1:
2: \documentclass[12pt,apj]{emulateapj}
3:
4: %% manuscript produces a one-column, double-spaced document:
5:
6: %\documentclass[manuscript]{aastex}
7:
8: %% preprint2 produces a double-column, single-spaced document:
9:
10: %\documentclass[preprint2]{aastex}
11:
12: %% Sometimes a paper's abstract is too long to fit on the
13: %% title page in preprint2 mode. When that is the case,
14: %% use the longabstract style option.
15:
16: %% \documentclass[preprint2,longabstract]{aastex}
17: %\documentclass[12pt,preprint]{aastex}
18:
19: %% If you want to create your own macros, you can do so
20: %% using \newcommand. Your macros should appear before
21: %% the \begin{document} command.
22: %%
23: %% If you are submitting to a journal that translates manuscripts
24: %% into SGML, you need to follow certain guidelines when preparing
25: %% your macros. See the AASTeX v5.x Author Guide
26: %% for information.
27:
28: %\newcommand{\vdag}{(v)^\dagger}
29: %\newcommand{\myemail}{skywalker@galaxy.far.far.away}
30:
31: %% You can insert a short comment on the title page using the command below.
32:
33: %\slugcomment{Not to appear in Nonlearned J., 45.}
34:
35: %% If you wish, you may supply running head information, although
36: %% this information may be modified by the editorial offices.
37: %% The left head contains a list of authors,
38: %% usually a maximum of three (otherwise use et al.). The right
39: %% head is a modified title of up to roughly 44 characters.
40: %% Running heads will not print in the manuscript style.
41:
42: %\shorttitle{Collapsed Cores in Globular Clusters}
43: %\shortauthors{Djorgovski et al.}
44:
45: %% This is the end of the preamble. Indicate the beginning of the
46: %% paper itself with \begin{document}.
47:
48: \begin{document}
49:
50: %% LaTeX will automatically break titles if they run longer than
51: %% one line. However, you may use \\ to force a line break if
52: %% you desire.
53:
54: \title{Evidence for Magnetic Flux Saturation in Rapidly Rotating M Stars}
55:
56: %% Use \author, \affil, and the \and command to format
57: %% author and affiliation information.
58: %% Note that \email has replaced the old \authoremail command
59: %% from AASTeX v4.0. You can use \email to mark an email address
60: %% anywhere in the paper, not just in the front matter.
61: %% As in the title, use \\ to force line breaks.
62:
63: \author{A. Reiners\altaffilmark{*}}
64: \affil{Institut f\"ur Astrophysik, Georg-August-Universit\"at, D-37077
65: G\"ottingen, Germany\email{Ansgar.Reiners@phys.uni-goettingen.de}}
66: \altaffiltext{*}{Emmy Noether Fellow}
67:
68: \and
69:
70: \author{G. Basri}
71: \affil{Astronomy Department, University of California, Berkeley, CA
72: 94720 \email{basri@berkeley.edu}}
73:
74: \and
75:
76: \author{M. Browning\altaffilmark{1}}
77: \affil{Astronomy Department, University of California, Berkeley, CA
78: 94720 \email{mbrowning@berkeley.edu}}
79: \altaffiltext{1}{Present address: Dept of Astronomy and Astrophysics,
80: University of Chicago, Chicago, IL 60637}
81:
82: %% Notice that each of these authors has alternate affiliations, which
83: %% are identified by the \altaffilmark after each name. Specify alternate
84: %% affiliation information with \altaffiltext, with one command per each
85: %% affiliation.
86:
87:
88: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
89: %% style, abstract will output a Received/Accepted line after the
90: %% title and affiliation information. No date will appear since the author
91: %% does not have this information. The dates will be filled in by the
92: %% editorial office after submission.
93:
94: \begin{abstract}
95: We present magnetic flux measurements in seven rapidly rotating M
96: dwarfs. Our sample stars have X-ray and H-alpha emission indicative
97: of saturated emission, i.e., emission at a high level independent of
98: rotation rate. Our measurements are made using near-infrared FeH
99: molecular spectra observed with HIRES at Keck. Because of their
100: large convective overturn times, the rotation velocity of M stars
101: with small Rossby numbers is relatively slow and does not hamper the
102: measurement of Zeeman splitting. The Rossby numbers of our sample
103: stars are as small as 0.01. All our sample stars exhibit magnetic
104: flux of kilo-Gauss strength. We find that the magnetic flux
105: saturates in the same regime as saturation of coronal and
106: chromospheric emission, at a critical Rossby number of around 0.1.
107: The filling factors of both field and emission are near unity by
108: then. We conclude that the strength of surface magnetic fields
109: remains independent of rotation rate below that; making the Rossby
110: number yet smaller by a factor of ten has little effect. These
111: saturated M-star dynamos generate an integrated magnetic flux of
112: roughly 3 kG, with a scatter of about 1 kG. The relation between
113: emission and flux also has substantial scatter.
114: \end{abstract}
115:
116:
117:
118:
119: %% Keywords should appear after the \end{abstract} command. The uncommented
120: %% example has been keyed in ApJ style. See the instructions to authors
121: %% for the journal to which you are submitting your paper to determine
122: %% what keyword punctuation is appropriate.
123:
124: %% Authors who wish to have the most important objects in their paper
125: %% linked in the electronic edition to a data center may do so in the
126: %% subject header. Objects should be in the appropriate "individual"
127: %% headers (e.g. quasars: individual, stars: individual, etc.) with the
128: %% additional provision that the total number of headers, including each
129: %% individual object, not exceed six. The \objectname{} macro, and its
130: %% alias \object{}, is used to mark each object. The macro takes the object
131: %% name as its primary argument. This name will appear in the paper
132: %% and serve as the link's anchor in the electronic edition if the name
133: %% is recognized by the data centers. The macro also takes an optional
134: %% argument in parentheses in cases where the data center identification
135: %% differs from what is to be printed in the paper.
136:
137: \keywords{stars: activity --- stars: late-type --- stars: magnetic
138: fields --- stars: rotation}
139:
140: %% From the front matter, we move on to the body of the paper.
141: %% In the first two sections, notice the use of the natbib \citep
142: %% and \citet commands to identify citations. The citations are
143: %% tied to the reference list via symbolic KEYs. The KEY corresponds
144: %% to the KEY in the \bibitem in the reference list below. We have
145: %% chosen the first three characters of the first author's name plus
146: %% the last two numeral of the year of publication as our KEY for
147: %% each reference.
148:
149: %\tableofcontents
150:
151: %\newpage
152:
153: \section{Introduction}
154:
155: Stellar magnetic activity is closely related to rotation in solar-type
156: stars. In slow rotators, activity scales with the rotation rate until
157: it becomes saturated at a certain velocity, which means that it does
158: not grow further regardless of rotation rate \citep{Noyes84,
159: Pizzolato03}. This effect is seen in a broad variety of activity
160: indicators \citep{Vilhu84, Vilhu87}. The threshold rate at which
161: saturation occurs depends on the spectral type of the star, with the
162: convective overturning time perhaps determining this threshold.
163: Saturation sets in where the value of the Rossby number $Ro =
164: P/\tau_{\rm conv} \la 0.1$, i.e. where the timescale of rotation is
165: significantly shorter than the timescale of typical convective eddies.
166:
167: Indicators of stellar activity are usually coronal or chromospheric
168: emission observed at X-ray, UV, optical, infrared, or radio
169: wavelengths. We know from the Sun that this emission is induced by
170: magnetic fields heating the upper layers of the solar atmosphere, and
171: by analogy we conclude that stellar activity is connected to magnetic
172: fields on the surface of other stars. At high rotation rates (or small
173: Rossby numbers), all activity indicators saturate, i.e. they do not
174: grow over a certain level regardless of higher rotation rates
175: \citep[see, e.g.,][]{James00}. Two possible explanations exist for the
176: saturation: 1) The stellar dynamo process saturates and stars cannot
177: produce magnetic fields stronger than the saturation value; or 2) The
178: magnetic fields continue to grow at more rapid rotation, but the
179: fraction of the surface filled with fields -- or the area covered by
180: spots -- reaches unity so that no more emitting plasma can be placed
181: on the star. The only way to decide which way the stars go is to
182: directly measure the magnetic field. Unfortunately, this is very
183: difficult and the picture -- particularly in stars with saturated
184: activity -- is not yet clear.
185:
186: \citet{Saar96} has collected measurements of magnetic flux $Bf$, of
187: the filling factor $f$, and of rotation periods to investigate the
188: behavior of magnetic fields on stars. It is important to realize that
189: the magnetic flux, or the magnetic field average over the whole
190: surface, is not the same as the local field. It is the (unsigned) mean
191: average of the magnetic field strength over the whole surface.
192: Furthermore, the strongest magnetic fields in cool spots may not be
193: fully captured because their contribution to the total flux is
194: dimished due to their low temperature. As on the Sun, magnetic flux
195: is probably concentrated in relatively small regions of strong fields
196: \citep[see, e.g.][]{JKV00}. \citet{Saar96} shows that in stars
197: rotating slower than the saturation threshold the magnetic flux $Bf$
198: as well as the filling factor $f$ show the same trend as all other
199: activity indicators: They grow with larger rotation rate. At high
200: rotation rates, \citet{Saar96} claims that saturation occurs in the
201: filling factor $f$ but not in the magnetic flux $Bf$. From this result
202: one would conclude that a star rotating at the saturation threshold is
203: completely covered with magnetism ($f=1$), and that the saturation
204: phenomenon is due to the saturation of the emission process while the
205: star's magnetic flux can grow further with higher rotation.
206: \citet{Saar01} has reinvestigated this issue with a few more data
207: points, noting that there is some indication for a saturation at $Bf
208: \sim 3$\,kG at small Rossby numbers. Much higher magnetic flux could
209: be in contradiction to the idea that magnetic fields in stellar
210: atmospheres cannot grow stronger than the equipartition field, i.e.
211: the field strength at which the magnetic pressure equals the gas
212: pressure. However, \citet{Solanki94} shows that the equipartition
213: field may not necessarily be a hard upper limit for the field strength
214: at $f=1$, so that more rapidly rotating stars could in principle have
215: much stronger fields.
216:
217: The only way to decide whether $Bf$ does saturate or not is to provide
218: direct measurements of magnetic flux in the regime of saturated
219: activity. Generally, the measurement of magnetic fields relies on the
220: splitting of spectral lines through the Zeeman effect
221: \citep[e.g.][]{Robinson80}. In rapid rotators, the subtle effect of
222: magnetic broadening is buried under the rotational line broadening so
223: that it is particularly difficult to directly measure the magnetic
224: flux in the regime of saturated magnetic activity. However, the
225: crucial datum for rotational line broadening is the projected rotation
226: velocity $v\,\sin{i}$ and not the Rossby number or the rotation period
227: (which are the numbers that appear to set the saturation threshold of
228: magnetic activity). Mainly because of the smaller radius of cooler
229: stars, the surface rotation velocity at which saturation sets in
230: depends on spectral type. In early G-type stars, the saturation
231: velocity in the transition region is on the order of 30\,km\,s$^{-1}$
232: \citep[and only 15\,km\,s$^{-1}$ in the corona,][]{Ayres99}. In M
233: dwarfs, it is less than 5\,km\,s$^{-1}$ \citep[see e.g.][]{Reiners07}.
234: Thus, in sun-like stars, the high surface rotation velocity required
235: for activity saturation hampers the measurement of magnetic flux, but
236: this regime can easily be probed in M stars.
237:
238: One potential problem with the use of M dwarfs for the investigation
239: of dynamo related phenomena is that the interior structure changes
240: around spectral type M3.5 -- stars cooler than that are completely
241: convective. Nevertheless, no change in activity is observed at the
242: threshold to complete convection. A rotation-activity connection is
243: observed in M stars down to spectral types M~8.5 \citep{Mohanty03,
244: Reiners07}.
245:
246: In this paper, we present direct measurements of magnetic flux in
247: several M stars. Some of them are very rapid rotators and clearly
248: belong to the regime of saturated activity. We aim to clarify whether
249: the magnetic flux $Bf$ saturates as H$\alpha$ and X-ray emission do,
250: or if $Bf$ continues to grow beyond $Bf \sim 3$\,kG.
251:
252: \section{Sample and Observations}
253:
254: For our sample we chose a number of mid-M stars with known X-ray
255: emission and presumably high rotation rates. The values $v\,\sin{i}$
256: and $\log{L_{\rm X}/L_{\rm bol}}$ are taken from \citet{Delfosse98}
257: except for GJ\,3379. This value is calculated from the X-ray
258: luminosity taken from \cite{NEXXUS}. With one exception, projected
259: rotation velocities $v\,\sin{i}$ were available for all targets. We
260: chose only stars in which $v\,\sin{i}$ was reported to be above
261: 5\,km\,s$^{-1}$ and which show saturated normalized X-ray emission.
262: For GJ\,3379 we are not aware of any former $v\,\sin{i}$ measurement,
263: but the high value of normalized X-ray emission is indicative of
264: saturation and we added the star to our sample.
265:
266: \begin{deluxetable}{lccc}
267: \tablecaption{\label{tab:observations} Log of observations.}
268: \tablewidth{0pt}
269: \tablehead{\colhead{Name} & \colhead{UTC Date} & \colhead{Exp.Time [s]} & \colhead{$\log{\frac{L_{\rm X}}{L_{\rm bol}}}$} }
270: \startdata
271:
272: GJ 3379 & 2007-09-30 & 200 & $-2.86$\\
273: GJ 2069\,B & 2008-01-24 & 600 & $-2.77$\\
274: Gl 493.1 & 2007-04-25 & 600 & $-3.31$\\
275: LHS 3376 & 2007-04-25 & 1800 & $-3.63$\\
276: GJ 1154\,A & 2007-04-25 & 600 & $-3.28$\\
277: GJ 1156 & 2007-04-25 & 600 & $-3.39$\\
278: Gl 412\,B & 2007-04-25 & 1200 & $-3.28$
279:
280: \enddata
281: \end{deluxetable}
282:
283: Data were taken at the W.M. Keck observatory with the HIRES
284: spectrograph. Our setup covers the wavelength range from below
285: H$\alpha$ (6560\,\AA) up to the molecular absorption band of FeH
286: around 1\,$\mu$m. We used a slit width of 1.15\,\arcsec achieving a
287: resolving power of about $R = 31\,000$. Our log of observations is
288: given in Table\,\ref{tab:observations}. Data were cosmic-ray
289: corrected, flatfielded, background subtracted, and wavelength
290: calibrated using a ThAr spectrum. Data reduction was carried out using
291: routines from the \texttt{echelle} package within the ESO/MIDAS
292: distribution. Fringing is not an issue in spectra taken with the new
293: HIRES CCD, even in very red spectral regions around 1\,$\mu$m.
294:
295: \section{Analysis}
296: \label{sect:results}
297:
298: \begin{deluxetable}{lcrrrr}
299: \tablecaption{\label{tab:results} Results of our analysis. Literature values are from \citet{Delfosse98}.}
300: \tablewidth{0pt}
301: \tablehead{\colhead{Name} & \colhead{SpT} & \colhead{$v\,\sin{i}$} & \colhead{$Bf$} & \colhead{$\log{\frac{L_{{\rm H}\alpha}}{L_{\rm bol}}}$} & $v\,\sin{i}_{\rm Lit}$\\
302: & & [km\,s$^{-1}$] & [G] & & [km\,s$^{-1}$]}
303: \startdata
304:
305: GJ 3379 & M3.5 & $<3$ & 2300 & $-3.35$ & \\
306: GJ 2069\,B & M4.0 & $ 6$ & 2700 & $-3.28$ & 9.2\\
307: Gl 493.1 & M4.5 & $18$ & 2100 & $-3.46$ & 16.8\\
308: LHS 3376 & M4.5 & $19$ & 2000 & $-3.73$ & 14.6\\
309: GJ 1154\,A & M5.0 & $ 6$ & 2100 & $-3.55$ & 5.2\\
310: GJ 1156 & M5.0 & $17$ & 2100 & $-3.53$ & 6.5\\
311: Gl 412\,B & M6.0 & $ 5$ & $>$3900 & $-3.72$ & 7.7
312: \enddata
313: \end{deluxetable}
314:
315:
316: The analysis of our spectra follows the strategy laid out in
317: \citet{RB06} and Reiners \& Basri (2007, RB07 in the following). We
318: measure the equivalent width of the H$\alpha$ emission and convert
319: this number to normalized H$\alpha$ luminosity using M-star
320: atmospheres calculated with the PHOENIX code \citep{Allard01}. To
321: measure the projected rotation velocity $v\,\sin{i}$ and the magnetic
322: flux $Bf$ of our sample stars, we utilize the absorption band of
323: molecular FeH close to 1\,$\mu$m. We compare our data to spectra of
324: the slowly rotating M-stars GJ\,1002 (M5.5) and
325: Gl\,873\footnote{Gl\,873 is rotating at $v\,\sin{i} \sim
326: 3$\,km\,$^{-1}$ (RB07), not at a higher velocity as reported in
327: \cite{Delfosse98}. The small (but detectable) rotation of Gl~873
328: does not affect our measurements.} (M3.5). In order to match the
329: absorption strength of the target spectra, the intensity of the FeH
330: absorption lines in the two comparison spectra is modified according
331: to an optical-depth scaling \citep[see][]{RB06}. In a first step, we
332: compare the artificially broadenend spectrum of GJ\,1002 to the target
333: spectra in the wavelength region at 9930--9960\,\AA\ to determine the
334: value of $v\,\sin{i}$ by $\chi^2$-minimization.
335:
336: For the determination of the magnetic flux $Bf$, we concentrate on
337: smaller wavelength regions that contain absorption lines particularly
338: useful for this purpose, i.e. regions that contain some magnetically
339: sensitive as well as magnetically insensitive lines. The magnetic flux
340: of Gl\,873 was measured to be 3.9\,kG \citep[using an atomic FeI
341: line;][]{JKV00}. For our measurement, we are using a spectrum that
342: contains both the FeI line and the FeH absorption band. The FeI line
343: in this spectrum is consistent with the same magnetic flux value as
344: found by \cite{JKV00} so that we can use the FeH pattern for the
345: calibration of magnetic flux measurements in other stars. This method
346: does not require theoretical models of the magnetic Zeeman splitting
347: of FeH lines, which are not available yet.
348:
349: We determine the magnetic flux of our target stars by comparison of
350: the spectral regions at 9895.5--9905.5\,\AA, 9937.5--9941.0\,\AA,
351: 9946.0--9956.0\,\AA, and 9971.5--9981.0\,\AA\ (for more details see
352: RB07). In Figs.\,\ref{fig:slow}--\ref{fig:rapid2} we show the data and
353: the quality of our fit in the top panels. Note that in rapid rotators
354: the difference between magnetic and non-magnetic stars is not
355: necessarily clearest at the exact location of magnetically sensitive
356: lines. The blending of lines through rotation pronounces differences
357: at wavelengths where the equivalent widths of lines differ the most
358: between active and inactive stars.
359:
360: For example, the two FeH lines at 9949.1\,\AA\ and 9951.7\,\AA\ are
361: magnetically not very sensitive \citep[see][]{RB06, Reiners08}, but it
362: appears that the 9949.1\AA\ line effectively gains a little in
363: equivalent width. Thus, at rotation rates as high as shown in
364: Figs.\,\ref{fig:rapid} and \ref{fig:rapid2}, the region around
365: 9949\,\AA\ becomes the one of largest difference. The reason for this
366: are the magnetically sensitive features next to insensitive lines
367: together with the effective gain in equivalent width, although such
368: differences are not necessarily overt for the observer once Doppler
369: broadening is introduced. On the other hand, it can also happen that
370: at wavelengths where the unrotated spectra are quite different,
371: rotational broadening averages in adjacent flux in such a way that the
372: magnetic differences end up erasing each other there (e.g. at
373: 9948\,\AA). Thus, it is necessary to carry out a spectral fitting
374: procedure after proper preparation instead of relying on appearances
375: in original template spectra at fixed wavelenghts.
376:
377: \section{Results}
378:
379: In the bottom panels of Figs.\,\ref{fig:slow}--\ref{fig:rapid2}, we
380: show the $\chi^2$-landscapes for all our targets as a function of
381: $v\,\sin{i}$ and $Bf$. Color-coding displays the quality of the fit.
382: In each $\chi^2$-landscape, the white contour marks the 3$\sigma$
383: region, i.e. $\chi^2 < \chi^2_{\rm min} + 9$ within this region. In
384: all cases, the mean deviation per degree of freedom is on the order of
385: 1 ($\chi_{\nu} \approx 1$) for the estimated signal-to-noise ratio.
386: Uncertainties in $v\,\sin{i}$ and $Bf$ are typically around
387: 1\,km\,s$^{-1}$ and a few hundred Gauss, respectively. We emphasize
388: that in particular in the case of $Bf$ systematic errors are a more
389: severe source of uncertainty so that the total uncertainty in $Bf$ is
390: more realistically in the 500--1000\,G range.
391:
392: The results of our analysis are given in Table\,\ref{tab:results};
393: projected rotation velocity $v\,\sin{i}$, magnetic flux or the mean
394: magnetic field $Bf$, and normalized H$\alpha$ activity $\log{L_{{\rm
395: H}\alpha}/L_{\rm bol}}$ are given in columns 3, 4, and 5,
396: respectively. All H$\alpha$ measurements confirm that our sample
397: targets are active stars close to the activity saturation level. Three
398: stars show very rapid rotation on the order of 20\,km\,s$^{-1}$, three
399: stars are rotating at a velocity around 6\,km\,s$^{-1}$. GJ~3376 shows
400: rotation below our detection limit of $v\,\sin{i} \approx
401: 3$\,km\,s$^{-1}$. For comparison, we include in column 6 measurements
402: of $v\,\sin{i}$ by \cite{Delfosse98}. In GJ\,1156, we measure a
403: rotational velocity three times higher than formerly reported, and in
404: GJ~2069\,B our new value of rotational broadening is $v\,\sin{i} =
405: 6$\,km\,s$^{-1}$ while \citet{Delfosse98} reports $v\,\sin{i} =
406: 9$\,km\,s$^{-1}$. We attribute the differences to higher data quality
407: in our sample and to our more sophisticated fitting procedure. The
408: uncertainty in $v\,\sin{i}$ also depends on the magnetic flux with a
409: higher uncertainty at higher magnetic flux. The crosstalk between
410: magnetic flux and rotation velocity may be weaker in other wavelength
411: regions (but so far this was not investigated in detail).
412:
413: Measuring magnetic flux is hampered if a target is rapidly rotating
414: because resolving individual lines becomes difficult in the presence
415: of rotation. In the three rapid rotators Gl~493.1, LHS~3376, and
416: GJ~1156, we used only the spectral regions 9946--9956\,\AA\ and
417: 9972--9981\,\AA, which are particularly well suited in rapid rotators
418: \citep{RB06}.
419:
420: In all seven stars, we detected mean magnetic fields of 2\,kG or
421: stronger. RB07 investigated 22 M dwarfs, 17 of them have spectral
422: types earlier than M7. All 8 of them with normalized H$\alpha$
423: luminosity larger than $\log{L_{{\rm H}\alpha}/L_{\rm bol}} = -4$ also
424: have mean magnetic fields on the order of 2\,kG or more. The most
425: active stars in that sample exhibit significant rotation, but the
426: sample contains only one star rotating more rapidly than $v\,\sin{i} =
427: 10$\,km\,s$^{-1}$. Our results are in good agreement with the
428: relation between mean magnetic fields and normalized H$\alpha$
429: activity found in RB07. The (projected) rotation velocities of the
430: three most rapidly rotating stars in our sample are at least a factor
431: of two higher, but we see no sign of normalized H$\alpha$ luminosity
432: higher than in RB07, and no exceptionally high value of $Bf$.
433: \emph{The main result of this work is that none of the three rapid
434: rotators with projected rotation velocities close to
435: 20\,km\,s$^{-1}$ shows a mean magnetic field above 3\,kG.}
436:
437: Can we really measure magnetic fields in stars rotating as rapidly as
438: $v\,\sin{i} = 20$\,km\,s$^{-1}$? In \citet{RB06} we have shown that
439: the magnetic sensitivity of FeH in principle allows the measurement of
440: magnetic flux in stars rotating as rapidly as $v\,\sin{i} =
441: 30$\,km\,s$^{-1}$. The limiting factor in rapid rotators is the
442: achievable signal-to-noise ratio. In our case of ``only'' $v\,\sin{i}
443: \approx 20$\,km\,s$^{-1}$ this is not a crucial problem. The
444: differences between stars with strong and weak magnetic fields in the
445: presence of rapid rotation can be seen in the spectra we show in
446: Figs.\,\ref{fig:rapid} and \ref{fig:rapid2}. The results of our
447: $\chi^2$ fits plotted in the lower panels of Fig.\,\ref{fig:rapid} and
448: \ref{fig:rapid2} show that in all three stars $\chi^2$ becomes
449: significantly larger at very low field strengths or if one allows for
450: mean fields as high as 4\,kG.
451:
452: Our active template star Gl~873, which we are using as comparison for
453: our targets, has a mean magnetic field of about 4\,kG. From comparison
454: to the spectrum of Gl\,873 we cannot measure magnetic flux in excess
455: of that value. In RB07 we show a spectrum of YZ~CMi (M4.5, $v\,\sin{i}
456: = 5$\,km\,s$^{-1}$) from which we inferred a mean field stronger than
457: 4\,kG. This spectrum shows that in the presence of stronger fields
458: the magnetically sensitive lines can become even more washed out
459: following the principles of Zeeman broadening.\footnote{We show this
460: spectrum in Fig.\,3 of RB07. In that plot, the ratio of the
461: magnetically sensitive to the insensitive lines is smaller in YZ~CMi
462: than in our magnetically active template.} The spectrum of Gl~412\,B
463: (WX~UMa) shows the same behavior at a rotation velocity of $v\,\sin{i}
464: = 5$\,km\,s$^{-1}$. Thus, although we lack a spectrum independently
465: calibrated to stronger magnetic flux to compare with, we see no way
466: that the spectra of the three most rapidly rotating stars can be
467: consistent with magnetic flux stronger than $Bf = 4$\,kG. We discuss
468: these stars further below.
469:
470:
471: \subsection{Saturation of the magnetic flux $Bf$}
472:
473: It is well accepted that coronal and chromospheric activity saturate
474: at high rotation rates. \citet{Mohanty03} showed that this relation is
475: still valid in stars as late as spectral type M8.5. In mid-M stars,
476: all stars with detected rotational broadening show H$\alpha$ emission
477: at the saturation level. In the left panel of
478: Fig.\,\ref{fig:vsini_alpha}, we plot $v\,\sin{i}$ vs. $\log{L_{{\rm
479: H}\alpha}/L_{\rm bol}}$ for all stars of spectral types earlier
480: than M7 from RB07 together with the stars from our new sample. This
481: shows that the rotation-activity relation is still valid in the
482: combined sample. In the right panel of Fig.\,\ref{fig:vsini_alpha}, we
483: plot measured magnetic flux $Bf$ as a function of $v\,\sin{i}$. If
484: magnetic flux did not saturate, we would expect the values of $Bf$ to
485: continue growing with higher rotation rate. This is not observed:
486: instead, the magnetic flux shows the same saturation effect as
487: H$\alpha$ emission. From this result we conclude that magnetic flux
488: generation does not grow any further in stars with saturated H$\alpha$
489: emission. In other words, magnetic flux saturates in roughly the same
490: fashion as activity, implying that activity does not saturate solely
491: because the whole area of the star is covered with fields ($f=1$).
492:
493:
494: \subsection{Comparison to hotter stars}
495:
496: \begin{deluxetable*}{llcccrrrrr}
497: \tablecaption{\label{tab:Rossby} Masses, radii, projected rotational velocities, periods, and Rossby numbers, magnetic flux, and normalized X-ray luminosities for our sample stars and the stars from RB07 (see text).}
498: \tablewidth{0pt}
499: \tablehead{\colhead{Name} & \colhead{Other} & \colhead{SpT} & \colhead{$M/M_\sun$} & \colhead{$R/R_\sun$} & \colhead{$v\,\sin{i}$} & \colhead{$P/\sin{i}$} & \colhead{$\log{(Ro/\sin{i})}$} & \colhead{$Bf$ [G]} &\colhead{$\log{\frac{L_{\rm X}}{L_{\rm bol}}}$} }
500: \startdata
501:
502: GJ 3379 & & M3.5 & 0.24 & 0.25 & $<$ 3 & $>$4.2\phantom{$^{a}$} & $>-1.2$ & 2000 & $ -2.86$ \\
503: GJ 2069\,B & CV Cnc & M4.0 & 0.24 & 0.25 & 6 & 2.3\phantom{$^{a}$} & $ -1.5$ & 2500 & $ -2.77$ \\
504: Gl 493.1 & FN Vir & M4.5 & 0.17 & 0.20 & 18 & 0.5\phantom{$^{a}$} & $ -2.1$ & 2100 & $ -3.31$ \\
505: LHS 3376 & & M4.5 & 0.14 & 0.17 & 16 & 0.5\phantom{$^{a}$} & $ -2.1$ & 2000 & $ -3.63$ \\
506: GJ 1154\,A & & M5.0 & 0.18 & 0.20 & 6 & 1.7\phantom{$^{a}$} & $ -1.6$ & 2000 & $ -3.28$ \\
507: GJ 1156 & GL Vir & M5.0 & 0.14 & 0.16 & 17 & 0.5\phantom{$^{a}$} & $ -2.1$ & 2100 & $ -3.39$ \\
508: Gl 412\,B & WX UMa & M6.0 & 0.11 & 0.13 & 5 & 1.4\phantom{$^{a}$} & $ -1.7$ & $>$ 3900 & $ -3.28$ \\
509: \cutinhead{Stars from RB07}\\
510: Gl 70 & & M2.0 & 0.35 & 0.33 & $<$ 3 & $>$5.6\phantom{$^{a}$} & $>-1.1$ & 0 & $<-4.44$ \\
511: Gl 729 & V1216 Sgr& M3.5 & 0.18 & 0.20 & 4 & 2.9\tablenotemark{a} & $ -1.4$ & 2200 & $ -3.50$ \\
512: Gl 873 & EV Lac & M3.5 & 0.33 & 0.31 & 3 & 4.4\tablenotemark{b} & $ -1.2$ & 3900 & $ -3.07$ \\
513: Gl 388 & AD Leo & M3.5 & 0.42 & 0.39 & 3 & 2.6\tablenotemark{b} & $ -1.4$ & 2900 & $ -3.02$ \\
514: Gl 876 & & M4.0 & 0.32 & 0.31 & $<$ 3 & $>$5.2\phantom{$^{a}$} & $>-1.1$ & 0 & $ -5.23$ \\
515: GJ 1005A & & M4.0 & 0.22 & 0.23 & $<$ 3 & $>$3.9\phantom{$^{a}$} & $>-1.3$ & 0 & $ -5.05$ \\
516: Gl 299 & & M4.5 & 0.15 & 0.18 & $<$ 3 & $>$3.0\phantom{$^{a}$} & $>-1.4$ & 500 & $<-5.55$ \\
517: GJ 1227 & & M4.5 & 0.17 & 0.19 & $<$ 3 & $>$3.2\phantom{$^{a}$} & $>-1.3$ & 0 & $<-3.86$ \\
518: GJ 1224 & & M4.5 & 0.15 & 0.18 & $<$ 3 & $>$3.0\phantom{$^{a}$} & $>-1.4$ & 2700 & $ -3.06$ \\
519: Gl 285 & YZ CMi & M4.5 & 0.31 & 0.30 & 5 & 2.8\tablenotemark{b} & $ -1.4$ & $>$ 3900 & $ -3.02$ \\
520: Gl 905 & HH And & M5.0 & 0.14 & 0.17 & $<$ 3 & $>$2.8\phantom{$^{a}$} & $>-1.4$ & 0 & $ -3.75$ \\
521: GJ 1057 & & M5.0 & 0.16 & 0.18 & $<$ 3 & $>$3.1\phantom{$^{a}$} & $>-1.4$ & 0 & $<-3.87$ \\
522: GJ 1245B & & M5.5 & 0.11 & 0.14 & 7 & 1.0\phantom{$^{a}$} & $ -1.8$ & 1700 & $ -3.58$ \\
523: GJ 1286 & & M5.5 & 0.12 & 0.14 & $<$ 3 & $>$2.4\phantom{$^{a}$} & $>-1.5$ & 400 & $<-3.77$ \\
524: GJ 1002 & & M5.5 & 0.11 & 0.13 & $<$ 3 & $>$2.3\phantom{$^{a}$} & $>-1.5$ & 0 & $<-5.24$ \\
525: Gl 406 & CN Leo & M5.5 & 0.10 & 0.13 & 3 & 1.9\phantom{$^{a}$} & $ -1.6$ & 2400 & $ -2.77$ \\
526: GJ 1111 & DX Cnc & M6.0 & 0.10 & 0.12 & 13 & 0.4\phantom{$^{a}$} & $ -2.2$ & 1700 & $ -3.88$ \\
527: \enddata
528: \tablenotetext{a}{\cite{Kiraga07}}
529: \tablenotetext{b}{\cite{Saar01}}
530: \end{deluxetable*}
531:
532: How does the saturation of magnetic flux in our M-dwarf sample fit
533: into the picture of rotation and activity in hotter stars? In order to
534: compare activity measurements in very different stars, the Rossby
535: number, i.e., the ratio of rotation period $P$ and convective overturn
536: time $\tau_{\rm conv}$, is the parameter of choice. We note, however,
537: that the main effect of the Rossby number is to compare rotation
538: periods, and that the influence of the convective overturn time is
539: still debatable. \citet{Kiraga07} provide a measurement of the
540: rotation period of Gl\,729. For Gl~873, AD~Leo, and YZ~CMi, $P$ can be
541: found in \citet{Saar01}. To obtain the (projected) Rossby number, $Ro
542: = P/\tau_{\rm conv}$, for the other stars, we calculate the
543: (projected) rotation periods for our sample and the stars from RB07
544: from the projected surface rotation velocity $v\,\sin{i}$ for which we
545: require the radii. Note that this only provides $Ro/\sin{i}$, which is
546: an upper limit of $Ro$. To determine the radius, we employed the
547: mass-luminosity relation from \citet{Delfosse00} and the mass-radius
548: relation at an age of 5\,Gyrs from \cite{Baraffe98}. To compute the
549: masses, we used J-magnitudes from \cite{2MASS} and distances from
550: \cite{Hawley96}. Parallaxes for Gl~299 and GJ~1286 are from
551: \cite{Harrington80} and \cite{Oppenheimer01}, respectively. For the
552: convective overturn time, we adopt a value of $\tau_{\rm conv} =
553: 70$\,d consistent with the values given for M dwarfs in
554: \citet{Saar01}, which are taken from \citet{Gilliland86}.
555: \citet{Kiraga07} have recently calculated empirical turnover times for
556: a sample of M stars. Their relation yields values about $\tau_c \sim
557: 50$--100\,d for the mass range considered in our sample. This is
558: roughly consistent with $\tau_c = 70$\,d; if we were using the
559: convective overturn times from \citet{Kiraga07}, the Rossby numbers
560: would be about 0.15\,dex larger (smaller) for the hottest (coolest)
561: stars. This difference does not affect our conclusions. Radii,
562: (projected) periods and Rossby numbers are given in
563: Table\,\ref{tab:Rossby}.
564:
565: We combine our measurements with data on additional stars from
566: \citet{Saar96} and \citet{Saar01}. For the stars from \citet{Saar96}
567: we adopt a value of $\tau_{\rm conv} = 20$\,d (spectral type G0--K5)
568: noting that the convective overturn time changes over this range of
569: stars but that the difference is not significant on the level
570: considered here.
571:
572: The behavior of $Bf$ with Rossby number is shown in
573: Fig.\,\ref{fig:Bf_Ro} where we plot all the mentioned samples
574: together. Data from this work are plotted as filled circles, data from
575: RB07 as open circles (M dwarfs), and data from \cite{Saar96} and
576: \cite{Saar01} as crosses (note that crosses are hotter stars). Stars
577: in which only upper limits of $v\,\sin{i}$ were measured are not shown
578: (non-detections of rotation). Fig.\,\ref{fig:Bf_Ro} clearly shows the
579: dependence of magnetic flux generation with Rossby number: At large
580: Rossby numbers ($Ro > 0.1$), i.e. in the non-saturated regime,
581: magnetic flux increases with decreasing $Ro$ (more rapid rotation). At
582: smaller Rossby number, i.e., in the regime where H$\alpha$ and X-ray
583: emission saturate, $Bf$ saturates as well.
584:
585: In Fig.\,\ref{fig:Bf_Ro}, the saturated regime mainly consists of M
586: dwarfs while the rising part of the relation (crosses) is populated by
587: hotter stars. \citet{Reiners07} showed that a drop in activity with
588: higher Rossby number also appears in slowly rotating M dwarfs (his
589: Fig.\,10). A substantial piece of evidence for an intact
590: rotation-activity relation among M dwarfs can be found in the data
591: presented by \citet{Kiraga07}. From their data, we show normalized
592: X-ray activity vs. Rossby number (again assuming $\tau_{\rm conv} =
593: 70$\,d) in Fig.\,\ref{fig:Kiraga}. Clearly, a rise of activity at
594: high Rossby number ($Ro = 0.1\dots1$) and a saturation plateau at
595: lower Rossby number appears among M dwarfs, too.
596:
597: We conclude that the dynamo process saturates at Rossby number of
598: about $Ro = 0.1$. No systematic increase of $Bf$ occurs if the Rossby
599: number is smaller by another order of magnitude.
600:
601:
602: \subsection{The regime of saturation}
603:
604: The level of magnetic flux in the saturated regime is between 2 and
605: 4\,kG. In two stars, we observe magnetic flux that may be higher than
606: 4\,kG. At least one of the two stars is not among those with the
607: smallest Rossby numbers in our sample. We can speculate whether the
608: Rossby numbers of both stars with the highest magnetic flux are
609: smaller than the Rossby numbers of the other stars in our sample,
610: i.e., whether the Rossby numbers of YZ~CMi and WX~UMa are also on the
611: order of $\log{\emph{Ro}} \approx -2$. In the case of WX~Uma, we only
612: have $Ro/\sin{i}$, and for a real value of $\log{\emph{Ro}} = -2$ the
613: star would be observed under an inclination angle of $i < 30^\circ$.
614: For WX~UMa, this is a viable option. On the other hand, a rotation
615: period of $P = 2.8$\,d is reported for YZ~CMi, which is in good
616: agreement with an inclination angle close to $i = 90^\circ$ given the
617: estimated radius and the measured rotation velocity. To push the value
618: of $\log{\emph{Ro}} \approx -1.4$ to $-2$, either the rotation
619: velocity must be a factor of 4 higher ($i < 15^\circ$ implying that
620: the rotation period is wrong), or the convective overturn time must be
621: longer by the same amount. Both options seem rather unlikely. We note
622: that \citet{Saar01} reported a magnetic flux of $Bf = 3.3$\,kG for
623: YZ~CMi, a value that is somewhat lower than our result. This may
624: indicate that our value does not reflect an unusually strong average
625: field strength in YZ~CMi, but that the magnetic flux shows rather
626: large scatter (either due to uncertainties in the measurements or
627: temporal fluctuations).
628:
629: The easiest explanation for the two very high values of $Bf > 4$\,kG
630: is that the scatter in the saturated magnetic flux level is fairly
631: large, and $Bf$ between 2 and 5\,kG might just be the allowed range at
632: small Rossby numbers (including observational effects). We have
633: searched for other parameters, such as exceptionally low gravity, that
634: could cause the high flux values in the two stars. We did not find
635: any particular stellar parameter that distingiushes YZ~CMi and WX~Uma
636: from other flare stars. In particular, age is probably of little
637: direct importance for the generation of very high magnetic flux:
638: WX~Uma is an old disk flare star while YZ~CMi is a member of the young
639: disk population \citep{Veeder74}.
640:
641: Even if the two strongest magnetic flux measurements are due to
642: exceptionally small Rossby numbers, this would not explain the
643: saturation of $Bf$ between $\log{\emph{Ro}} = -2$ and $-1$. Across an
644: order of magnitude in $Ro$, $Bf$ varies by at most a factor of a few.
645: This is in striking contrast to the nearly hundredfold increase in
646: $Bf$ in going from $Ro \approx 1$ to $Ro \approx 0.1$. We have
647: concluded that both magnetic flux and chromospheric H$\alpha$ (as well
648: as coronal X-ray) emission saturate at small $Ro$. With the large
649: uncertainties in $Bf$, we cannot reliably determine whether there is
650: any super-saturation effect, with magnetic flux declining in the most
651: rapid rotators.
652:
653: Although both emission and magnetic flux appear to saturate at small
654: $Ro$, we observe a large scatter of magnetic flux among the stars with
655: small $Ro$, and likewise some scatter in H$\alpha$ emission. We thus
656: examined whether a relation between H$\alpha$ emission and magnetic
657: flux still exists at very small Rossby numbers. In
658: Fig.\,\ref{fig:Bf_alpha}, we plot normalized H$\alpha$ emission vs.
659: magnetic flux for our stars. As expected, $\log{L_{{\rm
660: H}\alpha}/L_{\rm bol}}$ grows as $Bf$ grows from zero to 2\,kG.
661: Beyond that, i.e. in the regime of saturated magnetic flux, no
662: further increase in chromospheric emission is observed, although $Bf$
663: grows as large as 4\,kG. No obvious correlation exists between
664: chromospheric emission and magnetic flux in the regime of saturation.
665: This could mean that saturation of chromospheric emission occurs above
666: a certain level of integrated flux on the star (for example, because
667: the filling factor reached unity). The field could then grow stronger
668: without causing further heating. This is in fact not ruled out by our
669: poorly constrained understanding of chromospheric heating mechanisms.
670:
671: As of now, the available data do not permit us to distinguish reliably
672: between these two possibilities. But, as noted above, the most rapid
673: rotators in our sample probably do not possess magnetic fluxes higher
674: than among other stars in the saturated regime: in
675: Fig.\,\ref{fig:Bf_alpha}, stars with the smallest Rossby numbers lie
676: at the low-$Bf$ end of the plateau. Neither are they the stars with
677: the highest normalized chromspheric emission. Thus the lack of tight
678: correlation between emission and $Bf$ in the saturation regime should
679: \emph{not} be construed as evidence that magnetic flux continues to be
680: correlated with rotation, but decouples from emission. Instead, our
681: data suggest that any variation of H$\alpha$ emission or $Bf$ in the
682: saturation regime must either be random -- i.e. due to the
683: observational scatter -- or depend on parameters other than rotation.
684: At this stage, we note only that the plateau in
685: Fig.\,\ref{fig:Bf_alpha} is consistent with scatter alone.
686:
687:
688: \section{Summary}
689:
690: Using absorption lines of molecular FeH we have measured the magnetic
691: flux in seven M stars that are known X-ray sources. With one exception
692: they were known as rapid rotators as well. We reinvestigated the
693: projected rotational velocities $v\,\sin{i}$ and measured
694: chromospheric emission in H$\alpha$. All stars proved to be strong
695: H$\alpha$ emitters. In our analysis of $v\,\sin{i}$, we found some
696: inconsistencies with former literature, which we ascribe to our more
697: sophisticated (direct) fitting method and better data quality.
698:
699: All of our target stars show strong H$\alpha$ emission at the
700: saturation level ($\log{L_{{\rm H}\alpha}/L_{\rm bol}} > -4$). We
701: detected magnetic fields of kG-strength in all our targets. While in
702: less rapidly rotating stars, H$\alpha$ and magnetic flux correlate
703: with rotation rate, no such correlation is observed in our sample.
704:
705: In contrast to sun-like stars of spectral types G and K, M dwarfs
706: rotating at $v\,\sin{i} \approx 10$\,km\,s$^{-1}$ have very small
707: Rossby numbers. Thus, our sample targets add to the amount of stars
708: with measured magnetic flux in the regime of small Rossby number, i.e.
709: in the regime of saturated magnetic activity. Our primary goal was to
710: determine whether magnetic flux in this regime continues to grow with
711: more rapid rotation, as sometimes suggested \citep[e.g.,][]{Saar96},
712: or instead saturates in the same manner as coronal and chromospheric
713: emission.
714:
715: Our main conclusion is that around a Rossby number of $Ro \sim 0.1$,
716: magnetic flux saturates at approximately $Bf = 3$\,kG. Below $Ro \sim
717: 0.1$, $Bf$ does not grow stronger with decreasing Rossby number. In
718: looking at the effect of Rossby number ($Ro = P/\tau_{\rm conv}$), we
719: are primarily sensitive to the effects of rotation, because the
720: sampled rotations vary by a factor of more than ten, while the
721: convective overturn time probably changes less than a factor of two in
722: the range of M stars we are considering.
723:
724: In the regime of saturated magnetic flux and chromospheric emission,
725: we still observe a strong scatter in magnetic flux. The interpretation
726: of this feature is not clear. It may be that below the critical Rossby
727: number, normalized H$\alpha$ emission is not sensitive to changes in
728: $Bf$. Alternatively, in light of the large systematic errors of our
729: $Bf$ measurements, the scatter of $Bf$ in the saturated regime may be
730: fully explained by observational uncertainties.
731:
732: Our results indicate that the strengths of stellar magnetic fields,
733: and not merely their filling fractions on the surface alone, reach a
734: maximum at a certain rotation rate. The scatter in magnetic flux among
735: the stars with very small Rossby numbers, however, is substantial.
736: Typical values are between 2 and 4\,kG. Two stars show magnetic flux
737: stronger than 4\,kG, and we cannot exclude that individual stars can
738: generate magnetic flux stronger than that. Determining what sets the
739: maximum value -- whether equipartition with either the atmospheric
740: pressure or the turbulent velocity field, or some more subtle effect
741: -- remains a challenge for future observations and theories.
742:
743:
744: \acknowledgments
745:
746: This work is based on observations from the W.M. Keck Observatory,
747: which is operated as a scientific partnership among the California
748: Institute of Technology, the University of California and the National
749: Aeronautics and Space Administration. We would like to acknowledge the
750: great cultural significance of Mauna Kea for native Hawaiians and
751: express our gratitude for permission to observe from atop this
752: mountain. AR acknowledges research funding from the DFG under an Emmy
753: Noether Fellowship (RE 1664/4-1). G.B. acknowledges support from the
754: NSF through grant AST-0606748. M.B. was supported by an NSF Astronomy
755: and Astrophysics postdoctoral fellowship (AST-0502413).
756:
757: %We are grateful to a lot of people.
758:
759: %% To help institutions obtain information on the effectiveness of their
760: %% telescopes, the AAS Journals has created a group of keywords for telescope
761: %% facilities. A common set of keywords will make these types of searches
762: %% significantly easier and more accurate. In addition, they will also be
763: %% useful in linking papers together which utilize the same telescopes
764: %% within the framework of the National Virtual Observatory.
765: %% See the AASTeX Web site at http://www.journals.uchicago.edu/AAS/AASTeX
766: %% for information on obtaining the facility keywords.
767:
768: %% After the acknowledgments section, use the following syntax and the
769: %% \facility{} macro to list the keywords of facilities used in the research
770: %% for the paper. Each keyword will be checked against the master list during
771: %% copy editing. Individual instruments can be provided in parentheses,
772: %% after the keyword, but they will not be verified.
773:
774: %Facilities: \facility{Nickel}, \facility{HST(STIS)}, \facility{CXO(ASIS)}.
775:
776: %% Appendix material should be preceded with a single \appendix command.
777: %% There should be a \section command for each appendix. Mark appendix
778: %% subsections with the same markup you use in the main body of the paper.
779:
780: %% Each Appendix (indicated with \section) will be lettered A, B, C, etc.
781: %% The equation counter will reset when it encounters the \appendix
782: %% command and will number appendix equations (A1), (A2), etc.
783:
784: \begin{thebibliography}{}
785: \bibitem[Allard et al., 2001]{Allard01}Allard, F., Hauschildt, P.H.,
786: Alexander, D.R., Tamanai, A., \& Schweitzer, A., 2001, \apj, 556,
787: 357
788: \bibitem[Ayres, 1999]{Ayres99}Ayres, T., 1999, \apj, 525, 240
789: \bibitem[Baraffe et al., 1998]{Baraffe98}Baraffe, I., Chabrier, G.,
790: Allard, F., Hauschildt, P.H., 1998, A\&A, 337, 403
791: \bibitem[Cutri et al., 2003]{2MASS} Cutri et al., 2003, The 2MASS
792: All-Sky Catalog of Point Sources, University of Massachusetts and
793: Infrared Processing and Analysis Center; IPAC/California Institute
794: of Technology
795: \bibitem[Delfosse et al., 1998]{Delfosse98}Delfosse, X., Forveille,
796: T., Perrier, C., \& Mayor, M., 1998, \aap, 331, 581
797: \bibitem[Delfosse et al., 2000]{Delfosse00}Delfosse, X., Forveille,
798: T., S\'egransan, D., Beuzit, J.-L., Udry, S., Perrier, C., \& Mayor,
799: M., 2000, \aap, 364, 217
800: \bibitem[Harrington \& Dahn, 1980]{Harrington80}Harrington, R.S., \&
801: Dahn, C.C., 1980, AJ, 85, 454
802: \bibitem[Hawley et al., 1996]{Hawley96}{Hawley}, S.~L., {Gizis},
803: J.~E., \& {Reid}, I.~N., 1996, \aj, 112, 2799
804: \bibitem[James et al., 2000]{James00}James, D.J., Jardine, M.M.,
805: Jeffries, R.D., Randich, S., Collier Cameron, A., \& Ferreira, M.,
806: 2000, MNRAS, 318, 1217
807: \bibitem[Johns-Krull \& Valenti, 2000]{JKV00}Johns-Krull, C., \&
808: Valenti, J.A., 2000, ASPC, 198, 371
809: \bibitem[Gilliland, 1986]{Gilliland86}Gilliland, R.L., 1986, \apj,
810: 300, 339
811: \bibitem[Kenyon \& Hartmann, 1995]{Kenyon95}Kenyon, S.J., \& Hartmann,
812: L., 1995, ApJSS, 101,117
813: \bibitem[Kiraga \& St\c{e}pie\'n, 2007]{Kiraga07}Kiraga, M. \&
814: St\c{e}pie\'n, K., 2007, AcA, 57, 149
815: \bibitem[Mohanty \& Basri, 2003]{Mohanty03}Mohanty, S., \& Basri, G.,
816: 2003, \apj, 583, 451
817: \bibitem[Mohanty et al., 2005]{Mohanty05}Mohanty, S., Jayawardhana,
818: R., \& Basri, G., 2005, \apj, 626, 498
819: \bibitem[Noyes et al., 1984]{Noyes84}Noyes, R.W., Hartmann, L.W.,
820: Baliunas, S.L., Duncan, D.K., Vaughan, A.H. 1984, \apj, 279, 763
821: \bibitem[Oppenheimer et al., 2001]{Oppenheimer01}Oppenheimer, B.R.,
822: Golimowski, D.A., Kulkarni, S.R., Matthews, K., Nakajima, T.,
823: Creech-Eakman, M., \& Durrance, S.T., 2001, \aj, 121, 2189
824: \bibitem[Pizzolato et al., 2003]{Pizzolato03}Pizzolato, N., Maggio,
825: A., Micela, G., Sciortino, S., \& Ventura, P., 2003, \aap, 397, 147
826: \bibitem[Reiners, 2007]{Reiners07}Reiners, A., 2007, \aap, 467, 259
827: \bibitem[Reiners \& Basri, 2006]{RB06}Reiners, A., \& Basri, G., 2006,
828: \apj, 644, 497
829: \bibitem[Reiners \& Basri, 2007]{RB07}Reiners, A., \& Basri, G., 2007,
830: \apj, 656, 1121
831: \bibitem[Reiners \& Basri, 2008]{Reiners08}Reiners, A. \& Basri, G.,
832: 2008, A\&A, 489, L45
833: \bibitem[Robinson, 1980]{Robinson80}Robinson, R.D., \apj, 239, 961
834: \bibitem[Saar, 1996]{Saar96}Saar, S.H., 1996, IAU Symposium, 176, 237
835: \bibitem[Saar, 2001]{Saar01}Saar, S.H., 2001, ASP Conf. Ser., 223, 292
836: \bibitem[Schmitt \& Liefke, 2004]{NEXXUS}Schmitt, J.H.M.M., \& Liefke,
837: C., 2004, \aap, 417, 651
838: \bibitem[Solanki, 1994]{Solanki94}Solanki, S.K., 1994, ASP Conf. Ser.
839: 64, 477
840: \bibitem[Veeder, 1974]{Veeder74}Veeder, G.J., 1974, AJ, 79, 702
841: \bibitem[Vilhu, 1984]{Vilhu84}Vilhu, O., 1984, A\&A, 133, 117
842: \bibitem[Vilhu, 1987]{Vilhu87}Vilhu, O., 1987, Lecture Notes in
843: Physics, Vol. 291, eds. Linsky, J.L., \& Stencel, R.E., Springer,
844: 1987., p.110
845: \end{thebibliography}
846:
847:
848:
849: \begin{figure}
850: \centering
851: \includegraphics[width=.48\hsize]{f1a.eps}
852: \includegraphics[width=.48\hsize]{f1b.eps}\\
853: \includegraphics[width=.475\hsize,clip=,bbllx=20,bblly=0,bburx=600,bbury=425]{f1c.eps}
854: \includegraphics[width=.475\hsize,clip=,bbllx=20,bblly=0,bburx=600,bbury=425]{f1d.eps}
855: \caption{\label{fig:slow}Top panel: Data and fit of the two slow
856: rotators GJ 3379 (left) and GJ 2069B (right). The data are shown
857: in black. Our fit to the data for the case of no magnetic flux is
858: overplotted in blue, very strong magnetic flux ($Bf = 3.9\,kG$) in
859: red, and the best fit with intermediate flux values in green.
860: Bottom panel: $\chi^2$-landscapes showing the goodness of fit as a
861: function of $v\,\sin{i}$ and $Bf$. Dark/blue color indicates good
862: fit quality, bright/yellow color means bad fit. The white contour
863: marks the formal 3$\,\sigma$ region, i.e., $\chi^2 < \chi^2_{\rm
864: min} + 9$.}
865: \end{figure}
866:
867: \begin{figure}
868: \centering
869: \includegraphics[width=.48\hsize]{f2a.eps}
870: \includegraphics[width=.48\hsize]{f2b.eps}\\
871: \includegraphics[width=.475\hsize,clip=,bbllx=20,bblly=0,bburx=600,bbury=425]{f2c.eps}
872: \includegraphics[width=.475\hsize,clip=,bbllx=20,bblly=0,bburx=600,bbury=425]{f2d.eps}
873: \caption{\label{fig:intermediate}Date and fit (top panel) and
874: $\chi^2$-landscapes (bottom panel) as in Fig.\,\ref{fig:slow} for
875: the two intermediate rotators GJ 1154A (left) and GJ 412B
876: (right).}
877: \end{figure}
878:
879: \begin{figure}
880: \centering
881: \includegraphics[width=.46\hsize]{f3a.eps}\quad\,
882: \includegraphics[width=.46\hsize]{f3b.eps}\\
883: \includegraphics[width=.475\hsize,clip=,bbllx=20,bblly=0,bburx=600,bbury=425]{f3c.eps}
884: \includegraphics[width=.475\hsize,clip=,bbllx=20,bblly=0,bburx=600,bbury=425]{f3d.eps}
885: \caption{\label{fig:rapid}Data and fits (top panel) and
886: $\chi^2$-landscape as in Fig.\,\ref{fig:slow} for the rapid
887: rotators GJ 1156 (left) and Gl 493.1 (right).}
888: \end{figure}
889:
890: \begin{figure}
891: \includegraphics[width=.46\hsize]{f4a.eps}\\
892: \includegraphics[width=.475\hsize,clip=,bbllx=20,bblly=0,bburx=600,bbury=425]{f4b.eps}
893: \caption{\label{fig:rapid2}As Fig.\,\ref{fig:rapid} for the rapid
894: rotator LHS 3376.}
895: \end{figure}
896:
897:
898: \begin{figure}
899: \plottwo{f5a.eps}{f5b.eps}
900: \caption{\label{fig:vsini_alpha} Left panel: Normalized H$\alpha$
901: activity as a function of $v\,\sin{i}$ in M stars; right panel:
902: Magnetic flux $Bf$ as a function of $v\,\sin{i}$. Filled circles
903: are from this work, triangles come from RB07. Downward arrows
904: indicate upper limits, numbers give the number of multiple
905: measurements with the same results. The two lower limits of $Bf$
906: are indicated with upward arrows. }
907: \end{figure}
908:
909: \begin{figure}
910: \plotone{f6.eps}
911: \caption{\label{fig:Bf_Ro}Magnetic flux $Bf$ as a function of Rossby
912: number. Data are from \citet{Saar96}, \citet{Saar01}, RB07, and
913: from this work. 11 stars without measured rotation periods and
914: with no detection of rotational broadening ($v\,\sin{i} <
915: 3$\,km\,s$^{-1}$), i.e. lower limits of $Ro$, are not shown. They
916: would form a vertical line at about $\log{Ro} = -1$ but probably
917: lie on top of the rising part of the correlation
918: \citep[see][]{Reiners07}. Data from this work are plotted as
919: filled circles, data from RB07 as open circles (M stars), and data
920: from \cite{Saar96} and \cite{Saar01} as crosses (spectral types
921: G0--M2). Rotation rates from this work and from RB07 are
922: calculated from $v\,\sin{i}$ implying that open and filled circles
923: are $Ro/\sin{i}$ hence upper limits of $Ro$. }
924: \end{figure}
925:
926: \begin{figure}
927: \plotone{f7.eps}
928: \caption{\label{fig:Kiraga}Normalized X-ray activity as a function
929: of Rossby number in M stars (using $\tau_{\rm conv} = 70$\,d) from
930: \citet{Kiraga07}. This plot is essentially the same as their
931: Fig.\,7 but in logarithmic units so that the saturation plateau
932: becomes clear.}
933: \end{figure}
934:
935: \begin{figure}
936: \plotone{f8.eps}
937: \caption{\label{fig:Bf_alpha}The correlation between H$\alpha$
938: activity and magnetic flux $Bf$ in M stars. Symbol size scales
939: with inverse Rossby number (large symbols have small $Ro$).
940: Triangles are from RB07, circles are from this work.}
941: \end{figure}
942:
943: \end{document}
944:
945: %%
946: %% End of file `sample.tex'.
947: