Prospects for cool white dwarf science from Pan-STARRS
Nigel Hambly, Nick Rowell, John Tonry, Eugene Magnier, Christopher Stubbs
aa r X i v : . [ a s t r o - ph . I M ] M a r **Volume Title**ASP Conference Series, Vol. **Volume Number****Author** c (cid:13) **Copyright Year** Astronomical Society of the Pacific Prospects for cool white dwarf science from Pan-STARRS
Nigel Hambly, Nick Rowell, John Tonry, Eugene Magnier and ChristopherStubbs Scottish Universities Physics Alliance (SUPA), Institute for Astronomy, Schoolof Physics and Astronomy, University of Edinburgh, Royal Observatory,Blackford Hill, Edinburgh EH9 3HJ, UK Space Technology Centre, School of Computing, University of Dundee,Dundee DD1 4HN, UK Institute for Astronomy, University of Hawaii at Manoa, Honolulu, HI 96822,USA Department of Physics, Harvard University, Cambridge, MA 02138, USA
Abstract.
We discuss the prospects for new deep, wide–angle surveys of the Galacticcool white dwarf populations using data from Pan-STARRS: the Panoramic SurveyTelescope & Rapid Response System.
1. Introduction
Cool white dwarfs (WDs) provide the means to measure the star formation history(see the article by Rowell elsewhere in these proceedings) and age of the populationto which they belong, as well as laboratories for the study of matter under extremes ofpressure. Large samples cannot be culled from photometry alone — cool WDs presentcolours that are indistinguishable from those of the profusion of field F / G / K dwarfs.However, using proper motion µ as a proxy for distance d (where generally µ ∝ / d ),we can employ the technique of reduced proper motion (RPM) to distinguish the WDsfrom the much more luminous dwarfs provided we have a multi–epoch survey over asu ffi ciently large time baseline to yield accurate proper motions.To date, the widest angle studies have had to employ legacy Schmidt photographicsurveys to provide multi–epoch imaging. The photographic plates are limited in theirseeing (typically 2 to 3 arcsec) and depth (typically R ∼ ∼ ) of cool WDsthereby enabling determinations of the age of the disk population. However the depthof the resulting luminosity functions is as yet insu ffi cient to determine the ages of thethick disk and spheroid of the Galaxy, and the total space density of the coolest, oldestthick disk and spheroid stars remains unknown (e.g. Rowell & Hambly 2011, hereafterRH11). 1 Hambly, Rowell,Tonry,Magnier &Stubbs
2. Pan-STARRS
The Panoramic Survey Telescope and Rapid Response System prototype ‘PS1’ — seeMorganson et al. (2012) and references therein — is now well into it’s 3 year wide–angle observing campaign (the so–called ‘3 π ’ survey), aiming to survey 75% of thesky in the 5 grizy P1 passbands at least 4 times per year yielding a multi–epoch sur-vey of stunning proportions. With a typical seeing of ∼ π survey is essentially completely covered in all passbands with the ex-pected number of individual epochs for this stage in the project; overall observationalcompleteness stands at 59%. A fully world–public release of PS1 data is presentlyscheduled for the end of 2014; in the meantime, data access is restricted to scientistswithin those institutes that are part of the PS1 Science Consortium.
3. Depth and completeness
Figure 1 shows number–magnitude histograms for a 200 square degree equatorial strip(10 ◦ < α < ◦ and − . ◦ < δ < + . ◦ ) across the South Galactic Cap from the extantPan-STARRS 3 π survey. To be counted here, detections must be present in at leastgri P1 with at least 10 individual epochs. We compare also with the Rowell & Hamblysurvey sample by identifying the same objects in those catalogues employing the closestcorresponding passbands (BRI). The increase in depth a ff orded by Pan-STARRS isclear, as is the general incompleteness in the photographic data (up to ∼ Figure 1. Number-magnitude histograms of sources in the extant 3 π survey (solidlines labelled by passband) compared with similar from the SuperCOSMOS legacySchmidt photographic survey (dotted lines labelled by passband). http: // pan-starrs.ifa.hawaii.edu / public / ool WDsand Pan-STARRS 3
4. Reduced Proper Motion diagrams
Proper motions from the full 3yr Pan-STARRS 3 π survey are of course not yet available,but we can get a good idea of the potential gains by combining the new photometry withthe legacy photographic plate proper motions. Figure 2 shows 3 RPM diagrams: on theleft is that from the photographic plate survey data alone for the SGC subsample ofthe RH11 survey (R < . = ff . The tighter subdwarf sequence and cleaner separation of the WDs is clear whenusing Pan-STARRS, as is the enhanced depth (despite the necessity of employing stillthe photographic plate astrometry). Figure 2. RPM diagrams for the Rowell & Hambly sample using the photo-graphic photometry (left panel); the same sample but using PS1 photometry (middlepanel); and a sample employing the photographic astrometry and PS1 photometrywithout the R < .
75 cut employed by RH11 (right panel).
5. Discussion
Tonry et al. (2012) present new results from the Pan-STARRS Medium Deep Survey;however we suggest that it is the 3 π survey that will sample the greatest volume for thecoolest WDs, and will discover many objects amenable to spectroscopic follow-up on4 to 8m–class facilities. The prospective numbers of cool WDs that will be detected inthe 3 π survey can be calculated using simple scaling arguments from the RH11 sample. Hambly, Rowell,Tonry,Magnier &StubbsVolume sampled for a uniformly distributed population (e.g. the local spheroid popu-lation) will go as d for distance d limited by magnitude ( d P1 / d RH11 = (r P1 − r RH11 ) / )where r P1 is the magnitude limit for a sample drawn from the PS1 3 π survey with propermotion characteristics similar to RH11. They found that RPM discriminates usefullywith proper motions as low as 5 σ µ and employed a magnitude–dependent lower propermotion limit typically between 60 < σ µ <
100 mas / yr at the limit of the SuperCOS-MOS survey data (their Figure 2). To estimate PS1 proper motion precision, we notethe Irwin (1985) rule–of–thumb which states that in units of the scale size of well sam-pled faint point sources in uniform background noise (0.6 arcsec for PS1 correspondingto half–width at half–maximum), centroiding precision σ x is equal to relative flux pre-cision σ f / f . Hence 5 σ f detections will have centroids accurate to σ x ≈ .
12 arcsec,and given N =
60 measurements over ∆ t = . σ µ = √ × ( σ x / ∆ t )(12 / N ) . ≈
22 mas yr − where the factor 12 comesfrom the variance of a uniform distribution. For a lower proper motion limit compara-ble to RH11, say 5 σ µ =
80 mas / yr, we require σ µ =
16 mas yr − which corresponds to σ x ∼
87 mas. This should be achieved a factor 1.4 (or 0.35 mag) brighter than the 5 σ f detection limit, the latter corresponding to 50% completeness at r P1 ∼ . r P1 ∼ .
25. Noting the 50% incompleteness in RH11and that r
RH11 = .
75, the number of PS1 spheroid WDs will be ∼ × ≈
16 timeshigher. For the thin disk component, scale height e ff ects reduce the distance exponentto 2, so the factor increase is 2 × ≈
8; the thick disk increase will be somewherebetween the two. Not only will the sample be significantly larger, but the availabilityof 5–colour photometry will greatly improve the RH11 analysis which was limited to50% errors in photometric distances from 3–colour photographic photometry.
Acknowledgments.