The IPHAS-POSS-I proper motion survey of the Galactic Plane
N.R. Deacon, P.J. Groot, J.E. Drew, R. Greimel, N.C. Hambly, M.J. Irwin, A. Aungwerojwit, J. Drake, D. Steeghs
aa r X i v : . [ a s t r o - ph . GA ] M a y Mon. Not. R. Astron. Soc. , 1–18 (2005) Printed 21 November 2018 (MN L A TEX style file v2.2)
The IPHAS-POSS-I proper motion survey of theGalactic Plane
N.R. Deacon ⋆ , P.J. Groot , J.E. Drew , R. Greimel , , N.C. Hambly , M.J. Irwin ,A. Aungwerojwit , , J. Drake , D. Steeghs , , Department of Astrophysics, IMAPP, Radboud University Nijmegen, P.O. Box 9010, 6500 GL Nijmegen, The Netherlands Centre for Astrophysics Research, University of Hertfordshire, College Lane, Hatfield AL10 9AB Isaac Newton Group of Telescopes, Apartado de correos 321, E38700 Santa Cruz de La Palma, Tenerife, Spain Institut f¨ur Physik, Karl-Franzen Universit¨at Graz, Universit¨atsplatz 5, 8010 Graz, Austria SUPA † , Institute for Astronomy, School of Physics, University of Edinburgh, Royal Observatory Edinburgh,Blackford Hill, Edinburgh, EH9 3HJ Institute of Astronomy, Madingley Road, Cambridge CB3 0HA Department of Physics, Faculty of Science, Naresuan University, Phitsanulok, 65000, Thailand Department of Physics, University of Warwick, Coventry, CV4 7AL, UK Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138, USA
ABSTRACT
We present a proper motion survey of the Galactic plane, using IPHAS dataand POSS-I Schmidt plate data as a first epoch, that probes down to propermotions below 50 milliarcseconds per year. The IPHAS survey covers thenorthern plane ( | b | < ◦ ) with CCD photometry in the r , i and H α pass-bands. We examine roughly 1400 sq. deg. of the IPHAS survey area and drawup a catalogue containing 103058 objects with significant proper motions be-low 150 millarcseconds per year in the magnitude range 13.5 < r ′ <
19. Oursurvey sample contains large samples of white dwarfs and subdwarfs which canbe identified using a reduced proper motion diagram. We also found severalobjects with IPHAS colours suggesting H α emission and significant propermotions. One is the known cataclysmic variable GD552; two are known DBwhite dwarfs and five others are found to be non-DA (DB and DC) whitedwarfs, which were included in the H α emission line catalogue due to theirlack of absorption in the H α narrow-band. c (cid:13) N.R. Deacon et al.
Key words:
Astronomical data bases: Surveys – optical: stars – Astrometryand celestial mechanics: Astrometry – Stars
The INT Photometric H α Survey (IPHAS, Drew et al., 2005) is a deep ( r < r , i , H α ) covering 1800 sq. deg. of the northern Galactic Plane( | b | < ◦ ). IPHAS forms part of the European Galactic Plane Surveys (EGAPS), which alsoincludes the UKIRT Infrared Deep Sky Survey (UKIDSS) Galactic Plane Survey (Lawrenceet al., 2007, Lucas et al. 2008) covering 1800 sq. deg. of the plane in J , H , K to a depthof K =19 and the UV EXcess survey (UVEX, Groot et al. in prep.). UVEX is planned tocomplement IPHAS by covering the same area but in u , g and HeI 5875˚Awith an additional r band epoch. These surveys also have upcoming southern counterparts. With the numberdensity of stars highly concentrated on the Plane, IPHAS and EGAPS provide ideal toolsto study a whole range of stellar and Galactic research topics. They have already yieldedsignificant discoveries in fields such as cataclysmic variables (Witham et al., 2007), planetarynebulae (Mampaso et al., 2006, Wesson et al., 2008), young low mass objects (Valdivielso etal., 2009), star forming regions (Vink et al., 2008) and extinction in the Galactic plane (Saleet al., 2009). Large scale CCD-based astronomical surveys such as IPHAS provide accuratephotometric and astrometric data on large numbers of astronomical objects. In addition totheir main science goals, surveys such as EGAPS make their data public (see Gonzalez-Solares et al., 2008 for details on public IPHAS data) and they can be used by anyone inthe astronomical community to pursue their own research aims. Combining IPHAS datawith those from other surveys with different wavebands or epochs can lead to discoveries ofvariable objects and can also allow the parameter space of each object to be expanded toinclude not only magnitudes and positions but proper motions as well. Here we undertakethe first comprehensive, optical, wide field survey to identify proper motions below 0.1arcseconds per year in the Galactic Plane by cross-referencing the IPHAS database withSuperCOSMOS (Hambly et al., 2001) scans of the POSS-I plates taken in the 1950s. Thisgives us a proper motion baseline of approximately fifty years.Early proper motion surveys utilised blink comparators and exceptional patience to in- ⋆ E-mail: [email protected] † Scottish Universities’ Physics Alliance c (cid:13) , 1–18
PHAS-POSS-I Proper Motion Survey µ > | b | < ◦ at magnitudes fainter than V =16(Lepine & Shara, 2005). The main modern computational study is that of Lepine (2008).They used a sophisticated algorithm to degrade POSS-II images to the same quality asthe older POSS-I images. The two could then be subtracted and high proper motion starsidentified. This survey is complete to V=20 and µ =0.15”/yr. However the survey suffersfrom crowding in the Galactic Plane leading to a reduction in completeness. Lepine & Sharaestimate they are only 80-90% complete down to V=19 within 15 degrees of the GalacticPlane. Fedorov et al. (2009) predict their upcoming catalogue will cover low proper mo-tions in the Galactic Plane but will not provide a consistent proper motion range due toa varying maximum proper motion. Gould & Kollmeier (2004) used data from the SloanDigital Sky Survey photographic plate data to produce a proper motion survey below 100milliarcseconds per year. However this avoided the Galactic plane. The study of Folkes et al.(2007) attempts to fill in the Galactic Plane gap left by southern surveys such as Deacon &Hambly (2007), Pokorny et al. (2004) and Finch et al. (2007) (all of which avoid the Plane)by combining UKST and 2MASS data in a similar manner to Deacon & Hambly (2007) toidentify candidate low mass stars and brown dwarfs from their proper motion. In order to plan our proper motion survey we had to first consider the datasets available.Two datasets are available for use as a first epoch, both having been scanned using theSuperCOSMOS plate scanning machine (Hambly et al. 2001). As well as the POSS-I plates,the newer, higher quality POSS-II plates with better emulsion sensitivity and improvedresolution are also available. These provide better astrometric accuracy but a much shortertime baseline with respect to IPHAS (10-15 years compared to the IPHAS data versus theroughly 50 year epoch difference betweeen IPHAS and POSS-I). However a shorter baselinemeans less contamination due to spurious pairings; n spurious ∝ ( µ max ∆ t ) , where n spurious isthe number of spurious pairings, µ max is the maximum proper motion and ∆ t is the epoch c (cid:13) , 1–18 N.R. Deacon et al.
Figure 1.
The astrometric errors (in arcseconds) between the IPHAS and UVEX surveys. difference. n spurious is also proportional to the density of objects around the target. This isone of the reasons most proper motion surveys have avoided higher density areas of the skysuch as the Galactic Plane. Along with these data we also have the upcoming UV EXcess(UVEX) survey (Groot et al., in prep.) which will be a blue companion to IPHAS and asecond r epoch. This will provide us with CCD quality second epoch astrometry, observedby the same telescope and camera, reduced by the same pipeline but with only a 3-5 yearbaseline. Examining the positional errors between IPHAS and UVEX we found that theywere typically 40 milliarcseconds (see Figure 1), rising to 50 mas at r =19 (where POSS-Iplate astrometry becomes difficult, see Figure 2) and to roughly 100 mas, as the survey limit(r ∼
21) is approached. Hence we can assume that with a three year baseline, the minimum5 σ proper motion detectable between IPHAS and UVEX at the survey limit (r ∼
21) isroughly 166 mas/yr. At the limit at which astrometry on the POSS-I plates becomes difficult( R F =19) the minimum proper motion becomes 100 mas/yr. Hence below this latter limit(also below the µ =0.15”/yr lower proper motion limit of Lepine & Shara, 2008) there is c (cid:13) , 1–18 PHAS-POSS-I Proper Motion Survey r =19) and leavingsome overlap we decided on a maximum proper motion of 0.15”/yr. This means that evenwith the exceptionally long baseline between IPHAS and the POSS-I plates the maximumpairing radius is only ∼ δ ∼ . ◦ ,south of this we use SuperCOMSOS UK Schmidt Telescope R plates.Before beginning the proper motion survey it is important to have both surveys on thesame astrometric framework as our initial calculations will be based on the global astrometricframeworks of both surveys. IPHAS is tied to the 2MASS astrometric framework so weconverted the POSS-I astrometry to the 2MASS astrometric reference frame. This was donein an identical way to the transformation of UKST I plates to the 2MASS system describedin Section 2.1 of Deacon & Hambly (2007).In order to estimate the minimum significant positional shift that can be detected werobustly calculated the positional errors between the POSS-I plates and the IPHAS survey(the error estimates calculated between the IPHAS and UVEX surveys found in Figure 1were calculated in the same way). This was done by identifying the same objects in eachepoch and calculating the positional differences. These were then binned by magnitude andthe error calculated (Figure 2). After examining this plot, we determined the 5 σ positionalshift to be at one arcsecond, this was then used as our minimum positional shift. Thismeans our minimum proper motion will be roughly 20 mas/yr. As we will calculate relativeastrometric solutions for each object in our final catalogue this number may vary slightly.Our search methodology was as follows. Objects which were flagged as stellar sourcesor probable stellar sources (classification flags -1 and -2, Drew et al., 2005) in IPHAS wereselected. This excludes saturated sources and hence introduces a bright limit to our survey c (cid:13) , 1–18 N.R. Deacon et al.
Figure 2.
The astrometric errors (in arcseconds) between the IPHAS and POSS-I data. at approximately r =13.5. One initial problem encountered was the difference in the sizesof the Point Spread Functions of the two surveys. Often two stars with a small separationwhich are resolved in IPHAS will be blended together on the lower resolution POSS platesleading to the erroneous conclusion that one or both of them have moved. To remove thispotential source of contamination any IPHAS object of brightness r = x (where x is inmagnitudes) which had another IPHAS object brighter than x − r =9).Subsequently, IPHAS objects which were not affected by such crowding had their posi-tions compared with the POSS-I data to see if they had a companion within an arcsecond.If they did they were judged not to have a significant proper motion and hence were ex-cluded. Any potential POSS-I pair for these unpaired objects was then searched for. First c (cid:13) , 1–18 PHAS-POSS-I Proper Motion Survey r max = µ max ∆ t was searched. Any potential pair had to have a POSS-I R F magnitude within 3 σ (where σ is approximated from the values for measurement errors quoted in Hambly et al. 2001,roughly 0.2 magnitudes at best ) of the IPHAS r magnitude and had to be stellar sourceswhich had not been deblended and were not in close proximity to bright stars (note this 3 σ cut could exclude high proper motion variables). To ensure that the paired POSS-I objectdoes not have an IPHAS counterpart, the POSS-I positions were crosschecked with IPHASpositions and any object with an IPHAS pair within one arcsecond was excluded. In order to gain an insight into the local astrometric accuracy of each proper motion mea-surement, a local relative astrometry mapping was carried out for each candidate. To do thisall objects in the same IPHAS field as the target with brightnesses within one magnitude ofthe star in question were selected. These were then used to produce a 6 parameter plate-platefit using SlaLib routines (Wallace, 1998) to determine the astrometric differences betweenthe two reference frames and to estimate the random errors remaining once these differenceshave been corrected for. This fit was then applied and used to calculate a proper motionrelative to this reference frame. This also yielded measures of the positional errors for eachfield. However in cases with few reference stars (i.e. <
20) the error will be underestimated.To correct for this we carried out a series of simulations. Sets of reference stars on two differ-ent reference frames were created. These were given small random bulk offsets between thereference frames as well as individual random Gaussian errors. A fit between the referenceframes was carried out and the calculated positional error compared to the indiviual posi-tional errors used. It was found that for few reference stars the error was underestimated.We find that the correction factor is well fitted by the equation, σ true σ measured ≈ . n . ref (1)Where σ true is the actual error, σ measured the measured error and n ref the number of referencestars. We find this relation holds fairly well down to as few as six reference stars. Thiscorrection factor was used to ensure all our quoted errors are accurate. Where there werenot enough reference stars for any fit an error calculated from the global positional errorestimates shown in Figure 2 was used. Note as the IPHAS photometric errors are typically much smaller than POSS-I errors we ignore them in our error estimation.c (cid:13) , 1–18
N.R. Deacon et al.
Figure 3.
Two reduced proper motion diagrams for our dataset. (a) shows all objects in our sample, (b) shows only those with µ greater than 50mas/yr. The populations shown are as follows, the main locus is the main sequence, below and to the left arethe higher velocity and bluer subdwarfs and to the left of them are the intrinsically fainter white dwarfs. The large grey dotsrepresent the objects common between this catalogue and the catalogue of H α emitters from Witham et al. (2008). These areall plotted on both panels, regardless of their proper motions. The final catalogue consists of 103058 objects spread across 14126 IPHAS fields (includingoverlap fields) where the area of each field is roughly 0.3 sq.deg. These objects all have propermotions more significant than 5 σ where the proper motion errors were typically below 10milliarcseconds per year (i.e. µ min < c (cid:13) , 1–18 PHAS-POSS-I Proper Motion Survey Figure 4.
Distribution of objects in our sample across the Galactic plane. The coverage is in general good, however the coverageappears patchy in parts, particularly at low Galactic longditude ( l <
Figure 5.
The density of stellar sources in the IPHAS survey with black being most dense and white being less dense. Thelarger stellar density closer to the Galactic centre along with the patches of extinction close to the plane in this region can beclearly seen. H r = r + 5 log µ + 5 log (47 . H r = M r + 5 log d − v T − (4 . − d + 8 . H r = M r + 5 log v T (2)Where µ is the proper motion in arcseconds per year, d is the distance in parsecs and v T is the tangental velocity in km/s. The above definition of reduced proper motion is not themost commonly used but is useful as it removes the constants needed to convert betweenunits. Our reduced proper motion diagram is shown in Figure 3. The form is roughly whatwe would expect from a standard Galactic stellar population with clearly identifiable dwarf,subdwarf and white dwarf loci. However after we studied the spatial distribution of objects c (cid:13) , 1–18 N.R. Deacon et al.
Figure 6.
The density of stellar sources in the IPHAS survey for each field vs. the number of proper motion objects detectedin each field. The solid line shows the mean number of objects for fields binned by stellar density. Note the general trend, densefields have fewer detected objects. This is because the crowding confusion reduction algorithm removes more of the area ofcrowded fields.
Figure 7.
Colour-colour diagrams for the objects. The panel on the left (a) shows all the objects in our sample while the panelon the right (b) shows only those with proper motions greater than 50mas/yr. The main stellar locus runs from (0.0,0.1) to(2.0,1.0), this is a near-perfect unreddened main sequence (see Drew et al., 2005). Below and to the left lie the bluer whitedwarfs and above and to the left lie potential H α emitters. The large grey dots represent the objects common between thiscatalogue and the catalogue of H α emitters from Witham et al. (2008). These are all plotted on both panels, regardless of theirproper motions. c (cid:13) , 1–18 PHAS-POSS-I Proper Motion Survey Figure 8.
A histogram of the separations of common proper motion pairs in our sample. The trend for coincidence objectswould be N ∝ r . As we see no deviation from this trend at small seperations, we conclude that there is no significant populationof true common proper motion binaries in our sample. it was found that there were several fields with many (more than 250) objects. After someinvestigation it became clear that these fields had poor astrometric solutions in the IPHASdata (mostly due to poor observing conditions). When we examined a histogram of numberof detected objects per field it was found that these fields lay beyond the point where themain distribution had died away. Additionally when the reduced proper motion diagramsfor objects in these fields was examined it was found that it did not contain the expectedpopulation distributions, indicating that the proper motion determinations were not correct.Hence any object lying in these fields was excluded from the final catalogue. A plot of thespatial distribution of the remaining objects can be found in Figure 4. It shows that for themajority of the northern plane, the coverage is good with a few patches of incompleteness.However moving along the plane towards the Galactic centre the number of objects dropsoff dramatically. This is due to our selection criteria excluding crowded regions as well aslarge numbers of objects in these regions being blended with other images (again a result ofhigh stellar density). This can be seen in Figure 5 which shows the density of stellar sourcesin each IPHAS field: there are clearly fewer high proper motion objects detected in areas ofhigher stellar density . This is also shown by the inverse correleation between the densityof stellar sources in a field and the typical number of detected proper motion sources in thatfield (see Figure 6). Figure 7 shows an IPHAS colour-colour plot for our objects. The mainlocus is a clear, unreddened main sequence (see Drew et al., 2005), widened by the fact that The general trend towards more crowded fields towards the Galactic centre can be seen in Figure 3 of Gonzalez-Solares etal. 2008c (cid:13) , 1–18 N.R. Deacon et al. the IPHAS photometry is not yet globally calibrated. Approximately 96% of objects in thecatalogue lie on or close to this main sequence. There is also a white dwarf locus presentlying below and to the left of the main sequence. Many objects lie above and to the left of themain sequence. While this may suggest H α emission, it may also be due to poor photometryin a particular field. Hence rather than select all these as potential H α emitters, in the nextsection we will use the study of Witham et al. (2008) to identify objects which appear tohave significant H α emission relative to the main sequence on the particular field . Finally wechecked our sample for common proper motion binaries. To investigate if we had a distinctpopulation of common proper motion binaries, we plotted a histogram of the separationsof all the objects with proper motions within 2 σ of each other. The trend for coincidenceobjects should be N ∝ r and any excess above this at small separations would indicate aseparate population of physically bound common proper motion objects. Figure 8 shows ourhistogram, clearly there is no distinct population of common proper motion binaries present. As stated earlier, the IPHAS survey is currently being exploited for many different scientificgoals. One study utilising IPHAS photometry is that of Witham et al. (2008). Here IPHASphotometry is used to identify objects which lie significantly above the main stellar locuson a colour-colour diagram similar to Figure 7. As there will be offsets in the photometryfrom field to field, Witham et al. (2008) identifies potential H α emitters relative to thecolour-colour diagram for the field the object lies in. Hence objects which appear to be H α emitters due to the poor photometry of an individual field are not included in Witham etal’s sample. This allows us to treat this dataset as a clean sample of potential H α emitters.Cross-referencing this with our own proper motion sample will remove highly reddened (anddistant) Be stars from the Witham sample and should leave only potential CataclysmicVariables candidates, dMe stars and non-DA white dwarfs (ie. nearby stellar sources showingeither H α emission or less than expected H α absorption). In this cross-referencing, we alsoincluded objects found in our study with proper motions between 0.2 and 0.15 arcsecondsper year and objects with r magnitudes between 19 and 20. These were not included in thefinal catalogue as these objects were found to suffer from a high level of contamination.The thirty six crossmatches are shown in Table 1. Note eight crossmatches were excludedfrom this list and from Figures 3 and 7 after inspection of the images by eye found that c (cid:13) , 1–18 PHAS-POSS-I Proper Motion Survey Table 1.
Objects common between our catalogue and the H α catalogue of Witham et al. (2008). IPHASJ225040+632838 is theknown proper motion CV system GD 552 (Greenstein & Giclas 1978), IPHASJ043839+410931 is (GD 61 Giclas, Burham & Thomas,1965), IPHASJ210951+425705 is EGGR 334 (Greenstein, 1974) and IPHASJ032825+580645 is the known high proper motion starLSPM J0328+5806 (Lepine & Shara, 2005). spectral type sources WHT spectroscopy, FAST spectroscopy, 3 Giclas, Burham &Thomas (1965), Greenstein (1974), Greenstein & Giclas 1978Name Position µ α µ δ σ µ α σ µ δ r i H α SpT”/yr ”/yr ”/yr ”/yrIPHASJ000528+663951 00 05 28.05 +66 39 51.5 0.031 -0.051 0.009 0.009 IPHASJ034042+573053 03 40 42.96 +57 30 53.7 0.069 -0.033 0.006 0.006 IPHASJ040147+540650 04 01 47.07 +54 06 50.8 0.044 -0.058 0.006 0.007 IPHASJ045400+470031 04 54 00.68 +47 00 31.0 0.004 -0.023 0.006 0.003 IPHASJ055551+324150 05 55 51.14 +32 41 50.3 -0.037 -0.001 0.006 0.006 IPHASJ055752+274641 05 57 52.90 +27 46 41.8 0.025 -0.044 0.005 0.006 IPHASJ061409+171136 06 14 09.36 +17 11 36.0 0.003 -0.044 0.006 0.006 IPHASJ183523+014245 18 35 23.26 +01 42 45.4 0.154 0.015 0.006 0.006 −
04 03 04.3 0.051 -0.027 0.005 0.005 IPHASJ190142-043621 19 01 42.09 −
04 36 21.1 0.001 -0.034 0.007 0.006 −
02 52 32.4 0.032 -0.010 0.005 0.005 IPHASJ215029+554250 21 50 29.23 +55 42 50.6 0.027 0.006 0.005 0.005 IPHASJ225040+632838 22 50 40.03 +63 28 38.2 0.102 -0.037 0.005 0.006 IPHASJ232003+571736 23 20 03.28 +57 17 36.6 -0.035 0.000 0.007 0.007 they may be blended objects. Examining Figure 7 we can see that many of the grey dots(representing Witham et al.’s H α emitters with significant proper motions) fall along themain sequence. It is possible that these are true H α emitters and appear in this part ofthe diagram due to uncorrected field to field photometric offsets or some selection effect.Of these objects one (IPHASJ053015+251137) appears to share a common proper motionwith the nearby (separation 42”) star TYC 1852-777-1 (Hog et al., 1998). The two propermotions agree within one sigma implying these are a true bound pair or part of the samemoving group. Three other objects redder than r − i = 0 . α emission. The question remains as to why these objects appeared in c (cid:13) , 1–18 N.R. Deacon et al.
Witham et al.’s catalogue. Witham et al. fitted a curve to the unreddened main sequencein each field and identified emitters as objects which lay significantly above this curve. Thetwo non-emitting M dwarfs lie in the brightest selection bin of Witham et al.’s selectionprocess ( r < α emission. One moderately red object (IPHASJ191733+031937) appears to lie on thesubdwarf sequence.Fifteen of the cross matches objects which appear to lie on the white dwarf sequence inthe reduced proper motion diagram (Figure 3). Of these IPHASJ225040+632838 is the lowstate CV system GD 552 (Greenstein & Giclas 1978). Another two, IPHASJ043839+410931(GD 61, Giclas, Burham & Thomas, 1965) and IPHASJ210951+425705 (EGGR 334, Green-stein, 1974) are known DB white dwarfs. Additionally three objects had spectra taken inthe IPHAS spectroscopic follow-up programme with the FAST spectrograph on the 1.5mTillinghast telescope on Mount Hopkins. Of the remaining nine objects, three had spectrataken using the ISIS spectrograph on the William Herschel Telescope (WHT) on La Palma.These spectra were used to provide rough spectral classifications which can be found in Ta-ble 1. Seven of the eight spectrally classified objects which lie bluewards of r − i =0.4 in thecolour-colour diagram (excluding the known CV GD 552) are non-DA white dwarfs. Hencewe believe the remaining objects are good non-DA white dwarf candidates.Valdivielso et al. (2008) have produced a sample of young, low mass objects using IPHASdata. Clearly identifying the proper motions of such objects could establish a connection witha known star forming association or moving group. Unfortunately none of these objectsappear in our catalogue. In order to provide a rough estimate of our completeness, we plotted a cumulative propermotion histogram. This is shown in Figure 9. Assuming uniform spatial and velocity dis-tributions and a fully complete survey, the distribution should scale as N ∝ µ − . This isrepresented by the solid line in the plot. It is clear that we begin to become incomplete be-low 60 milliarcseconds per year. This is due to a combination of our limiting magnitude andsome objects falling in fields with poor astrometry (hence having proper motions which are c (cid:13)000
Witham et al.’s catalogue. Witham et al. fitted a curve to the unreddened main sequencein each field and identified emitters as objects which lay significantly above this curve. Thetwo non-emitting M dwarfs lie in the brightest selection bin of Witham et al.’s selectionprocess ( r < α emission. One moderately red object (IPHASJ191733+031937) appears to lie on thesubdwarf sequence.Fifteen of the cross matches objects which appear to lie on the white dwarf sequence inthe reduced proper motion diagram (Figure 3). Of these IPHASJ225040+632838 is the lowstate CV system GD 552 (Greenstein & Giclas 1978). Another two, IPHASJ043839+410931(GD 61, Giclas, Burham & Thomas, 1965) and IPHASJ210951+425705 (EGGR 334, Green-stein, 1974) are known DB white dwarfs. Additionally three objects had spectra taken inthe IPHAS spectroscopic follow-up programme with the FAST spectrograph on the 1.5mTillinghast telescope on Mount Hopkins. Of the remaining nine objects, three had spectrataken using the ISIS spectrograph on the William Herschel Telescope (WHT) on La Palma.These spectra were used to provide rough spectral classifications which can be found in Ta-ble 1. Seven of the eight spectrally classified objects which lie bluewards of r − i =0.4 in thecolour-colour diagram (excluding the known CV GD 552) are non-DA white dwarfs. Hencewe believe the remaining objects are good non-DA white dwarf candidates.Valdivielso et al. (2008) have produced a sample of young, low mass objects using IPHASdata. Clearly identifying the proper motions of such objects could establish a connection witha known star forming association or moving group. Unfortunately none of these objectsappear in our catalogue. In order to provide a rough estimate of our completeness, we plotted a cumulative propermotion histogram. This is shown in Figure 9. Assuming uniform spatial and velocity dis-tributions and a fully complete survey, the distribution should scale as N ∝ µ − . This isrepresented by the solid line in the plot. It is clear that we begin to become incomplete be-low 60 milliarcseconds per year. This is due to a combination of our limiting magnitude andsome objects falling in fields with poor astrometry (hence having proper motions which are c (cid:13)000 , 1–18 PHAS-POSS-I Proper Motion Survey Figure 9.
A cumulative proper motion histogram for our the objects in our catalogue. The solid line represents the N ∝ µ − relation that would be expected with no incompleteness. It is clear that our survey begins to become incomplete below about60 milliarcseconds per year and that below 20 millarcseconds per year there are virtually no objects. The dotted line representsthe study of Gould & Kollmeier (2004). Clearly their study is complete to lower proper motions than ours. In the region whereour survey is most complete there is a factor of 2 difference between the numbers. not significant enough). Below about 25 mas/yr it is clear the distribution flattens off andwe can say we have no significant population below this mark. We have also compared ourresults with those in Gould & Kollmeier (2004). Figure 9 shows that in the proper motionrange where both surveys have similar proper motion completeness, we have half the numberof objects that Gould & Kollmeier have. This is despite the two surveys having similar areas(both around 1400 sq. deg.). However our survey covers a much more crowded area thantheirs. Deacon, Hambly & Cooke (2005) calculated the area lost to bright and blended starsacross the southern sky. Examining their Figure 10, it is clear that in the southern regions ofthe sky at similar Galactic latitude to ours, the completeness is often 50% or worse. Hencewe believe this difference in numbers is due the more crowded nature of our survey area.The IPHAS survey consists of 15270 pointings, which between them cover the 1800square-degree survey area twice or more. Hence simply taking the size of the detector andmultiplying it by the number of fields our survey covers (12362) will not yield an accurateestimate of our current survey area. A rough estimate can be provided by multiplying thefraction of the fields we cover (approximately 81%) by the total final survey area of 1800 c (cid:13) , 1–18 N.R. Deacon et al. sq. deg. This yields and approximate area for our proper motion survey of 1457 sq. deg.However as stated above, due to crowding we are only likely to identify proper motionobjects in roughly half this total area. Once data from the few unobserved IPHAS fieldshave been relaesed we will apply the same method to the remaining fields, completing ourproper motion survey.In calculating our astrometric solutions we use sets of reference stars. These may havesmall bulk motions. Additionally for the objects where we have too few reference stars theraw IPHAS positions are used. These are tied to the 2MASS (Skrutskie et al. 2006) systemusing reference stars. Hence we will measure proper motions relative to these reference starsrather than absolute proper motions. Lepine (2008) also encountered this problem. Theyconcluded that the difference between absolute and relative proper motions was typicallyless than their measurement errors. As our measurement errors are similar to theirs (typicallbelow their quoted global errors of 8mas/yr in each axis), we deduce that any offset betweenthe relative and absolute proper motions of our sample will also be below our calculatederrors.
We have completed the first comprehensive wide field proper motion survey of the northernGalactic plane ( | b | < ◦ ) covering proper motions between 150 and approximately 30 arcsec-onds per year. This sample covers a large section (1457 sq. deg.) of the northern plane andcontains 57249 objects with significant proper motions. We also identify seventeen objectsin common between our catalogue and the H α emission catalogue of Witham et al. (2008).These objects fell in to two distinct groups, a blue group dominated by non-DA white dwarfsand a red group dominated by maginally selected ordinary main sequence objects. This sam-ple will clearly be useful in the study of populations such as white dwarfs and subdwarfsin the Galactic plane. We will seek to complete the catalogue for the full survey area andwill use the upcoming UVEX data to extend it to higher proper motions above the currentimposed limit of 0.15 arcseconds per year. ACKNOWLEDGMENTS
This paper uses data from the SuperCOSMOS Sky Survey and from the INT Photometric H α Survey of the northern Galactic plane (IPHAS) carried out at the Isaac Newton Telescope c (cid:13) , 1–18 PHAS-POSS-I Proper Motion Survey
REFERENCES
Deacon, N.R., Hambly, N.C., Cooke, J.A., 2005, A&A, 435, 363Deacon, N.R., Hambly, N.C., 2007, A&A, 468, 163Drew, J.E., Greimel, R., Irwin, M.J., et al., 2005, MNRAS, 362, 753Fedorov, P.N., Myznikov, A.A., Akhmetov, V.S., 2009, MNRAS, 393, 133Finch, C.T., Henry, T.J., Subasavage, J.P., et al., 2007, AJ, 133, 2898Folkes, S.L., Pinfield, D.J., Kendall, T.R., Jones, H.R.A., 2007, MNRAS, 378, 901Giclas, H.L., Burnham, R., Thomas, N.G., 1965, LowOB, 6, 155Gonzalez-Solares, E.A., Walton, N.A., Greimel, R., et al., 2008, 2008, MNRAS, 388, 89Gould, A., Kollmeier, J.A, 2004, ApJS, 152, 103Greenstein, J.L., Giclas, H., 1978, PASP, 90, 460Greenstein, J.L., 1974, ApJ, 189, 131Groot, P.J., et al., in prep.Hambly, N.C., MacGillivray, H.T., Read, M.A., Tritton, S.B., Thomson, E.B., Kelly, B.D.,Morgan, D.H., Smith, R.E., Driver, S.P., Williamson, J., Parker, Q.A., Hawkins, M.R.S.,Williams, P.M., Lawrence, A.,MNRAS, 326, 4, 1279, 2001Hog, E., et al., 1998, A&A, 335, 65Lawrence, A., Warren, S.J., Almaini, O., et al., 2007, MNRAS.379, 1599Lepine, S., Shara, M.M., 2005, AJ, 129, 3, 1483Lepine, S., 2008, AJ, 135, 2177 c (cid:13) , 1–18 N.R. Deacon et al. [h]
Table A1.
An example of the data tables available electronically for this paper. indicates astrometric solutions calculated fromreference stars, implies the astrometic errors are drawn from global error estimates. The Modified Julian date (MJD) and theposition are both taken from the IPHAS observations.Name Position µ α µ δ σ µ α σ µ δ r i H α MJDJ2000 ”/yr ”/yr ”/yr ”/yrIPHASJ000001+575210 00 00 01.48 +57 52 10.9 -0.009 -0.034 0.008 0.004 Lucas, P.W., Hoare, M.G., Longmore, A., et al., 2008, MNRAS, 391, 136Luyten, W.J., 1918, Lick Observatory Bulletin, vol. 10, pp.135-140Luyten Half Arcsecond Catalogue, Luyten, W.J., University of Minnesota, Minneaplois,1979New Luyten Two Tenths Catalogue, Luyten, W.J., University of Minnesota, Minneapolis,1979Mampaso, A., Corradi, R.L.M., Viironen, K., et al., 2006, A&A, 458, 203Pokorny, R.S., Jones, H.R.A., Hambly, N.C., 2003, A&A, 397, 575Sale, S.E., Drew, J.E., Unruh, Y.C., 2009, MNRAS, 392, 497Skrutskie, M.F., Cutri, R.M., Stiening, R., et al., 2006, AJ, 131, 1163Valdivielso, L., Martin, E.L., Bouy, H., et al., 2009, A&A, 497, 973Vink, J.S., Drew, J.E., Steeghs, D., et al., 2008, MNRAS, 387, 308Wallace, P.T., Starlink User Note No. 67.42: SLALIB: Postitional Astronomy Library,CCLRC/Rutherford Appleton Laboratory, PPARC, 1998Wesson, R., Barlow, M.J., Corradi, R.L.M., 2008 ApJ, 688, 21Witham, A.R., Knigge, C., Drew, J.E., et al., 2008, MNRAS, 384, 1277
APPENDIX A: EXAMPLE DATA TABLE
The data tables for this paper will be available electronically. Here we give an example ofone of the data tables. c (cid:13)000