1: \documentclass{emulateapj}
2: \submitted{{\it accepted for publication in AJ}}
3: %\usepackage{deluxetable}
4:
5: %\documentstyle[12pt,aasms4,psfig]{article}
6: %\documentstyle[emulateapj,psfig,apjfonts]{article}
7: %\documentstyle[11pt,preprint,psfig]{aastex}
8: \usepackage{graphicx}
9: \shortauthors{West et al.}
10: \shorttitle{DR5 Low-Mass Stars}
11: \begin{document}
12:
13: \title{Constraining the Age-Activity Relation for Cool Stars: The SDSS
14: DR5 Low-Mass Star Spectroscopic Sample}
15:
16: \author{Andrew A. West\altaffilmark{1,2},
17: Suzanne L. Hawley\altaffilmark{3},
18: John J. Bochanski\altaffilmark{3},
19: Kevin R. Covey\altaffilmark{4,5},
20: I. Neill Reid\altaffilmark{6},
21: Saurav Dhital\altaffilmark{7},
22: Eric J. Hilton\altaffilmark{3},
23: Michael Masuda\altaffilmark{3}}
24:
25:
26: \altaffiltext{1}{Corresponding author: awest@astro.berkeley.edu}
27: \altaffiltext{2}{Astronomy Department, University of California, 601 Campbell Hall, Berkeley, CA 94720-3411}
28: \altaffiltext{3}{Department of Astronomy, University of Washington, Box 351580,
29: Seattle, WA 98195}
30: \altaffiltext{4}{Harvard-Smithsonian Center for Astrophysics, 60
31: Garden Street, Cambridge MA 02138}
32: \altaffiltext{5}{Spitzer Fellow}
33: \altaffiltext{6}{Space Telescope Science Institute, 3700 San Martin Drive,Baltimore, MD 21218}
34: \altaffiltext{7}{Department of Physics \& Astronomy, Vanderbilt
35: University, Nashville, TN 37235}
36:
37: \begin{abstract}
38: We present a spectroscopic analysis of over 38,000 low-mass stars from
39: the Sloan Digital Sky Survey (SDSS) Data Release 5 (DR5). Analysis of
40: this unprecedentedly large sample confirms the previously detected
41: decrease in the fraction of magnetically active stars (as traced by
42: H$\alpha$ emission) as a function of vertical distance from the
43: Galactic Plane. The magnitude and slope of this effect varies as a
44: function of spectral type. Using simple 1-D dynamical models, we
45: demonstrate that the drop in activity fraction can be explained by
46: thin disk dynamical heating and a rapid decrease in magnetic
47: activity. The timescale for this rapid activity decrease changes
48: according to the spectral type. By comparing our data to the
49: simulations, we calibrate the age-activity relation at each M dwarf
50: spectral type. We also present evidence for a possible decrease in
51: the metallicity as a function of height above the Galactic
52: Plane. In addition to our activity analysis, we provide line
53: measurements, molecular band indices, colors, radial velocities, 3-D
54: space motions and mean properties as a function of spectral type for
55: the SDSS DR5 low-mass star sample.
56: \end{abstract}
57:
58: \keywords{solar neighborhood --- stars: low-mass, brown dwarfs ---
59: stars: activity --- stars: late-type --- Galaxy: structure --- Galaxy:
60: kinematics and dynamics}
61:
62: \section{Introduction}
63:
64: As the most numerous stellar constituents of the Milky Way, M dwarfs
65: are an ideal population for tracing the structure and evolution of the
66: stellar thin disk. A number of studies have utilized the ubiquity of
67: M dwarfs to study the dynamics and distribution of stars in the Solar
68: neighborhood (Wielen 1977; Weis \& Upgren 1995; Reid et al. 1995;
69: Hawley et al. 1996; Bochanski et al. 2007b). Many M dwarfs also
70: host intense magnetic dynamos that give rise to chromospheric and
71: coronal heating, producing emission from the x-ray to the radio. Over
72: 30 years ago Wilson and Woolley (1970) found a link between
73: magnetic activity (as traced by the CaII emission strength) and the
74: orbits of more than 300 nearby late-type dwarfs. Their study
75: concluded that magnetic activity in these stars was directly related to
76: their age. Subsequent studies over the following decades have found a
77: similar connection between age and activity in low-mass stars (Wielen
78: 1977; Giampapa \& Liebert 1986; Soderblom, Duncan \& Johnson 1991;
79: Hawley et al. 1996; Hawley, Tourtellot \& Reid 1999; Hawley, Reid \&
80: Tourtellot 2000).
81:
82: The mechanism that controls magnetic activity in M dwarfs is still
83: unknown. In the Sun, magnetic field generation has a strong
84: connection to the Sun's rotation. Helioseismology has indicated that
85: the boundary between the convective and the radiative zones (known as
86: the tachocline) creates a rotational sheer; the convective zone
87: undergoes differential rotation, while the radiative zone rotates like
88: a solid body (Parker 1993; Ossendrijver 2003; Thompson et al. 2003).
89: The tachocline allows magnetic fields to be generated, stored and
90: ultimately rise to the surface where they emerge as magnetic loops.
91: These loops drive the heating of the stellar chromosphere and corona,
92: resulting in dramatic stellar flares and lower level quiescent magnetic activity. Rotation in solar type stars slows with
93: time and as a result, activity decreases. Skumanich (1972) found that
94: both activity (as measured by Ca II emission) and rotation decrease
95: over time as a power law t$^{-0.5}$. Subsequent studies confirmed the
96: Skumanich age/rotation-activity relation but found that the slope of
97: the power law may change as a function of spectral type and that
98: non-power law functions are not ruled out (Barry 1988; Soderblom et
99: al. 1991). However, after a spectral type of $\sim$M3 (0.35
100: M$_{\odot}$; Reid \& Hawley 2005; Chabrier \& Baraffe 1997), stars become fully
101: convective and the tachocline presumably disappears. This transition marks an important change in the
102: stellar interior that must affect the production and storage of
103: internal magnetic fields.
104:
105: %There is strong evidence that the rotation-activity relation extends
106: %from stars more massive than the Sun to cooler M dwarfs (Delfosse et
107: %al. 1998; Pizzolato et al. 2003; Reimers 2007; Kiraga \& Stepien
108: %2007). However, at a spectral type of $\sim$M3, stars become fully
109: %convective and the tachocline presumably disappears. The lack of a
110: %large sample of high quality data for late-type stars complicates the
111: %situation. However, a few studies have uncovered evidence of a
112: %rotation-activity relation extending past the M3 convective transition
113: %and into the brown dwarf regime (Delfosse et al. 1998; Mohanty \&
114: %Basri 2002). Recent computer simulations of magnetic dynamos in fully
115: %convective stars also find evidence for magnetic field generation in
116: %the absence of a convective/radiative boundary (Browning 2007). In
117: %contrast, Bercik et al. (2005) found that a non-rotating convective
118: %medium may be able to create small scale activity in fully convective
119: %stars. In addition, a new study using high resolution Keck
120: %spectroscopy (West \& Basri in prep) has found evidence for rapid
121: %rotation in a sample inactive M6 dwarfs. Although rotation may play a
122: %role in activity generation, in this paper we will only address on how
123: %activity varies with age.
124:
125: Many studies have found evidence that the age-activity relation
126: extends into the M dwarf regime. Eggen (1990) observed a Skumanich
127: type power law decay in activity strength as a function of age.
128: Larger samples of M dwarfs have added further evidence to the
129: hypothesis that the magnetic activity in M dwarfs slowly decreases
130: over time (Fleming, Schmitt \& Giampapa 1995; Gizis, Reid \& Hawley
131: 2002). There are also data that support the hypothesis that activity
132: in M dwarfs may have a finite lifetime. Stauffer et al. (1994)
133: suggested that activity may not be present in the more massive M
134: dwarfs in the Pleiades. Hawley et al. (2000) confirmed the Stauffer
135: et al. (1994) claim by observing a sample of clusters that spanned
136: several Gyrs in age. They were able to calibrate the activity
137: lifetimes for early type M dwarfs by observing the color at which
138: activity (as traced by spectroscopically observed H$\alpha$ emission)
139: was no longer present. Because of the small sample size, the derived
140: age-activity relation serves as a lower limit at a given color or
141: spectral type. Silvestri et al. (2006) also argued for finite
142: activity lifetimes by observing M dwarfs in binary systems with white
143: dwarfs, finding no activity in the companions of the coolest, and
144: therefore oldest white dwarfs. However, accurate activity lifetimes
145: could not be calculated for the M dwarfs due to uncertainties in the
146: pre-white dwarf ages, binary effects and small sample sizes. A larger
147: spectroscopic sample is required to further constrain the activity-age
148: relation in both nearby clusters and the field.
149:
150: The advent of large surveys such as the Sloan Digital Sky Survey
151: (SDSS; York et al. 2000) and the Two Micron All Sky Survey (2MASS;
152: Cutri et al. 2003) has created optical and infrared catalogs of
153: several million M dwarfs. In addition to the photometric data, the
154: SDSS has obtained spectra for over 50,000 M dwarfs and the numbers
155: continue to increase as SDSS-II (Adelman-McCarthy et al. 2007b) pushes
156: to lower Galactic latitudes. Recently these surveys have been
157: utilized to examine the statistical properties of M dwarfs in a
158: variety of capacities, including analyses of mean M dwarf colors and
159: spectroscopic properties (Hawley et al. 2002; West et al. 2004;
160: hereafter W04; West, Walkowicz \& Hawley 2005; Bochanski et al. 2007a;
161: Covey et al. 2007), computing the dynamics of the Milky Way (West et
162: al. 2006; hereafter W06; Bochanski et al. 2007b) and the detailed
163: investigation of chromospheric magnetic activity (W04; W06; Silvestri
164: et al. 2006; Bochanski et al. 2007a). The activity studies of W04 and
165: W06 used SDSS spectra to show that the fraction of active M7 stars
166: decreases with vertical distance from the Galactic Plane, further
167: evidence of an age-activity relation. By comparing the data to simple
168: dynamical simulations, W06 argued that the observed trend results from
169: the combination of a rapid decrease in magnetic activity after 6-7
170: Gyrs and the correlation between age and height above the Galactic
171: Plane introduced by dynamical heating of the M7 stars. A larger sample
172: of low-mass star spectra allows for the obvious extension of the W06
173: study to all M dwarf spectral types. From this new statistical
174: platform it is now possible to derive an age-activity relation for the
175: entire M dwarf sequence and to investigate how this relation changes
176: across the M3-M4 convective boundary.
177:
178: In this paper we present a catalog of more than 44,000 spectroscopically
179: confirmed M dwarfs from the SDSS. We describe the data, which are
180: publically available, in \S2. We present the mean sample characteristics and
181: an analysis of the magnetic activity in \S3. In \S4, we compare our data to
182: a series of dynamical simulations and derive an age-activity relation
183: for low-mass dwarfs. We summarize these results in \S5.
184:
185:
186: \section{Data}
187:
188: The SDSS (Gunn et al. 1998; Fukugita et al. 1996; York et al. 2000;
189: Hogg et al. 2001; Gunn et al. 2006; Ivezi{\'c} et al. 2004; Pier et
190: al. 2003; Smith et al. 2002; Stoughton et al. 2002; Tucker et al. 2006) provides excellent
191: quality spectroscopic and photometric data for examining the
192: properties of low-mass dwarfs. Our sample was drawn from the SDSS Data
193: Release 5 (DR5; Adelman-McCarthy 2007a) and includes spectra from
194: several of the ``special plates'' that are not part of the standard
195: SDSS survey\footnote{see
196: http://www.sdss.org/dr5/products/spectra/special.html}. A total of
197: 49933 stars were selected from the SDSS spectroscopic database based
198: on the colors typical of M dwarfs ($r-i >$ 0.53 and $i-z >$ 0.3; West
199: et al. 2005) and a standard set of SDSS processing flags (SATURATED,
200: BRIGHT, NODEBLEND, INTERP\_CENTER, BAD\_COUNTS\_ERROR, PEAKCENTER,
201: NOTCHECKED, and NOPROFILE were all required to be unset).\footnote{See
202: http://www.sdss.org/dr5/products/catalogs/flags.html for more
203: information on SDSS flags} Each spectrum was processed using the
204: HAMMER stellar spectral-typing facility (Covey et al. 2007). The
205: HAMMER determines the spectral type of each star using measurements of
206: molecular bands and line strengths. We examined a subset of
207: the spectra by eye and confirmed that the spectral types given by the
208: HAMMER were accurate to within 1 subtype. We also limited our study
209: to spectral types M0-L0. The HAMMER was
210: able to successfully spectral type 44084 of the stars. Most of the
211: spectra that were not typed were either low signal-to-noise or bad
212: spectra. The spectral
213: type distribution of the DR5 low-mass star
214: spectroscopic sample is shown in Figure
215: \ref{figure:dr5dist}.
216:
217: To ensure a statistically robust sample, we removed stars with
218: 1) line-of-sight $r$-band extinction $>$ 0.5; 2) colors consistent with
219: having a white dwarf companion (Smol{\v c}i{\'c} et al. 2004); and 3) measured radial
220: velocities $>$ 500 km s$^{-1}$ (spurious velocities). The resulting high quality dataset
221: contains 38835 stars, roughly 4 times the number of stars used in the
222: W04 study.
223:
224:
225: \begin{figure}
226: \plotone{f1.eps}
227: \caption{Spectral type distribution of the SDSS DR5 low-mass star
228: spectroscopic sample. The total sample (solid) contains over 44,000
229: spectroscopically confirmed M and L-type dwarfs. Our
230: analysis is confined to a high quality subsample (dashed) and a high quality
231: subsample with measured proper motions (dot-dashed) which have over
232: 38,000 and 20,000 stars respectively. See text for more details.}
233: \label{figure:dr5dist}
234: \end{figure}
235:
236:
237:
238: The magnetic activity of each star was assigned based on its measured
239: H$\alpha$ emission line strength. Although other emission lines can
240: be used as activity indicators (e.g. CaII H \& K), our study defines
241: an active star as one that has H$\alpha$ emission in its spectrum. H$\alpha$
242: EWs were measured by the HAMMER, which now includes most of the
243: magnetic activity analysis described in W04. We assessed the accuracy
244: of our activity analysis by running Monte Carlo simulations to test
245: how well we could measure the H$\alpha$ emission line at a specific
246: spectral type and for a given signal-to-noise spectrum. Using the
247: Bochanski inactive templates (Bochanski et al. 2007a), we added noise
248: and an emission line to each spectral type and then used our software
249: to measure the H$\alpha$ EW 1000 times for each variation of noise,
250: spectral type and emission strength. We found that at a
251: signal-to-noise ratio (measured near H$\alpha$) of 3 and an H$\alpha$
252: EW of 1\AA\ (the same parameters used in previous studies; e.g. W04,
253: W06), we could accurately recover the H$\alpha$ emission over 96\% of
254: the time for all spectral types. This agrees with the less than 4\%
255: error in assigning activity status that was estimated in W04.
256:
257: After running the stars through the HAMMER, we
258: used the method of Walkowicz, Hawley \& West (2004) to compute the
259: ratio of luminosity in the H$\alpha$ line to the bolometric luminosity
260: (L$_{\rm{H\alpha}}$/L$_{bol}$) for all stars in the sample. We
261: adopted the color derived ($i-z$) $\chi$ values for this
262: calculation. Distances to all stars were calculated using the
263: photometric parallax methods of West et al. (2005) and Davenport et
264: al. (2006). Using the positions and distances of each star, Galactic
265: height was computed assuming the Sun is 15 pc above the Plane (Cohen
266: 1995; Ng et al. 1997; Binney et al. 1997).
267:
268: \begin{deluxetable*}{lrrccccc}
269: \tablewidth{0pt}
270: \tablewidth{0pt}
271: \tablecolumns{8}
272: \tabletypesize{\scriptsize}
273: \tablecaption{Colors of Late-Type Dwarfs}
274: \renewcommand{\arraystretch}{.6}
275: \tablehead{
276: \colhead{Spectral Type}&
277: \colhead{N$_{\rm{SDSS}}$\tablenotemark{a}}&
278: \colhead{N$_{\rm{2MASS}}$\tablenotemark{a}}&
279: \colhead{$r-i$}&
280: \colhead{$i-z$}&
281: \colhead{$z-J$}&
282: \colhead{$J-H$}&
283: \colhead{$H-K$}}
284: \startdata
285: M0 & 4945 & 1962 & 0.66 (0.12) &0.38 (0.07) &1.23 (0.19) &0.65 (0.10) &0.18 (0.14)\\
286: M1 & 3721 &1504 &0.82 (0.18) &0.46 (0.07) &1.27 (0.21) &0.64 (0.10) &0.21 (0.13)\\
287: M2 & 6105 & 2472 &1.00 (0.13) &0.55 (0.08) &1.31 (0.17) &0.62 (0.10) &0.22 (0.12)\\
288: M3 & 8609 & 3752& 1.21 (0.16) &0.65 (0.08) &1.37 (0.17) &0.60 (0.10) &0.24 (0.14)\\
289: M4 & 4984 & 2126&1.46 (0.15) &0.79 (0.09) &1.45 (0.16) &0.59 (0.10) &0.27 (0.13)\\
290: M5 & 3608 & 879& 1.91 (0.13) &1.06 (0.07) &1.61 (0.11) &0.59 (0.11) &0.31 (0.14)\\
291: M6 & 2674 & 570&2.11 (0.14) &1.16 (0.07) &1.71 (0.10) &0.60 (0.11) &0.33 (0.14)\\
292: M7 & 382 & 306&2.50 (0.18) &1.37 (0.09) &1.85 (0.10) &0.62 (0.09) &0.36 (0.11)\\
293: M8 & 74 & 147&2.73 (0.25) &1.58 (0.15) &2.04 (0.12) &0.66 (0.06) &0.40 (0.06)\\
294: M9 & 13 & 69&2.81 (0.21) &1.76 (0.08) &2.25 (0.09) &0.70 (0.08) &0.42 (0.06)\\
295: L0 & 10 & 31&2.49 (0.09) &1.84 (0.06) &2.45 (0.11) &0.77 (0.06) &0.50 (0.04)\\
296: \enddata
297: \tablecomments{The median and rms scatter (in parentheses) of the
298: SDSS-2MASS colors at each spectral type. Because the stars used to
299: compute these values have small measurement errors, the scatter
300: represents the intrinsic scatter of the stellar locus and not the
301: uncertainties in the median measurement. SDSS spectroscopic
302: targeting does not uniformly sample the stellar locus, possibly
303: introducing a bias to our derived colors. The $u-g$ and $g-r$ colors
304: are not shown because of the large influence that metallicity has on
305: the $g-r$ colors (West et al. 2004; Bochanski et al. 2007a) and the
306: large uncertainties in the $u$-band fluxes for most of the sample
307: stars.} \tablenotetext{a}{Only stars with magnitude uncertainties
308: smaller than 0.05 were used to compute the median colors}
309: \label{table:colors}
310: \end{deluxetable*}
311:
312: We used the SDSS/USNO-B matched catalog (Munn et al. 2004) to obtain
313: proper motions. The SDSS/USNO-B catalog uses SDSS galaxies to recalibrate
314: the USNO-B positions as well as includes SDSS stellar astrometry as an
315: additional epoch for improved proper motion measurements. The
316: USNO-B proper motions were obtained for 27285 of the SDSS stars (many
317: stars were not matched due to the relative shallowness of the USNO-B catalog).
318: Applying the same quality cuts as above, we arrived at a proper motion
319: sample of 20160 stars. Radial velocities were measured for all SDSS
320: stars using the method described in Bochanski et al. (2007a),
321: employing co-added, zero-velocity SDSS spectra which were used to
322: create high signal-to-noise ratio radial velocity templates.
323: Our derived radial velocities are accurate to within $\sim$5 km
324: s$^{-1}$ (Bochanski et al. 2007a). Using these proper motions the
325: derived distances and the radial velocities, space
326: motions were computed for all SDSS stars with measured proper motions.
327:
328: We also matched our spectroscopic sample to the 2MASS point source
329: catalog and found matches for 37845 of our stars. Although the
330: infrared colors are not a vital part of our analysis, we include them
331: in order to compute statistically robust 2MASS colors as a function of
332: spectral type and will include them in the online electronic tables (see below).
333:
334: \begin{figure}
335: \plotone{f2.eps}
336: \caption{The color-apparent magnitude distribution for the sample.
337: The blue ``edge'' is our color selection described above. The range
338: of magnitudes is constrained by both the bright and faint limits of
339: SDSS spectroscopy (Stoughton et al. 2002; Strauss et al. 2002) as well as our requirement
340: that none of the stars be saturated in the photometric data. The
341: gaps and overdense regions in color-magnitude space do not represent
342: the actual distribution of stars but are a reflection of the SDSS
343: spectroscopic targeting selection criteria, which are not designed
344: to sample stars in a complete, or even uniform, manner.}
345: \label{figure:cmd}
346: \end{figure}
347:
348:
349: The measured quantities for the SDSS DR5 low-mass star spectroscopic
350: sample are available for electronic download for use by the
351: astronomical community\footnote{Measured quantities can be obtained
352: electronically using the CDS Vizier database
353: http://vizier.u-strasbg.fr/viz-bin/VizieR}. The individual spectra
354: are available at the SDSS catalog archive
355: server\footnote{http://cas.sdss.org/astrodr5/en}. It is important to
356: note that these data do not represent a complete sample and that
357: spectral targeting introduces a variety of selection effects.
358: However, the sample covers a large range of values for many of the
359: physical attributes of the stars, including activity, metallicity and
360: Galactic motion, making an accurate activity analysis with this sample
361: possible. In addition, because most of the derived quantities are
362: computed by automatic routines, values for a small percentage of
363: individual stars will be incorrect. Large statistical results should
364: not be affected by these small errors. Users are nevertheless
365: cautioned to understand the origins of the data before using them
366: indiscriminately.
367:
368:
369: \section{Observational Results}
370:
371: Table \ref{table:colors} gives the median and rms scatter (in
372: parentheses) of the SDSS and 2MASS colors at each spectral type. The
373: stars used to compute these values were selected to have small
374: measurement errors and thus, the scatter represents the intrinsic
375: scatter of the stellar locus and not the uncertainties in the median
376: measurement. SDSS spectroscopic targeting does not uniformly sample
377: the stellar locus, possibly introducing a bias to our derived colors.
378: The $u-g$ and $g-r$ colors were not shown because of the
379: large influence that metallicity has on the $g-r$ colors (West et
380: al. 2004; Bochanski et al. 2007a) and the large uncertainties in the
381: $u$-band fluxes for most of the sample stars. The SDSS colors are in
382: excellent agreement with the previous results found in West et
383: al. (2005) and Bochanski et al. (2007a). Because of the larger sample
384: size, our 2MASS colors supersede those reported in Hawley et
385: al. (2002), West et al. (2005) and Covey et al. (2007). Figure
386: \ref{figure:cmd} shows the color-apparent magnitude distribution for
387: the sample. The blue ``edge'' is our color selection described
388: above. The range of magnitudes is constrained by both the bright and
389: faint limits of SDSS spectroscopy (Stoughton et al. 2002; Strauss et al. 2002) as well as
390: our requirement that none of the stars be saturated in the photometric
391: data. The gaps and overdense regions in color-magnitude space
392: do not
393: represent the actual distribution of stars but are a reflection of
394: the SDSS spectroscopic targeting selection criteria, which are
395: not designed to sample stars in a complete, or even uniform, manner.
396: Figure \ref{figure:cmd} also helps to demonstrate why some of the
397: computed colors from Table \ref{table:colors} may be affected by selection biases.
398: Clearly, an analysis of the luminosity function of low mass stars cannot
399: be carried out with these data, but is currently being
400: addressed using the SDSS photometric data (Covey et
401: al. 2008; Bochanski et al. 2008).
402:
403:
404:
405:
406: \subsection{Activity}
407:
408: Applying the aforementioned signal-to-noise and EW cuts (3 and 1\AA\
409: respectively) to our activity analysis yields an updated distribution
410: of activity fraction as a function of spectral type (Figure
411: \ref{figure:actfrac}). At earlier types ($<$M7) this distribution
412: resembles that of previous studies (Hawley et al. 1996; Gizis et
413: al. 2000; W04). However, the later types do not show the same
414: decrease in activity fraction seen previously (W04; Schmidt et
415: al. 2007). This is likely a result of our strict quality cuts, which
416: removes the distant, late-type stars from the sample. The more
417: distant stars are more likely to be older and therefore
418: inactive. However, the low signal-to-noise in those spectra makes
419: their activity state uncertain, and they are therefore excluded from
420: the analysis.
421:
422: \begin{figure}
423: \plotone{f3.eps}
424: \caption{Fraction of active stars as a function of spectral type for
425: the DR5 spectroscopic sample (diamonds). At earlier types ($<$M7)
426: this distribution resembles that of previous studies (triangles;
427: Hawley et al. 1996; Gizis et al. 2000; W04). However, we do not see
428: a turnover in the activity fraction in the latest types as seen
429: previously. This is likely a result of our strict quality cuts,
430: which remove the distant, late-type stars from the sample. The more
431: distant stars are more likely to be older and therefore
432: inactive. However, the low signal-to-noise in those spectra does not
433: allow us to be certain of their activity state and they are
434: therefore excluded from the analysis.}
435: \label{figure:actfrac}
436: \end{figure}
437:
438:
439:
440: Figure \ref{figure:actfrac} highlights an important point about
441: the selection effects surrounding previous determinations of the
442: activity fraction as a function of spectral type. Because the
443: activity fraction at a given spectral type changes as a function of
444: Galactic height (W04; W06), the sample selection will
445: bias the observed activity fraction. We demonstrate this effect by
446: plotting the activity fraction as a function of spectral type and
447: constraining the absolute Galactic height of the stars to 9 different
448: regions (Figure \ref{figure:disteff}). It is clear that the volume
449: in which the data are observed plays a large role in the resulting activity fractions.
450:
451:
452: \begin{figure}
453: \plotone{f4.eps}
454: \caption{The fraction of active stars as a function of spectral type
455: for 9 different regions in vertical distance from the Plane. By
456: tracing individual spectral types over each distance bin, we see
457: that distance plays an important role in determining the shape of
458: Figure \ref{figure:actfrac}; most of the stars are active in the
459: nearby bins and the activity fraction declines as we examine more
460: distant stars.}
461: \label{figure:disteff}
462: \end{figure}
463:
464:
465:
466: The distance effect can be clearly seen by expanding on the study of
467: W06 and examining how the activity fraction changes as
468: a function of vertical height for each spectral type. Figure
469: \ref{figure:allfracgrid} shows the activity fraction as a function of
470: absolute height above the Plane for spectral types M0-M8 (data have
471: been folded across the Plane to increase the signal). There are
472: not enough stars in the later-type bins to justify their analysis. In
473: almost every case, the stars closer to the Galactic Plane have larger activity
474: fractions than those further from the Plane. The amplitudes and
475: slopes of the decrease are different for each spectral type.
476: Dynamical models with activity lifetimes are
477: fit to these data and discussed in \S4.
478:
479: \begin{figure*}
480: \epsscale{1}
481: \plotone{f5.eps}
482: \caption{The fraction of active stars as a function of absolute
483: distance from the Galactic Plane for M0-M8 dwarfs. The activity
484: fraction decreases with Galactic height at all spectral types. The
485: magnitude and slope of the decline varies as a function of spectral
486: type. This effect is likely caused by the combination of thin disk
487: dynamical heating and a rapid decrease of magnetic activity after a
488: characteristic time.}
489: \label{figure:allfracgrid}
490: \end{figure*}
491:
492:
493:
494: As previously seen in M7 dwarfs (W06), most of the spectral types show
495: some evidence that the level of activity (quantified by
496: L$_{\rm{H\alpha}}$/L$_{bol}$; Hawley et al. 1996) also decreases as a
497: function of vertical height above the Plane. Figure
498: \ref{figure:lbolgrid} shows the mean L$_{\rm{H\alpha}}$/L$_{bol}$
499: distribution for M2-M7 stars as a function of vertical distance from
500: the Plane. The other types were excluded because of the small number
501: of active stars at early spectral types and small vertical extent at
502: late spectral types. Although the spread of the distribution is large
503: (large error bars) in each bin, the uncertainty in the mean (small
504: error bars) is small, and indicates a small but significant decrease
505: in activity with distance. This trend runs opposite to what one might
506: expect from a magnitude-limited selection effect; stars with more
507: activity should be easier to see further away. This decrease is
508: marginally consistent with a Skumanich (1972) type power law
509: age-activity relation (t$^{-\alpha}$). However, a unique value of
510: $\alpha$ cannot explain all of the trends, and the activity decrease
511: is likely more complicated.
512:
513:
514: \begin{figure}
515: \plotone{f6.eps}
516: \caption{Median L$_{\rm{H\alpha}}$/L$_{bol}$ as a function of absolute
517: vertical distance from the Plane. The narrow error bars indicate
518: the spread of L$_{\rm{H\alpha}}$/L$_{bol}$ values while the wide
519: error bars are the uncertainties in the median relations. Although
520: the spread in L$_{\rm{H\alpha}}$/L$_{bol}$ is large at each bin,
521: there is a slight but significant decreasing trend with distance
522: from the Galactic Plane. The decrease is marginally consistent
523: with a Skumanich (1972) type power law age-activity relation
524: (t$^{-\alpha}$). However, a unique value of $\alpha$ cannot explain
525: all of the trends and the activity decrease is likely more complicated.}
526: \label{figure:lbolgrid}
527: \end{figure}
528:
529:
530:
531:
532: In addition, the lack of stars with small L$_{\rm{H\alpha}}$/L$_{bol}$
533: at small distances from the Plane suggests that activity does not
534: continue a slow decline forever, but undergoes a rapid decrease at
535: some point during a star's lifetime. There are stars in the DR5
536: sample that have H$\alpha$ EW $<$ 1\AA\ but do not get included in our
537: main study and have smaller L$_{\rm{H\alpha}}$/L$_{bol}$ values.
538: However, there is still a population of nearby, high signal-to-noise
539: stars in which we could detect H$\alpha$ activity if it were there -
540: but we don't. Our ability to observe inactive stars is confirmed by
541: other studies including the high resolution spectroscopic study of
542: Rauscher \& Marcy (2006) and the Bochanski et al. (2007a) spectral
543: templates, where thousands of our inactive SDSS spectra were co-added
544: to produce extremely high signal-to-noise template spectra. The
545: inactive templates as well as many of the Rauscher \& Marcy (2006)
546: stars show no signs of H$\alpha$ activity, confirming our ability to
547: measure a star when it is in the inactive phase.
548:
549: \subsection{Metallicity}
550: In addition to having less activity and greater dynamical heating,
551: stars formed earlier in the Galaxy's history should have fewer metals;
552: we should see a decrease in metallicity as a function of vertical
553: distance from the Plane. Currently, there are no robust techniques
554: for translating low resolution spectra of M dwarfs to an absolute
555: metallicity scale. Woolf and Wallerstein (2006) attempted to
556: calibrate M dwarf metallicities using the molecular indices described
557: in in previous studies (Reid, Hawley \& Gizis 1995; Gizis 1997), but
558: were thwarted by small sample size. High resolution follow-up of
559: recent large surveys of cool, nearby stars that span a range of
560: metallicities (e.g. this sample; Marshall 2007), will allow for larger
561: datasets to properly calibrate the metallicity scale. In the
562: meantime, we can use molecular indices to determine relative metal
563: content. Recent studies have used a combination of the Gizis
564: (1997) indices (CaH2, CaH3 and TiO5) to identify low mass subdwarfs
565: and extreme subdwarfs (Lepine, Shara \& Rich 2003; Burgasser \&
566: Kirkpatrick 2006) . We adapt the Lepine two-dimensional metallicity
567: space to a single ratio, (CaH2+CaH3)/TiO5, and use this as a proxy for
568: metallicity at a given spectral type (a similar technique was used in
569: Bochanski et al. 2007b).
570:
571: Figure \ref{figure:met} shows the metal sensitive ratio
572: ((CaH2+CaH3)/TiO5) as a function of vertical distance above the Plane
573: for M0-M8 dwarfs. A decrease in metallicity as a function of Galactic
574: height is clearly evident, indicating that the more distant stars have
575: fewer metals. For the latest-type dwarfs, the small number of usable
576: stars does not allow for an accurate analysis, but for the early-mid
577: types the metallicity trend is both pronounced and significant. In
578: the future, when molecular indices can be reliably tied to an absolute
579: metallicity scale, we will be able to use each spectral type
580: independently to calibrate the metallicity evolution of the Milky Way
581: disk.
582:
583:
584: \begin{figure*}
585: \plotone{f7.eps}
586: \caption{The metal sensitive ratio (CaH2+CaH3)/TiO5 as a function of
587: absolute vertical distance from the Plane for M0-M8 dwarfs. Stars
588: further from the Galactic Plane are more metal-poor. Error bars
589: represent the uncertainties in the mean values for each bin. Although we
590: cannot yet tie this ratio to an absolute metallicity scale,
591: our results demonstrate that future M dwarf studies should be able
592: to determine the metallicity evolution of the Milky Way thin disk in
593: a statistically robust way.}
594: \label{figure:met}
595: \end{figure*}
596:
597:
598: \section{Simulations}
599:
600: We used the simple 1-D dynamical model described in W06 to trace the
601: vertical dynamics of stars as a function of time. For each spectral
602: type, we assumed a constant star formation rate, and injected a new
603: population of 50 stars at the Galactic midplane every 200 Myr, for a
604: total simulation time of 10 Gyrs. Each new group of stars began with
605: a randomly drawn velocity dispersion of 8 km s$^{-1}$ (Binney et
606: al. 2000) and had a new position and velocity computed every 0.1 Myr.
607: We used the ``leap-frog'' integration technique (Press et al. 1992)
608: and the vertical Galactic potential from Kuijken \& Gilmore (1989) and
609: Siebert et al. (2003) to track the vertical dynamics of each star in
610: the simulation.
611:
612: We simulated dynamical heating by altering the velocities (energies)
613: of stars such that their new velocity dispersions would match
614: a $\sigma \propto t^{0.5}$ relation (Wielen 1977;
615: Fuchs et al. 2001; H\"anninen \& Flynn 2002). The velocities were
616: also scaled by the square-root of the mass appropriate for each
617: type (as defined by Reid \& Hawley 2005). We varied the heating
618: energy about this mean value and added randomly drawn velocities
619: when stars were within a given distance from the Plane. This
620: distance (or ``region of influence''; see W06) is a way to
621: parameterize the cross section of interactions with molecular clouds
622: during a vertical Plane crossing and it is modeled to be symmetric about
623: the Galactic midplane. In a given orbit,
624: a star may have no interactions with molecular clouds, or it may run
625: into the full extent of a molecular region. Thus, to first order,
626: ``the regions of influence'' can be thought of as the product of the
627: scale height of molecular gas and the fraction of Plane crossings that
628: result in a dynamical kick. The size of the ``region of influence'' is varied from $\pm$0.5 pc to $\pm$5 pc in
629: intervals of 0.5 pc. In total we ran 70 simulations for each M dwarf
630: spectral type. Our simulations tracked the velocity, position and age
631: of each star.
632:
633: \begin{figure}
634: \epsscale{1}
635: \plotone{f8.eps}
636: \caption{Active fraction as a function of absolute vertical distance
637: from the Plane (diamonds) compared with the best-fit simulation
638: data (open circles). All simulated data uses a ``region of
639: influence'' of 1 pc. The simple 1-D simulations fit the observed
640: trends well at all spectral types.}
641: \label{figure:allfracgridmod}
642: \end{figure}
643:
644: After the dynamical simulations were complete, we introduced a
645: timescale for magnetic activity to the simulated data. Specifically,
646: we assigned an activity lifetime to every star of a given spectral
647: type. We varied these lifetimes from 0 Gyr to 9 Gyr in 0.1 Gyr
648: intervals for each individual simulation. From the resulting data, we
649: can derive activity fractions as a function of vertical distance from
650: the Plane and examine how the simulated properties compare to the
651: observed data.
652:
653: We compared the simulated activity fractions to the observed activity
654: fraction as a function of galactic height and performed a chi-squared minimization, fitting for
655: ``region of influence'' size, dynamical heating energy and activity
656: lifetime. Because the ``region of influence'' should be the same for
657: all spectral types, we first solved for the best-fit value and found it to
658: be 1 pc, slightly smaller than the 3 pc found in W06. Although 1 pc
659: is close to the edge of the explored grid (0.5 pc), it produces a
660: significantly better fit to the data as compared to the runs that use
661: 0.5 or 1.5 pc values. Then, using a fixed
662: ``region of influence'' of 1 pc, the energies and
663: activity lifetimes were determined for each spectral type.
664: Figure \ref{figure:allfracgridmod}
665: shows the observed activity fraction as a function of vertical
666: distance from the Plane with the simulated trends
667: overplotted for the best-fit simulations.
668:
669:
670:
671: \begin{figure}
672: \epsscale{1}
673: \plotone{f9.eps}
674: \caption{Vertical velocity dispersions as a function of vertical
675: distance from the Plane for the DR5 sample (diamonds) compared to
676: the simulations (open circles). Only spectral types
677: that have large numbers of nearby ($<$ 500 pc) stars with measured
678: proper motions are shown. Our simple dynamical simulations are
679: able to reproduce nearly all of the velocity data.}
680: \label{figure:allvelgridmod}
681: \end{figure}
682:
683: To ensure that our simulations were producing dynamically accurate
684: results, we compared the vertical velocity dispersions from our
685: proper motion matched sample to the simulated velocity
686: dispersions. The velocity dispersions were measured using the
687: probability analysis described in Bochanski et al. 2007b (see
688: also Reid et al. 1995).
689: Figure \ref{figure:allvelgridmod} shows the velocity dispersions
690: as a function of absolute vertical distance from the Plane for both
691: the DR5 sample and the simulations. For several
692: spectral types, there are a limited number of stars with good proper
693: motions. Therefore, only the M1-M6 spectral types have been included.
694: The data and model velocities are in excellent agreement for most of
695: the measured velocity bins. Deviations from the models are likely due
696: to the simplicity of our dynamical models, which do not allow for
697: such known complications as dynamical substructure.
698:
699: \begin{deluxetable}{lccccc}
700: \tablewidth{80pt}
701: \tablecolumns{3}
702: \tabletypesize{\small}
703: \tablecaption{Activity Lifetimes of M dwarfs}
704: \renewcommand{\arraystretch}{.6}
705: \tablehead{
706: \colhead{Spectral Type}&
707: \colhead{Age (Gyr)}}
708: \startdata
709: \vspace{1mm}
710: M0&$0.8\pm^{\ 0.6}_{\ 0.6}$\\
711: \vspace{1mm}
712: M1&$0.4\pm^{\ 0.4}_{\ 0.4}$\\
713: \vspace{1mm}
714: M2&$1.2\pm^{\ 0.4}_{\ 0.4}$\\
715: \vspace{1mm}
716: M3&$2.0\pm^{\ 0.5}_{\ 0.5}$\\
717: \vspace{1mm}
718: M4&$4.5\pm^{\ 0.5}_{\ 1.0}$\\
719: \vspace{1mm}
720: M5&$7.0\pm^{\ 0.5}_{\ 0.5}$\\
721: \vspace{1mm}
722: M6&$7.0\pm^{\ 0.5}_{\ 0.5}$\\
723: \vspace{1mm}
724: M7&$8.0\pm^{\ 0.5}_{\ 1.0}$\\
725: \enddata
726: \label{table:activeage}
727: \end{deluxetable}
728:
729: The final result of our analysis is an estimate of the activity
730: lifetime at each spectral type. The values are given in
731: Table \ref{table:activeage}
732: and shown in Figure \ref{figure:activeage}.
733: The error bars indicate the range of
734: activity lifetimes for all of the simulations at each spectral type.
735: The dotted line is the Hawley et al. (2000) activity-age relation.
736: The activity lifetimes increase to later spectral types and as predicted, our
737: results fall above the Hawley relation and
738: extend to later spectral types. Due to the small sample size beyond a
739: spectral type of M7, we only include lifetimes for M0-M7 dwarfs.
740: There is a large jump in the activity lifetime at the fully convective boundary between types M3-M5, suggesting that a change in the magnetic dynamo mechanism at this boundary leads to much longer lasting magnetic activity in later type M dwarfs. The derived trend also increases our understanding of
741: the selection effects responsible for the activity fractions shown in
742: Figures \ref{figure:actfrac} and \ref{figure:disteff}; later type
743: stars are seen to be more active because they are both closer
744: \emph{and} their active lifetimes are longer.
745:
746:
747: \begin{figure*}
748: \plotone{f10.eps}
749: \caption{The activity lifetimes of M dwarfs as determined by comparing
750: the DR5 spectroscopic data to the 1-D simulations (stars). The
751: Hawley et al. (2000) age-activity relation is overplotted for
752: comparison (dotted). As predicted, the Hawley relation provides a lower
753: limit to the ages. We find that there is
754: a significant increase in activity lifetimes between spectral types
755: M3 to M5, possibly
756: indicating a physical change in the production of magnetic fields
757: as full convection sets in. The uncertainties indicate the spread
758: of the model fits at each spectral type. }
759: \label{figure:activeage}
760: \end{figure*}
761:
762:
763: \section{Summary}
764:
765: Using over 44,000 spectroscopically confirmed M and L dwarfs, we have
766: expanded the analysis of W06 and completed a detailed study of the
767: magnetic activity, metallicity and dynamics of cool stars in the
768: SDSS. The results of our analysis are summarized below.
769:
770: 1. We have compiled the largest spectroscopic sample of low mass stars
771: ever assembled. By matching our spectra to SDSS and 2MASS photometry,
772: we compute the mean optical/IR colors for each spectral type with
773: unprecedented robustness. Although these colors represent the best
774: determination of optical/IR colors for cool stars to date, they may
775: still be affected by the non-uniform sample selection.
776:
777: 2. We show that activity fractions decrease as a function of absolute
778: height above the Galactic Plane for all M dwarf spectral types and
779: that the magnitude and
780: slope of this effect is different for each spectral type.
781:
782: 3. The amount of activity as quantified by
783: L$_{\rm{H}\alpha}$/L$_{bol}$ decreases slightly as a function of
784: vertical distance from the Plane. However,
785: L$_{\rm{H}\alpha}$/L$_{bol}$ does not continue to decrease to the
786: smallest detectable values, providing evidence for a rapid decrease of
787: magnetic activity after some characteristic lifetime.
788:
789: 4. The metallicity sensitive ratio (CaH2+CaH3)/TiO5 also decreases
790: with increased distance from the Plane. Although we cannot yet calibrate
791: this ratio to an absolute metallicity scale, we suggest that future
792: studies will be able to use data of this type to determine the
793: metallicity evolution of the Milky Way using low mass dwarfs.
794:
795: 5. We use a 1-D dynamical simulation to create a realistic model
796: of M dwarf motions and positions over the lifetime of the Galaxy and
797: show that our models match the observed activity and velocity trends.
798:
799: 6. Comparing our activity data to the dynamical simulations, we
800: derive an age-activity relation for M dwarfs for spectral types
801: M0-M7. We find that there is a strong increase in the active lifetimes
802: of stars in the spectral type range from M3-M5, which may be due to the
803: onset of full convection and subsequent change in the mechanism for
804: producing and storing magnetic fields in the stellar interior.
805:
806:
807: This last point is the main result of the paper and deserved further
808: elucidation. The sharp increase in activity lifetime after spectral
809: type M3 (where the stars are predicted to become fully convective) may
810: be an empirical indication of the change from a solar-like,
811: rotationally-dependent magnetic dynamo in the earlier type stars, to a
812: turbulent dynamo (that may or may not be dependent on rotation) in the
813: later type stars. The short activity lifetimes of the early type M
814: dwarfs would then be naturally explained by an extension of solar type
815: rotation-activity relations (e.g. Skumanich) to these stars, and the
816: timescale would be set by their spindown time. Although previous
817: studies have claimed evidence for a rotation-activity relation
818: extending past the M3 convective transition and into the brown dwarf
819: regime (Delfosse et al. 1998; Mohanty \& Basri 2003), these studies
820: may be biased by samples of nearby, young, active M dwarfs. Future rotation
821: studies of inactive late-type M dwarfs are needed to confirm these
822: results.
823:
824: The long activity lifetimes we have found for the fully convective
825: stars, where the magnetic fields are presumably generated by a
826: turbulent dynamo mechanism, set significant constraints on the
827: operation of the dynamo and the emergence of the surface fields to
828: power the activity. Future models must successfully predict long
829: (several Gyr) lifetimes, and the lifetimes must be mass dependent
830: (increasing with lower mass).
831:
832: Our age-activity relation is consistent with previous findings for
833: early type M dwarfs and we extend the relation to stars near the M dwarf/brown
834: dwarf frontier. The faintness of M dwarfs in old star clusters limits our
835: ability to observe this relation directly. However, future
836: spectroscopic studies should be able to test our age-activity relation with data for clusters of known (older) ages.
837:
838: Many other studies are possible with this large spectroscopic sample.
839: Future papers will focus on the detailed dynamics of M dwarfs as probes
840: of Galactic structure (Hawley et al. 2008; West et al. 2008),
841: investigate the flare rate of active stars (Hilton et al. 2008) and
842: extend our DR5 sample to the lower mass L dwarfs (Schmidt
843: et al. 2008).
844:
845: The advent of large surveys such as SDSS and 2MASS has allowed for
846: many statistically robust studies of M dwarfs, the most numerous yet
847: faintest stellar constituents of the Milky Way. Future large surveys
848: such as Pan-STARRS and LSST will add much to photometric studies of
849: low-mass stars but will not have spectroscopic components. Therefore,
850: this SDSS low-mass star spectroscopic sample should remain useful
851: for the foreseeable future.
852:
853:
854: \section{Acknowledgments}
855: AAW acknowledges the support of NSF grant 0540567. SLH and JJB are
856: supported by NSF grant AST02-05875 and AST067644. KRC is supported by
857: a Spitzer Postdoctoral Fellowship. The authors would like to thank
858: Matthew Browning, Lucianne Walkowicz, Tom Quinn, Adam Burgasser, and
859: Gibor Basri for useful discussions while completing this project.
860:
861: Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, the U.S. Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, the Max Planck Society, and the Higher Education Funding Council for England. The SDSS Web Site is http://www.sdss.org/.
862:
863: The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. The Participating Institutions are the American Museum of Natural History, Astrophysical Institute Potsdam, University of Basel, Cambridge University, Case Western Reserve University, University of Chicago, Drexel University, Fermilab, the Institute for Advanced Study, the Japan Participation Group, Johns Hopkins University, the Joint Institute for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and Cosmology, the Korean Scientist Group, the Chinese Academy of Sciences (LAMOST), Los Alamos National Laboratory, the Max-Planck-Institute for Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New Mexico State University, Ohio State University, University of Pittsburgh, University of Portsmouth, Princeton University, the United States Naval Observatory, and the University of Washington.
864:
865: \bibliographystyle{apj}
866:
867: %\begin{thebibliography}
868: %\nocite{*}
869: %\bibliography{ms}
870:
871: %\end{thebibliography}
872:
873: \begin{thebibliography}{71}
874: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
875:
876: \bibitem[{{Adelman-McCarthy}
877: {et~al.}(2007{\natexlab{a}})}]{2007ApJS..172..634A}
878: {Adelman-McCarthy}, J.~K., {et~al.} 2007{\natexlab{a}}, \apjs, 172, 634
879:
880: \bibitem[{{Adelman-McCarthy}
881: {et~al.}(2007{\natexlab{b}})}]{2007arXiv0707.3413A}
882: ---. 2007{\natexlab{b}}, \apjs, submitted
883:
884: \bibitem[{{Barry}(1988)}]{1988ApJ...334..436B}
885: {Barry}, D.~C. 1988, \apj, 334, 436
886:
887:
888: \bibitem[{{Binney} {et~al.}(2000){Binney}, {Dehnen}, \&
889: {Bertelli}}]{2000MNRAS.318..658B}
890: {Binney}, J., {Dehnen}, W., \& {Bertelli}, G. 2000, \mnras, 318, 658
891:
892: \bibitem[{{Binney} {et~al.}(1997){Binney}, {Gerhard}, \&
893: {Spergel}}]{1997MNRAS.288..365B}
894: {Binney}, J., {Gerhard}, O., \& {Spergel}, D. 1997, \mnras, 288, 365
895:
896: \bibitem[{{Bochanski} {et~al.}(2007{\natexlab{a}}){Bochanski}, {Munn},
897: {Hawley}, {West}, {Covey}, \& {Schneider}}]{2007arXiv0708.0044B}
898: {Bochanski}, J.~J., {Munn}, J.~A., {Hawley}, S.~L., {West}, A.~A., {Covey},
899: K.~R., \& {Schneider}, D.~P. 2007{\natexlab{a}}, \aj, in press
900:
901: \bibitem[{{Bochanski} {et~al.}(2007{\natexlab{b}}){Bochanski}, {West},
902: {Hawley}, \& {Covey}}]{2007AJ....133..531B}
903: {Bochanski}, J.~J., {West}, A.~A., {Hawley}, S.~L., \& {Covey}, K.~R.
904: 2007{\natexlab{b}}, \aj, 133, 531
905:
906: \bibitem[{{Bochanski} {et~al.}(2008)}]{}
907: {Bochanski}, J.~J., {et~al.} 2008, \aj, in prep
908:
909: \bibitem[{{Burgasser} \& {Kirkpatrick}(2006)}]{2006ApJ...645.1485B}
910: {Burgasser}, A.~J., \& {Kirkpatrick}, J.~D. 2006, \apj, 645, 1485
911:
912: \bibitem[{{Chabrier} \& {Baraffe}(1997)}]{1997A&A...327.1039C}
913: {Chabrier}, G., \& {Baraffe}, I. 1997, \aap, 327, 1039
914:
915: \bibitem[{{Cohen}(1995)}]{1995ApJ...444..874C}
916: {Cohen}, M. 1995, \apj, 444, 874
917:
918: \bibitem[{{Covey} {et~al.}(2007)}]{2007arXiv0707.4473C}
919: {Covey}, K.~R., {et~al.} 2007, \aj, in press
920:
921: \bibitem[{{Covey} {et~al.}(2008)}]{2008}
922: ---. 2008, \aj, in prep
923:
924: \bibitem[{{Cutri} {et~al.}(2003)}]{2003tmc..book.....C}
925: {Cutri}, R.~M., {et~al.} 2003, {2MASS All Sky Catalog of point sources.} (The
926: IRSA 2MASS All-Sky Point Source Catalog, NASA/IPAC Infrared Science
927: Archive.~http://irsa.ipac.caltech.edu/applications/Gator/)
928:
929: \bibitem[{{Davenport} {et~al.}(2006){Davenport}, {West}, {Matthiesen},
930: {Schmieding}, \& {Kobelski}}]{2006PASP..118.1679D}
931: {Davenport}, J.~R.~A., {West}, A.~A., {Matthiesen}, C.~K., {Schmieding}, M., \&
932: {Kobelski}, A. 2006, \pasp, 118, 1679
933:
934: \bibitem[{{Delfosse} {et~al.}(1998){Delfosse}, {Forveille}, {Perrier}, \&
935: {Mayor}}]{1998A&A...331..581D}
936: {Delfosse}, X., {Forveille}, T., {Perrier}, C., \& {Mayor}, M. 1998, \aap, 331,
937: 581
938:
939: \bibitem[{{Eggen}(1990)}]{1990PASP..102..166E}
940: {Eggen}, O.~J. 1990, \pasp, 102, 166
941:
942: \bibitem[{{Fleming} {et~al.}(1995){Fleming}, {Schmitt}, \&
943: {Giampapa}}]{1995ApJ...450..401F}
944: {Fleming}, T.~A., {Schmitt}, J.~H.~M.~M., \& {Giampapa}, M.~S. 1995, \apj, 450,
945: 401
946:
947: \bibitem[{{Fuchs} {et~al.}(2001){Fuchs}, {Dettbarn}, {Jahrei{\ss}}, \&
948: {Wielen}}]{2001ASPC..228..235F}
949: {Fuchs}, B., {Dettbarn}, C., {Jahrei{\ss}}, H., \& {Wielen}, R. 2001, in ASP
950: Conf. Ser. 228: Dynamics of Star Clusters and the Milky Way, ed.
951: S.~{Deiters}, B.~{Fuchs}, A.~{Just}, R.~{Spurzem}, \& R.~{Wielen}, 235
952:
953: \bibitem[{{Fukugita} {et~al.}(1996){Fukugita}, {Ichikawa}, {Gunn}, {Doi},
954: {Shimasaku}, \& {Schneider}}]{1996AJ....111.1748F}
955: {Fukugita}, M., {Ichikawa}, T., {Gunn}, J.~E., {Doi}, M., {Shimasaku}, K., \&
956: {Schneider}, D.~P. 1996, \aj, 111, 1748
957:
958: \bibitem[{{Giampapa} \& {Liebert}(1986)}]{1986ApJ...305..784G}
959: {Giampapa}, M.~S., \& {Liebert}, J. 1986, \apj, 305, 784
960:
961: \bibitem[{{Gizis}(1997)}]{1997AJ....113..806G}
962: {Gizis}, J.~E. 1997, \aj, 113, 806
963:
964: \bibitem[{{Gizis} {et~al.}(2000){Gizis}, {Monet}, {Reid}, {Kirkpatrick},
965: {Liebert}, \& {Williams}}]{2000AJ....120.1085G}
966: {Gizis}, J.~E., {Monet}, D.~G., {Reid}, I.~N., {Kirkpatrick}, J.~D., {Liebert},
967: J., \& {Williams}, R.~J. 2000, \aj, 120, 1085
968:
969: \bibitem[{{Gizis} {et~al.}(2002){Gizis}, {Reid}, \&
970: {Hawley}}]{2002AJ....123.3356G}
971: {Gizis}, J.~E., {Reid}, I.~N., \& {Hawley}, S.~L. 2002, \aj, 123, 3356
972:
973: \bibitem[{{Gunn} {et~al.}(1998)}]{1998AJ....116.3040G}
974: {Gunn}, J.~E., {et~al.} 1998, \aj, 116, 3040
975:
976: \bibitem[{{Gunn} {et~al.}(2006)}]{2006AJ....131.2332G}
977: ---. 2006, \aj, 131, 2332
978:
979: \bibitem[{{H{\"a}nninen} \& {Flynn}(2002)}]{2002MNRAS.337..731H}
980: {H{\"a}nninen}, J., \& {Flynn}, C. 2002, \mnras, 337, 731
981:
982: \bibitem[{{Hawley} {et~al.}(2002){Hawley}, {Covey},
983: {et~al.}}]{2002AJ....123.3409H}
984: {Hawley}, S.~L., {Covey}, K.~R., {et~al.} 2002, \aj, 123, 3409
985:
986: \bibitem[{{Hawley} {et~al.}(1996){Hawley}, {Gizis}, \&
987: {Reid}}]{1996AJ....112.2799H}
988: {Hawley}, S.~L., {Gizis}, J.~E., \& {Reid}, I.~N. 1996, \aj, 112, 2799
989:
990: \bibitem[{{Hawley} {et~al.}(2000){Hawley}, {Reid}, \&
991: {Tourtellot}}]{2000vlms.conf..109H}
992: {Hawley}, S.~L., {Reid}, I.~N., \& {Tourtellot}, J.~G. 2000, in Very Low-mass
993: Stars and Brown Dwarfs, Edited by R. Rebolo and M. R. Zapatero-Osorio.
994: Published by the Cambridge University Press, UK, 2000., p.109, ed.
995: R.~{Rebolo} \& M.~R. {Zapatero-Osorio}, 109--+
996:
997: \bibitem[{{Hawley} {et~al.}(1999){Hawley}, {Tourtellot}, \&
998: {Reid}}]{1999AJ....117.1341H}
999: {Hawley}, S.~L., {Tourtellot}, J.~G., \& {Reid}, I.~N. 1999, \aj, 117, 1341
1000:
1001: \bibitem[{{Hawley} {et~al.}(2008)}]{2008as}
1002: {Hawley}, S.~L., {et~al.} 2008, \aj, in prep
1003:
1004: \bibitem[{{Hilton} {et~al.}(2008)}]{2008AJ..409H}
1005: {Hilton}, E.~J., {et~al.} 2008, \aj, in prep
1006:
1007: \bibitem[{{Hogg} {et~al.}(2001){Hogg}, {Finkbeiner}, {Schlegel}, \&
1008: {Gunn}}]{2001AJ....122.2129H}
1009: {Hogg}, D.~W., {Finkbeiner}, D.~P., {Schlegel}, D.~J., \& {Gunn}, J.~E. 2001,
1010: \aj, 122, 2129
1011:
1012: \bibitem[{{Ivezi{\'c}} {et~al.}(2004)}]{2004AN....325..583I}
1013: {Ivezi{\'c}}, {\v Z}., {et~al.} 2004, Astronomische Nachrichten, 325, 583
1014:
1015: \bibitem[{{Kuijken} \& {Gilmore}(1989)}]{1989MNRAS.239..605K}
1016: {Kuijken}, K., \& {Gilmore}, G. 1989, \mnras, 239, 605
1017:
1018: \bibitem[{{L{\'e}pine} {et~al.}(2003){L{\'e}pine}, {Shara}, \&
1019: {Rich}}]{2003ApJ...585L..69L}
1020: {L{\'e}pine}, S., {Shara}, M.~M., \& {Rich}, R.~M. 2003, \apjl, 585, L69
1021:
1022: \bibitem[{{Marshall}(2007)}]{2007AJ....134..778M}
1023: {Marshall}, J.~L. 2007, \aj, 134, 778
1024:
1025: \bibitem[{{Mohanty} \& {Basri}(2003)}]{2003ApJ...583..451M}
1026: {Mohanty}, S., \& {Basri}, G. 2003, \apj, 583, 451
1027:
1028: \bibitem[{{Munn} {et~al.}(2004)}]{2004AJ....127.3042G}
1029: {Munn}, J.~A., {et~al.} 2004, \aj, 127, 3042
1030:
1031:
1032: \bibitem[{{Ng} {et~al.}(1997){Ng}, {Bertelli}, {Chiosi}, \&
1033: {Bressan}}]{1997A&A...324...65N}
1034: {Ng}, Y.~K., {Bertelli}, G., {Chiosi}, C., \& {Bressan}, A. 1997, \aap, 324, 65
1035:
1036: \bibitem[{{Ossendrijver}(2003)}]{2003A&ARv..11..287O}
1037: {Ossendrijver}, M. 2003, \aapr, 11, 287
1038:
1039: \bibitem[{{Parker}(1993)}]{1993ApJ...408..707P}
1040: {Parker}, E.~N. 1993, \apj, 408, 707
1041:
1042: \bibitem[{{Pier} {et~al.}(2003){Pier}, {Munn}, {Hindsley}, {Hennessy}, {Kent},
1043: {Lupton}, \& {Ivezi{\'c}}}]{2003AJ....125.1559P}
1044: {Pier}, J.~R., {Munn}, J.~A., {Hindsley}, R.~B., {Hennessy}, G.~S., {Kent},
1045: S.~M., {Lupton}, R.~H., \& {Ivezi{\'c}}, {\v Z}. 2003, \aj, 125, 1559
1046:
1047: \bibitem[{{Press} {et~al.}(1992){Press}, {Teukolsky}, {Vetterling}, \&
1048: {Flannery}}]{1992nrca.book.....P}
1049: {Press}, W.~H., {Teukolsky}, S.~A., {Vetterling}, W.~T., \& {Flannery}, B.~P.
1050: 1992, {Numerical recipes in C. The art of scientific computing} (Cambridge:
1051: University Press, |c1992, 2nd ed.)
1052:
1053: \bibitem[{{Rauscher} \& {Marcy}(2006)}]{2006PASP...118..617}
1054: {Rauscher}, E. \& {Marcy}, G.~W. 2006, \pasp, 118, 617
1055:
1056: \bibitem[{{Reid} {et~al.}(1995){Reid}, {Hawley}, \&
1057: {Gizis}}]{1995AJ....110.1838R}
1058: {Reid}, I.~N., {Hawley}, S.~L., \& {Gizis}, J.~E. 1995, \aj, 110, 1838
1059:
1060: \bibitem[{{Reid} \& {Hawley}(2005)}]{2005nlds.conf.....R}
1061: {Reid}, N., \& {Hawley}, S.~L., eds. 2005, {New light on dark stars : red
1062: dwarfs, low mass stars, brown dwarfs}, ed. N.~{Reid} \& S.~L. {Hawley}
1063:
1064: \bibitem[{{Schmidt} {et~al.}(2008)}]{2008.123.3409H}
1065: {Schmidt}, S.~J., {et~al.} 2008, \aj, in prep
1066:
1067: \bibitem[{{Schmidt} {et~al.}(2007)}]{2007AJ....133.2258S}
1068: {Schmidt}, S.~J., {Cruz}, K.~L., {Bongiorno}, B.~J., {Liebert}, J. \&
1069: {Reid}, I.~N. 2007, \aj, 133, 2258
1070:
1071: \bibitem[{{Siebert} {et~al.}(2003){Siebert}, {Bienaym{\' e}}, \&
1072: {Soubiran}}]{2003A&A...399..531S}
1073: {Siebert}, A., {Bienaym{\' e}}, O., \& {Soubiran}, C. 2003, \aap, 399, 531
1074:
1075: \bibitem[{{Silvestri} {et~al.}(2006)}]{2006AJ....131.1674S}
1076: {Silvestri}, N.~M., {et~al.} 2006, \aj, 131, 1674
1077:
1078: \bibitem[{{Skumanich}(1972)}]{1972ApJ...171..565S}
1079: {Skumanich}, A. 1972, \apj, 171, 565
1080:
1081: \bibitem[{{Smith} {et~al.}(2002)}]{2002AJ....123.2121S}
1082: {Smith}, J.~A., {et~al.} 2002, \aj, 123, 2121
1083:
1084: \bibitem[{{Smol{\v c}i{\'c}} {et~al.}(2004)}]{2004ApJ...615L.141S}
1085: {Smol{\v c}i{\'c}}, V., {et~al.} 2004, \apjl, 615, L141
1086:
1087: \bibitem[{{Soderblom} {et~al.}(1991){Soderblom}, {Duncan}, \&
1088: {Johnson}}]{1991ApJ...375..722S}
1089: {Soderblom}, D.~R., {Duncan}, D.~K., \& {Johnson}, D.~R.~H. 1991, \apj, 375,
1090: 722
1091:
1092: \bibitem[{{Stauffer} {et~al.}(1994){Stauffer}, {Liebert}, {Giampapa},
1093: {Macintosh}, {Reid}, \& {Hamilton}}]{1994AJ....108..160S}
1094: {Stauffer}, J.~R., {Liebert}, J., {Giampapa}, M., {Macintosh}, B., {Reid}, N.,
1095: \& {Hamilton}, D. 1994, \aj, 108, 160
1096:
1097: \bibitem[{{Stoughton} {et~al.}(2002)}]{2002AJ....123..485S}
1098: {Stoughton}, C., {et~al.} 2002, \aj, 123, 485
1099:
1100: \bibitem[{{Strauss} {et~al.}(2002)}]{2002AJ....124.1810S}
1101: {Strauss}, M.~A., {et~al.} 2002, \aj, 124, 1810
1102:
1103: \bibitem[{{Thompson} {et~al.}(2003){Thompson}, {Christensen-Dalsgaard},
1104: {Miesch}, \& {Toomre}}]{2003ARA&A..41..599T}
1105: {Thompson}, M.~J., {Christensen-Dalsgaard}, J., {Miesch}, M.~S., \& {Toomre},
1106: J. 2003, \araa, 41, 599
1107:
1108: \bibitem[{{Tucker} \& others.(2006)}]{2006AN....327..821T}
1109: {Tucker}, D.~L., \& others. 2006, Astronomische Nachrichten, 327, 821
1110:
1111: \bibitem[{{Walkowicz} {et~al.}(2004){Walkowicz}, {Hawley}, \&
1112: {West}}]{2004PASP..116.1105W}
1113: {Walkowicz}, L.~M., {Hawley}, S.~L., \& {West}, A.~A. 2004, \pasp, 116, 1105
1114:
1115: \bibitem[{{Weis} \& {Upgren}(1995)}]{1995AJ....109..812W}
1116: {Weis}, E.~W., \& {Upgren}, A.~R. 1995, \aj, 109, 812
1117:
1118:
1119: \bibitem[{{West} {et~al.}(2006){West}, {Bochanski}, {Hawley}, {Cruz}, {Covey},
1120: {Silvestri}, {Reid}, \& {Liebert}}]{2006AJ....132.2507W}
1121: {West}, A.~A., {Bochanski}, J.~J., {Hawley}, S.~L., {Cruz}, K.~L., {Covey},
1122: K.~R., {Silvestri}, N.~M., {Reid}, I.~N., \& {Liebert}, J. 2006, \aj, 132,
1123: 2507
1124:
1125: \bibitem[{{West} {et~al.}(2004){West}, {Hawley}, {Walkowicz}, {Covey},
1126: {Silvestri}, {Raymond}, {Harris}, {Munn}, {McGehee}, {Ivezi{\' c}}, \&
1127: {Brinkmann}}]{2004AJ....128..426W}
1128: {West}, A.~A., {Hawley}, S.~L., {Walkowicz}, L.~M., {Covey}, K.~R.,
1129: {Silvestri}, N.~M., {Raymond}, S.~N., {Harris}, H.~C., {Munn}, J.~A.,
1130: {McGehee}, P.~M., {Ivezi{\' c}}, {\v Z}., \& {Brinkmann}, J. 2004, \aj, 128,
1131: 426
1132:
1133: \bibitem[{{West} {et~al.}(2005){West}, {Walkowicz}, \&
1134: {Hawley}}]{2005PASP..117..706W}
1135: {West}, A.~A., {Walkowicz}, L.~M., \& {Hawley}, S.~L. 2005, \pasp, 117, 706
1136:
1137: \bibitem[{{West} {et~al.}(2008)}]{2006AJ....132.}
1138: {West}, A.~A., {et~al.} 2008, \aj, in prep
1139:
1140: \bibitem[{{Wielen}(1977)}]{1977A&A....60..263W}
1141: {Wielen}, R. 1977, \aap, 60, 263
1142:
1143: \bibitem[{{Wilson} \& {Woolley}(1970)}]{1970MNRAS.148..463W}
1144: {Wilson}, O., \& {Woolley}, R. 1970, \mnras, 148, 463
1145:
1146: \bibitem[{{Woolf} \& {Wallerstein}(2006)}]{2006PASP..118..218W}
1147: {Woolf}, V.~M., \& {Wallerstein}, G. 2006, \pasp, 118, 218
1148:
1149: \bibitem[{{York} {et~al.}(2000)}]{2000AJ....120.1579Y}
1150: {York}, D.~G., {et~al.} 2000, \aj, 120, 1579
1151:
1152: \end{thebibliography}
1153:
1154:
1155: %table
1156:
1157:
1158:
1159:
1160: %\clearpage
1161:
1162: %\begin{thebibliography}
1163: %\nocite{*}
1164: %\bibliography{west}
1165:
1166: %\end{thebibliography}
1167:
1168:
1169:
1170:
1171:
1172:
1173:
1174:
1175: \end{document}
1176: