 subdwarf according to the metallicityindex ζ TiO / CaH defined by L´epine, Rich, & Shara (2007).The spectrum of SDSS J1535B shown in the middle of Fig.16 is similar to that of SDSS J1457A but with shallower CaH
Figure 16.
SDSS spectra of three carbon subdwarf SDSSJ1457A, SDSS J1535B and J1614A. SDSS best-fitting spectra(red, carbon dwarf) and spectrum of SDSS J0923 (top panel,blue, esdM7), SDSS J1448 (middle panel, blue, esdM7) and SDSSJ1134 (bottom panel, blue, esdM7.5) are over-plotted for compar-ison. Spectra are normalized at 7500 ˚A and binned by 11 pixelsexcept SDSS best-fitting carbon dwarf spectra. indices and C swan bands. The best-fitting carbon dwarfspectrum does not provide a good fit for the CaH indicesat 6700-7000 ˚A. SDSS J1535B compares well with an esdK7type subdwarf SDSS J144819.31+363400.4 (SDSS J1448) inthe spectral range 6400-7900 ˚A. The carbon dwarf emis-sion index at 6750-6950 ˚A explains why the CaH indices ofSDSS J1535A are slightly shallower than those of the esdK7subdwarf. We classify SDSS J1535B as a K7 type carbonsubdwarf. SDSS J1535A is a K5.5 subdwarf according to c (cid:13) , 1–25 ed subdwarfs in binary systems Fig. 3. SDSS J1535AB is an esdK5.5 + esdK7 carbon sub-dwarf system according to its CaH and TiO indices. Theymay have higher metallicity than a typical esdK becausethe TiO strength will also be affected by a higher carbonabundance. Fig. 15 shows that SDSS J1535A probably hasmetallicity between esdK7 and sdK7.
SDSS J161454.33+145314.7 (SDSS J1614A) was clas-sified as an usdK7 subdwarf according to the metallicity in-dex ζ TiO / CaH defined by L´epine, Rich, & Shara (2007). Thespectrum of SDSS J1614A is shown at the bottom of Fig. 16.The best-fitting SDSS spectrum of SDSS J1614A is a carbondwarf. The spectrum of the carbon dwarf does not providea good fit for the CaH indices at 6700-7000 ˚A which is themajor spectroscopic feature of an RSD. The spectrum ofSDSS J1614A compares well with that of an esdK7.5 sub-dwarf SDSS J113419.66+345807.8 (SDSS J1134), and wethus classify it as a K7.5 type carbon subdwarf. Its faintercompanion, SDSS J161519.36+145719.9 which is 437.85 arc-sec away, has a spectral type of esdM2.5 according to Fig. 3.We also note that the colours of SDSS J1614B are consistentwith an esdM2.5 subdwarf (Fig. 15).The classification system for dwarf carbon stars has notbeen established. Spectral types assigned to these three car-bon subdwarfs are equivalents of RSDs, do not have infor-mation of carbon abundance. The C swan bands in thesecarbon subdwarf spectra are somewhat weaker than in nor-mal carbon dwarf spectra, suggesting that they have lowercarbon abundance than normal carbon dwarfs. This couldexplain why the CaH indices of these late-type K subdwarfsare not as strong as we might expect. There are three pos-sibilities to explain why these carbon subdwarfs have lesscarbon abundance than carbon dwarfs. (1) The progenitorsof their WD companions had lower mass leading to a lowerlevel of carbon accretion on to the secondary. Subdwarfs areolder than dwarfs and thus have enough time for solar-masscompanions to evolve through the red giant, AGB and WDstages. (2) These carbon subdwarfs have wider separationfrom their unseen WD companions than younger carbondwarfs, again leading to lower levels of accretion. (3) Themetal-poor WD progenitors have a lower carbon abundance. WDs provide important constraints on Galactic time-scales(Schmidt 1959) because their age can be well estimatedfrom WD cooling time-scales combined with the evolution-ary life-times of their progenitors. Binary systems containingold WDs and subsolar metallicity RSD components couldprovide a link between age and chemical abundance. Two oldWD companions to early type K dwarfs with low metallicity([ M /H] ∼ − .
5) have been identified (Jao et al. 2003, 2005).Monteiro et al. (2006) measured the age of these WDs tobe 6-9 Gyr, concluding that they were not likely to be mem-bers of the halo because they are younger than the canonicalhalo age of 12-14 Gyr (Gilmore, Wyse, & Kuijken 1989).The ratio between the strength of TiO and CaHbands near 7000 ˚A for RSDs is a metallicity diagnos-tic (Bessell 1982; Allard & Hauschildt 1995). Thus WDswith RSD companions have advantages for the study ofchemical enhancement and the early formation history ofthe Galaxy. Four M and two late-type K subdwarfs inour sample are found to be companions to probable WDs (see Table 7). Fig. 13 shows spectra of two RSD com-panions to WDs (SDSS J124559.97+300325.2 and SDSSJ141055.98+450222.6). WD companions are identified us-ing reduced PMs (Fig. 11) and SDSS colours (Fig. 12). Thereduced PMs and r − z colour of these six binaries are plottedin Fig. 11 and joined with cyan lines. These six WD com-panions are associated with confirmed WDs on the left ofthe plot. The RSD companions are located on the sequenceof RSDs in Fig. 11. The g − r versus r − z plot in Fig. 12also suggests that these six binary systems contain WD andRSD components. WDs are located at the bottom left in the g − r versus r − z plot (see Fig. 15), while cool WDs overlapwith the hot tail of the main sequence. Most of RSDs inour sample are beyond 200 pc and therefore very cool WDcompanions will be missed by our search. These six RSDswith WD companions are classified as sdM or sdK could bemembers of either the thick disc or inner halo of the Galaxy. Fig. 17 shows the spectrum of SDSS J163340.83+133417.0(SDSS J1633), which is classified as an esdK7 subdwarf. Ithas a significant flux excess in the blue band when comparedto the normal esdK7 subdwarf SDSS J100849.85+200923.4(SDSS J1008). When we remove the spectrum of SDSSJ1008 from that of SDSS J1633, a typical WD spectrumremained. So SDSS J1633 is actually a WD+esdK7 spectro-scopic binary system. It has a PM of µ RA = − . ± . · yr − ; µ Dec = − . ± .
61 mas · yr − and a radial ve-locity of − . ± .
45 km · s − . SDSS r, i and z band ab-solute magnitudes of K7 subdwarfs are 9 . < M r < . . < M i < .
42 and 8 . < M z < .
08 respectively,based on sdK7 and esdK7 subdwarfs with parallax mea-surements. The distance of SDSS J1633 has been estimated(by averaging the absolute magnitudes in M i and M z ) as592 +229 − pc. Its tangential velocity is 356 +138 − km s − , andthe resulting space velocity is U = − . +22 . − . km s − , V = − . +72 . − . km s − and W = 14 . +45 . − . km s − .We checked the SDSS images of SDSS J1633, to see ifany companion was resolved. Fig. 18 shows the SDSS u-,g-, r-, i- band images of SDSS J1633. It is less likely tobe resolved in the SDSS u, i and z bands because WDs aretoo faint in i, z bands and RSDs are relatively faint in the u band. There is no evidence that the system is resolved in the urz bands. The SDSS g band image shows a slightly ellipticalprofile; however, all stars around SDSS J1633 in this g − bandimage show elliptical profiles so this must be a characteristicof this particular SDSS image. Thus we conclude that theWD+esdK7 binary is not resolved in the SDSS images, andthat the separation of this binary is less than ∼ ∼ T eff and log g ) of theWD component were derived by performing a fit of the ob-served Balmer lines to hydrogen-rich WD models (Koester,private communication), following the procedure describedin Garc´es, Catal´an, & Ribas (2011). The Balmer lines insuch WD models were calculated with the modified Starkbroadening profiles of Tremblay & Bergeron (2009). For theline fitting we used the code fitsb2 (Napiwotzki et al. 2004),which follows a procedure based on χ minimization. H α wasnot included in the fit, since it was not clearly visible in thespectrum, probably due to the contribution of the subdwarf c (cid:13) , 1–25 Z. H. Zhang et al.
Figure 17.
The spectrum of SDSS J1633. Black: spectrum ofSDSS J1633 smoothed by 11 pixels. Red: SDSS J1008. Blue: thedifference between SDSS J1613 and SDSS J1008. The spectra ofSDSS J1633 and SDSS J1008 are normalized at 8000 ˚A.
Figure 18.
SDSS u,g,r,i images of SDSS J1633. All images areobserved on 2004 June 12. Each image has a size of 6 arcsec × companion. The atmospheric parameters obtained were: T eff = 12980 ±
770 K and log g = 8.4 ± β is poor, which can also be due to thecontribution of the companion. Considering these param-eters and using the WD cooling sequences of Salaris et al.(2000) we determined the mass and cooling time of this WD,obtaining 0.86 ± M ⊙ and 0.53 ± Figure 19.
Fits of the observed Balmer lines for the WD compan-ion of SDSS J1633. Dotted lines correspond to the observationsand the red line corresponds to the best-fitting (WD models).Balmer lines range from H β (bottom) to H ǫ (top). the two stars may have interacted in the past. It is difficultthen to obtain the total age for this WD since the progen-itor lifetime could range from 0.2 to 6 Gyr if we considerprogenitor masses from 4 to 1 M ⊙ and the stellar tracksof Dominguez et al. (1999) for Z =0.004. It is worth notingthat the temperature obtained, 12980K, is also quite highfor a halo WD with such a large mass. Typical halo WDswith T eff > M ⊙ (derivedfrom the ESO SN Ia progenitor survey project; Pauli et al.2006). However, we note that the mass of the WD may havechanged as a member of a close binary. Jao et al. (2005) dis-covered two WDs in systems with RSDs, LHS 193AB andLHS 300AB. These two systems have large tangential veloc-ities and are likely members of the thick disc population ofthe Galaxy. Monteiro et al. (2006) estimated ages of thesetwo WDs of 6-9 Gyr. If SDSS J1633 is a halo object, it shouldhave an age of > We have selected ∼ · yr − . 42 of these objects are late-type M sub-dwarfs with spectral type of > M6, 30 of them are new ones.We fitted an absolute magnitude – spectral type relation-ship (in the r, i, z, J, H, K bands) for M and L subdwarfs,showing that subdwarfs have different sequences to M andL dwarfs. Metal-poor cool dwarfs are subdwarfs only forspectral types of M5, and become ‘super’ dwarfs for latertypes ( > M5). We estimated distances of our M subdwarfs c (cid:13) , 1–25 ed subdwarfs in binary systems using the absolute magnitude – spectral type relationships,and placed constraints on the U, V, W space velocities. Oursample shows that halo and disc populations have overlapsof metallicity and kinematics.Five M ultra subdwarfs are found to have considerablyhigher gravity than normal M subdwarfs. Their CaH absorp-tion features are significantly deeper than normal M subd-warfs (whose spectra are similar in other respects). Theseobjects provide a good tests for how surface gravity effectsthe spectra of cool stars. These high-gravity features areonly found in M ultra subdwarfs which may reveal the rolethat metallicity plays in the formation and evolution his-tory of low-mass stars. We also identify fourteen carbon richRSDs which represent a new population of carbon subd-warfs. These objects can help us to study carbon star pop-ulations over a much greater age range.We have presented 45 RSDs in wide binary systems ( >
100 au) containing sources with SDSS spectroscopy, con-firming associations through common PMs. Their separa-tions range from 0.4 arcsec to 9 arcmin, and the secondarieshave spectral types ranging from late-type K to late-typeM. 30 are wide and 15 are partially resolved binary sys-tems. G 224–58AB is one of our widest binary systems, andcontains an esdK2- and an esdM5.5-type subdwarf. SDSSJ210105.37–065633.0AB is a closer binary system that con-tains an esdM1- and an esdK5.5-type subdwarf. We foundone spectroscopic and six wide WD + RSD binary systems.With age constraints from the WD companions in these sys-tems, we can study the chemical evolution of the Galaxy.Three metal-poor carbon dwarfs are found in binary systemswith subdwarf companions. Kinematics and radial velocityfollow up would be very useful to better understand thephysics of carbon subdwarfs. Although our binary search isnot complete, our sample shows that the binary fraction ofRSDs goes down with decreasing mass and metallicity, andwe estimate a RSD binary fraction of &
5% for separation >
100 au and &
10% for all separation distances.In the future it will be possible to use UCSD binarysystems (e.g. G 224–58AB) as benchmarks to test metal-poor ultracool atmospheric models. It will also be possibleto use M subdwarf binaries systems (e.g. SDSS J2101AB,SDSS J143305.04+301727.6AB) to test M subdwarf clas-sification methods, particularly gravity effects. We can alsomeasure the metallicity of carbon subdwarfs most effectivelyif we are able to study their subdwarf companions (whichdo not suffer from carbon pollution). A larger sample ofcarbon dwarfs/subdwarfs in wide binaries is expected in thefuture, from surveys/facilities such as Pan-STARRS, LAM-OST, Gaia and LSST, providing the potential to identifylarge numbers of new nearby M, L and T subdwarf multiplesystems.
ACKNOWLEDGEMENTS ‘Junta para la Ampliaci´on de Estudios’ . SC acknowl-edges financial support from the European Commission inthe form of a Marie Curie Intra European Fellowship (PIEF-GA-2009-237718). The authors thank the referee, Dr NigelHambly for the useful and constructive comments.
REFERENCES
Abazajian K. N., et al., 2009, ApJS, 182, 543Adelman-McCarthy J. K., et al., 2008, ApJS, 175, 297Aihara H., et al., 2011, ApJS, 193, 29Allard F., Hauschildt P. H., 1995, ApJ, 445, 433Andrei A. H., et al., 2011, AJ, 141, 54Baade W., 1944, ApJ, 100, 137Baraffe I., Chabrier G., Allard F., Hauschildt P. H., 1997,A&A, 327, 1054Baraffe I., Chabrier G., Barman T. S., Allard F.,Hauschildt P. H., 2003, A&A, 402, 701Bessell M. S., 1982, PASAu, 4, 417Bessell M. S., 1991, AJ, 101, 662Bowler B. P., Liu M. C., Cushing M. C., 2009, ApJ, 706,1114Brott I., Hauschildt P. H., 2005, ESASP, 576, 565Burgasser A. J., et al., 2003, ApJ, 592, 1186Burgasser A. J., 2004, ApJ, 614, L73Burgasser A. J., Burrows A., Kirkpatrick J. D., 2006a, ApJ,639, 1095Burgasser A. J., Kirkpatrick J. D., 2006, ApJ, 645, 1485 c (cid:13) , 1–25 Z. H. Zhang et al.
Burgasser A. J., Cruz K. L., Kirkpatrick J. D., 2007, ApJ,657, 494Burgasser A. J., Vrba F. J., L´epine S., Munn J. A., Lugin-buhl C. B., Henden A. A., Guetter H. H., Canzian B. C.,2008, ApJ, 672, 1159Burgasser A. J., Witte S., Helling C., Sanderson R. E.,Bochanski J. J., Hauschildt P. H., 2009, ApJ, 697, 148Burgasser A. J., L´epine S., Lodieu N., Scholz R.-D., De-lorme P., Jao W.-C., Swift B. J., Cushing M. C., 2009,AIPC, 1094, 242Burningham B., et al., 2009, MNRAS, 395, 1237Burningham B., et al., 2010, MNRAS, 404, 1952Burningham B., et al., 2011, MNRAS, 414, 3590Burrows A., Hubbard W. B., Lunine J. I., Liebert J., 2001,RvMP, 73, 719Catal´an S., Isern J., Garc´ıa-Berro E., Ribas I., 2008, MN-RAS, 387, 1693Chabrier G., 2003, PASP, 115, 763Costa E., M´endez R. A., Jao W.-C., Henry T. J., Subasav-age J. P., Brown M. A., Ianna P. A., Bartlett J., 2005,AJ, 130, 337Costa E., M´endez R. A., Jao W.-C., Henry T. J., Subasav-age J. P., Ianna P. A., 2006, AJ, 132, 1234Cushing M. C., Looper D., Burgasser A. J., KirkpatrickJ. D., Faherty J., Cruz K. L., Sweet A., Sanderson R. E.,2009, ApJ, 696, 986Dahn C. C., Liebert J., Kron R. G., Spinrad H., HintzenP. M., 1977, ApJ, 216, 757Dahn C. C., et al., 2002, AJ, 124, 1170Dahn C. C., et al., 2008, ApJ, 686, 548Day-Jones A. C., et al., 2011, MNRAS, 410, 705Dearborn D. S. P., Liebert J., Aaronson M., Dahn C. C.,Harrington R., Mould J., Greenstein J. L., 1986, ApJ, 300,314Dhital S., West A. A., Stassun K. G., Bochanski J. J.,Massey A. P., Bastien F. A., 2012, AJ, 143, 67Dominguez I., Chieffi A., Limongi M., Straniero O., 1999,ApJ, 524, 226Downes R. A., et al., 2004, AJ, 127, 2838Dupuy T. J., Liu M. C., Ireland M. J., 2009, ApJ, 692, 729Dupuy T. J., Liu M. C., 2012, ApJS, 201, 19Duquennoy A., Mayor M., 1991, A&A, 248, 485Faherty J. K., Burgasser A. J., Cruz K. L., Shara M. M.,Walter F. M., Gelino C. R., 2009, AJ, 137, 1Faherty J. K., Burgasser A. J., West A. A., Bochanski J. J.,Cruz K. L., Shara M. M., Walter F. M., 2010, AJ, 139,176Faherty J. K., et al., 2012, ApJ, 752, 56Fischer D. A., Marcy G. W., 1992, ApJ, 396, 178Garc´es A., Catal´an S., Ribas I., 2011, A&A, 531, A7Gass H., Wehrse R., Liebert J., 1988, A&A, 189, 194Gatewood G., Coban L., 2009, AJ, 137, 402Gilmore G., Wyse R. F. G., Kuijken K., 1989, ARA&A,27, 555Gizis J. E., 1997, AJ, 113, 806Gizis J. E., Harvin J., 2006, AJ, 132, 2372Gliese W., Jahreiß H., 1991, adc..rept,Han Z., Podsiadlowski P., Lynas-Gray A. E., 2007, MN-RAS, 380, 1098Harrington R. S., Dahn C. C., 1980, AJ, 85, 454Helling C., et al., 2008, MNRAS, 391, 1854Henry T. J., Jao W.-C., Subasavage J. P., Beaulieu T. D., Ianna P. A., Costa E., M´endez R. A., 2006, AJ, 132, 2360Jao W.-C., Henry T. J., Subasavage J. P., Bean J. L., CostaE., Ianna P. A., M´endez R. A., 2003, AJ, 125, 332Jao W.-C., Henry T. J., Subasavage J. P., Brown M. A.,Ianna P. A., Bartlett J. L., Costa E., M´endez R. A., 2005,AJ, 129, 1954Jao W.-C., Henry T. J., Beaulieu T. D., Subasavage J. P.,2008, AJ, 136, 840Jao W.-C., Mason B. D., Hartkopf W. I., Henry T. J.,Ramos S. N., 2009, AJ, 137, 3800Jao W.-C., Henry T. J., Subasavage J. P., Winters J. G.,Riedel A. R., Ianna P. A., 2011, AJ, 141, 117Jenkins L. F., 1952, QB813,Kaltenegger L., Traub W. A., 2009, ApJ, 698, 519Kirkpatrick J. D., Henry T. J., Irwin M. J., 1997, AJ, 113,1421Kirkpatrick J. D., et al., 2010, ApJS, 190, 100Kirkpatrick J. D., et al., 2012, ApJ, 753, 156Kuiper G. P., 1939, ApJ, 89, 548Lawrence A., et al., 2007, MNRAS, 379, 1599Leggett S. K., et al., 2008, ApJ, 682, 1256L´epine S., Rich R. M., Shara M. M., 2003a, AJ, 125, 1598L´epine S., Shara M. M., Rich R. M., 2003b, ApJ, 585, L69L´epine S., Rich R. M., Shara M. M., 2003c, ApJ, 591, L49L´epine S., Shara M. M., Rich R. M., 2004, ApJ, 602, L125L´epine S., Rich R. M., Shara M. M., 2007, ApJ, 669, 1235L´epine S., Scholz R.-D., 2008, ApJ, 681, L33L´epine S., Thorstensen J. R., Shara M. M., Rich R. M.,2009, AJ, 137, 4109Lodieu N., Zapatero Osorio M. R., Mart´ın E. L., 2009,A&A, 499, 729Lodieu N., Zapatero Osorio M. R., Mart´ın E. L., SolanoE., Aberasturi M., 2010, ApJ, 708, L107Lodieu N., Espinoza Contreras M., Zapatero Osorio M. R.,Solano E., Aberasturi M., Mart´ın E. L., 2012, A&A, 542,A105Lowrance P. J., Kirkpatrick J. D., Reid I. N., Cruz K. L.,Liebert J., 2003, ApJ, 584, L95Luhman K. L., et al., 2007, ApJ, 654, 570Margon B., et al., 2002, AJ, 124, 1651Marley M. S., Seager S., Saumon D., Lodders K., AckermanA. S., Freedman R. S., Fan X., 2002, ApJ, 568, 335Marshall J. L., 2008, AJ, 135, 1000Montalb´an J., D’Antona F., Mazzitelli I., 2000, A&A, 360,935Monteiro H., Jao W.-C., Henry T., Subasavage J., BeaulieuT., 2006, ApJ, 638, 446Munn J. A., et al., 2004, AJ, 127, 3034Murray D. N., et al., 2011, MNRAS, 414, 575Napiwotzki R., et al., 2004, ASPC, 318, 402Neves V., Bonfils X., Santos N. C., Delfosse X., ForveilleT., Allard F., Udry S., 2013, A&A, 551, A36¨Onehag A., Heiter U., Gustafsson B., Piskunov N., Plez B.,Reiners A., 2012, A&A, 542, A33Pauli E.-M., Napiwotzki R., Heber U., Altmann M.,Odenkirchen M., 2006, A&A, 447, 173Perryman M. A. C., et al., 1997, A&A, 323, L49Pinfield D. J., et al., 2012, MNRAS, 422, 1922Plez B., Cohen J. G., 2005, A&A, 434, 1117Prugniel P., Soubiran C., 2001, A&A, 369, 1048Reid I. N., Hawley S. L., 2005, nlds.book,Riaz B., Gizis J. E., Samaddar D., 2008, ApJ, 672, 1153 c (cid:13) , 1–25 ed subdwarfs in binary systems Riedel A. R., et al., 2010, AJ, 140, 897Rojas-Ayala B., Covey K. R., Muirhead P. S., Lloyd J. P.,2010, ApJ, 720, L113Rojas-Ayala B., Covey K. R., Muirhead P. S., Lloyd J. P.,2012, ApJ, 748, 93Roman N. G., 1950, ApJ, 112, 554Roman N. G., 1952, ApJ, 116, 122Roman N. G., 1954, AJ, 59, 307Salaris M., Garc´ıa-Berro E., Hernanz M., Isern J., SaumonD., 2000, ApJ, 544, 1036Saumon D., Bergeron P., Lunine J. I., Hubbard W. B.,Burrows A., 1994, ApJ, 424, 333Schilbach E., R¨oser S., Scholz R.-D., 2009, A&A, 493, L27Schmidt M., 1959, ApJ, 129, 243Scholz R.-D., Lehmann I., Matute I., Zinnecker H., 2004,A&A, 425, 519Scholz R.-D., Lodieu N., McCaughrean M. J., 2004, A&A,428, L25Schweitzer A., Scholz R.-D., Stauffer J., Irwin M., Mc-Caughrean M. J., 1999, A&A, 350, L62Sivarani T., L´epine S., Kembhavi A. K., Gupchup J., 2009,ApJ, 694, L140Smart R. L., Ioannidis G., Jones H. R. A., Bucciarelli B.,Lattanzi M. G., 2010, A&A, 514, A84Spagna A., Lattanzi M. G., Re Fiorentin P., Smart R. L.,2010, A&A, 510, L4Steinhardt C. L., Sasselov D. D., 2005, astro,arXiv:astro-ph/0502152Terrien R. C., Mahadevan S., Bender C. F., Deshpande R.,Ramsey L. W., Bochanski J. J., 2012, ApJ, 747, L38Tremblay P.-E., Bergeron P., 2009, ApJ, 696, 1755Udry S., Mayor M., Queloz D., 1999, ASPC, 185, 367van Altena W. F., Lee J. T., Hoffleit E. D., 1995, gcts.book,van Dokkum P. G., Conroy C., 2010, Natur, 468, 940van Leeuwen F., 2007, A&A, 474, 653Vrba F. J., et al., 2004, AJ, 127, 2948Witte S., Helling C., Hauschildt P. H., 2009, A&A, 506,1367Woolf V. M., L´epine S., Wallerstein G., 2009, PASP, 121,117York D. G., et al., 2000, AJ, 120, 1579Zhang Z. H., et al., 2010, MNRAS, 404, 1817Zhang Z. H., et al., 2009, A&A, 497, 619Zhang Z. H., Pinfield D. J., Burningham B., Jones H. R. A.,Day-Jones A. C., Marocco F., Gomes J., Galvez-OrtizM. C., 2013, EPJWC, 47, 6007
APPENDIX A: BYPRODUCT
This paper has been typeset from a TEX/ L A TEX file preparedby the author. c (cid:13) , 1–25 Z. H. Zhang et al.
Table A1.
37 carbon dwarfs and subdwarfsSDSS Name SDSS g SDSS r SDSS i SDSS z µ RA (mas/yr) µ Dec (mas/yr) SpT a J012518.66 − ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± − ± ± ± ± − ± − ± − ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± ± ± ± ± ± − ± ± − ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± ± − ± − ± ± ± ± ± − ± a The sdC represents carbon subdwarfs, dC represents carbon dwarfs, dCf represents subdwarfs with weak carbon features.
Table A2.
Nine common proper motion pairs of ultracool dwarfs.Comp SDSS Name SDSS g SDSS r SDSS i SDSS z µ RA (mas/yr) µ Dec (mas/yr) Sep( ′ ) SpT01A J042604.36+170714.3 16.52 ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± ± − ± ± ± ± ± ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± − ± ± ± ± ± − ± ± ± ± ± ± − ± ± (cid:13)000