High Proper Motion Objects from the UKIDSS Galactic Plane Survey
L. Smith, P. W. Lucas, R. Bunce, B. Burningham, H. R. A. Jones, R. L. Smart, N. Skrzypek, D. R. Rodriguez, J. Faherty, G. Barentsen, J. E. Drew, A. H. Andrei, S. Catalán, D. J. Pinfield, D. Redburn
aa r X i v : . [ a s t r o - ph . S R ] J un Mon. Not. R. Astron. Soc. , 000–000 (0000) Printed 24 July 2018 (MN L A TEX style file v2.2)
High Proper Motion Objects from the UKIDSS Galactic PlaneSurvey
Leigh Smith ⋆ , P.W. Lucas , R. Bunce , B. Burningham , H.R.A. Jones ,R.L. Smart , N. Skrzypek , D.R. Rodriguez , J. Faherty , , G. Barentsen ,J.E. Drew , A.H. Andrei , , , S. Catal´an , D.J. Pinfield , D. Redburn Centre for Astrophysics Research, Science and Technology Research Institute, University of Hertfordshire, Hatfield AL10 9AB, UK Istituto Nazionale di Astrofisica, Osservatorio Astrofisico di Torino, Strada Osservatorio 20, 10025 Pino Torinese, Italy Astrophysics Group, Imperial College London, Blackett Laboratory, Prince Consort Road, London SW7 2AZ, UK Departamento de Astronomia, Universidad de Chile, Casilla 36-D, Correo Central, Santiago, Chile Department of Terrestrial Magnetism, Carnegie Institution of Washington, Washington, DC 20015, USA Hubble Fellow Observat´orio Nacional/MCTI, R. General Jos´e Cristino 77, CEP 20921-400 Rio de Janeiro - RJ, Brazil Observat´orio do Valongo/UFRJ, Ladeira do Pedro Antˆonio 43, CEP 20080-090 Rio de Janeiro - RJ, Brazil Department of Physics, University of Warwick, Coventry, CV4 7AL
June, 2014
ABSTRACT
The UKIDSS Galactic Plane Survey (GPS) began in 2005 as a 7 year effort to survey ∼ of the northern Galactic plane in the J, H, and K passbands. The survey included a sec-ond epoch of K band data, with a baseline of 2 to 8 years, for the purpose of investigatingvariability and measuring proper motions. We have calculated proper motions for 167 Millionsources in a 900 deg area located at l > ◦ in order to search for new high proper motionobjects. Visual inspection has verified 617 high proper motion sources ( > mas yr − )down to K =
17, of which 153 are new discoveries. Among these we have a new spectroscop-ically confirmed T5 dwarf, an additional T dwarf with estimated type T6, 13 new L dwarfcandidates, and two new common proper motion systems containing ultracool dwarf candi-dates. We provide improved proper motions for an additional 12 high proper motion stars thatwere independently discovered in the WISE dataset during the course of this investigation.
Key words: catalogues - proper motions - binaries:general - brown dwarfs - stars:low mass
Source confusion in the Galactic plane reduces the completenessof searches for nearby stars and brown dwarfs and high propermotion sources in general. The two epochs of high resolutionUKIDSS GPS data (Lawrence et al. 2007, Lucas et al. 2008) pro-vide a new resource to search for previously missed high propermotion objects, especially brown dwarfs which would typicallyhave been undetected in previous optical searches. It also allowsfor identification of new high amplitude infrared variable stars(Contreras Pe˜na et al. 2014).The thick veil of dust in the Galactic plane is something of a bene-fit in identifying nearby high proper motion objects in optical sur-veys since it obscures more distant stars and lessens the problemof source confusion (see Boyd et al. 2011a, Figure 6; Boyd et al.2011b, Figure 7). However, many of the nearest objects are rela-tively faint at optical wavelengths and we must turn to the near in- ⋆ E-mail: [email protected] frared where they are brighter. Extinction is less of a problem in thenear infrared, which leads us back to a greater problem of sourceconfusion. This has allowed many nearby objects to go unidentifieduntil very recently.Looper et al. (2007) reported 11 T dwarf discoveries in the 2MASSdataset (Skrutskie et al. 2006), three of which were found in asearch for mid-late T dwarfs at low Galactic latitudes in the 2MASSPoint Source Catalogue. Phan-Bao et al. (2008) detected 26 newultracool dwarfs (UCDs, generally regarded as M7 and later) ina photometric and proper motion search at low Galactic latitudesin the DEep Near-Infrared Survey of the Southern sky (DENIS,Epchtein et al. 1997). Lucas et al. (2010) photometrically identi-fied a very cool T dwarf in the UKIDSS Galactic Plane Survey(GPS, (Lucas et al. 2008)). Burningham et al. (2011) identified afurther two mid-late T dwarfs in the GPS using a similar method.Artigau et al. (2010) identified DENIS J081730.0-615520, a T6dwarf at 4.9pc and b ∼ -14 ◦ in DENIS, as an unmatched sourcebetween the DENIS and 2MASS catalogues due to it’s high propermotion. Gizis et al. (2011) and Castro & Gizis (2012) identified 2 Ldwarfs within 10 pc of the sun at low Galactic latitudes by search- c (cid:13) L. Smith et al. ing for detections in WISE with no corresponding detection in2MASS, indicating a high proper motion. Castro et al. (2013) iden-tified a further 4 L dwarfs by the same method, one of which WISEJ040418.01+412735.6 is close to the Galactic plane and a mem-ber of the small subclass of unusually red L dwarfs. Folkes et al.(2012) identified 246 new UCDs with detectable proper motion intheir search for UCDs at low Galactic latitudes using the SUPER-COSMOS and 2MASS surveys. Beam´ın et al. (2013) identified anunusually blue L5 dwarf less than 5 ◦ from the Galactic centre at17.5 pc distance due to its high proper motion evident in the VistaVariables in the Via Lactea Survey (VVV). Mace et al. (2013a)and Cushing et al. (2014) discovered numerous late T dwarfs inthe WISE survey, including WISE J192841.35+235604.9 (T6) andWISE J200050.19+362950.1 (T8) both of which are bright ob-jects in the GPS footprint that are likely to be within 8 pc of thesun (see Section 5.4). Scholz (2014) used WISE data to identifya 5-7 pc (taking into account the possibility of multiplicity) ∼ M9type UCD in the Galactic plane through a photometric selectionof candidates followed by identification of those where the nearest2MASS source was >
1” from the WISE position.Recently there have been two new all sky proper mo-tion searches using the WISE database by Luhman (2014)and Luhman & Sheppard (2014), and Kirkpatrick et al. (2014).Luhman (2013) identified WISE J104915.57-531906.1, a binarybrown dwarf system at 2 pc . Given their relative brightness, manyof the objects listed above could have been identified in previoussurveys but for the effect of source confusion on both colour-basedand proper motion-based searches. Given the success of the recentsearches with the low resolution 2 epoch WISE dataset, a searchfor high proper motion stars using the 2 epoch high resolution GPSdataset could be expected to reveal many previously unidentifiedobjects in the solar neighbourhood.The initial search described here is limited to objects withK <
17 mag, l > ◦ and includes only data taken up to March31st 2013.This paper is organised as follows. In Section 2 we describethe available data. In Section 3 we briefly describe the proper mo-tion calculation method. In Section 4 we determine the accuracyand reliability of the catalogue. In Section 5 we outline searchesundertaken for objects of interest within the catalogue. In Section 6we summarise. The UKIDSS GPS covers 1868 deg in J, H, and K passbands toan approximate 5 σ depth of 18.1 in K. It included a second epochof K band observations two or more years after the initial epoch.Most of the second epoch K band data are not yet available in thecurrent (8th) GPS data release and have not been processed fullyby the WFCAM Science Archive (WSA) team. However, it is re-trievable through their Archive Listing tool. Using this we obtainedUKIDSS GPS K band FITS file catalogues from observations takenbetween May 2005 (the start of UKIDSS) and March 31st 2013and converted them to ASCII format using software provided bythe UKIDSS pipeline team based at the Cambridge Astronom-ical Survey Unit. The observations from this date range give usepoch baselines of between 1.9 and 6.4 years (see Figure 1). More available at: http://casu.ast.cam.ac.uk/surveys-projects/wfcam/technical Proper Motion Limit (" yr −1 ) A r e a ( d e g ) Epoch Baseline (years)
Figure 1.
The area distribution of epoch baselines and maximum propermotion limits of the catalogue. recent data, from March 31st 2013 to the end of 2013, take thefinal maximum epoch baseline to approximately 8 years. The cat-alogues were matched using their telescope pointing positions. Inmost cases there were two observations per pointing separated bygreater than 1.8 years. In these cases we took the earlier observa-tion as the first epoch and the later as the second epoch. In somecases where there were more than two observations per pointing,this is usually due to extra observations that had to be repeated onsubsequent nights (e.g. due to low image quality). In such cases weseparated those pointing groups into two further groups separatedby > l > ◦ , in order to keep the number of high proper motioncandidates to a manageable level. Since we have a fixed matchingradius and epoch baselines ranging from 1.9 to 6.4 years we alsohave upper proper motion detection limits of between 3.75 and 12.6” yr − , see Figure 1. The resultant pipeline input catalogue con-tained ∼
167 million sources and covered approximately 900 deg ,see Figure 2. Note that this method of data collection and initialprocessing is very similar to that previously used in Smith et al.(2014) to create a proper motion catalogue for the UKIDSS LargeArea Survey. The only differences are an increase in the matchingradius and a decrease in the magnitude difference tolerance. Theincreased matching radius allows us to probe for higher proper mo-tion sources, at the expense of an increase in the rate of mismatches,while not negatively impacting our sensitivity towards low propermotion sources. Since the source density is much greater in theGPS relative to the UKIDSS Large Area Survey (LAS) the rate of c (cid:13) , 000–000 GPS High PM Sources Right Ascension D e c li n a t i o n Right Ascension −20−100102030405060 D e c li n a t i o n Known New
Figure 2.
The coverage of the two epochs of K band data is shown here in grey. Overplotted are the visually verified high proper motion sources. Empty circlesare previously identified by other authors, filled circles are new discoveries. mismatches is also much higher, the decrease in the magnitude dif-ference tolerance is an effort to reduce the rate somewhat.
The proper motion calculation method remains almost identical tothe method used for the UKIDSS Large Area Survey (LAS) de-scribed by Smith et al. (2014), with a few alterations which werenecessary to deal with the higher source density and passband dif-ference.We select preliminary reference sources that meet the follow-ing criteria based on the data in the FITS file catalogues:Stellar image profile at both epochs;K at both epochs between 12.25 and 17.00;Ellipticity at both epochs < < = = = = σ detection limit in the GPS K band thanthe LAS J band, and limits the number of mismatches selected asreference sources since they become more common at fainter mag-nitudes. The new lower magnitude limit is a conservative cut ofslightly saturated sources and also allows us to retain distant lumi-nous stars as reference points. In the LAS and elsewhere at highGalactic latitude the brightest sources are nearby stars with largemotions, owing to the small scale height of the Galactic disc. Thesemake poor astrometric reference stars. In the Galactic plane how-ever, the brightest sources are usually very luminous distant starswith relatively small motions which make good reference stars.Hence it was advantageous to reject bright sources as referencestars in the LAS but the opposite is true for the GPS. Using thesepreliminary reference sources initial motions were calculated as de-scribed in Smith et al. (2014) and we then rejected all referencestars with initial motion > σ . We rejected all frames (WFCAM arrays) that contained fewer than 100 remaining reference sourcessince inspection indicated that these contained bad data.We perform a transformation of second epoch array positionsto first epoch array positions using one of two second order polyno-mial transformations. A transformation matrix is calculated either‘globally’, i.e. using reference sources across each whole frame;or ‘locally’, i.e. using reference sources from only a portion of theframe local to the target. Where possible we use the local trans-formation (see Smith et al. 2014 Figure 2 for a justification of thispreference), which selects reference sources from within a radiusgoverned by the local density and distribution of reference sources.More specifically the radius of reference source selection is thesmallest radius in which we find at least 5 reference sources in eachquadrant around the target, rounded up to the nearest 20”. We seta minimum of 300” on this value. If there are insufficient nearbyreference sources then we use the global transformation method in-stead.These motions, calculated as the residual to the transforma-tion are relative to the reference stars used. Since these stars arealso moving (albeit very slowly) the motion calculated is not abso-lute. For the LAS we were able to account for this by subtractingthe calculated median motion of extragalactic sources relative to thereference stars. The Smith et al. (2014) relative to absolute propermotion correction relied on reasonably trustworthy morphologicalclassification. Lucas et al. (2008) found that in the GPS the highsource density and resultant high frequency of stellar blends meanerroneous galaxy classifications based on a merge of the J, H, andK band classifications are common. In the FITS file data we usehere we have morphological classifications based on a single banddetection, which is even less reliable. Furthermore, the factor of ∼
10 increase in the number of sources meant that it was neces-sary to split the input catalogue into some 20 subsets and run thosethrough the pipeline individually. The Smith et al. method of rela-tive to absolute correction involved selecting extragalactic sourcesfrom within 3 ◦ of each target frame, meaning that fields near theedges of the dataset would be corrected using fewer extragalacticsources than fields > ◦ from the edge. Splitting the input cata-logue created many more field edges than would be the case when c (cid:13) , 000–000 L. Smith et al. processing the catalogue as a whole. For these reasons we chosenot to perform the relative to absolute conversion.Despite the lack of relative to absolute correction, the rela-tively large distance to most stars in the GPS and their consequentlyvery small proper motions should provide a zero point that is fairlynear to absolute. We calculated sample relative to absolute correc-tions using the Besanc¸on models (Robin et al. 2003) and find thatthe correction is always mas yr − in the area covered by thispaper. Due to the high incidence of false high proper motion detectionswe found visual verification to be essential. To reduce the numberof candidates to a manageable quantity for visual verification weadopted the following quality criteria based on tests on a subset ofthe data spread throughout the plane. l > ◦ ;Proper motion > mas yr − ;Classified as stellar at both epochs;Ellipticity at both epochs < < at either epoch; andFewer than 10 other candidates in the same 13.65’ × <
17 pass the class and ellipticitycuts, and we therefore adopt this figure as an estimate of the com-pleteness of this selection. We note that the region at l < ◦ couldbe investigated if the search is limited to bright but unsaturated starsin the 12 < K <
14 magnitude range.To identify genuine high proper motion sources we blinkedthe candidates in sequence by calling DS9 in blink mode. Regionswere overlaid showing the position of each source at the first epoch,the calculated position of the source at the second epoch (Fig. 3),and the radius of first order cross-talk. During array readout elec-tronic cross-talk can cause fainter duplicate images of saturated ornear-saturated sources at a distance of 256 n pixels (in the case of2 × x or y direction depending on the readout direction.Dye et al. (2006) discuss cross-talk and other data artefacts presentin WFCAM data. The overlaid regions made identification of mis-matches and cross-talk, which were the dominant source of falsedetections, straightforward. To add to the visual cross-talk identi-fication we also used software designed to identify possible cross-talk in WFCAM data using the positions of bright nearby 2MASSsources. This identified a small number of additional cross-talksources that had initially been missed. We identified 617 genuinehigh proper motion sources from within this sample which gives anoverall ratio of false to genuine candidates of 9:1. Figure 4 showsthe distribution of false and genuine high proper motion detectionsin Galactic coordinates. The fraction of false positives increases Figure 3.
First epoch K (top) and second epoch K blinking im-ages generated by DS9 for the known high proper motion object2MASS J19483064+2321473. Our proper motion for this source is 295 mas yr − . The images are separated by 5.15 years. The regions are placedat the same position on both images relative to their WCS. The × shows thefirst epoch K band position of the source, the circle shows the radius of mo-tion between the image epochs, the line shows the direction of travel andits length corresponds to 10 years of motion. The intersection of the circleand line therefore shows the expected position of the source at the secondepoch. There is another circle with a radius of 256 pixels centred on thefirst epoch position of the target, which falls beyond the boundaries of theseimages, corresponding to the distance of a saturated or near-saturated starthat could cause first order cross-talk at the target position. The bar belowthe source in each image shows 10” and they are oriented north up east left. rapidly with decreasing Galactic longitude, the ratio of false to gen-uine candidates is ∼ l = ◦ . The density of genuine highproper motion sources appears fairly uniform.Figure 5 shows that the fraction of false positive proper mo-tions is much greater at the high proper motion end. This is un-surprising since the large radius of apparent motion between theimages allows for a higher incidence of mismatches. For the samereason sources with an apparent low proper motion are expected tobe more reliable than those with high proper motion. At low propermotion the major source of false detections shifts to crosstalk. Thiscan be identified and removed fairly reliably by searching for bright2MASS sources at a ∼
51” radius from the target. We find that re-moval of crosstalk sources in this way increases the genuine frac-tion amongst relatively low PM candidates (200-300 mas yr − ) to ∼ c (cid:13) , 000–000 GPS High PM Sources
60 80 100 120 140 160 180 200 220−6−4−20246 b False Genuine
60 80 100 120 140 160 180 200 220 l C o un t Figure 4.
The distribution in Galactic coordinates of high proper motioncandidates identified as false (grey) and genuine (black). The upper panelshows the coverage in Galactic coordinates and that the general distributionof genuine high proper motion objects is fairly uniform across the field, incontrast to the increase in false high proper motion objects with decreasingGalactic longitude. The lower panel shows the distribution in Galactic lon-gitude alone. Increased coverage width in Galactic latitude corresponds toand accounts for the peaks in the distribution, which is otherwise fairly flat.The large number of false high proper motion candidates at low Galacticlongitudes is driven by the large increase in source density relative to highGalactic longitudes which increases the frequency of mismatches.
200 400 600 800 1000 1200 1400 1600 1800 2000 µ total (mas yr −1 ) G e nu i n e F r a c t i o n Figure 5.
Histogram showing the decrease in the fraction of genuine highproper motion detections as total proper motion increases. The sample ofsources described in Section 4.1 were sorted by total proper motion andbinned into groups of 200 sources. Note that the x axis continues to 16 arcsec yr − , though the genuine fraction remains at zero past what isshown here. The fraction of genuine high proper motion sources increasesrapidly with decreasing proper motion.
200 300 400 500 600 700 800
UGPS µ total (mas yr −1 ) A l t e r n a t i v e µ t o t a l ( m a s y r − ) LSPMSCR
Figure 6.
Our total proper motion compared to those of alternative sources.The alternative sources are given by the legend. The length of the armsof the crosses indicate the uncertainty in the proper motion measure-ment. LSPM proper motion uncertainties are taken as 8 mas yr − in RAand Dec. We estimate the Boyd et al. total proper motion uncertainties at10 mas yr − . To evaluate the accuracy of the proper motions we compared themto the long epoch baseline optical catalogues of L´epine & Shara(2005, LSPM; covering the north) and Boyd et al. (2011a, 2011b;covering the south). We identified 406 sources common to theLSPM catalogue and 15 common to the Boyd et al. catalogues. Fig-ure 6 shows a comparison of the total proper motions of the 421sources between the catalogues, for which Pearson’s r correlationcoefficient is 0.996. Of the sources in common with the LSPM,69% of the proper motions agree within their 1 σ uncertainties, weomitted the Boyd et al. proper motions from this calculation sincethe authors do not provide an estimate of their uncertainty.To evaluate the accuracy at the lower end of the proper motionscale we selected sources with proper motion < mas yr − , l > ◦ , K magnitude at either epoch <
17, K magnitude un-certainty at both epochs < < < µ < mas yr − which is expected since we are largely sam-pling distant main sequence stars and giants with very small propermotion. This tends to suggest that the reliability of the catalogue re-mains high at µ < mas yr − , continuing the trend indicatedin Figure 5. By contrast, in the LAS (i.e. outside the Galactic plane)we are sampling relatively nearby stars which might be expected tohave (marginally) measurable proper motions, which could explainwhy the distribution in Figure 7 is less strongly peaked towardszero. We tested this by generating a catalogue sample for the sameregion and magnitude range as the LAS dataset using the Besanc¸onmodels. We found that the typical proper motions of the model sam-ple were in the range of 0 to 20 mas yr − and consistent with thedata in Figure 7 after making allowance for the uncertainties in theLAS proper motions.Note that we still produce proper motions for mildly saturated c (cid:13) , 000–000 L. Smith et al. µ total (relative, mas yr −1 ) N o r m a li s e d C o un t Figure 7.
The distribution of a sample of low proper motion sources in theGPS (black) and the LAS (grey). objects as the morphological classification flag for saturation in theFITS catalogues is not always reliable. As a result we would en-courage that proper motions for sources brighter than K ∼
12 beused with caution.
In order to characterise the new high proper motion objects we col-lected r and i band photometry from the INT/WFC Photometric H α Survey (IPHAS, Drew et al. 2005), specifically a preliminary ver-sion of the DR2 catalogue (Barentsen et al. 2014), and the J and Hband photometry from either the UKIDSS DR8 GPS catalogue ortheir FITS file catalogues which were retrieved from the WSA us-ing their archive listing service. Additionally, an IPHAS detectionat a third epoch allows us to safely rule out the possibility that twotransient objects (e.g. solar system objects) produced an apparentproper motion of a single object.The r and i band merged catalogues are available in the IPHASDR2 catalogue. A small number of the high proper motion sourceswere in IPHAS fields observed in poor weather and for this reasonare not present in the DR2 catalogue. However, the photometricuncertainties take the poor observing conditions into account andwe included them where necessary. The coordinate of each IPHASsource is taken from the r band. Where an r band detection is miss-ing the i band is used. We took the proper motion of each sourceinto account by identifying the nearest IPHAS source to the GPSsource’s expected position at the IPHAS epoch. We visually in-spected all IPHAS images to ensure that the catalogue photometrywas not compromised by blending or astrometric errors. We found17 matches located >
1” from their expected position, only one ofwhich was a genuine IPHAS match, UGPS J211859.26+433801.3(see Sections 5.2 and 5.3.1). The other 16 were mismatches usuallydue to a non-detection of the target in the i band.We matched to the UKIDSS DR8 GPS catalogue using a ra-dius which took into account the possibility that our position epoch(the first K band epoch), is not always the same as the epochfrom which the WSA take positions. We found that 444 of thehigh proper motion sources were matched to only one UKIDSSDR8 source, with a further 43 instances where there were multiple matches to the same high proper motion source. We believe this tobe due to missed matches in the creation of the band-merged cata-logues by the WSA since these high proper motion sources almostall have motions between the epochs greater than the 1” matchingradius used by the WSA for GPS data. Additionally, the groups aremostly pairs where the first match is a K1 detection only and theother is a K2 detection only (K1 being defined in the WSA as theK epoch contemporaneous with the J and H data).We then matched to the J and H band FITS file cataloguestaking into account the proper motion of the source to calculate theexpected position of the target in these images.Where the catalogue photometry might possibly be unsatisfac-tory (e.g. an IPHAS non-detection, or a high contrast binary with asmall separation where significant PSF overlap was likely to haveoccured) and the target was deemed interesting (e.g. a UCD candi-date), it was necessary to perform additional photometry. We usedthe IRAF DAOPHOT package in these cases and the targets inquestion are identified in their tables and/or text.
To identify those high proper motion sources already in the litera-ture we cross checked against both SIMBAD and VizieR. Betweenthem these services contain several catalogues of verified highproper motion sources (e.g. the LSPM catalogue, L´epine & Shara2005; the search by Boyd et al., 2011a, 2011b) which are likely tohave previously identified many of the same sources from the 617that we identified.We used the SIMBAD script service to compile a list of allstars in their database with proper motion > mas yr − . Tothis we matched the epoch 2000 positions of our high proper mo-tion detections using a 15” matching radius, keeping only the clos-est match. We considered these matches genuine and the sourceknown if the J, H, and K photometry taken from the 2MASS PointSource Catalogue (Skrutskie et al. 2006) did not differ by morethan one magnitude. We note that most matches with differingphotometry also had large proper motion differences. We iden-tified 426 of our sources in the SIMBAD database, leaving 191unknown at this stage. We identified all catalogues in the VizieRdatabase which contain any source within 15” of the position ofeach of our remaining high proper motion candidates. We dis-missed identifications from the catalogues which were repeatedlyidentified but do not contain visually verified high proper motionsource discoveries. These include proper motion catalogues suchas the USNO-B1.0 Catalog, (Monet et al. 2003); the PPMXL Cat-alog (Roeser et al. 2010) and photometric catalogues such as theWISE All-Sky Data Release (Wright et al. 2010); the UKIDSS-DR6 Galactic Plane Survey (Lucas et al. 2008). Of the 191 sourceschecked, 29 were identified in other surveys (the majority identifiedby Boyd et al., 2011a, 2011b, which are not present in the SIMBADdatabase) and 162 had no corresponding sources in any VizieR cat-alogues other than non-verified proper motion catalogues and sin-gle epoch photometric catalogues. Subsequently we identified eightmore sources in common with the search by Luhman (2014) andanother with Luhman & Sheppard (2014). The remaining 153 highproper motion sources we consider to be new discoveries and canbe found in Table A1. We note that the known high proper mo-tion sources include twelve very recent WISE-based proper motiondiscoveries from Kirkpatrick et al. (2014), and Luhman (2014) andLuhman & Sheppard (2014) which had relatively poor astrometricprecision. In Table A2 we provide GPS proper motions for these c (cid:13) , 000–000 GPS High PM Sources
10 11 12 13 14 15 16 17 K µ t o t a l ( m a s y r − ) KnownNew
Figure 8.
The distribution of known (grey) and newly discovered (black)high proper motion sources identified by this work. objects, which benefit from higher resolution data and longer timebaselines.
Brown dwarfs have insufficient mass to support nuclear hydrogenburning. Since they lack a significant internal heating mechanismthey cool over time through the brown dwarf sequence. The browndwarf sequence begins with very young examples of late M typedwarfs and progresses on through L and T type dwarfs into Ydwarfs. An L type dwarf can also be either a low mass star or abrown dwarf depending on its mass and age (see e.g. Burrows et al.2001). L dwarfs occupy the 2250 K to 1400 K temperature range(Kirkpatrick et al. 1999), which allows dust to form in their pho-tospheres producing many spectral features not present in warmerobjects. Almost a thousand L dwarfs have been identified to datebut more discoveries of unusual L dwarfs are still needed to aid thedevelopment of evolutionary models and model atmospheres. Forexample relatively few L dwarfs are known in astrometric binaries,yet these are crucial for accurate mass determination. L dwarfs withunusual colours, low metallicity and halo kinematics, and low sur-face gravity (see e.g. Kirkpatrick et al. 2010, Faherty et al. 2012)are also poorly sampled.Since detectable L dwarfs are relatively nearby and as suchtend to exhibit large proper motions, we searched our high propermotion sample for previously undetected examples. L dwarfs ex-hibit a range of near infrared colours dependant on a number offactors. They are relatively faint in optical bandpasses and for thisreason optical to infrared colours are very useful for identifyingthem.We classified sources as L dwarf candidates if they displayedi-J > − K of 0.94. We note however that the IPHASfield containing this object was observed in poor weather, and thephotometry is unreliable as a result. This object also falls withinthe area of the Galactic plane covered by the SDSS. The SDSSi band magnitude of 19.08 gives us an i-J colour of 3.9 which is roughly consistent with a late M dwarf and would match the rela-tively blue J-K colour. The WISE W1 − W2 colour for this objectis 0.29 ± − K colour greater than its J − Hcolour. Another candidate, UGPS J054457.43+370504.1, is likelyto be within the 25 pc volume limited sample. We estimate its dis-tance at 16 to 23 pc assuming a spectral type of L2.5 ± Ultracool dwarfs as binary companions to other objects (e.g. mainsequence stars or white dwarfs) offer an opportunity to test theproperties predicted for them by atmospheric models and henceevaluate and refine the models themselves. Age and metallicity aredifficult to constrain observationally in UCDs and these propertiescan sometimes be measured for a companion and then adopted forthe UCD since they will usually have formed from the same molec-ular cloud at a similar time (Pinfield et al. 2006).We undertook a search for new benchmark UCD candidatesusing two methods: The first was a straightforward search of cur-rent proper motion catalogues for companions to the 153 previouslyunidentified high proper motion sources. The second method is awide search of the full 167 million source results table for commonproper motion companions to all 617 genuine high proper motionsources.
For the search of existing proper motion catalogues we usedthe LSPM catalogue for the northern sources, which also con-tains the Tycho-2 Catalogue of 2.5 Million Bright Stars (Tycho-2, Høg et al. 2000) and the All-sky Compiled Catalogue of 2.5million stars (ASCC-2.5, Kharchenko 2001), and the cataloguecreated by Boyd et al. (2011a, 2011b) for the small number ofour objects in the south. We performed a 1000” sky match, re-turning all matches with proper motion difference significance in α cos δ and δ combined < σ . The Boyd et al. search yielded noresults. Table 2 shows the seven candidate pairs identified in theLSPM search, two of which are matched to the same GPS source(UGPS J211859.26+433801.3) and are therefore a candidate triplesystem. Below we discuss the two systems in which the newly dis-covered high proper motion object may be an ultracool dwarf.UGPS J211859.26+433801.3 is a close, faint ( ∆ J ≃ = µ = ± mas yr − ) companion to the bright M dwarf LP 234-2220. LP 234-2220 was classified as an M3.5 dwarf with an esti-mated distance of 53.3 ± c (cid:13) , 000–000 L. Smith et al.
Table 1.
L dwarf candidates based on red optical to infrared colours and a minimum χ fit of theavailable photometry to brown dwarf colour templates. The i band photometry is from the IPHASsurvey and is on the Vega system. We note that these spectral types are approximations based onavailable photometry, which was limited in some cases.RA Dec K i − J J − H H − K SpTy est. Note03:53:04.59 +47:55:45.6 14.59 3.8 0.61 0.58 L0 a ∼ d a f f ∼ d > b a > b d > b d a a ∼ c c c e a aa Spectral type based on a photometric typing by Skrzypek et al. (2014, in prep) using availableoptical and NIR photometry. b Not visible in i band image, lower limit on i band photometry taken as the magnitude of a 3 σ detection. c Performed own photometry in i, J, H, and K. Near infrared colours have ± d Spectral type based on a photometric typing by Skrzypek et al. (2014, in prep), though missingseveral bands. e In a binary or triple system, see Section 5.3. f Based on IPHAS observations in poor weather, this object has SDSS coverage giving i-J of3.9 and an revised spectral type estimate of M8.5.
Table 2.
The seven candidate companions identified in a search of the LSPM catalogue, two of which are matched to the same GPS source andare therefore together a candidate triple system. The first three columns are those of the GPS source, the next four are of the LSPM candidatecompanion. The coordinates given are at epoch 2000.0. The proper motions of the candidate companion identified by an asterisk are those ofthe Tycho-2 catalogue (as indicated in the LSPM catalogue by the astrometric flag), otherwise they are those of the LSPM. All proper motionsare in units of mas yr − . K s is the 2MASS short K band magnitude obtained from the LSPM catalogue. α δ K i-J Name µ α cos δ µ δ K s Separation ∆ µ (”) Significance ( σ )04:02:29.42 +48:12:56.6 16.53 0.64 LSPM J0402+4812 139 -223 15.49 2.1 0.6604:35:19.94 +43:06:09.4 11.13 1.94 LSPM J0435+4305 ∗
155 -163 7.62 67.3 1.3521:11:04.39 +48:00:21.9 11.87 1.89 LSPM J2109+4811 183 127 11.25 949.2 a ∼ ∼ a a Owing to their very large angular separations these are likely chance alignments, see Section 5.3.3. on LP 234-2220. The uncertainties in the new photometry are rel-atively large, as much as 0.3 mag in each case. For this reason werelied more on the magnitude of the contrast than optical or infraredcolours as an indicator of spectral type since the uncertainty on thecontrast is dependant on the uncertainty of only one photometricmeasurement of the secondary, rather than colour which is depen-dent on two. A companion i − J colour of ∼ J ∼ ∆ J ≃ σ . The angularseparation gives a projected separation of order 50,000AU at 53pc.Such a system is unlikely to have survived for any significantlength of time and the IPHAS narrow band photometry indicatesthat neither component has any excess H α emission (whichwould have indicated youth). We find in Section 5.3.3 that weexpect to find several pairs of sources in our sample with suchlarge angular separations and similar proper motions that arenot physically associated. We therefore conclude that the similar c (cid:13) , 000–000 GPS High PM Sources proper motions of LP 234-2220 and 2MASS J2119+4352 are mostlikely coincidental.UGPS J214115.07+564012.9 (UGPS J2141+5640B hereafter) isa µ = ± mas yr − common proper motion companionto G 232-30 with a separation of ∼ = For the internal search of the full GPS proper motion results ta-ble we search for candidate common proper motion ( ∆ µ < mas yr − ) companions to all 617 identified high proper motionGPS sources with separations up to 30’. We applied no furtherquality control or brightness selection criteria to the candidate listin order that we not reject any potentially valuable sources dueto e.g. a profile misclassification at a single epoch, this selectionreturned 1032 candidates. A visual inspection yielded 41 genuinehigh proper motion objects within this sample after removal of 5duplicate sources from frame overlap regions. Among these wefind 11 instances where both components are among the original617 GPS high proper motion sources and hence they produce a re-versed pair (i.e. two instances with switched components), removalof these pairs left us with 19 candidate common proper motion pairswhich we show in Table 3. Since all the companions did not meetthe original high proper motion source candidate selection criteriatheir astrometry is likely compromised and the uncertainty on theproper motion will be underestimated, as a result the stated signifi-cance of the proper motion difference should be regarded as a lowerlimit.Based on the i-J colours of the original GPS high proper mo-tion sources from Table 3 and their K band contrasts we identifiedpairs 1, 2, and 12 as candidates for new UCD benchmark objects.Based on the positions of the three candidate primaries in a K bandreduced proper motion (H K ) vs. i-J plot, the primary in pair 2 ap-pears to be a white dwarf (faint in H K , blue in i-J) while the re-maining two candidate primaries appear to be main sequence starswith the primary of pair 12 just on the edge of the subdwarf locus(unremarkable i-J, faint in H K ). Inspection of the IPHAS i band im-ages and catalogues showed that the secondary in pair 2 is equal in iband brightness to the primary (17.39) and it is therefore likely thatthey are a pair of equal mass white dwarfs given their almost identi-cal i-K colour. The secondaries in pairs 1 and 12 are non-detectionsin the IPHAS i band images and are promising UCD candidates asa result.UGPS J034214.85+541019.6 AB: - pair 1 in Table 3,UGPS J0342+5410 AB hereafter. UGPS J0342+5410 B is an µ = ± mas yr − IPHAS i band non-detection, the 3 σ de-tection limit of the field is 21.2. The pair are separated by 2.9”.Flux from the primary at this radius only increases the backgroundcount level by of order 30% in the i band image so the detection limit should still be reasonably accurate. UGPS J0342+5410 B hasJ = ± = ± = = ± µ = ± mas yr − )and UGPS J0654+0400 ( µ = ± mas yr − ) hereafter.UGPS J0654+0400 is an IPHAS i band non-detection, the 3 σ detection limit of the field is 20.3. It has J and H band magni-tudes of 18.86 ± ± = = Here we evaluate the probability that the common proper motioncompanions discussed above are chance alignments. Our estimatesare drawn from large simulations made with the online Besanc¸onsynthetic stellar population tool (Robin et al. 2003). The catalogueswere generated with a K magnitude range equivalent to our ourhigh proper motion sample, using the Galactic position of eachcandidate. We generated ”small field” simulated catalogues withseveral million stars (equivalent to a 1500 deg area but with prop-erties fixed for the precise Galactic location) for each of the binarycandidates discussed. The catalogue simulations generate realisticproper motions for each source but do not produce physically as-sociated systems (such as moving groups or binaries). All com-mon proper motion companions in the sample are therefore purelychance alignments. We then identified all sources within each sim-ulated catalogue that have a proper motion consistent with the GPScomponent of the binary candidate, within the 2 σ uncertainty on theproper motion difference. The approximate probability of a sourceappearing within a given angular separation r and with a commonproper motion to one of our GPS high proper motion sources istherefore the number of matches in the simulated catalogue multi-plied by the area of a circle with radius r , divided by 1500 deg .We must also take into account that we have 617 high proper mo-tion objects, and therefore 617 chances of finding such a chancealignment. For each binary candidate discussed above we treatedthis using a simple multiplicative factor. In fact there is some vari-ation in the number of proper motion matches in the simulated cat-alogues, depending on Galactic coordinates and the direction ofproper motion but tests indicate that this is less than a factor of 4.In the cases discussed above with separations <
1’ the probabilityis always less than 10 − so the factor of 4 is not significant. Note c (cid:13) , 000–000 L. Smith et al.
Table 3.
Internal GPS common proper motion companion candidates. Columns 2 to 4 refer to the original GPS high proper motion source and columns 5 to8 refer to the GPS source which did not meet the original high proper motion candidate selection criteria. The coordinates given are at epoch 2000.0. Thedropout note indicates the reason the companion was not selected as an initial high proper motion candidate. Dropout note key: a - flagged as saturated ateither epoch; b - flagged as a galaxy at either epoch; c - bad pixel within 2” aperture flag at either epoch; d - ellipticity > µ tot justbelow original selection criteria (200 mas yr − ); f - neither epoch K band magnitudes below 17. In the ‘Known’ column the left tick/cross corresponds tothe original GPS source, the right tick/cross corresponds to the new GPS companion candidate.Pair α δ K α δ K Dropout Separation ∆ µ KnownNote (”) Significance ( σ )1 03:42:14.85 +54:10:19.6 11.14 03:42:14.52 +54:10:18.8 15.69 bd 2.9 1.10 ✓ ✗ ✗ ✗ ✓ ✗ ✓ ✓ ✓ ✗ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✗
10 05:47:33.08 +38:03:05.5 15.02 05:47:32.93 +38:03:05.2 16.91 b 1.8 0.44 ✗ ✗
11 06:48:20.56 +05:40:33.5 11.20 06:48:20.30 +05:40:30.1 9.09 a 5.1 2.12 ✓ ✓
12 06:53:11.66 +03:47:50.7 12.44 06:54:20.86 +04:00:56.5 17.11 cf 1300.1 ✓ ✗
13 20:40:04.51 +42:21:07.1 12.03 20:39:52.64 +42:20:33.8 10.91 d 135.7 0.64 ✓ ✓
14 20:55:51.57 +43:27:48.1 13.22 20:56:56.34 +43:08:08.4 11.19 a 1375.4 ✓ ✓
15 21:18:59.26 +43:38:01.3 13.50 21:18:59.30 +43:38:03.5 10.04 a 2.2 1.83 ✗ ✓
16 21:18:59.26 +43:38:01.3 13.50 21:19:30.94 +43:52:26.8 9.89 a 931.2 ✗ ✓
17 21:23:42.21 +44:19:17.1 12.76 21:23:43.45 +44:19:28.0 10.59 a 17.2 0.13 ✓ ✓
18 22:21:29.14 +55:56:00.1 14.20 22:23:34.94 +56:10:05.1 10.39 a 1350.9 ✗ ✓
19 22:37:06.22 +55:54:40.8 10.82 22:37:05.95 +55:54:44.3 9.33 a 4.2 0.55 ✓ ✓ Known high proper motion object is actually a blend of both components, previously unresolved. Owing to their very large angular separations these are likely chance alignments, see Section 5.3.3. that for each candidate companion the Galactic coordinates, propermotions and angular separations are different and they require aunique calculation as a result.In Tables 2 and 3 we listed six very widely separated ( > T type brown dwarfs are later in the brown dwarf sequence than Ldwarfs. As they age and cool through the brown dwarf sequence thedust present in L dwarf atmospheres sinks below the photosphereand its effect on their spectra disappears. The cooler temperaturesof T dwarfs allow molecules such as methane and water to form,which are responsible for the deep absorption features observed intheir spectra. They emit most of their radiation in the near infraredand for this reason recent large scale near infrared sky surveys (e.g.2MASS, Skrutskie et al. 2006; DENIS, Epchtein et al. 1997; CF-BDS, Delorme et al. 2010; UKIDSS, Lawrence et al. 2007) are re-sponsible for the majority of current T dwarf discoveries. Later typeT dwarfs are cooler still and the WISE mission (Wright et al. 2010)in the mid infrared becomes more sensitive to them at around T6and later. To date several hundred T dwarfs have been identified.T dwarfs are extremely faint, even in the near infrared and areonly detectable by the current generation of large scale NIR surveysout to of order 100pc. Due to the close proximity of detectable T dwarfs they tend to exhibit relatively large proper motions. For thisreason proper motion searches such as this could be expected toidentify many examples of T dwarfs. We have found two new ex-amples of T dwarfs, which we describe below, among the 153 pre-viously unidentified high proper motion sources due to their char-acteristic blue J-H and H-K colors. We also recover UGPS J0722-05 amongst the 617 high proper motion sources. The two GPS Tdwarfs identified by Burningham et al. (2011) are fainter than our K =
17 cut and were not recovered as a result. Two very nearby brightT dwarfs in the GPS footprint were identified by the WISE team re-cently. WISE J192841.35+235604.9 (Mace et al. 2013) lies outsidethe area covered by this paper and WISE J200050.19+362950.1(Cushing et al. 2014) was excluded due to a high ellipticity and aprofile misclassification in the second epoch K band observation.Our proper motion for WISE J200050.19+362950.1 is 75 ± ± mas yr − in α cos δ and δ respectively.UGPS J20480024+503821.9 (UGPS J2048+5038 hereafter) wasidentified as a J = µ = ± mas yr − , IPHAS i bandnon-detection. The 3 σ IPHAS i band detection limit of this fieldis 20.4 ± > − H andH − K (-0.07 and 0.05 respectively) and the H − W2 and W1 − W2colours ( ∼ ∼ − W2 and W1 − W2 colours ( ∼ ∼ c (cid:13) , 000–000 GPS High PM Sources Wavelength (µm) F λ ( − W m − µ m − ) Figure 9.
An IRTF SpeX spectrum of the previously unidentified T5 dwarfUGPS J2048+5038.
We obtained a NASA Infrared Telescope Facility (IRTF)SpeX (Rayner et al. 2003) spectrum of UGPS J2048+5038 (seeFigure 9) on the 30th of September 2013 using the 0.8” slit inprism mode with six AB nod cycles of 200s per nod, althoughthe target drifted out of the slit for the final two and these werediscarded as a result. This gave a total on source time of 1600s. Wealso observed the spectral standard HD199217 for use in removingtelluric features from the target spectrum and flux calibration. Wecombined and reduced the spectrum using the standard reductiontool: SpeXTool (Cushing et al. 2004).We have classified UGPS J2048+5038 following the spectral typingscheme laid out by Burgasser et al. (2006) for T dwarfs. In Figure10 we show our SpeX YJHK spectrum of UGPS J2048+5038compared to the T4 (2MASS J22541892+3123498) andT5 (2MASS J15031961+2525196) spectral templates ofBurgasser et al. (2006). The new source matches the T5 tem-plate very closely in the J and H band flux peaks, but showsless flux in the Y band peak, and enhanced flux in the K band.In Table 4 we give the spectral flux ratios used for index basedclassification in the Burgasser et al. (2006) scheme. These valuesfurther support the classification of UGPS J2048+5038 as a T5,and we thus adopt this classification for this object (T5 ± M J of14.44 for an isolated T5 dwarf, which puts UGPS J2048+5038(J = µ = ± mas yr − ) found theclosest match to have a separation >
1” of the expected position ofthe target at the IPHAS epoch. Subsequent visual inspection of theIPHAS i band image confirmed a mismatch or blend with a back-ground source. We attempted a subtraction of the PSF of the back-ground source using the standard IRAF
ALLSTAR program andsubsequent inspection of the residual image showed no remainingsign of either the background source or the target. This suggestedeither the pair were so close they were completely unresolved, ormore likely given the expected ∼ Figure 10.
Our
Y JHK spectrum of UGPS J2048+5038 compared to theT4 (2MASS J15031961+2525196 and T5 (2MASS J15031961+2525196)spectral templates defined in Burgasser et al. (2006).
Table 4.
The spectral flux ratios used for classifying UGPS J2048+5038.Index Ratio Value TypeH O-J R . . f ( λ ) dλ R . . f ( λ ) dλ . ± . T5CH -J R . . f ( λ ) dλ R . . f ( λ ) dλ . ± . T5H O- H R . . f ( λ ) dλ R . . f ( λ ) dλ . ± . T4/5CH - H R . . f ( λ ) dλ R . . f ( λ ) dλ . ± . T5CH -K R . . f ( λ ) dλ R . . f ( λ ) dλ . ± . T5 An approximate 3 σ IPHAS i band detection limit of this fieldis 20.5 magnitudes, suggesting the target is approximately L0 orlater type. Given the blue nature of UGPS J0355+4743 in the nearinfrared, (J-H = -0.37 and H-K = − W2 and W1 − W2of ∼ ∼ M J of 14.78 for an isolated T6dwarf, which puts UGPS J0355+4743 (J = UGPS J04514383+4549580 ( µ = ± mas yr − ) is a faintcompanion to LHS 1708, a G1 type main sequence star, with a sep-aration of 5.6” and a ∆ J of 5.9 magnitudes. It was identified in oursearch for new candidate UCD benchmark objects. Our proper mo-tion differs by 2.1 σ from the Hipparcos proper motion of LHS 1708and as a result the pair did not show in the < σ candidate com-panion list in Section 5.3.1. Given the K band brightness (10.8), theuncertainty on our proper motion for UGPS J04514383+4549580( ± mas yr − ) is likely underestimated by the pipeline and the c (cid:13) , 000–000 L. Smith et al.
Table 5.
Parameters of the two previously unidentified T dwarfs which wedescribe in Section 5.4. UGPS J2048+5038 UGPS J0355+4743Right Ascension 20:48:00.24 03:55:32.00Declination +50:38:21.9 +47:43:58.8Spectral Type T5 ∼ T6J 16.30 16.20 µ total mas yr − mas yr − Distance ∼ pc ∼ pc J − H -0.07 -0.37H − K 0.05 0.08W1 − W2 ∼ − W2 ∼ significance of the similarity in the proper motion of this pair istherefore also underestimated. Based on its IPHAS optical and GPSinfrared colours UGPS J04514383+4549580 is either a mid-M typemain sequence star or a white dwarf. LHS 1708 does have an en-try in the Washington Double star Catalogue (WDS, Mason et al.2001), although the stated separations (101.6” and 85.80” for the1909 and 1989 epochs respectively) are higher than the object wehave identified. The position of the WDS secondary at the 1909and 1989 epochs, which we calculated from the stated separationsand position angles relative to the position of LHS 1708 at thetwo epochs, is consistent with a bright source on the GPS imagewhich shows no proper motion. We conclude that the secondarygiven in the WDS does not share a proper motion with LHS 1708and is therefore not a genuine companion. However, our objectUGPS J04514383+4549580, is a genuine common proper motioncompanion to LHS 1708. We present the results of a search for high proper motion objects inthe UKIDSS Galactic Plane Survey. We selected 5,655 high propermotion ( µ > mas yr − ) candidates from 900 deg of sky at l > ◦ and K <
17 for visual verification and found 617 to begenuine, 153 of which were previously unidentified. Among thenew high proper motion discoveries we identified two new mid Tdwarfs that are likely to be within 25 pc, a further thirteen new Ldwarf candidates and two ultracool dwarf binary candidates.The large 24” matching radius we adopted in an effort to de-tect objects with very high proper motions at the expense of a largenumber of mismatches gave an overall ratio of false to genuine can-didates of 9:1. At high galactic longitudes, where the source densityis much lower, we found this to be less of a problem; the ratio offalse to genuine candidates at l = ◦ is ∼ <
16 is 6.6 mas yr − . Our propermotions for sources in common with existing long epoch baselineoptical catalogues are in good agreement within their uncertainties.Proper motions were calculated for 167 million sources in to-tal and we plan to extend our search to objects with lower but not in-significant motions. We also plan to extend the selection presentedhere to data taken after March 31st 2013, search for brighter highproper motion objects at l < ◦ , and search for high proper motionobjects at K >
17 with the aid of colour selections.
ACKNOWLEDGEMENTS
We would like to thank the referee, John Gizis, for a very positivereview. LS acknowledges a studentship funded by the Science &Technology Facilities Research Council (STFC) of the UK; PWL,PP, DJP, GB, JED acknowledge the support of a consolidated grant(ST/J001333/1) also funded by STFC. This work is based in part ondata obtained as part of the UKIRT Infrared Deep Sky Survey. Theauthors would like to acknowledge the Marie Curie 7th EuropeanCommunity Framework Programme grant n.247593 Interpretationand Parameterization of Extremely Red COOL dwarfs (IPER-COOL) International Research Staff Exchange Scheme. A.H.A. ac-knowledges CNPq grant PQ306775/2009-3. D.R.R. acknowledgessupport from FONDECYT grant 3130520. This research has madeuse of the SIMBAD database and VizieR catalogue access tool,operated at CDS, Strasbourg, France. This research has made useof NASA’s Astrophysics Data System Bibliographic Services. Thisresearch has made use of SAOImage DS9, developed by Smith-sonian Astrophysical Observatory. This research has made use ofthe SpeX Spectrograph and Imager on the NASA Infrared Tele-scope Facility, see Rayner et al. (2003). The authors would like toacknowledge Setpoint Hertfordshire and the Nuffield Foundationfor organising and funding the research placement of R. Bunce.
REFERENCES
Artigau ´E., Radigan J., Folkes S., Jayawardhana R., Kurtev R.,Lafreni`ere D., Doyon R., Borissova J., 2010, ApJ, 718, L38Barentsen G., Farnhill H. J., Drew J. E., et al. 2014, MNRASBeam´ın J. C., Minniti D., Gromadzki M., Kurtev R., Ivanov V. D.,Beletsky Y., Lucas P., Saito R. K., Borissova J., 2013, A&A, 557,L8Boyd M. R., Henry T. J., Jao W.-C., Subasavage J. P., HamblyN. C., 2011, AJ, 142, 92Boyd M. R., Winters J. G., Henry T. J., Jao W.-C., Finch C. T.,Subasavage J. P., Hambly N. C., 2011, AJ, 142, 10Burgasser A. J., Geballe T. R., Leggett S. K., Kirkpatrick J. D.,Golimowski D. A., 2006, ApJ, 637, 1067Burningham B., Leggett S. K., Homeier D., Saumon D., LucasP. W., Pinfield D. J., Tinney C. G., Allard F., Marley M. S., JonesH. R. A., Murray D. N., Ishii M., Day-Jones A., Gomes J., ZhangZ. H., 2011, MNRAS, 414, 3590Burningham B., Lucas P. W., Leggett S. K., Smart R., Baker D.,Pinfield D. J., Tinney C. G., Homeier D., Allard F., Zhang Z. H.,Gomes J., Day-Jones A. C., Jones H. R. A., Kov´acs G., LodieuN., Marocco F., Murray D. N., Sip˝ocz B., 2011, MNRAS, 414,L90Burrows A., Hubbard W. B., Lunine J. I., Liebert J., 2001, Re-views of Modern Physics, 73, 719Castro P. J., Gizis J. E., 2012, ApJ, 746, 3Castro P. J., Gizis J. E., Harris H. C., Mace G. N., KirkpatrickJ. D., McLean I. S., Pattarakijwanich P., Skrutskie M. F., 2013,ApJ, 776, 126Contreras Pe˜na C., Lucas P. W., Froebrich D., Kumar M. S. N.,Goldstein J., Drew J. E., Adamson A., Davis C. J., Barentsen G.,Wright N. J., 2014, MNRAS, 439, 1829Cushing M. C., Kirkpatrick J. D., Gelino C. R., Mace G. N., Skrut-skie M. F., Gould A., 2014, ArXiv e-printsCushing M. C., Vacca W. D., Rayner J. T., 2004, PASP, 116, 362Day-Jones A. C., Marocco F., Pinfield D. J., Zhang Z. H., Burn-ingham B., Deacon N., Ruiz M. T., Gallardo J., Jones H. R. A., c (cid:13) , 000–000 GPS High PM Sources Lucas P. W. L., Jenkins J. S., Gomes J., Folkes S. L., ClarkeJ. R. A., 2013, MNRAS, 430, 1171Delorme P., Albert L., Forveille T., Artigau E., Delfosse X., Reyl´eC., Willott C. J., Bertin E., Wilkins S. M., Allard F., Arzouma-nian D., 2010, A&A, 518, A39Drew J. E., Greimel R., Irwin M. J., Aungwerojwit A., BarlowM. J., Corradi R. L. M., Drake J. J., G¨ansicke B. T., Groot P.,Hales A., Hopewell E. C., Irwin J., Knigge C., Leisy P., et al.2005, MNRAS, 362, 753Dupuy T. J., Liu M. C., 2012, ApJ, 201, 19Dye S., Warren S. J., Hambly N. C., Cross N. J. G., Hodgkin S. T.,Irwin M. J., Lawrence A., Adamson A. J., Almaini O., EdgeA. C., Hirst P., Jameson R. F., Lucas P. W., van Breukelen C.,Bryant J., Casali M., Collins R. S. e. a., 2006, MNRAS, 372,1227Epchtein N., de Batz B., Capoani L., Chevallier L., Copet E.,Fouqu´e P., Lacombe P., Le Bertre T., Pau S., Rouan D., RuphyS., Simon G., Tiph`ene D., Burton W. B., Bertin E., et al. 1997,The Messenger, 87, 27Faherty J. K., Burgasser A. J., Walter F. M., Van der Bliek N.,Shara M. M., Cruz K. L., West A. A., Vrba F. J., Anglada-Escud´eG., 2012, ApJ, 752, 56Folkes S. L., Pinfield D. J., Jones H. R. A., Kurtev R., ZhangZ., G´alvez-Ortiz M. C., Marocco F., Day-Jones A. C., ClarkeJ. R. A., 2012, MNRAS, 427, 3280Gizis J. E., Burgasser A. J., Faherty J. K., Castro P. J., SharaM. M., 2011, AJ, 142, 171Høg E., Fabricius C., Makarov V. V., Urban S., Corbin T., WycoffG., Bastian U., Schwekendiek P., Wicenec A., 2000, A&A, 355,L27Kharchenko N. V., 2001, Kinematika i Fizika Nebesnykh Tel, 17,409Kirkpatrick J. D., Cushing M. C., Gelino C. R., Griffith R. L.,Skrutskie M. F., Marsh K. A., Wright e., 2011, ApJ, 197, 19Kirkpatrick J. D., Looper D. L., Burgasser A. J., Schurr S. D.,Cutri R. M., Cushing M. C., Cruz K. L., Sweet A. C., KnappG. R., Barman T. S., Bochanski J. J., Roellig T. L., McLean I. S.,McGovern M. R., Rice E. L., 2010, ApJ, 190, 100Kirkpatrick J. D., Reid I. N., Liebert J., Cutri R. M., Nelson B.,Beichman C. A., Dahn C. C., Monet D. G., Gizis J. E., SkrutskieM. F., 1999, ApJ, 519, 802Kirkpatrick J. D., Schneider A., Fajardo-Acosta S., Gelino C. R.,Mace G. N., Wright E. L., Logsdon S. E., McLean I. S., CushingM. C., Skrutskie M. F., Eisenhardt P. R., Stern D., Balokovi´c M.,Burgasser A. J., Faherty J. K., et al. 2014, ApJ, 783, 122Lawrence A., Warren S. J., Almaini O., Edge A. C., Hambly N. C.,Jameson R. F., Lucas P., Casali e., 2007, MNRAS, 379, 1599L´epine S., Shara M. M., 2005, AJ, 129, 1483Looper D. L., Kirkpatrick J. D., Burgasser A. J., 2007, AJ, 134,1162Lucas P. W., Hoare M. G., Longmore A., Schr¨oder A. C., DavisC. J., Adamson A., Bandyopadhyay R. M., de Grijs R., SmithM., Gosling A., Mitchison S., G´asp´ar A., Coe M., Tamura M., etal.. 2008, MNRAS, 391, 136Lucas P. W., Tinney C. G., Burningham B., Leggett S. K., Pin-field D. J., Smart R., Jones H. R. A., Marocco F., Barber R. J.,Yurchenko S. N., Tennyson J., Ishii M., Tamura M., Day-JonesA. C., Adamson A., Allard F., Homeier D., 2010, MNRAS, 408,L56Luhman K. L., 2013, ApJ, 767, L1Luhman K. L., 2014, ApJ, 781, 4Luhman K. L., Sheppard S. S., 2014, ArXiv e-prints Mace G. N., Kirkpatrick J. D., Cushing M. C., Gelino C. R., Grif-fith R. L., Skrutskie M. F., Marsh K. A., Wright E. L., EisenhardtP. R., McLean I. S., Thompson M. A., Mix K., Bailey V., Beich-man C. A., et al. 2013, ApJ, 205, 6Mason B. D., Wycoff G. L., Hartkopf W. I., Douglass G. G., Wor-ley C. E., 2001, AJ, 122, 3466Monet D. G., Levine S. E., Canzian B., Ables H. D., Bird A. R.,Dahn C. C., Guetter e., 2003, AJ, 125, 984Phan-Bao N., Bessell M. S., Mart´ın E. L., Simon G., BorsenbergerJ., Tata R., Guibert J., Crifo F., Forveille T., Delfosse X., Lim J.,de Batz B., 2008, MNRAS, 383, 831Pinfield D. J., Jones H. R. A., Lucas P. W., Kendall T. R., FolkesS. L., Day-Jones A. C., Chappelle R. J., Steele I. A., 2006, MN-RAS, 368, 1281Rayner J. T., Toomey D. W., Onaka P. M., Denault A. J.,Stahlberger W. E., Vacca W. D., Cushing M. C., Wang S., 2003,PASP, 115, 362Reid I. N., Cruz K. L., Allen P., Mungall F., Kilkenny D., LiebertJ., Hawley S. L., Fraser O. J., Covey K. R., Lowrance P., Kirk-patrick J. D., Burgasser A. J., 2004, AJ, 128, 463Robin A. C., Reyl´e C., Derri`ere S., Picaud S., 2003, A&A, 409,523Roeser S., Demleitner M., Schilbach E., 2010, AJ, 139, 2440Scholz R.-D., 2014, A&A, 561, A113Skrutskie M. F., Cutri R. M., Stiening R., Weinberg M. D., Schnei-der S., Carpenter J. M., Beichman C. e., 2006, AJ, 131, 1163Smith L., Lucas P. W., Burningham B., Jones H. R. A., SmartR. L., Andrei A. H., Catal´an S., Pinfield D. J., 2014, MNRAS,437, 3603Taylor M. B., 2006, in Gabriel C., Arviset C., Ponz D., EnriqueS., eds, Astronomical Data Analysis Software and Systems XVVol. 351 of Astronomical Society of the Pacific Conference Se-ries, STILTS - A Package for Command-Line Processing of Tab-ular Data. p. 666Wright E. L., Eisenhardt P. R. M., Mainzer A. K., Ressler M. E.,Cutri R. M., Jarrett T., Kirkpatrick J. D., Padgett D., McMillane., 2010, AJ, 140, 1868
APPENDIX A: c (cid:13) , 000–000 L. Smith et al.
Table A1.
The high proper motion sources identified for the first time in this publication. The coordinates given are at epoch 2000.0 α δ µ α cos δ µ δ r i J H K03:20:43.87 +59:18:23.2 146 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
12 -168 ±
11 16.20 ± ± ± ±
10 -224 ±
10 18.00 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ae ± ± ± ± ± ± ± ± ± ± ± a ± ± ± ± ± ± ± b ± b ± b ±
13 -133 ±
13 19.48 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± > d ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± e ± e ± ± ± ±
11 -189 ±
11 20.21 ± ± ± ± ± ± ± ± ± ± ± ± ±
10 -243 ± ± ± ± ± ± ± ± ± ± ± ± ±
13 -280 ± > d ± ± ± ±
10 -200 ±
10 19.9 ± a ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± > d ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± a Based on our own aperture photometry due to the source being a non-detection in IPHAS i band catalogue but visible in image. b Based on our own aperture photometry, survey photometry was deemed unreliable due to nearby bright companion. c Undetected in IPHAS source identification. Profile fit photometry performed to measure this magnitude. d Undetected in IPHAS i band, 3 σ limit of field given. e Based on IPHAS observations in poor weather, this photometry is unreliable. c (cid:13) , 000–000
GPS High PM Sources Table A1 – continued α δ µ α cos δ µ δ r i J H K06:33:33.37 +10:01:27.4 184 ±
15 -214 ±
12 19.28 ± e ± e ± ± ± ± ± ± ± ± ± ± ± ae ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± a ± ± ± ± ± ± ± ± ± ± ± e ± e ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
10 -96 ± > d ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
11 -199 ±
12 19.85 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± > d ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
23 -189 ±
20 16.47 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± > d ± ± ± ± ± ± ± ± ± ± (cid:13)000
20 16.47 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± > d ± ± ± ± ± ± ± ± ± ± (cid:13)000 , 000–000 L. Smith et al.
Table A1 – continued α δ µ α cos δ µ δ r i J H K21:05:57.74 +47:01:44.5 126 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± c ± b ± b ± b ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
13 182 ±
13 17.78 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± (cid:13) , 000–000 GPS High PM Sources Table A2.
UKIDSS GPS proper motions and epoch 2000.0 coordinates for twelve recent WISE discoveries. α δ µ α cos δ µ δ K Discoverer03:23:01.53 +56:26:00.8 289 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± (cid:13)000