1: \documentstyle[11pt,newpasp,twoside,epsf]{article}
2: \markboth{Hansen, B.}{Old White Dwarfs}
3: \pagestyle{myheadings}
4: \nofiles
5:
6: % Some definitions I use in these instructions.
7:
8: \def\emphasize#1{{\sl#1\/}}
9: \def\arg#1{{\it#1\/}}
10: \let\prog=\arg
11:
12: \def\edcomment#1{\iffalse\marginpar{\raggedright\sl#1\/}\else\relax\fi}
13: \marginparwidth 1.25in
14: \marginparsep .125in
15: \marginparpush .25in
16: \reversemarginpar
17:
18: \begin{document}
19: \title{Old White Dwarfs as a Microlensing Population}
20: \author{Brad M. S. Hansen}
21: \affil{Department of Astrophysical Sciences, Peyton Hall,
22: Princeton University,
23: Princeton, NJ, 08544, USA}
24: %\author{Ima Co-Author}
25: %\affil{The Name of My Institution, The Full Address of My Institution}
26:
27: \begin{abstract}
28: A popular interpretation of recent microlensing studies of the line
29: of sight towards the Large Magellanic Cloud invokes a population of
30: old white dwarf stars in the Galactic halo. Below I review the basic
31: properties of old white dwarf stars and the ongoing efforts to detect
32: this population directly.
33: \end{abstract}
34:
35: \section{Introduction}
36:
37: Other authors in this volume cover the microlensing motivations much
38: better than I can, so I shall suffice to remind you that one possible
39: explanation of the microlensing events towards the LMC invokes a population
40: of objects in the range $0.3-0.8$M$_{\odot}$. Potentially these could be
41: either normal hydrogen-burning stars or white dwarfs, the burnt-out remnants
42: of stellar evolution. To distinguish these populations, we turn to direct
43: searches at optical wavelengths, since the latter population is
44: $10^{-3}-10^{-4}$ as bright as the former. The number counts of faint
45: red stars suggest that hydrogen burning stars cannot account for the
46: microlensing population (Bahcall et al 1994; Graff \& Freese 1996).
47: The question I wish to
48: address is how well one can constrain the white dwarf hypothesis by
49: similar means.
50:
51:
52:
53: \section{White Dwarf Cooling}
54:
55: To derive a constraint from direct optical searches, we need to know how to
56: recognise white dwarfs. White dwarfs of moderate age ($<10$~Gyr) emit
57: approximately as black bodies. However, once the white dwarf cools to
58: effective temperatures below $\sim 5000$~K, the atmospheric hydrogen
59: resides in molecular form. This has dramatic consequences for the appearance
60: of the white dwarf (Hansen 1998, 1999a; Saumon \& Jacobsen 1999), because the
61: dominant opacity source in such an atmosphere is collisionally induced
62: absorption by H$_2$, which absorbs primarily in the near infra-red.
63: For black bodies of effective temperatures $\sim 3000-5000$~K the peak
64: of the spectrum lies in the same wavelength region, so that the increased
65: absorption leads to dramatic deviations from the traditional assumption
66: of black body colours. The general trend of the colour evolution is that
67: the flux is forced to emerge preferentially blueward of $\sim 0.8 \mu$m.
68: With the evolution of the black body flux redward, the peak of the spectrum
69: for old white dwarfs lies around $0.6 \mu$m.
70:
71: A correct quantitative analysis of the atmospheric conditions is not only
72: important for determining the colours of the white dwarf, but is also of
73: critical importance for the cooling evolution. For white dwarfs cooler than
74: $T_{\rm eff} \sim 6000$~K, the internal structure has no radiative zones. The
75: bulk of the white dwarf mass resides in a degenerate core which is approximately
76: isothermal due to the efficient thermal conduction by the electrons near the
77: surface of the Fermi sea. This isothermal core is joined to a thin convective
78: envelope which extends all the way to the photosphere. Thus, the entire white
79: dwarf structure is critically dependant on the surface boundary condition and
80: hence the atmospheric conditions. It must be noted that, while sophisticated
81: analyses of white dwarf atmospheres have been around for several years
82: (e.g. Bergeron, Saumon \& Wesemael 1995), the white dwarf cooling calculations
83: have lagged significantly behind, employing grey atmosphere boundary conditions.
84: Thus, while one may trust the masses and gravities inferred in recent atmospheric
85: analyses, one cannot completely trust the ages inferred for the older white dwarfs
86: in most recent papers. To my knowledge, the only cooling calculations using
87: boundary conditions based on proper atmosphere models are those of Hansen (1999a)
88: and Salaris et al (2000). For isochrones and luminosity functions
89: using the Hansen (1999a) models, see Richer et al (2000).
90:
91: The models described above concern white dwarfs with pure hydrogen atmospheres.
92: Empirically, a significant fraction of known white dwarfs appear to have no
93: hydrogen in their atmospheres (Liebert et al 1979; Bergeron, Ruiz \& Leggett 1997).
94: When pure helium atmospheres reach effective temperatures $<5000$~K, they have
95: no molecular component to provide significant opacity and, as a result, the
96: photosphere lies at much higher densities, where the beginnings of pressure
97: ionization start to provide a small ionized component. This higher density and
98: the concomitant difference in the boundary condition results in much faster
99: cooling for old white dwarfs with helium atmospheres. Furthermore, these
100: objects are expected to appear approximately as black bodies (contrary to the
101: hydrogen atmosphere case) but will be hard to detect because they cool much
102: faster.
103:
104: The matter of a component of the white dwarf population sporting helium
105: atmospheres is one that is often not mentioned in the recent studies attempting
106: to tie direct searches to microlensing populations. It is important to note
107: that this represents an essentially unobservable (due to their much more rapid
108: cooling) component and, as such, will always introduce some uncertainty into
109: the comparison.
110:
111: \section{Observational Progress}
112:
113: \subsection{The story thus far}
114:
115: \begin{figure}
116: \plotone{HDF4.ps}
117: \caption{The source of all the fuss. The five points are the magnitudes \& colours
118: of the 5 proper motion objects of Ibata et al (1999). The solid curves describe the
119: cooling of a 0.6 solar mass white dwarf with hydrogen atmosphere located at 1 and 2
120: kpc. Representative ages are indicated along the lower curve. The dashed curves
121: are the same but for a pure helium atmosphere. The much more rapid cooling is
122: evident.
123: }
124: \end{figure}
125:
126: The last couple of years have been fun for those interested in the question
127: of halo white dwarfs. The realisation that old white dwarfs would appear
128: somewhat bluer than black bodies meant that previous constraints on the local
129: number density based on Hubble Deep Field point source counts were somewhat overstated.
130: Much excitement was generated by the detection of proper motions of several
131: faint blue point sources in the HDF North (Ibata et al 1999). The colours and
132: magnitudes of these objects are consistent with old white dwarfs $\sim 12$~Gyr
133: old at distances $\sim 1$~kpc. Given the radical consequences of this observation,
134: some kind of confirmation would be nice. A third epoch, to confirm the measurement,
135: has been approved but was postponed due to the HST gyro problems in December 1999.
136: Nevertheless, the veracity of the HDF proper motions should be addressed within
137: the next year or so.
138:
139:
140: In the interim, several other interesting results have come to light. Most
141: importantly, Hodgkin et al (2000) presented a near-infrared spectrum of a nearby
142: cool white dwarf, demonstrating the reality of the flux-suppression due to
143: molecular hydrogen absorption. This provides a welcome check that the
144: theory behind the predicted colour change is at least qualitatively correct.
145: The high proper motion of this object also suggests membership of the Galactic halo.
146: Another object showing such a long wavelength depression is LHS~3250 (Harris et al
147: 1999)
148:
149: A further interesting indirect argument supporting the white dwarf hypothesis
150: comes from Mendez \& Minniti (2000), who assert that the number counts of faint blue
151: point sources (the class amongst which Ibata et al discovered proper motions)
152: in the Hubble Deep Field South is approximately twice that in the HDF North. This
153: is consistent with the objects (whatever they are) being Galactic in origin, since
154: the southern field looks in a direction closer to the Galactic centre and consequently
155: the stellar density is expected to be higher.
156:
157: A note of caution, however, is sounded by Flynn et al (2000), who find little evidence
158: for a population equivalent to that of Ibata et al in the Luyten proper motion
159: survey. They conclude that hydrogen atmosphere white dwarfs cannot comprise a
160: significant fraction of the Galactic halo. On the other hand, a new search of
161: wide field photographic plates by Ibata et al (2000) has uncovered several
162: high proper motion objects, at least two of which are spectroscopically confirmed
163: to be white dwarfs and with some modicum of IR flux suppression.
164:
165: In my not-completely-unbiased interpretation of the above, the evidence in
166: favour of the white dwarf hypothesis is encouraging, although not conclusive.
167: The recent results of Hodgkin et al and Ibata et al (2000) have
168: demonstrated at least the existence of a population of high proper motion
169: (and thus probably halo) white dwarfs. The question is now a matter of number
170: density. It is on this point that we must await further data.
171:
172: \subsection{Future prospects}
173:
174: While the above results indicate that some white dwarfs do show
175: halo kinematics, the fact
176: that there is little evidence for a large halo white
177: dwarf population in existing large-scale proper motion
178: surveys casts some doubt that this population may be responsible
179: for the microlensing results. This last statement relies heavily
180: on the data set of Luyten (1979), so it would be nice to check this
181: with another large scale survey.
182:
183: Towards this end, a program led by Peter Stetson has begun at CFHT
184: to image almost 20 square degrees to $V \sim 25$. Repeated with
185: an interval of 2-3 years, this program should provide a conclusive
186: test of the white dwarf halo hypothesis. Using the new MACHO mass
187: fraction estimate ($\sim 20\%$) and assuming only half of all white
188: dwarfs have hydrogen atmospheres, this program is still expected to
189: find about 8 white dwarfs per square degree.
190:
191: \begin{figure}
192: \plottwo{gri.ps}{ri.ps}
193: \caption{The left hand panel shows the SDSS stellar locus
194: (York et al 2000) in the
195: g-r/r-i plane, with the evolution of a representative white dwarf
196: shown by the dashed line. Thus, it is possible to select very cool
197: white dwarfs by colour alone. However, objects that cool are also
198: faint. The right hand panel shows the apparent magnitudes of the
199: stars with g-r$>$0.8 as well as the dashed line for the cooling of a 0.6~$M_{\odot}$
200: white dwarf at 10~pc (i.e. the absolute magnitude). The dotted line shows
201: the apparent magnitude of a 3500~K white dwarf at various distances.}
202: \end{figure}
203:
204: Another fascinating prospect is the hope of finding such old white
205: dwarfs in the Sloan Digital Sky Survey (SDSS). Although this comprehensive
206: large scale survey will not provide proper motions as a matter of course,
207: some proper motion information is potentially available from comparison
208: with Palomar Sky Survey plates. Furthermore, there will be a strip of sky
209: in the southern hemisphere that will be multiply imaged. Even without
210: proper motions, very old white dwarfs are potentially detectable by
211: colours alone in the SDSS five-band photometric system. Photometric
212: selection alone is unlikely to provide a complete census as only the
213: oldest white dwarfs will stand out from the locus of other stars
214: (the problem is that even the reddest bandpass doesn't extend significantly
215: beyond $1\mu$m). Nevertheless, very old white dwarfs are interesting
216: in their own right.
217:
218: Figure~2 shows a representative stellar locus and how one
219: might select old white dwarfs based on colours. The colour selection
220: suggests that only white dwarfs cooler than 3500~K can be selected in
221: this manner. Using an i-band detection limit of 18, we find such objects
222: can be detected to distances $\sim 30$~pc. If the entire MACHO fraction
223: (taken to be 20\%) were in 0.5$M_{\odot}$ white dwarfs, 50\% of whom
224: had hydrogen atmospheres, and all had 3500~K temperatures, we would expect
225: approximately 43 white dwarfs in the Sloan survey, or 1 every
226: 480 square degrees. This is probably an overestimate. Hotter white dwarfs
227: cannot be distinguished from the stellar locus (it is worth noting that
228: the Hodgkin et al and Ibata et al detections all correspond to
229: $T_{\rm eff} \sim 3500-4000$~K, so in the marginal region)
230: and cooler objects are fainter and thus the
231: effective volume is smaller. Nevertheless, this simple estimate offers
232: encouragement that at least some white dwarfs are potentially detectable
233: in this manner. More detailed calculations are underway to provide a more
234: robust estimate.
235:
236: Another large scale proper motion survey that should offer interesting
237: constraints is that being undertaken by the EROS project (Goldman 1998).
238: With a survey area of 350 square degrees and an I-limit of 20.5, we can
239: use the same approximate scenario as above to estimate $\sim 23$ white
240: dwarfs in their sample.
241: This is a somewhat more realistic ballpark number in this case as the
242: proper motion selection will allow detection of hotter objects.
243:
244: The bottom line of the above is that, although uncertainty about mass distributions
245: and chemical compositions make prognostication difficult, the fact that naive
246: estimates lead to predictions of 10-100 detections in ongoing surveys suggests
247: that we should see something if the hypothesis of a significant white dwarf
248: halo is correct.
249:
250: \section{Beige Dwarfs}
251:
252: Although it was not my intention to address this subject at the conference,
253: it was raised a couple of times, so I will conclude with a few remarks about
254: ``Beige'' dwarfs (Hansen 1999b). Given that brown dwarfs are strongly ruled out
255: by the microlensing timescales, red dwarfs by direct observations and white
256: dwarfs (as well as neutron stars)
257: are an uncomfortable fit due to issues of chemical pollution, it appears
258: that there are no surviving baryonic candidates for halo MACHOs. However, there
259: is one remaining possibility. Lenzuni, Chernoff \& Salpeter (1992) demonstrated
260: that one could circumvent the traditional hydrogen burning limit by starting
261: with a brown dwarf and accreting material slowly enough that it could cool.
262: In this fashion one could construct a degenerate hydrogen/helium object with
263: mass $>0.1 M_{\odot}$. If one could build such objects to masses $>0.3 M_{\odot}$,
264: they would make ideal MACHO candidates as they would be faint and would have
265: no nuclear burning history and thus no chemical pollution problem. The term
266: ``Beige'' dwarf comes from the superposition of brown dwarf and white dwarf
267: characteristics.
268: The primary
269: problem for this scenario is that there is little evidence that this mode of
270: ``star'' formation was ever important.
271:
272: The existence of Beige dwarfs is observationally testable. They have similar
273: radii to white dwarfs but can only exist at effective temperatures
274: $T_{\rm eff} < 2000$~K, so the detection of hotter objects would suggest white
275: dwarfs. Indeed, the spectroscopically confirmed objects of Hambly et al
276: and Ibata et al (2000) indicate temperatures $\sim 3500-4000$~K. Therefore
277: these are unlikely to be Beige dwarfs.
278:
279: \acknowledgements
280: I would like to thank the conference organisers for a thoroughly enjoyable
281: (and efficiently run) meeting.
282: Support for this work was provided by NASA through Hubble Fellowship grant
283: \#HF-01120.01-99A,
284: from the Space Telescope Science Institute, which is operated by the
285: Association of Universities
286: for Research in Astronomy, Inc., under NASA contract NAS5-26555.
287:
288: %\clearpage
289:
290:
291: \begin{references}
292: \reference{Bahcall, J. N., Flynn, C., Gould, A. \& Kirhakos, S. 1994, ApJ, 435, L51}
293: \reference{Bergeron, P., Ruiz, M.-T. \& Leggett, S. K. 1997, ApJS, 108, 339}
294: \reference{Bergeron, P., Saumon, D. \& Wesemael, F. 1995, ApJ, 443, 764}
295: \reference{Flynn, C., Sommer-Larsen, J., Fuchs, B., Graff, D. S. \& Salim, S. 2000,
296: astro-ph/9912264}
297: \reference{Goldman, B., in The Third Stromlo Symposium: The Galactic Halo, eds.
298: Gibson, B. K., Axelrod, T. S. \& Putman, M. E., ASP Conference Series Vol 165,
299: p 413; San Francisco}
300: \reference{ Graff, D.S. \& Freese, K. 1996, ApJ, 435, L516}
301: \reference{ Hansen, B. M. S. 1998, Nature, 394, 860}
302: \reference{ Hansen, B. M. S., 1999a, ApJ, 520, 680}
303: \reference{ Hansen, B. M. S., 1999b, ApJ, 517, L39}
304: \reference{ Harris, H. C., Dahn, C. C., Vrba, F. J., Henden, A. A., Liebert, J.,
305: Schmidt, G. D. \& Reid, I. N. 1999, ApJ, 524, 1000}
306: \reference{ Hodgkin, S. T., Oppenheimer, B. R., Hambly, N. C., Jameson, R. F.,
307: Smartt, S. J. \& Steele, I. A. 2000, Nature, 403, 57}
308: \reference{ Ibata, R. A., Richer, H. B., Gilliland, R. L. \& Scott, D. 1999, ApJ, 524, L95}
309: \reference{ Ibata, R. A., Irwin, M., Bienayme, O., Scholz, R. \& Guibert, J. 2000,
310: ApJ, 532, L41}
311: \reference{Lenzuni, P., Chernoff, D. F. \& Salpeter, E. E., 1992, ApJ, 393, 232}
312: \reference{ Liebert, J., Dahn, C. C., Gresham, M. \& Strittmatter, P. A. 1979,
313: ApJ, 233, 226}
314: \reference{Luyten, W. J. 1979, {\em LHS Catalogue: A catalogue of stars with proper motions
315: exceeding 0.5" annually}, University of Minnesota, Minneapolis}
316: \reference{Mendez, R. A. \& Minniti, D. 2000, ApJ, 529, 911}
317: \reference{ Richer, H. B., Hansen, B., Limongi, M., Chieffi, A., Straniero, O.
318: \& Fahlman, G. G. 2000, ApJ, 529, 318}
319: \reference{Salaris, M., Garcia-Berro, E., Hernanz, M., Isern, J. \& Saumon, D. 2000, submitted to ApJ}
320: \reference{Saumon, D. \& Jacobsen, S. B. 1999, ApJ 511, L107}
321: \reference{York, D. et al 2000, AJ submitted}
322: \end{references}
323:
324: \end{document}
325: