1: \documentclass[12pt,preprint]{aastex}
2:
3: \shorttitle{Late-Type Red Supergiants}
4: \shortauthors{Levesque et al}
5: %\slugcomment{DRAFT FEB 23 2007}
6:
7: \begin{document}
8:
9: \title{Late-Type Red Supergiants: Too Cool for the Magellanic Clouds?}
10:
11: \author{Emily M. Levesque\altaffilmark{1}}
12: \affil{Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive,
13: Honolulu, HI 96822}
14: \email{emsque@ifa.hawaii.edu}
15:
16: \author {Philip Massey\altaffilmark{1}}
17: \affil{Lowell Observatory, 1400 W. Mars Hill Road, Flagstaff, AZ 86001}
18: \email{Phil.Massey@lowell.edu}
19:
20: \author{K. A. G. Olsen}
21: \affil{Cerro Tololo Inter-American Observatory, National Optical Astronomy Observatory,
22: Casilla 603, La Serena, Chile}
23: \email{kolsen@noao.edu}
24:
25: \author{Bertrand Plez}
26: \affil{GRAAL, Universite Montpellier II, CNRS, 34095 Montpellier, France}
27: \email{Bertrand.Plez@graal.univ-montp2.fr}
28:
29: \altaffiltext{1}{Visiting Astronomer, Cerro Tololo Inter-American Observatory (CTIO),
30: National Optical Astronomy Observatory (NOAO), which is operated by the Association of
31: Universities for Research in Astronomy (AURA), Inc., under cooperative agreement with the
32: National Science Foundation (NSF).}
33:
34:
35: \begin{abstract}
36:
37: We have identified seven red supergiants (RSGs) in the Large Magellanic Cloud (LMC) and four
38: RSGs in the Small Magellanic Cloud (SMC), all of which have spectral types that are considerably
39: later than the average type observed in their parent galaxy.
40: Using moderate-resolution optical spectrophotometry and the MARCS stellar
41: atmosphere models, we determine their physical properties and place them on the H-R diagram
42: for comparison with the predictions of current stellar evolutionary tracks.
43: The radial velocities of these stars suggest that they are likely
44: all members of the Clouds rather than foreground dwarfs or halo giants. Their
45: locations in the H-R diagram also show us that these stars are cooler than the current evolutionary
46: tracks allow, appearing to the right of the Hayashi limit, a region in which
47: stars are no longer in hydrodynamic equilibrium. These stars exhibit considerable variability in their $V$ magnitudes, and three of these
48: stars also show changes in their effective temperatures (and spectral
49: types) on the time-scales of months. One of these stars, [M2002] SMC 055188, was caught in
50: an M4.5 I state, as late as that seen in HV 11423 at its recent extreme:
51: considerably later, and cooler, than any other supergiant in the SMC. In
52: addition, we find evidence of variable extinction due to circumstellar dust
53: and changes in the stars' luminosities,
54: also consistent with our recent findings for HV 11423 - when
55: these stars are hotter they are also dustier and more luminous.
56: We suggest that these
57: stars have unusual properties because they are in an unstable (and short-lived)
58: evolutionary phase.
59: \end{abstract}
60:
61: \keywords{stars: late-type---stars: evolution--stars: mass loss---supergiants}
62:
63:
64: \section{Introduction}
65: \label{Sec-intro}
66:
67: Red supergiants (RSGs) are a He-burning phase in the evolution of moderately high mass
68: stars (10-25$M_\odot$).
69: Until recently, the location of RSGs on the Hertzsprung-Russell diagram was at odds with predictions
70: of stellar evolutionary tracks. Levesque et al.\ (2005, hereafter Paper I) fitted the new generation
71: of MARCS atmosphere models (Gustafsson et al.\ 1975, 2003; Plez et al.\ 1992; Plez 2003) to
72: moderate-resolution optical spectrophotometry of Galactic RSGs. The physical parameters derived
73: from this work brought the stars into much better agreement with the Geneva evolutionary tracks
74: for solar metallicity (Meynet \& Maeder 2003). Subsequently we performed a similar analysis
75: (Levesque et al.\ 2006, hereafter Paper II)
76: for RSGs in the Magellanic Clouds, where the metallicity is significantly
77: lower ($Z/Z_\odot=0.5$ for the LMC, and $Z/Z_\odot$=0.2 for the SMC; see Westerlund 1997), with
78: a similar improvement seen.
79:
80: Paper~II emphasized that, on average,
81: RSGs in the Magellanic Clouds were not as cool as Galactic RSGs,
82: in accordance with the shifting of the right-most extension (the Hayashi limit, as described in Hayashi \& Hoshi 1961) of the evolutionary tracks
83: to warmer effective temperatures at lower metallicities. Stars in this region are fully convective,
84: and cooler stars would not be in hydrostatic equilibrium. This fact is reflected in the
85: shifting of the observed average spectral subtypes of RSGs in these
86: galaxies, from M2~I in the Milky Way to M1~I in the LMC and
87: K5-7~I in the SMC (Massey \& Olsen 2003), in accordance with the observation and explanation originally
88: put forth by Elias et al.\ (1985).
89:
90: However, Massey \& Olsen (2003) did find a few SMC and LMC RSGs that were
91: considerably later in type than average. Radial velocities (Massey \& Olsen 2003) of these
92: spectral outliers
93: had suggested these could not be foreground Galactic dwarfs, but could
94: not rule out the possibility that they were Galactic halo giants, as their
95: radial velocities would be similar to confirmed SMC and LMC
96: members. Were these Magellanic Cloud members? Or did they present a challenge to evolutionary
97: theory? We identified additional late-type RSG candidates from our
98: own previously unpublished spectroscopy (from late 2004, described below) and/or
99: broad-band colors ($V-K$, primarily) and decided to re-investigate their membership
100: and physical properties.
101:
102: Here we present moderate-resolution spectrophotometry of seven late-type RSGs in the LMC
103: and four late-type RSGs in the SMC (\S~\ref{Sec-obs}). In analyzing the
104: data (\S~\ref{Sec-analysis}), we begin by first considering the question of whether
105: or not these stars are members
106: in the Clouds (\S~\ref{Sec-RVs}). Next, we determine their spectral types and
107: physical properties
108: (\S\ref{Sec-spec}) from our spectrophotometry.
109: As a ``reality check", we also derive the physical properties from broad-band
110: $V-K$ photometry (\S\ref{Sec-bb}). In order to better understand the nature of
111: these objects, we also examine their photometric and spectral
112: variability (\S\ref{Sec-var}).
113: In \S\ref{Sec-results} we place these stars on the HR diagram, and also revise our
114: metallicity-dependent effective
115: temperature calibration of RSGs based on the new data. In \S\ref{Sec-conclude} we summarize our findings and discuss our future work on the subject.
116:
117: \section{Observations and Analysis}
118: \subsection{Observations}
119: \label{Sec-obs}
120:
121: We list our sample of late-type Magellanic Cloud RSGs in Table~\ref{tab:stars}.
122: Observations were made on 20-23 December 2005 using the RC
123: Spectrograph on the 4-m
124: Blanco telescope. The detector was a Loral CCD (3000 x 1000 pixel).
125: We used a 316 line mm$^{-1}$
126: grating (KPGL2) in first order, which gave us 2\AA\ pixel$^{-1}$. A 225 $\mu$m
127: (1.5") slit was used, yielding a spectral resolution of 7.5\AA\ (3.8 pixels). We
128: observed with two different grating tilts, one covering 3400-6200\AA, and one
129: covering 5300-9000\AA. For the blue observations we used a BG-39
130: blocking filter to eliminate any possibility of scattered red light affecting our
131: observations, particularly in the far blue and near-UV, where the stellar flux is
132: small; in the red, we used a WG-495 filter to block out second-order blue light.
133: Exposure times ranged from 200-900 seconds in the blue, and 150-600 seconds in the red.
134:
135: Observations were made at the parallactic angle, to ensure good flux calibration,
136: and numerous spectrophotometric standard stars (chosen from the
137: list of Hamuy et al.\ 1992) were also observed. We observed several ``featureless"
138: stars in order to remove telluric absorption, following Bessell (1990). The reduced blue and red
139: spectra were combined in order to bring the flux levels into agreement.
140:
141: For six of our stars (SMC 046662, SMC 055188, LMC 143035, LMC 150020, LMC 158646, and LMC 170452)
142: we also had full or partial spectrophotometry obtained during November and December 2004 with the
143: same instrument. (The details of that instrumental setup are given in Paper II.) For SMC 046662
144: and SMC 055188 the data are complete from 4100\AA\ to 9100\AA. For the other four stars
145: observed in 2004, we have
146: only partial data in the ``blue'' (4100-6450\AA). Although we could not use the latter for
147: determining physical properties (as the baseline was insufficient for accurate extinction
148: determinations), these data nevertheless proved very useful, as it provided a means for
149: checking for spectral variability and determining the relevant timescales. For one of our stars
150: (SMC 083593) we had incomplete data during both observing runs: the blue was observed in 2004,
151: and the red was obtained during this paper's primary observing run in 2005. The slopes and line depths
152: agreed well in the region of overlap, and so we combined the two halves for our analysis of the spectra.
153:
154: \section{Analysis}
155: \label{Sec-analysis}
156: We classify all of the stars in this sample (from the 2005 data and, when available, the 2004 data), based
157: primarily on the TiO band depths. While all
158: of the TiO bands are examined, the
159: TiO bands at $\lambda \lambda$ 5847, 6158, and 6658 all become
160: distinctively and progressively stronger at later spectral types, and
161: as a result are the primary features considered in our classification. For most of the 2004 observations
162: our data extend only up to 6450\AA, and so we made do with the first two bands.
163: Our method of classification was consistent with that used in Papers I and II.
164:
165: \subsection{Membership in the Clouds}
166: \label{Sec-RVs}
167: A magnitude- and color-selected sample of RSG candidates will be contaminated by foreground Galactic
168: dwarfs and potentially by more distant halo dwarfs and giants.
169: Of these, only the foreground dwarfs can be easily recognized on the basis
170: of radial velocities, but these are by far the major contaminant in studying RSGs
171: in Local Group galaxies (see, for instance, Massey 1998). Massey \&
172: Olsen (2003) obtained precision radial velocities, with the CTIO 4-m and Hydra
173: fiber positioner feeding a bench-mounted
174: spectrograph, of a sample of red stars seen towards the Clouds.
175: Indeed most of the stars in their sample had radial velocities consistent with those
176: of the Magellanic Clouds, plus a small fraction (11\% for the SMC, and 5.3\% for the LMC) which
177: had much smaller radial velocities. The latter are readily identified as foreground dwarfs,
178: while the former are tentatively identified as
179: {\it bona fide} RSGs. However, since most
180: of the SMC's and LMC's apparent radial velocity is simply a reflection of
181: the sun's motion, the sample of RSGs could be contaminated by red stars
182: in the Milky Way's halo. In Paper~II we estimated that this would be a few percent or
183: less, but here we reconsider the issue.
184:
185: The sample of stars in Paper~II and here mostly have $12<V<14$, with a few fainter stars.
186: Their $B-V$ colors are greater than 1.6 (see Table 1 of Paper~II and Table 1 of the present
187: paper). According to an updated version of the Bahcall \& Soniera (1980) model, kindly
188: provided by Heather Morrison, we expect a surface density of halo stars (all giants)
189: of about $0.2\pm0.15$ deg$^{-2}$ in this magnitude/color range towards either the LMC
190: or the SMC, where the uncertainty reflects
191: the effects of different assumptions. The area of the Massey (2002) survey was 14.5 deg$^2$
192: towards the LMC, and 7.2 deg$^2$ towards the SMC, so we might expect $2.9\pm2.1$ halo
193: giants (0.6\% of the bright and red stars) seen towards the LMC, and $1.4\pm1.1$ (0.9\%)
194: seen towards the SMC in his catalog. Thus we expect only a fraction of a star in the entire
195: spectroscopic sample of 84 stars (73 discussed in Paper~II and 11 discussed here).
196: Of the 11 stars in Table~1, 1\% (0.1 stars) is probably a large overestimate, as the vast majority
197: of these few halo contaminates have $B-V<1.8$, while all but two of the stars in Table~1 have
198: $B-V>1.8$.
199:
200: Finally, we can use a more exacting test of the kinematics than the simple
201: radial velocity cutoff used by Massey \& Olsen (2003), and
202: ask if the LMC stars discussed here follow the radial velocities of other RSGs as a function of
203: spatial position in the LMC.
204: The kinematics of the SMC are quite complex, and so we restrict the argument here to the LMC, where the kinematics
205: are relatively well understood (Olsen \& Massey 2007). Recall that differences in the radial velocities
206: within the LMC are dominated by the transverse motion of the LMC coupled to
207: the change in position factor due to the LMC's large angular extent, and that other effects,
208: such as rotation, are a relatively minor perturbation (see, for example, Meatheringham et al. 1988, Schommer et al. 1992, and van der Marel et al. 2002). Fig.~\ref{fig:kinematics} shows a histogram of LMC RSG velocities analyzed in Olsen \& Massey (2007). The LMC RSGs discussed here follow the kinematics
209: of the galaxy, something we do not expect to be true of halo giants. We conclude that the sample we discuss here
210: is unlikely to contain foreground objects.
211:
212:
213: \subsection{Modeling the Spectrophotometry}
214: \label{Sec-spec}
215:
216: The observed spectral energy distributions of the sample stars were compared to MARCS
217: stellar atmosphere models of metallicity $Z/Z_{\odot}=0.2$ for the SMC and $Z/Z_{\odot}=0.5$ for the LMC.
218: The models ranged from 3000 to 4500 K in increments of 100 K, and were interpolated for
219: intermediate temperatures at 25 K increments. The log $g$ values for the models ranged from
220: $-1$ to $+1$ in increments of 0.5 dex.
221:
222: When fitting the data, we reddened the models using a Cardelli et al.\ (1989) reddening
223: law with $R_V = 3.1$. The reddening and $T_{\rm eff}$ for each object was determined
224: by finding the best by-eye fit to both the spectral features and continuum, initially using
225: a model with log $g = 0.0$. For these later-type
226: stars with distinct TiO features, our precision was about 50 K, while for the earlier
227: K-type stars from Paper I, our precision was approximately 100 K. The extinction values $A_V$ are
228: determined to 0.15 mag. Upon fitting the reddening and $T_{\rm eff}$, we then examined the appropriate
229: log $g$ value for the star: the bolometric corrections from the models were used with the reddening
230: and photometry to compute the bolometric luminosity, assuming distance moduli of 18.9 for the SMC
231: and 18.5 for the LMC (van den Bergh 2000). From the bolometric luminosity and $T_{\rm eff}$, the physical log g was determined.
232: If the resulting value indicated that a model with log $g = -0.5$ or $+0.5$ would be more
233: appopriate than the initial estimate of log $g = 0.0$, the star was refitted with the more
234: appropriate surface gravity value and the process was repeated. A changed log $g$ value did not
235: affect the $T_{\rm eff}$ determination, but did slightly affect the extinction value. Typically this
236: process converged upon a satisfactory log $g$ choice for the model after two or three iterations.
237:
238: Our fits are shown in Figure~\ref{fig:fits}. For the purposes of scaling, we have truncated the spectra to 4000\AA. In general the models show excellent agreement with our
239: observed spectral energy distributions. The excess flux in the near-UV region of the spectra is
240: likely due to circumstellar dust, scattering the light from the star into the line of sight; see Massey
241: et al.\ (2005). The synthetic spectra has to be smoothed to match the
242: resolution of the data. While a small mismatch can lead to discrepancies
243: in the comparison of atomic lines (such as the superimposing of the Mg I triplet near 5200\AA, a feature only
244: partially resolved in our spectra, onto TiO $\lambda$5167), the details of
245: the smoothing are unimportant for comparison of
246: the broader TiO molecular bands.
247:
248: Since these stars are considerably variable in $V$ (\S~\ref{Sec-var}) we chose to
249: use our spectrophotometry to provide a contemporaneous measurement.
250: We obtained these values from our spectrophotometry following the
251: procedures of Bessell et al.\ (1998). We also computed contemporary $B-V$
252: values the same way.
253: These values are listed in Table~\ref{tab:stars}. This then allows the bolometric luminosity
254: we derive to be directly related to the reddening and effective temperatures we
255: derive, both of which may vary with time (Massey et al.\ 2007). To derive
256: $M_{\rm bol}$, we first calculated $M_V$ through the simple equation
257: $$M_V = V - A_V - {\rm DM}$$
258: assuming true distance moduli (DM) of 18.9 for the SMC and 18.5 for the LMC. The bolometric correction
259: $BC_V$ was calculated as a function of effective temperature based on fits made to the MARCS models;
260: the equations are given in Section 3.2 of Paper~II.
261: With these quantities, $M_{\rm bol}$ is then simply
262: $$M_{\rm bol} = M_V + {\rm BC}_V$$
263: From the bolometric luminosity we can derive the ratio $L/L_{\odot}$ by
264: $$(L/L_{\odot}) = 10^{(M_{\rm bol} - 4.74)/(-2.5)}$$
265: and use the luminosity-radius-temperature relation:
266: $$R/R_{\odot} = (L/L_{\odot})^{0.5} (T_{\rm eff}/5770)^{-2}$$
267: to obtain the stellar radii.
268:
269:
270: \subsection{Alternative Method Using K-band Photometry}
271: \label{Sec-bb}
272:
273: In Papers I and II we found it was very useful to have some (nearly) independent
274: check on our results by using the existing $K$-band photometry. In general,
275: the variability at $K$ is less than that at $V$ (Josselin et al.\ 2000); we find
276: below (\S~\ref{Sec-var}) that this is true for these stars as well.
277:
278: In order to derive (nearly)
279: independent values of $T_{\rm eff}$ and $M_{\rm bol}$ from the K-band
280: photometry, we dereddened the $V-K$ values from Table~\ref{tab:stars} using
281: $(V-K)_0 = V - K - 0.88A_V$, where the numerical value was derived in Paper II
282: (and is in agreement with that of Schlegel et al.\ 1998), and the values for
283: $A_V$ are taken from Table~\ref{tab:results}.
284:
285: We derive the effective temperatures from $(V-K)_0$ using relations
286: derived from the models and given in Section 3.3.1 of Paper~II.
287:
288: To derive the bolometric luminosity using these new temperatures,
289: we first calculated $M_K$ by:
290: $$M_K = K - A_K -{\rm DM}$$
291: where $A_K = A_V \times 0.12$. We then derived the
292: bolometric corrections to the K-band (BC$_K$) using
293: the K-band effective temperatures and the relations
294: given in Paper~II.
295: The bolometric luminosity is then simply:
296: $$M_{\rm bol} = M_K + {\rm BC}_K$$
297: and the stellar radius follows as shown in \S~\ref{Sec-spec}. The values
298: for BC$_K$ came from adopting the $(V-K)_0$ temperatures and
299: using a relation derived from the MARCS models in Paper II.
300:
301: We include these values in Table~\ref{tab:results}, and they show good general
302: agreement with the values obtained from spectrophotometry. Systematically, the
303: K-band temperatures yield a median difference of $-$200 K in temperature
304: for the 4 SMC stars and $-$96 K in temperature
305: for the 7 LMC stars, in the sense that the K-band values are larger.
306: These systematic differences are similar to those found from
307: the larger sample considered in Paper~II, where we
308: found median differences of $-$170~K and $-$105~K between the K-band photometry
309: and spectral fitting for the SMC and LMC, respectively. The median K-band
310: luminosities are lower, 0.41~mag for the SMC, and 0.18~mag for the LMC.
311: In Paper~II we attributed the differences to the inherent limitations of the 1-D
312: atmosphere models.
313:
314:
315:
316: \subsection{Variability}
317: \label{Sec-var}
318:
319: \subsubsection{Photometric Variability}
320: Many of the objects in our sample demonstrate large variability in their $V$ magnitudes. Photometry was obtained from the All Sky Automated Survey (ASAS) project (Pojmanski 2002).
321: To this, we added some additional data of our own. First, we used the CCD photometry
322: of Massey (2002) and obtained individual measurements for each of our stars,
323: rather than the averages given in that paper. These new values are given
324: in Table~\ref{tab:phot}. Included as well are the values we derive from our
325: spectrophotometry, including that of the 2004 observing run described in
326: Paper II. We also list the extinction $A_V$ we derive from our spectral
327: fitting, in order to supplement the limited color information available. The photometry derived from the spectrophotometry
328: was derived using the $V$ band curve of Bessell (1990) and the zero-points
329: given by Bessell et al.\ (1998).
330:
331: The light curves are given in
332: Fig.~\ref{fig:phot} for nine of our eleven objects
333: (SMC 055188 and LMC 170452 lacked sufficient ASAS data). While normal RSGs
334: are known to be variable in $V$ (Josselin et al.\ 2001), each of the late-type
335: stars described here show larger variations. The maximum changes we see
336: for the stars discussed here average to a $\Delta V$ of 1.3 for the SMC sample
337: and 1.6 for the LMC sample. For comparison, the ASAS photometry of seventy
338: SMC and LMC RSGs studied in Paper II average to a $\Delta V$ of 0.9 mag.
339: Many of the late-type RSG lightcurves also seem to suggest some level of
340: quasi-periodicity.
341:
342: While SMC 055188 and LMC 170452 are not included in Fig.~\ref{fig:phot}, our
343: observations in Table~\ref{tab:phot} show that they are also quite variable in
344: $V$. In addition, NOMAD (Zacharias et al.\ 2004) gives $V$ magnitudes of 13.82
345: and 14.91 for these stars, respectively. So, it appears that these stars show
346: $V$-band variability of 1.2-1.3 mag.
347:
348: In contrast, we expected that the variability at $K$ would be less; at least, this is typical
349: of other RSGs (Josselin et al.\ 2000). We checked the 2MASS values
350: against DENIS (Kimeswenger et al.\ 2004),
351: and found much lower
352: variability, with an average difference at $K$ between the two surveys
353: of 0.08 mag.
354:
355: Finally, we note that for four of the stars,
356: the $B-V$ photometry we derive from our (2005) spectrophotometry
357: differs significantly
358: ($>0.1$ mag) from the 1999/2000 values of Massey (2002). We list both
359: the new and old values
360: in Table~\ref{tab:phot}. In general the agreement is good, but for
361: four of our stars (SMC 083593, LMC 148035,
362: LMC 162635, and LMC 170452) we see differences of several tenths in
363: $B-V$. As discussed in our
364: study of HV 11423 (Massey et al.\ 2007) $B-V$ is not very sensitive
365: to effective temperatures for RSGs.
366: Instead we suggest that the variations we see in $B-V$ are indicative
367: of changes in the amount of circumstellar
368: dust causing differences in the reddening. (As discussed by Massey
369: et al.\ 2005, dust around
370: RSGs {\it should}, and apparently does, result in a significant
371: amount of circumstellar extinction; in the extreme
372: cases in the Milky Way, amounting to several magnitudes.) We saw
373: similar changes in $B-V$ for HV 11423,
374: and suggests episodic dust ejection on the time scale of a few years,
375: consistent with the study of Danchi et
376: al. (1994).
377:
378: Note that these changes in $B-V$ are not correlated in some simple way with
379: changes in $V$: a redder color does not necessarily mean that the star is
380: fainter. This demonstrates that the $V$-band variability is {\it not} simply
381: caused by changes in the amount of circumstellar extinction, but that
382: rather real changes in the star (such as effective temperature) are
383: responsible for the change in $V$. For HV 11423 we argued that these
384: physical changes were also triggering bursts of enhanced dust production,
385: further complicating the light curve.
386:
387: \subsubsection{Spectral Variability}
388:
389: We were intrigued by the large discrepancies in spectral subtypes assigned to some of our stars
390: by previous studies and ourselves (Table~\ref{tab:stars}). Although assigning spectral types is a
391: little subjective, we did not see many such differences in the comparison of Massey \& Olsen (2003)
392: types with ours in Paper~II. Yet, spectral variability of a type or more is virtually unknown for RSGs.
393: The notable past exception is the SMC RSG star HV~11423, recently found by Massey et al.\ (2007) to have varied several times between K0-1~I and M4.5-5~I in the past several years. At its coolest,
394: it is the latest type supergiant in the SMC. HV~11423 shows large photometric variability at $V$, due in part
395: to the substantial change in effective temperature (while holding relatively constant in bolometric luminosity), and
396: in part due to the increased circumstellar extinction, presumably from outbursts of dust formation resulting from mass loss.
397:
398: We were therefore interested to follow up the question about whether or not the stars discussed
399: here are truly variable or not in spectral type. We have included spectral
400: types for these stars in Table~\ref{tab:phot} from Massey \& Olsen (2003) and this study,
401: as well as types from our 2004 data. While in most cases we do not have
402: sufficient wavelength coverage to determine physical properties, the data
403: we do have is sufficient to determine accurate spectral types, and therefore
404: useful for evaluating the spectral variability of these stars over the past
405: few years.
406:
407: From examining the data in Table~\ref{tab:stars}, we find
408: four stars that show significant differences in their listed spectral types: SMC 046662 (M2~I to K2-3~I), SMC 055188 (M2~I to M4.5~I), LMC 148035 (M4~I to M2.5~I), and LMC 170452 (M4.5-5~I to M1.5~I). We compare
409: their October 2001 spectra (from Massey \& Olsen 2003) with those from late
410: 2004 and December 2005
411: in Fig.~\ref{fig:changes}. We see that indeed the spectral
412: changes for SMC 046662, SMC 055188, and LMC 170452 are real, and that the line strengths have indeed changed dramatically throughout the past few
413: observations. In the case of LMC 148035 the differences are much less
414: significant, and Massey \& Olsen (2003) have assigned too late a spectral type.
415:
416: In the case of HV 11423 (Massey et al.\ 2007) we
417: were able to combine the changes in the photometric and
418: spectrophotometric properties of the star
419: to present a picture of the physical changes that were taking place
420: within the star. HV 11423 varied
421: in effective temperature from 4300 to 3500 K on a time-scale of
422: months. When the star is as cool as it
423: gets, it has a very late spectral type, M4.5~I or so, much later than
424: other supergiants that were known in the SMC (prior to this study),
425: and far beyond the region where the star is stable hydrodynamically.
426: This 800 K change in effective temperature was reflected in the
427: star's changing $V$ and luminosity: when the star was hot, it was
428: also brighter and slightly more luminous, with the differences
429: amounting to -0.6 mag in both cases. Although the
430: differences in $V$ and $M_{\rm bol}$ are the same, this is coincidental,
431: as the absolute {\it visual} magnitude $M_V$ changed by -1.9 mag. The
432: change in $V$ was smaller because our analysis also showed that the amount of
433: visual extinction also changed by 1.3 mag, due presumably to additional
434: circumstellar dust that forms when the star is cool. Of course, with only two or three epochs of
435: observations it is difficult to sort out what
436: changes sporadically rather than systematically.
437:
438: Still, a very similar picture seems to emerge here. For SMC 046662,
439: SMC 055188, and LMC 170452 we see changes in
440: the spectral types and effective temperatures on the time scale of a
441: year, albeit by lesser amounts. When these stars are hottest they are also
442: at their brightest. For SMC 046662 and
443: SMC 055188 we can also conclude that when
444: the stars are hottest they are also more luminous (Table~\ref{tab:results}). The difference that we observed were smaller, with the changes in
445: effective temperature and bolometric luminosity amounting to
446: 125 K and -0.2 mag, respectively, for SMC 046662. At the extremes we observed, the changes for SMC 055188 amounted to 200 K and -0.5 mag. Futhermore, we can estimate
447: the amount of extinction for these two stars, and find $A_V$ is larger when the star is
448: hotter. All of this behavior is remarkably similar to what we see in HV 11423.
449:
450: \section{Results}
451: \label{Sec-results}
452:
453: \subsection{Placement on the H-R Diagram}
454: In Figure~\ref{fig:HRD} we place our LMC and SMC sample stars on stellar evolutionary tracks of the
455: appropriate metallicity. It is clear that stellar evolutionary theory is not in agreement with
456: the observed parameters of these
457: late-type stars. Specifically, the evolutionary tracks do not extend to cool enough temperatures
458: to accomodate these stars in their current states. The location of the RSGs as derived from
459: spectral fitting is, on average, 275 K cooler than the tracks allow for the LMC and 541 K cooler
460: for the SMC. The agreement is better for
461: the physical properties derived from $V-K$ photometry, although even here the
462: tracks do not extend to cool enough temperatures - the RSG locations are 205 K too cool for the LMC
463: and 216 K too cool for the SMC. However, this improved agreement is largely
464: due to the fact that, as discussed in (\S\ref{Sec-bb}), the temperatures derived from $V-K$
465: photometry are
466: also generally warmer than those derived from spectral fitting. The foreshadowing of this result
467: can be seen in Figure 8 of
468: Paper II, where while agreement for most SMC and LMC RSGs is good, disagreement with
469: evolutionary theory is visible for the coolest SMC RSGs. It appears, therefore,
470: that the location of these evolutionary tracks does not accomodate
471: the full range of RSG properties in this low-metallicity environment.
472:
473: The discrepancy appears slightly worse for the SMC than the LMC, particularly in the case of the
474: HRD positions derived from spectral fitting. Recall that
475: at low metallicity rotation plays an enhanced effect on the luminosities of the evolutionary
476: tracks (Maeder \& Meynet 2001) due to the effects of mixing.
477: Still, the current evolutionary models do not show much of a difference with the location in
478: temperature due to rotation, as is evident by comparing the black (no-rotation) and red
479: (high rotation) tracks in Fig.~\ref{fig:HRD}. For the time being, this increased discrepancy seen
480: in the SMC remains unexplained.
481:
482: \subsection{Revisions to the Effective Temperature Scale}
483: In Paper~II we compared the effective temperatures of stars of the same spectral subtype
484: in the SMC, LMC, and Milky Way. Because the metallicity is less in the SMC than in the LMC
485: or Milky Way, we expect that a given band strength of TiO (the basis for the spectral classification)
486: will require a cooler temperature in the SMC than in the LMC or Milky Way, and indeed we
487: found such a progression: an M1 star would have an effective temperature of 3625 K in the SMC,
488: 3695 in the LMC, and 3745 in the Milky Way. Put another way, stars with an effective temperature
489: of 3550 K would have TiO band strengths corresponding (roughly) to an M1.5~I in the SMC,
490: M2.5~I in the LMC, and M3.5~I in the Milky Way.
491:
492: With the present data we can improve the comparison for the latest types. In Table~\ref{tab:latetscale}
493: we update Table 4 of Paper~II.
494: Fig.~\ref{fig:TeffScale} shows updated effective temperature scales for the LMC and SMC with these new late-type members
495: included. We have also included our results on HV 11423 (Massey et al.\ 2007). The error bars at the upper right show our estimate of the uncertainty when measuring the temperature of a single
496: star - 100~K for the early-K type stars and 50~K for the later-type RSGs, as described in \S\ref{Sec-spec}. At M0-M2 we find that
497: a star in the LMC is about 50~K cooler than in the Milky Way, while a star in the SMC in
498: about 130~K cooler than in the Milky Way.
499:
500:
501:
502:
503: \section{Discussion}
504: \label{Sec-conclude}
505:
506: We have determined the physical properties of a sample of seven late-type RSGs
507: in the LMC and 4 late-type RSGs in the SMC; one additional star with extreme properties, HV 11423,
508: has been
509: studied separately (Massey et al.\ 2007). We argue that these stars are likely all
510: members of the Magellanic Clouds and not foreground objects, and have found
511: that these stars possess photometric variability at $V$ that is larger than RSGs of
512: earlier spectral types. Although four of our stars show significant
513: variability in $B-V$, suggesting changes in the amount of circumstellar
514: reddening, the variable V-band magnitudes are not correlated with the color
515: changes and therefore must be due to physical changes in the star itself.
516: Consistent with this, three of these stars - SMC 046662, SMC 055188, and LMC 170452 - have demonstrated spectral
517: variability of several subtypes. Other than the SMC star
518: HV 11423, this behavior is unknown for
519: RSGs. These late-type RSGs are significantly cooler than the evolutionary tracks allow,
520: with the discrepancy larger for the SMC than the LMC. Naively
521: we would argue that for the most part these stars are in the Hayashi
522: forbidden zone of the HRD, which is also true of HV 11423 when it is in its
523: cool state.
524:
525:
526: The extinction observed around most of these stars (Tables~\ref{tab:results} and~\ref{tab:phot}) is higher than what is typically seen for OB stars in the
527: Clouds, for which $A_V = 0.28$ (SMC), and $A_V = 0.40$ (LMC), where the values
528: come from Massey et al. (1995). This is similar to what was found by Massey
529: et al. (2005) for Galactic RSGs. Both HV~11423 and SMC 055188 are among the only four known
530: SMC RSGs that are IRAS sources (Massey et al.\ 2007), indicating that
531: we are seeing thermal emission from circumstellar dust. In the case of
532: HV~11423, we found evidence of a variable amount of visual extinction, which
533: we argued was connected with the sporadic production of dust, and we now
534: find similar evidence for sporadic dust production when comparing the 2004
535: and 2005 results of model fitting for SMC 055188 and SMC 046662.
536: Like HV~11423, SMC 055188 probably produces dust quite sporadically:
537: despite its presence in the IRAS source catalog, the star was not in the Midcourse
538: Space Experiment ({\it MSX}) 10$\mu$m flux MSXC6 catalog (Egan et al.\ 2003)
539: despite the fact
540: that MSX was considerably more sensitive than IRAS at this wavelength.
541: Similar sporadic dust production may be true of other stars in this sample,
542: given that we find that four of our stars, (SMC 083593, LMC 148035, LMC 162635, and LMC
543: 170452) show a change of several tenths of a magnitude in $B-V$ colors.
544: This could account for some of the $V$-band variability
545: we observe, but clearly not all, as in some cases $V$ has gotten larger while
546: $B-V$ has gotten smaller (e.g., SMC 083593, LMC 162635), or stayed the same
547: despite changes in $B-V$ (LMC 170452). Thus physical changes in the star are
548: primarily responsible for the variability in $V$, although changes in the
549: circumstellar extinction (as evidence by the $B-V$ changes) probably
550: complicated the light curves as well.
551:
552: Most interestingly, HV~11423, originally thought to be a unique and
553: extreme case, has now been joined by three
554: fellow RSGs exhibiting similar behavior: cool stars, inhabiting the
555: Hayashi forbidden zone, that show large
556: variability in spectral type, $V$ magnitudes, and extinction, presumed to
557: be from circumstellar dust. These stars suffer changes in effective
558: temperature and bolometric luminosity on timescales of months; when
559: they are at their hottest, they are also brighter, dustier, and
560: more luminous. As
561: described above, one would expect stars in the Hayashi forbidden region to be
562: unstable hydrodynamically, which we expect to lead to this variability
563: and behavior. Further
564: monitoring of these stars, both photometrically and spectroscopically, may
565: lead to an improved understanding of this phase of massive star evolution.
566:
567:
568:
569: \acknowledgements
570: We are grateful to CTIO for the hospitality and assistance provided during our observations. Constructive comments from the referee, Chris Evans, improved the
571: clarity and presentation of this manuscript. This
572: paper made use of data from the Two Micron All Sky Survey, which is a joint project of the University
573: of Massachusetts and the Infrared Processes and Analysis Center/California Institute of Technology,
574: funded by the National Aeronautics and Space Administration and the National Science Foundation. This
575: work was supported by the National Science Foundation through AST-0604569 to PM.
576:
577:
578: \begin{references}
579:
580: \reference {} Bahcall, J. N., \& Soneira, R. M. 1980, ApJS, 44, 73
581:
582: \reference {} Bessell, M. S. 1990, PASP, 102, 1181
583:
584: \reference {} Bessell, M. S., \& Brett, J. M. 1988, PASP, 100, 1134
585:
586: \reference {} Bessell, M. S., Castelli, F., \& Plez, B. 1998, A\&A, 333, 231
587:
588: \reference {} Cardelli, J. A., Clayton, G. C., \& Mathis, J. S. 1989, ApJ, 345, 245
589:
590: \reference {} Carpenter, J. M. 2001, AJ, 121, 2851
591:
592: \reference {} Charbonnel, C., Meynet, G., Maeder, A., Schaller, G., \& Schaerer, D. 1993,
593: A\&AS, 101, 415
594:
595: \reference {} Danchi, W. C., Bester, M., Degiacomi, C. G., Greenhill, L. J., \& Townes, C. H. 1994, AJ, 107, 1469
596:
597: \reference {} Egan, M. P. et al.\ 2003, The Midcourse Space Experiment Point Source Catalog
598: Version 2.3, Air Force Research Laboratory Technical Report AFRL-VS-TR-2003-1589
599:
600: \reference {} Elias, J. H., Frogel, J. A., \& Humphreys, R. M. 1985, ApJS, 57, 91
601:
602: \reference {} Gustafsson, B., Bell, R. A., Eriksson, K., \& Nordlund, \AA 1975, A\&A, 42, 407
603:
604: \reference {} Gustafsson, B., Edvardsson, B., Eriksson, K., Mizumo-Wiedner, M., Jorgensen,
605: U. G., \& Plez, B. 2003, in ASP Conf.\ ser.\ 288, Stellar Atmosphere
606: Modeling, ed. I. Hubeny, D., Mihalas, \& K. Werner (San Francisco: ASP), 331
607:
608: \reference {} Hamuy, M., Walder, A. R., Suntzeff, N. B., Gigoux, P., Heathcote, S. R., \& Phillips, M. M. 1992,
609: PASP, 104, 553
610:
611: \reference {} Hayashi, C., \& Hoshi, R. 1961, PASJ, 13, 552.
612:
613: \reference {} Josselin, E., Blommaert, J. A. D. L., Groenewegen, M. A. T., Omont, A., \& Li, F. L. 2000,
614: A\&A, 357, 225
615:
616: \reference{} Kimeswenger, S., et al.\ 2004, A\&A, 413, 1037
617:
618: \reference {} Levesque, E. M., Massey, P., Olsen, K. A. G., Plez, B.,
619: Josselin, E., Maeder, A., \& Meynet, G. 2005, ApJ, 628, 973 (Paper~I)
620:
621: \reference {} Levesque, E. M., Massey, P., Olsen, K. A. G., Plez, B., Meynet, G., \& Maeder, A. 2006,
622: ApJ, 645, 1102 (Paper~II)
623:
624:
625:
626: \reference {} Maeder, A., \& Meynet, G. 2001, A\&A, 373, 555
627:
628: \reference {} Massey, P. 1998, ApJ, 501, 153
629:
630: \reference {} Massey, P. 2002, ApJS, 141, 81
631:
632: \reference {} Massey, P., Lang, C. C., DeGioia-Eastwood, K., \& Garmany, C. D. 1995, ApJ, 438, 188
633:
634: \reference {} Massey, P., Levesque, E. M., Olsen, K. A. G., Plez, B., \& Skiff, B. A., 2007, ApJ, in press
635:
636: \reference {} Massey, P., \& Olsen, K. A. G. 2003, AJ 126, 2867
637:
638: \reference {} Massey, P., Plez, B., Levesque, E. M., Olsen, K. A. G., Clayton, G. C., \& Josselin, E. 2005,
639: ApJ, 634, 1286
640:
641: \reference {} Meatheringham, S. J, Dopita, M. A., Ford, H. C., \& Webster, B. L. 1988, ApJ, 327, 651
642:
643: \reference {} Meynet, G., \& Maeder, A. 2003, \&A, 404, 975
644:
645: \reference {} Meynet, G., \& Maeder, A. 2005, A\&A, 429, 581
646:
647: \reference {} Olsen, K. A. G., \& Massey, P. 2007, ApJ, 656, L61
648:
649: \reference {} Plez, B. 2003, in ASP Conf.\ Ser. \ 298, {\it GAIA} Spectroscopy: Science and
650: Technology, ed. U. Munari (San Francisco: ASP), 189
651:
652: \reference {} Plez, B., Brett, J.M., \& Nordlund, \AA. 1992, A\&A, 256, 551
653:
654: \reference {} Pojmanski, G. 2002, Acta Astronomica, 52, 397
655:
656: \reference {} Schaerer, D., Meynet, G., Maeder, A., \& Schaller, G. 1993, A\&AS, 98, 523
657:
658: \reference {} Schlegel, D. J., Finkbeiner, D. P., \& Davis, M. 1998, ApJ, 500, 525
659:
660: \reference {} Schommer, R. A., Suntzeff, N. B., Olszewski, E. W., \& Harris, H. C. 1992, AJ, 103, 447
661:
662: \reference {} Turnshek, D. E., Turnshek, D. A., Craine, E. R., \& Boeshaar, P. C. 1985,
663: An Atlas of Digital Spectra of Cool Stars (Tucson, Western Research Company)
664:
665: \reference {} van den Bergh, S. 2000, The Galaxies of the Local Group (Cambridge: Cambridge Univ. Press)
666:
667: \reference {} van der Marel, R. P., Alves, D. R., Hardy, E., \& Suntzeff, N. B. 2002, AJ, 124, 2639
668:
669: \reference {} Westerlund, B. E. 1997, The Magellanic Clouds (Cambridge: Cambridge Univ. Press)
670:
671: \reference {} Zacharias, N., Monet, D. G., Levine, S. E., Urban, S. E., Gaume, R., \& Wycoff, G. L. 2004, AAS,
672: 205.4815
673:
674:
675:
676:
677: \end{references}
678:
679: \clearpage
680:
681: \begin{figure}
682: \epsscale{1.0}
683: \plotone{f1.eps}
684: \caption{\label{fig:kinematics}Kinematics of the LMC. Figure (a) shows a histogram of LMC RSG velocities analyzed in Olsen \& Massey (2007), in bins of 5 km s$^{-1}$, with our seven LMC stars overplotted as a solid histogram, showing that the velocities are consistent with those of other RSGs in the LMC. Figure (b) shows the velocity residuals of our RSGs after subtracting a kinematic model fit to the Olsen \& Massey (2007) RSG sample, with our LMC stars again plotted as a solid histogram. Standard deviation of the velocity residuals is 12 km s$^{-1}$ for the larger sample and 14 km s$^{-1}$ for our stars.}
685: \end{figure}
686:
687: %these are currently the color figures; b&w figures are named f2abw.eps, etc.
688: \clearpage
689: \begin{figure}
690: \epsscale{0.3}
691: \plotone{f2a_color.eps}
692: \plotone{f2b_color.eps}
693: \plotone{f2c_color.eps}
694: \plotone{f2d_color.eps}
695: \plotone{f2e_color.eps}
696: \plotone{f2f_color.eps}
697: \plotone{f2g_color.eps}
698: \plotone{f2h_color.eps}
699: \plotone{f2i_color.eps}
700: \plotone{f2j_color.eps}
701: \plotone{f2k_color.eps}
702: \plotone{f2l_color.eps}
703: \plotone{f2m_color.eps}
704: \caption{\label{fig:fits} Fits of the MARCS stellar atmosphere models
705: to our spectrophotometry. The data are shown as a solid black line, while
706: the reddened MARCS models are shown as a dotted gray line. In the on-line version,
707: the model fits are shown in red. Note the additional fits for the 2004
708: spectrophotometry in the cases of SMC 046662 and SMC 055188.}
709: \end{figure}
710:
711: \clearpage
712: \begin{figure}
713: \epsscale{0.3}
714: \plotone{f3a.eps}
715: \plotone{f3b.eps}
716: \plotone{f3c.eps}
717: \plotone{f3d.eps}
718: \plotone{f3e.eps}
719: \plotone{f3f.eps}
720: \plotone{f3g.eps}
721: \plotone{f3h.eps}
722: \plotone{f3i.eps}
723: \caption{\label{fig:phot} Photometric variability. We show the photometry of nine of our stars, where
724: the small points come from the ASAS photometry, and the large points come
725: from Table~\ref{tab:phot}. Very few data points were available for SMC 055188 and LMC 170452, and these stars are not included in the figure, although we argue in the
726: text that these two stars also show large photometric variability.}
727: \end{figure}
728:
729: \clearpage
730: \begin{figure}
731: \epsscale{0.4}
732: \plotone{f4a.eps}
733: \plotone{f4b.eps}
734: \plotone{f4c.eps}
735: \plotone{f4d.eps}
736: \caption{\label{fig:changes} Spectral varaibility. We show a comparison of the
737: October 2001 spectra (from Massey \& Olsen 2003) with those obtained
738: in December 2004 and December 2005 (from the present paper), where the spectra have been
739: normalized, and nebular emission has been removed from the October 2001
740: spectrum of LMC 170452 to make the comparison easier.
741: Prominent absorption features are
742: indicated, based upon the identifications given in Turnshek et al.\ (1985).
743: SMC 046662, SMC 055188, and LMC 170452 clearly show dramatic changes in the
744: absorption line intensities, indicating a change in spectral type (and hence
745: effective temperature); the changes in LMC 148035 are relatively unremarkable.
746: }
747: \end{figure}
748:
749:
750: \clearpage
751: \begin{figure}
752: \epsscale{1.2}
753: \plotone{f5.eps}
754: \vskip -250pt
755: \caption{\label{fig:HRD} Location of the late-type MC RSGs discussed here compared to the evolutionary tracks. We show the
756: location of the RSGs in the H-R diagram of the SMC (left) and LMC (right), where the effective temperatures and bolometric
757: luminosities come from fitting the MARCS models to the optical spectrophotometry (top) and from the K-band photometry
758: (bottom). The older, nonrotation evolutionary
759: tracks that include the effects of overshooting are shown in green and come
760: from Charbonnel et al.\ (1993) for the SMC, and from Schaerer et al.\ (1993) for the LMC. The newer evolutionary
761: tracks (when available) are shown in black (zero rotation) and in
762: red (300 km s$^{-1}$ initial rotation) and come from Maeder \& Meynet (2001) for the SMC, and
763: Meynet \& Maeder (2005) for the LMC. Like-colored dots are used to link the
764: 2004 and 2005 observations of the same star (purple for SMC 046662 and
765: light blue for SMC 055188).
766: }
767: \end{figure}
768:
769: \clearpage
770: \begin{figure}
771: \epsscale{0.8}
772: \plotone{f6.eps}
773: \caption{\label{fig:TeffScale} Modified full temperature scales for the
774: Milky Way (black), LMC (red), and SMC (green), incorporating the eleven stars
775: from this sample. The
776: Magellanic Cloud scales now include more stars at the later spectral types,
777: which allows us to examine the metallicity effects at lower temperatures.
778: For example, a 3550 K star may appear as an M3.5 I in the Milky Way, an M2.5 I
779: in the LMC, and an M1.5 I in the SMC.}
780: \end{figure}
781:
782: \clearpage
783: \begin{deluxetable}{l c c c c c c c l l l l c }
784: \rotate
785: \tabletypesize{\scriptsize}
786: \tablewidth{0pc}
787: \tablenum{1}
788: \tablecolumns{13}
789: \tablecaption{\label{tab:stars} Program Stars}
790: \tablehead{
791: \colhead{Star}
792: & \colhead{$\alpha_{\rm 2000}$}
793: & \colhead{$\delta_{\rm 2000}$}
794: & \multicolumn{4}{c}{Photometry\tablenotemark{a}}
795: & \colhead{}
796: & \multicolumn{4}{c}{Spectral Type}
797: & \colhead{Radial Velocity\tablenotemark{c}} \\ \cline{4-7} \cline{9-12}
798: \colhead{[M2002]}
799: &\multicolumn{2}{c}{}
800: & \colhead{$V$}
801: & \colhead{$B-V$}
802: & \colhead{$K_S$}
803: & \colhead{$V-K$}
804: & \colhead{}
805: & \colhead{EFH85\tablenotemark{b}}
806: & \colhead{MO03\tablenotemark{c}}
807: & \colhead{New (2004)}
808: & \colhead{New (2005)}
809: & \colhead{(km s$^{-1}$)}
810: }
811: \startdata
812: SMC 046662 &00 59 35.04 &-72 04 06.2 &12.75 & 1.95 &8.35 &4.36 & &M0 Ia &M2 I &K2-3 I &M0 I & 180.2\\
813: SMC 052334 &01 01 54.16 &-71 52 18.8 &12.87 & 1.91 &8.69 &4.14 & &M0 Iab &K7 I &\nodata &K5-M0 I & 165.5 \\
814: SMC 055188 &01 03 02.38 &-72 01 52.9 &14.23 & 2.22 &8.62 &5.57 & &\nodata &M2 I &M4.5 I &M3-4 I & 176.8 \\
815: SMC 083593 &01 30 33.92 &-73 18 41.9 &12.83\tablenotemark{d} & 1.60 &8.60 &4.19 & &M2 Ia &\nodata &\nodata &M2 I\tablenotemark{d} & 180.1 \\
816: LMC 143035 &05 29 03.58 &-69 06 46.3 &14.10 & 1.93 &7.90 &6.16 & &\nodata &M3-4.5 I &M4.5-5 I &M4 I & 268.3\\
817: LMC 148035 &05 30 35.61 &-68 59 23.6 &13.42 & 1.98 &7.55 &5.83 & &\nodata &M4 I &\nodata &M2.5 I & 284.5\\
818: LMC 150040 &05 31 09.35 &-67 25 55.1 &13.09 & 1.90 &7.63 &5.42 & &\nodata &M4 I &M3-4 I &M3-4 I & 280.0\\
819: LMC 158646 &05 33 52.26 &-69 11 13.2 &13.80 & 2.18 &7.90 &5.86 & &\nodata &M3-4 I &\nodata &M2 I & 288.3 \\
820: LMC 162635 &05 35 24.61 &-69 04 03.2 &14.58 & 2.20 &8.66 &5.88 & &\nodata &M1 I &\nodata &M2 I & 296.9\\
821: LMC 168757 &05 37 36.96 &-69 29 23.5 &14.62 & 1.80 &8.21 &6.37 & &\nodata &M3 I &\nodata &M3-4 I & 272.3\\
822: LMC 170452 &05 38 16.10 &-69 10 10.9 &13.99 & 2.24 &7.87 &6.08 & &\nodata &M4.5-5 I &M4.5-5 I &M1.5 I & 289.4 \\
823: \enddata
824: \tablenotetext{a}{$V$-band photometry is derived from the spectrophotometry
825: itself, while $K_S$ values are from 2MASS. The $B-V$ values are from
826: Massey (2002). The $V-K$ values have been
827: computed by transforming $K_S$ to the Bessell \& Brett 1988 ``standard" K
828: by using the Carpenter (2001) relationship $K=K_S+0.04$.}
829: \tablenotetext{b}{From Elias et al.\ 1985}
830: \tablenotetext{c}{From Massey \& Olsen 2003}
831: \tablenotetext{d}{This value was determined from the merged spectrum of our
832: 2004 and 2005 spectra.}
833: \end{deluxetable}
834:
835:
836: \begin{deluxetable}{l c c c c c c c c c c c c c}
837: \rotate
838: \tabletypesize{\scriptsize}
839: \tablewidth{0pc}
840: \tablenum{2}
841: \tablecolumns{14}
842: \tablecaption{\label{tab:results} Physical Properties}
843: \tablehead{
844: \colhead{Star} &
845: \colhead{HJD - 2,450,000} &
846: \colhead{Type} &
847: \colhead{$V$} &
848: \colhead{$M_V$} &
849: \colhead{$A_V$} &
850: \multicolumn{4}{c}{Spectral Fitting} &
851: \colhead{} &
852: \multicolumn{3}{c}{$(V-K)_0$} \\ \cline{7-10} \cline{12-14}
853: \multicolumn{6}{c}{} &
854: \colhead{$T_{\rm eff}$} &
855: \colhead{log $g$ (Model)} &
856: \colhead{$R/R_\odot$} &
857: \colhead{$M_{\rm bol}$} &
858: \colhead{} &
859: \colhead{$T_{\rm eff}$} &
860: \colhead{$R/R_\odot$} &
861: \colhead{$M_{\rm bol}$}
862: }
863: \startdata
864: SMC 046662 &3340.58 &K2-3 I &12.38 &-7.14 &0.62 &3775 &0.0 &1000 &-8.41 & &4050 &730 &-8.03 \\
865: &3725.63 &M0 I &12.75 &-6.65 &0.50 &3650 &0.0 &960 &-8.19 & &3829 &760 &-7.89 \\
866: SMC 052334 &3725.67 &K5-M0 I &12.87 &-6.34 &0.31 &3675 &0.0 &800 &-7.82 & &3851 &640 &-7.55 \\
867: SMC 055188 &3341.59 &M4.5 I &14.99 &-4.87 &0.96 &3325 &0.0 &870 &-7.57 & &3355 &800 &-7.43 \\
868: &3725.68 &M3-4 I &14.23 &-6.07 &1.40 &3450 &0.0 &1100 &-8.23 & &3667 &730 &-7.62 \\
869: SMC 083593 &\nodata\tablenotemark{a} &M2 I &12.83 &-6.16 &0.09 &3525 &0.0 &970 &-8.06 & &3751 &680 &-7.54 \\
870: LMC 143035 &3725.81 &M4 I &14.10 &-5.64 &1.24 &3450 &-0.5 &1020 &-8.06 & &3482 &870 &-7.78 \\
871: LMC 148035 &3725.81 &M2.5 I &13.42 &-6.57 &1.49 &3575 &0.0 &1130 &-8.44 & &3612 &1010 &-8.26 \\
872: LMC 150040 &3725.82 &M3-4 I &13.09 &-6.03 &0.62 &3525 &-0.5 &990 &-8.11 & &3523 &950 &-8.01 \\
873: LMC 158646 &3725.82 &M2 I &13.80 &-6.25 &1.55 &3625 &-0.5 &870 &-7.94 & &3619 &870 &-7.92 \\
874: LMC 162635 &3725.83 &M2 I &14.58 &-5.78 &1.86 &3600 &0.0 &740 &-7.56 & &3696 &610 &-7.25 \\
875: LMC 168757 &3727.84 &M3-4 I &14.62 &-4.90 &1.02 &3425 &0.0 &780 &-7.45 & &3411 &760 &-7.39 \\
876: LMC 170452 &3727.83 &M1.5 I &13.99 &-6.68 &2.17 &3625 &0.0 &1060 &-8.37 & &3720 &890 &-8.10 \\
877: \enddata
878: \tablenotetext{a}{From merged 2004 and 2005 data.}
879: \end{deluxetable}
880:
881:
882: \begin{deluxetable}{l c c c c c}
883: \tablewidth{0pc}
884: \tabletypesize{\scriptsize}
885: \tablenum{3}
886: \tablecolumns{6}
887: \tablecaption{\label{tab:phot} Time Resolved Data\tablenotemark{a}}
888: \tablehead{
889: \colhead{HJD-2,450,000}
890: &\colhead{$V$\tablenotemark{b}}
891: &\colhead{$A_V$}
892: &\colhead{$B-V$}
893: &\colhead{$T_{\rm eff}$}
894: &\colhead{Spectral Type}
895: }
896: \startdata
897: \cutinhead{[M2002] SMC 046662}
898: 1186.58 & 12.90 &\nodata & 1.88 &\nodata &\nodata \\
899: 2188.56 & 13.32 &\nodata &\nodata &\nodata &M2 I \\ %2186
900: 3340.58 & 12.38 &0.62 & \nodata &3775 & K2-3 I\\
901: 3725.63 & 12.75 &0.50 & 1.95 &3650 & M0 I\\
902: \cutinhead{[M2002] SMC 052334}
903: 1186.58 & 12.89 &\nodata & 1.94 &\nodata &\nodata \\
904: 2187.56 & 12.72 &\nodata &\nodata &\nodata &K7 I \\ %2187
905: 3725.67 & 12.87 &0.31 & 1.91 &3675 &K5-M0 I\\
906: \cutinhead{[M2002] SMC 055188}
907: 1186.58 & 14.96 &\nodata & 2.25 &\nodata &\nodata\\
908: 2188.56 &\nodata &\nodata &\nodata &\nodata &M2 I \\ %nodata
909: 3341.59 & 14.99 &0.96 & \nodata &3325 &M4.5 I\\
910: 3725.68 & 14.23 &1.40 & 2.22 &3450 &M3-4 I\\
911: \cutinhead{[M2002] SMC 083593}
912: 1186.59 & 12.64 &\nodata & 1.87 &\nodata &\nodata\\
913: 3334.60 & 12.81 &\nodata & \nodata &\nodata &\nodata\\
914: 3724.60\tablenotemark{c} & 12.83 &0.09 & 1.60 &3525 &M2 I\\
915: \cutinhead{[M2002] LMC 143035}
916: 1186.67 & 13.53 &\nodata & 1.93 &\nodata &\nodata\\
917: 1998.51 & 13.51 &\nodata & 1.93 &\nodata &\nodata\\
918: 2188.72 & 14.24 &\nodata &\nodata &\nodata &M3-4.5 I \\ %2184
919: 3334.60 & 13.99 &\nodata & \nodata &\nodata &M4.5-5 I\\
920: 3725.81 & 14.10 &1.24 & 1.93 &3450 &M4 I\\
921: \cutinhead{[M2002] LMC 148035}
922: 1186.67 & 12.94 &\nodata & 1.82 &\nodata &\nodata\\
923: 1996.53 & 13.89 &\nodata & 1.71 &\nodata &\nodata \\
924: 1998.51 & 13.87 &\nodata & 1.62 &\nodata &\nodata \\
925: 2188.72 & 12.94 &\nodata &\nodata &\nodata &M4 I \\ %2187
926: 3725.81 & 13.42 &1.49 & 1.98 &3575 &M2.5 I\\
927: \cutinhead{[M2002] LMC 150040}
928: 1186.71 & 12.86 &\nodata & 1.97 &\nodata &\nodata\\
929: 1997.54 & 12.81 &\nodata & 1.96 &\nodata &\nodata\\
930: 2188.82 & 12.93 &\nodata &\nodata &\nodata &M4 I \\ %2187
931: 3334.60 & 12.60 &\nodata & \nodata &\nodata &M3-4 I\\
932: 3725.82 & 13.09 &0.62 & 1.90 &3525 &M3-4 I\\
933: \cutinhead{[M2002] LMC 158646}
934: 1186.71 & 12.66 &\nodata & 2.19 &\nodata &\nodata\\
935: 1996.53 & 13.10 &\nodata & 2.23 &\nodata &\nodata\\
936: 2188.79 & 13.54 &\nodata &\nodata &\nodata &M3-4 I \\ %2187
937: 3725.82 & 13.80 &1.55 & 2.18 &3625 &M2 I\\
938: \cutinhead{[M2002] LMC 162635}
939: 1996.53 & 14.23 &\nodata & 2.33 &\nodata &\nodata\\
940: 2188.79 & 13.52 &\nodata &\nodata &\nodata &M1 I \\ %2186
941: 3725.83 & 14.58 &1.86 & 2.20 &3600 &M2 I \\
942: \cutinhead{[M2002] LMC 168757}
943: 1996.53 & 14.08 &\nodata & 1.77 &\nodata &\nodata\\
944: 2188.76 & 13.23 &\nodata &\nodata &\nodata &M3 I \\ %2186
945: 3727.84 & 14.62 &1.02 & 1.80 &3425 &M3-4 I \\
946: \cutinhead{[M2002] LMC 170452}
947: 1996.53 & 13.99 &\nodata & 2.39 &\nodata &\nodata\\
948: 2188.76 &\nodata &\nodata &\nodata &\nodata &M4.5-5 I \\ %nodata
949: 3341.70 & 15.31 &\nodata & \nodata &\nodata &M4.5-5 I \\
950: 3727.83 & 13.99 &2.17 & 2.24 &3625 &M1.5 I \\
951: \enddata
952: \tablenotetext{a}{Measurements from 2,451,186-2,451,999 were obtained
953: using the CCD images described in Massey (2002). The uncertainty in
954: these values are 0.01~mag or smaller. Later measurements come
955: from our 2004 November/December and 2005 December spectrophotometry.}
956: \tablenotetext{b}{The $V$ magnitudes quoted for the 2,452,188 HJD's
957: comes from ASAS observations on the closest approximate date
958: to the observations, from 2,452,184-2,452,187.}
959: \tablenotetext{c}{From merged 2004 and 2005 data.}
960: \end{deluxetable}
961:
962: %\begin{deluxetable} {l c c c}
963: %\tabletypesize{\footnotesize}
964: %\tablewidth{0pc}
965: %\tablenum{4}
966: %\tablecolumns{4}
967: %\tablecaption{\label{tab:deltas} Title?}
968: %\tablehead{
969: %\colhead{HJD} &
970: %\colhead{$V$} &
971: %\colhead{$(B-V)$} &
972: %\colhead{Spectral Type} \\
973: %}
974: %\startdata
975: %\cutinhead{SMC 046662}
976: %1187 & 12.90 &1.88 &\nodata \\
977: %2187 & ASAS &\nodata &M0~I \\
978: %3341 & 12.38 &\nodata &K5-M0~I \\
979: %3725 & 12.75 &1.95 &M2~I \\
980: %\cutinhead{LMC 170452}
981: %1997 & 13.99 &2.38 &\nodata \\
982: %2187 & ASAS &\nodata &M4.5-5~I \\
983: %3342 & 15.31 &\nodata &M4.5-5~I \\
984: %3728 & 13.99 &2.24 &M1.5~I \\
985: %\enddata
986: %\end{deluxetable}
987:
988: \begin{deluxetable} {l c c c c c c c c c c c}
989: \tabletypesize{\footnotesize}
990: \tablewidth{0pc}
991: \tablenum{4}
992: \tablecolumns{12}
993: \tablecaption{\label{tab:latetscale}Effective Temperature Scales with New Data}
994: \tablehead{
995: \colhead{} &
996: \multicolumn{3}{c}{SMC} & &
997: \multicolumn{3}{c}{LMC} &&
998: \multicolumn{3}{c}{Milky Way\tablenotemark{a}} \\ \cline{2-4} \cline{6-8} \cline{10-12}
999: \colhead{Spectral Type} &
1000: \colhead{$T_{\rm eff}$ (K)} &
1001: \colhead{$\sigma_\mu$\tablenotemark{b}} &
1002: \colhead{$N$} & &
1003: \colhead{$T_{\rm eff}$ (K)} &
1004: \colhead{$\sigma_\mu$\tablenotemark{b}} &
1005: \colhead{$N$} & &
1006: \colhead{$T_{\rm eff}$ (K)} &
1007: \colhead{$\sigma_\mu$\tablenotemark{b}} &
1008: \colhead{$N$} \\
1009: }
1010: \startdata
1011: K1-K1.5 I & 4210 & 39 & 7 && 4300 & \nodata & 1 && 4100 & 100 & 3 \\
1012: K2-K3 I & 4003 & 56 & 15 && 4050 & 62 & 3 && 4015 & 40 & 7 \\
1013: K5-M0 I & 3777 & 34 & 11 && 3850 & 18 & 2 && 3840 & 30 & 3 \\
1014: M0 I & 3629 & 16 & 6 && 3738 & 11 & 4 && 3790 & 13 & 4 \\
1015: M1 I & 3625 & \nodata & 1 && 3695 & 8 & 5 && 3745 & 17 & 7 \\
1016: M1.5 I & \nodata & \nodata & \nodata && 3650 & 13 & 7 && 3710 & 8 & 6 \\
1017: M2 I & 3500 & 18 & 2 && 3621 & 6 & 7 && 3660 & 7 & 17 \\
1018: M2.5 I & \nodata & \nodata & \nodata && 3550 & 12 & 6 && 3615 & 10 & 5 \\
1019: M3 I & \nodata & \nodata & \nodata && 3542 & 18 & 3 && 3605 & 4 & 9 \\
1020: M3.5 I & 3450 & \nodata & 1 && 3475 & 35 & 2 && 3550 & 11 & 6 \\
1021: M4-M4.5 I & 3413 & 62 & 2 && 3450 & 18 & 3 && 3535 & 8 & 6 \\
1022: M5 I & \nodata & \nodata & \nodata && \nodata & \nodata & \nodata && 3450 & \nodata & 1 \\
1023:
1024: \enddata
1025: \tablenotetext{a}{From Paper I.}
1026: \tablenotetext{b}{Standard deviation of the mean.}
1027: \end{deluxetable}
1028:
1029:
1030:
1031: \end{document}
1032:
1033:
1034: