1: \documentclass[12pt,preprint]{aastex}

2: 

3: 

4: 

5: \begin{document}

6: \shorttitle{X-ray flares in ONC stars} \shortauthors{Getman et

7: al.}

8: 

9: 

10: 

11: \slugcomment{Accepted for publication in the Astrophysical Journal 07/17/08}

12: 

13: 

14: 

15: 

16: \title{X-ray flares in Orion young stars. II. Flares, magnetospheres,

17: and protoplanetary disks}

18: 

19: 

20: 

21: 

22: \author{Konstantin V.\ Getman\altaffilmark{1}, Eric D.\

23: Feigelson\altaffilmark{1}, Giusi Micela\altaffilmark{2}, Moira M.\

24: Jardine\altaffilmark{3}, Scott G. Gregory\altaffilmark{3},  Gordon

25: P.\ Garmire\altaffilmark{1}}

26: 

27: 

28: 

29: \altaffiltext{1}{Department of Astronomy \& Astrophysics, 525

30: Davey Laboratory, Pennsylvania State University, University Park

31: PA 16802} \altaffiltext{2}{INAF, Osservatorio Astronomico di

32: Palermo G. S. Vaiana, Piazza del Parlamento 1, I-90134 Palermo,

33: Italy} \altaffiltext{3}{SUPA, School of Physics and Astronomy,

34: North Haugh, St Andrews, Fife, KY16 9SS, Scotland, UK}

35: 

36: 

37: 

38: \email{gkosta@astro.psu.edu}

39: 

40: 

41: 

42: 

43: \begin{abstract}

44: 

45: 

46: We study the properties of powerful X-ray flares from 161 pre-main

47: sequence (PMS) stars observed with the Chandra X-ray Observatory

48: in the Orion Nebula region.  Relationships between flare

49: properties, protoplanetary disks and accretion are examined in

50: detail to test models of star-disk interactions at the inner edge

51: of the accretion disks.  Previous studies had found no differences

52: in flaring between diskfree and accreting systems other than a

53: small overall diminution of X-ray luminosity in accreting systems.

54: 

55: 

56: The most important finding is that X-ray coronal extents in

57: fast-rotating diskfree stars can significantly exceed the

58: Keplerian corotation radius, whereas X-ray loop sizes in disky and

59: accreting systems do not exceed the corotation radius.  This is

60: consistent with models of star-disk magnetic interaction where the

61: inner disk truncates and confines the PMS stellar magnetosphere.

62: 

63: 

64: We also find two differences between flares in accreting and

65: diskfree PMS stars.   First, a subclass of super-hot flares with

66: peak plasma temperatures exceeding 100~MK are preferentially

67: present in accreting systems.   Second, we tentatively find that

68: accreting stars produce flares with shorter durations.  Both

69: results may be consequences of the distortion and destabilization

70: of the stellar magnetosphere by the interacting disk.  Finally, we

71: find no evidence that any flare types, even slow-rise flat-top

72: flares are produced in star-disk magnetic loops.  All are

73: consistent with enhanced solar long-duration events with both

74: footprints anchored in the stellar surface.

75: 

76: 

77: \end{abstract}

78: 

79: 

80: 

81: 

82: \keywords{open clusters and associations: individual (Orion Nebula

83: Cluster) - planetary systems: protoplanetary disks  - stars: flare

84: - stars: magnetic fields - stars: pre-main sequence - X-rays:

85: stars}

86: 

87: 

88: 

89: 

90: \section{INTRODUCTION \label{introduction_section}}

91: 

92: 

93: 

94: A broad consensus has emerged in the past decade concerning the

95: structure and astrophysics of pre-main sequence (PMS) stars.  The

96: angular momentum inherited from the collapsing interstellar

97: material is mostly distributed into a rapidly rotating protostar,

98: a Keplerian protoplanetary disk, and frequently the orbital motion

99: of multiple stellar components. The young star interacts with its

100: disk in a complex fashion with accretion and ejection of

101: collimated outflow.  It is widely believed that strong magnetic

102: fields generated within the young star mediate this star-disk

103: interaction, truncating the disk and funneling accretion onto

104: limited portions of the stellar surface.  These issues are

105: developed in a variety of reviews \citep[e.g.,][]{Hartmann98,

106: Bouvier07a, Bouvier07b}.

107: 

108: 

109: 

110: It has been difficult, however, to elucidate in detail the nature

111: of the PMS stellar magnetospheres and their role in facilitating

112: star-disk interactions and accretion.  Well-developed analytical

113: theory assumes that a dipolar field extends to several stellar

114: radii where the Keplerian orbits and stellar fields are in

115: co-rotation, interacts with the disk magnetic field,  and guides

116: both accretion onto the surface and into a high-velocity

117: collimated outflow \citep[e.g.,][]{Shu94, Lovelace95}.  However,

118: considerable evidence has emerged that the magnetic fields on the

119: surfaces of PMS stars are concentrated in complex multipolar

120: active regions similar to those on the Sun, rather than a simple

121: dipolar.  This emerges from photometric, Doppler imaging, circular

122: polarization and spectroscopy of Zeeman-sensitive lines from PMS

123: photospheres \citep[e.g.,][]{Daou06,Yang07, Johns-Krull07,

124: Donati07}.  Recently, theoretical efforts have begun to calculate

125: the resulting complex PMS magnetosphere and accretion process

126: \citep{Jardine06, Gregory06a, Long07}.

127: 

128: 

129: 

130: X-ray studies are a potentially useful tool for investigating

131: these issues.  It has long been known that late-type stars exhibit

132: their highest levels of X-ray emission, arising mostly from

133: violent magnetic reconnection events, during their PMS phase

134: \citep[e.g., reviews by][]{Feigelson99, Feigelson07}.  Recent

135: X-ray surveys of nearby PMS stellar populations give detailed

136: insights into PMS magnetic flaring; these include the Chandra

137: Orion Ultradeep Project \citep[COUP,][]{Getman05} and the

138: XMM/Newton Extended Survey of Taurus \citep[XEST,][]{Gudel07}.

139: Both astrophysical study of the properties of individual flares,

140: and statistical study of many flares, from the COUP and XEST

141: observations reveal that most events are similar to solar magnetic

142: flaring but with X-ray luminosities enhanced $10^3-10^5$ fold in

143: intensity \citep[e.g.][]{Favata05, Wolk05, Flaccomio05, Stassun06,

144: Maggio07, Caramazza07, Arzner07, Stelzer07, Franciosini07}.

145: 

146: 

147: 

148: The decay of PMS X-ray flares on timescales of $10^3-10^5$~s have

149: proved particularly amenable to astrophysical modeling.   One

150: favored model, developed by \citet{Reale97} and extensively

151: applied to solar and stellar X-ray flares, considers the X-rays

152: produced by thermal plasma at $T \sim 10^7$~K confined in a

153: cylindrical loop cooling by radiation and conduction but subject

154: to possible reheating by later magnetic reconnection events. Using

155: such models, inferences can be made concerning magnetic fields

156: responsible for the flare, including loop length and magnetic

157: field strength.  For the most powerful PMS flares, inferred loop

158: lengths reach $4-20$~R$_\star$, comparable to the expected inner

159: edges of protoplanetary disks \citep{Favata05}\footnote{Recall,

160: however, that the vast majority of weaker PMS flares arise in

161: magnetic loops no larger than the star

162: \citep[e.g.][]{Imanishi03,Wolk05, Franciosini07}.}. Thus, X-ray

163: flaring potentially probes the region of star-disk interaction.

164: Links are also emerging between other features of PMS X-ray

165: emission, such as the statistical suppression of flares in

166: accreting systems and the rotational modulation of X-rays, and

167: models of multipolar PMS magnetospheres \citep{Gregory06,

168: Gregory07}.

169: 

170: 

171: 

172: The present study extends the analysis of powerful flares from the

173: COUP made by \citet{Favata05}.  They modeled 32 flares using a

174: traditional technique of time-resolved spectroscopy. In

175: \citet[][Paper I]{Getman08} and \citet{Getman06}, we introduce a

176: more sensitive data analysis method based on adaptively smoothed

177: median energies which permits modeling of 216 COUP flares.  This

178: is a sufficiently large sample that permits new investigation of

179: possible links between flaring, PMS magnetospheres and

180: protoplanetary disks. Here, we present observational evidence that

181: X-ray emitting coronal structures are in fact truncated by inner

182: disks around the Keplerian corotation radius, just as predicted by

183: PMS theoretical models outlined above.  We also report a variety

184: of other results, both positive and negative, linking magnetic

185: flare properties to the presence or absence of disks.

186: 

187: 

188: 

189: The paper is organized as follows.  Preliminary work and recovery

190: of some established results are given in \S \ref{disk_section}.

191: The main results of our study appear in the following three

192: sections: the absence of strong links between some flare

193: properties and disks (\S \ref{flare_prop_section}); clear evidence

194: that PMS magnetospheres do not extend beyond disk inner edges (\S

195: \ref{truncation_section}); and possible relations between

196: super-hot flares,  accretion and non-dipolar magnetic fields (\S

197: \ref{superhot_section} and \ref{magnetic_fields.section}).

198: Discussion follows in section \S \ref{discussion_section}. Readers

199: are encouraged to consult Paper I for details on the selection,

200: modeling, and properties of the 216 COUP flares discussed here.

201: 

202: 

203: 

204: 

205: \section{PRELIMINARY CONSIDERATIONS \label{disk_section}}

206: 

207: 

208: 

209: \subsection{Disk and accretion indicators \label{indicators_section}}

210: 

211: 

212: 

213: We follow a well-established path in defining the presence of

214: disks and accretion in PMS stars. The $K_s$-[3.6]-[4.5]

215: color-color diagram on Figure \ref{fig_Ksch1_vs_ch12} shows 98

216: sources with available IR photometry (both $\Delta(H-K_s)$ and

217: $[3.6]-[4.5]$; see Table 4 of Paper~I). This diagram is known to

218: provide good discrimination between Class~I (protostellar),

219: Class~II (classical T Tauri) and Class~III (weak-lined T Tauri)

220: systems \citep[e.g.][]{Hartmann05}. The reddening vector of $A_V

221: \sim 30$~mag for a diskfree Vega-like spectrum assuming the

222: reddening law of \citet{Mathis90} demarcates the locus of

223: Class~III objects to the left from the Class~II and Class~I

224: systems to the right. All COUP 98 objects, except for three with

225: their color $[3.6]-[4.5]>0.5$, have their inferred X-ray column

226: densities $\log N_H < 22.3$~cm$^{-2}$ (Table~4 of Paper~I). For

227: normal interstellar gas-to-dust ratios, this corresponds to a

228: visual absorption of $<10-12$~mag \citep{Vuong03} and thus to the

229: limiting color of $[3.6]-[4.5]<0.2$ for Class~III objects. Several

230: stars to the left of the reddening vector and with $[3.6]-[4.5]$

231: color in excess of 0.2 may be systems with either higher

232: photometric errors and/or special reddening conditions in their

233: individual star-disk systems. Nevertheless, the rough mid-infrared

234: (MIR) color criterion of $[3.6]-[4.5]=0.2$ effectively

235: discriminates between Class~III and Class~II systems, and we use

236: this as our main MIR disk indicator. A few COUP sources with

237: $K_s-[3.6] \sim 2$ may be Class~I or transitional Class~I/II

238: systems.  One of them, COUP \#570, is classified as 0/Ib

239: protostellar candidate in \citet{Prisinzano07}.

240: 

241: 

242: $\Delta(H-K_s)$ near-infrared (NIR) excess is an indicator of a

243: heated inner dusty circumstellar disk.  It is measured from the

244: reddening vector on $J-H$ versus $H-K_s$ diagram of COUP sources

245: \citep[e.g. Figure 5$a$ in][]{Favata05} adopting photospheric

246: colors associated with PMS stars at age 1~Myr using the models of

247: \citet{Siess00}, and considering a reddening vector applied to

248: $0.1$~M$_{\odot}$ stars. Only four sources in our sample have

249: masses $M<0.2$~M$_{\odot}$, but 40 have $0.2<M<0.4$~M$_{\odot}$.

250: In order to allow better discrimination of inner disks, we can

251: relax the criterion $\Delta(H-K_s)=0$~mag to the value in the

252: range of $(-1,-0.06)$~mag corresponding to reddening vectors at

253: the mass range of $0.2-0.4$~M$_{\odot}$.

254: 

255: 

256: Using this NIR color excess measure, we find that a criterion of

257: $\Delta(H-K_s) = -0.06$~mag is a good discriminator between

258: diskfree and disky stars established with the MIR color criterion

259: of $[3.6]-[4.5]=0.2$; 95\% of sources classified as Class~III

260: using this MIR color criterion are also Class~III using our NIR

261: color criterion (Figure \ref{fig_Ksch1_vs_ch12}).  About 17 stars

262: are classified as Class~II systems using MIR colors but Class~III

263: using NIR colors;  these systems likely have evolved disks with

264: inner holes.

265: 

266: 

267: Discrimination between accreting and non-accreting objects employs

268: the equivalent width of the IR 8542~\AA\ Ca~II line, ${\rm

269: EW(Ca~II)}$, measured from low-resolution spectroscopy by

270: \citet{Hillenbrand97}.  In Figure \ref{fig_dhks_vs_ewca}, we adopt

271: the classification criterion used by \citet{Flaccomio03}: stars

272: with the line in emission with ${\rm EW(Ca~II}) < -1$~\AA\ are

273: accretors while stars with the line in absorption and equivalent

274: width of ${\rm EW(Ca~II}) > 1$~\AA\ are considered to be

275: non-accretors. Stars with intermediate values $-1 < {\rm

276: EW(Ca~II}) < 1$~\AA\ have an indeterminate classification. Figure

277: \ref{fig_dhks_vs_ewca} shows that this ${\rm EW(Ca~II)}$ accretion

278: indicator agrees in most cases with the $\Delta(H-K_s)$ NIR inner

279: disk indicator: inner disk photometric excess is seen in 15 of the

280: 19 accretors and no photometric excess is seen in 32 of the 41

281: non-accretors. COUP flaring sources with their conservatively

282: chosen ${\rm EW(Ca~II}) < -2$~\AA\ will be further classified in

283: the text as high-accretors\footnote{The association between

284: accreting stars and MIR-excess stars in our sample is not perfect.

285: Four high accretors located at the outskirts of the $Chandra$

286: field are not part of the MIR disk sample because they lie outside

287: the $Spitzer$ IRAC fields from which we obtained MIR photometry.

288: An additional high accretor lies projected against the

289: infrared-bright BN/KL region which is classified as a diskfree

290: star based on its (possibly erroneous) MIR photometry. See also \S

291: \ref{cloud_section} and Figure \ref{fig_ultrahot_3}.

292: \label{footnote_accretors}}.

293: 

294: 

295: Recall that photometric and inferred properties (such as age, mass

296: and rotation) for these COUP stars are tabulated by

297: \citet{Getman05}, and tables in Paper~I present various observed

298: and inferred stellar and flare quantities. Tables

299: \ref{tbl_correl_known} and \ref{tbl_correl_new} of the current

300: paper give general statistical properties of those quantities.

301: They also provide probabilities $P_{KS}$ from two-sample

302: univariate Kolmogorov-Smirnov tests comparing the distributions of

303: various quantities in diskfree and disky stars. Significant disk

304: effects may be present when $P_{KS} \la 0.05$. This is relevant

305: for such quantities as rotational period, Keplerian corotation

306: radius, peak flare plasma temperature, coronal loop size relative

307: to stellar radius, and X-ray coronal extent relative to corotation

308: radius.

309: 

310: 

311: 

312: 

313: \subsection{Disks and stellar rotation \label{rotation_section}}

314: 

315: 

316: 

317: Using our sample of X-ray bright COUP stars, we confirm the

318: well-established result that Orion Nebula Cluster PMS stars with

319: disks rotate slower than those without disks

320: \citep[e.g.,][]{Herbst02, Rebull06}. Our rotational periods are

321: obtained from published sources but our MIR photometry was derived

322: independently from {\it Spitzer Space Telescope} data as described

323: in Paper~I and our classification of disky versus diskfree stars

324: was determined as described above.

325: 

326: 

327: This expected result is shown in Figure \ref{fig_rotation} where

328: stars with MIR disks (blue circles) and accretion (green boxes)

329: systematically have longer periods and larger corotation

330: radii\footnote{Recall that Keplerian corotation radii for stars

331: with known rotational periods $P$ and masses $M$ are calculated in

332: Paper~I as $R_{cor} = (G M P^2/4 \pi^2)^{1/3}$.} than diskfree

333: stars (red circles). This supports the explanation that slow PMS

334: rotation is due to the loss of a stellar angular momentum through

335: a magnetic star-disk interaction \citep[see review

336: by][]{Bouvier07a}. Table \ref{tbl_correl_known} shows that the

337: rotational periods of our bright X-ray flaring stars are

338: well-separated with a median of 9.0 days for those with MIR disks

339: compared to 3.5 days for those without MIR disks. Rotation of our

340: accreting systems are not distinguishable from other stars with

341: MIR disk.  It is useful to note that the range of COUP rotational

342: periods of $[1-10]$~days translates into the range of corotation

343: radii of $[2 - 10]$~R$_{\star}$ with the majority of COUP stellar

344: radii in $[1.4 - 3]$~R$_{\odot}$ range (Figure

345: \ref{fig_rotation}).

346: 

347: 

348: 

349: 

350: \subsection{Disks and location in the cloud \label{cloud_section}}

351: 

352: 

353: 

354: Figure \ref{fig_ultrahot_3} shows the spatial distribution of our

355: stars with and without MIR disks in the Orion Nebula region. The

356: region is complicated: the rich, optically bright Orion Nebula

357: Cluster has low absorption and lies in front of the two OMC 1

358: molecular cloud cores, as well as widely distributed molecular

359: cloud material, where PMS stars are highly absorbed.  In the

360: figure, the stars are coded both by their disk properties and by

361: their absorption measured from the soft X-ray absorption in their

362: COUP spectra \citep{Getman05}. The spatial distribution of the

363: COUP stars is further discussed by \citet{Feigelson05} and

364: \citet{Prisinzano07}.

365: 

366: 

367: Bright flaring COUP stars with MIR disks and the highest

368: absorption (blue circles in panel $a$) are largely concentrated

369: around OMC 1 molecular filament to the north of the

370: Becklin-Neugebauer star forming region. Those with intermediate

371: absorption have a somewhat broader distribution centered

372: north-east of the OMC cloud cores. The diskfree stars in our

373: sample are generally less absorbed than those with MIR disks with

374: a dispersed spatial distribution similar to the disky stars with

375: intermediate absorption (Figure \ref{fig_ultrahot_3}$b$). For both

376: samples, stars harboring super-hot ($T_{obs,pk} > 100$~MK, see

377: \S~\ref{superhot_section}) flares are localized within the

378: north-eastern part of the cloud. The high accretors appear widely

379: dispersed (Figure \ref{fig_ultrahot_3}$c$), but this is a

380: selection effect - only 12 high accretors have their observed

381: X-ray net counts above the count threshold of our flare analysis

382: ($NC = 4000$~counts, see Paper~I and footnote

383: \ref{footnote_accretors}). The total known population of COUP high

384: accretors have spatial distribution similar to that of the bright

385: flaring COUP MIR disk stars with intermediate absorption (Figure

386: \ref{fig_ultrahot_3}$d$).

387: 

388: 

389: 

390: \section{FLARE PROPERTIES AND PROTOPLANETARY DISKS \label{flare_prop_section}}

391: 

392: Paper I presents in detail our analysis of the selected COUP flares and the

393: derivation of flare properties used in the analysis here: rise and decay

394: timescales, ``characteristic'' pre-flare and peak flare X-ray luminosities, peak

395: flare plasma temperature and emission measure, flare morphology, and

396: magnetic loop length responsible for the flare.  These

397: quantities are derived using the astrophysical model of flare decays developed

398: by \citet{Reale97}. This is a single loop model and is simplistic in a number of

399: ways (see its limitations in sections 2.2 and 2.5 of Paper~I). However, even in

400: the case of complex flares, it is appropriate to apply the model to a lightcurve

401: segment if there is an indication for the presence of a single ``dominant''

402: flaring structure (see sections 2.5 and 2.6 of Paper~I).

403: 

404: Adopting the model of Reale et al., \citet{Favata05} have analysed the strongest 32 COUP

405: flares using the long-standing flare analysis method of time-resolved

406: spectroscopy (TRS). To extend the flare sample of \citet{Favata05}, Paper~I

407: utilizes a more sensitive technique of flare analysis, the ``method of

408: adaptively smoothed median energy'' (MASME), introduced by \citet{Getman06}.

409: Instead of performing classical TRS with XSPEC over only a few characteristic

410: flare intervals, we employ an adaptively smoothed estimator of the median

411: energy of flare counts and count rate to infer the evolution of plasma

412: temperature and emission measure at dozens of time points along the decay phase

413: of a flare. This is achieved by calibrating median energies and count rate to

414: temperatures and emission measures through simulations of high

415: signal-to-noise spectra at fixed source's column density. The method permits

416: modeling of flare spectra on more rapid timescales  and with weaker signals

417: than was possible using traditional spectral fitting. Readers are encouraged to

418: consult section 2.2 and Appendices A and B of Paper~I on details of the method.

419: The result is that Paper~I  characterizes 216 COUP flares for analysis here,

420: in contrast to 32 flares analyzed by \citet{Favata05}.

421: 

422: 

423: 

424: 

425: \subsection{Disks have no effect on flare morphology \label{morphology_section}}

426: 

427: Paper I describes our qualitative classification of X-ray flares

428: by their lightcurves during the flare.  The four classes are:

429: `typical' flares with fast rises and slower decays characteristic

430: of most solar flares; `step' flares with extra emission during the

431: decay attributable to a reheating or a secondary reconnection

432: event;  `double' flares with two peaks suggestive of two nearly

433: simultaneous reconnection events; and `slow rise, top flat' (SRTF)

434: flares.  The first three classes are commonly seen in solar

435: flares.  For example, Figure~11 of Paper I shows solar flares with

436: secondary events during the decay of powerful long-duration

437: events.  Such events, scaled up several orders of magnitude, could

438: be classified as step or double flares in COUP lightcurves.

439: 

440: 

441: The SRTF flares are more unusual, and one may speculate that they

442: may be selectively formed in certain PMS stars.  For example, they

443: might be reconnection events associated with sheared star-disk

444: magnetic fields \citep{Montmerle00, Isobe03} rather than events

445: associated with field lines attached to the star. Flare morphology

446: is coded by different symbols in Figures \ref{fig_dhks_vs_ewca},

447: \ref{fig_L_vs_dHKs}, \ref{fig_L_vs_ch12} and \ref{fig_L_vs_ewca}.

448: We examine here the distribution of flare types along the abscissa

449: measuring the strength of accretion, NIR inner disk, and MIR disk

450: indicators. No pattern between flare morphology and disk

451: indicators is seen. In paricular, SRTF flares are seen in systems

452: with and without MIR and NIR disks.  However, SRTF flares are not

453: seen in high-accretion systems; this can be an indication of a

454: real physical effect (\S \ref{energetics_section}) or simply

455: attributed to the very limited statistics of both accreting and

456: SRTF flare samples.

457: 

458: 

459: 

460: 

461: \subsection{Disks are unrelated to flare energetics \label{energetics_section}}

462: 

463: 

464: 

465: We examine here whether any relationships are present between

466: disks and quantities associated with the strength of the X-ray

467: flares:  rise and decay timescales, peak luminosities, and total

468: energies in the X-ray bands.

469: 

470: 

471: 

472: Figure \ref{fig_energy_contrib} $a$-$b$ show the distributions of

473: flare rise and decay times stratified by disk and accretion

474: properties.  There is a hint that high-accretors have

475: systematically shorter flare timescales compared to the rest of

476: the bright COUP flare sample.  The evidence is only suggestive

477: because our sample of high-accretors is small (17 flares from 12

478: sources) and it is only marginally significant when measured by a

479: KS test ($P = 0.06$ when rise times for high-accretors are

480: compared to diskfree systems).  Figure \ref{fig_energy_contrib}

481: $c$ shows that the distribution of flare peak luminosities are

482: indistinguishable for diskfree, MIR disk, and accretion disk

483: systems.

484: 

485: 

486: 

487: We evaluate the total energies of each of the 216 flares as the

488: difference between the time-integrated flare energy $E_{flare}$

489: and the energy from the non-flare ``characteristic'' state

490: $E_{char}$ within the duration of the flare, ${\rm t}_{flare2}-

491: {\rm t}_{flare1}$.  We estimate $E_{flare} \approx L_{X,pk} \times

492: \tau_{decay2}$ and $E_{char} = L_{X,char} \times ({\rm

493: t}_{flare2}-{\rm t}_{flare1})$, where X-ray luminosity from the

494: ``characteristic'' state, $L_{X,char}$, was taken from our TRS

495: analysis (Paper I). $E_{char}$ is systematically lower in flares

496: from high-accretors (panel $d$);  this is due to the shorter flare

497: durations (above) and lower $L_{X,char}$ in accreting

498: systems\footnote{In our flare sample the effect of lower

499: $L_{X,char}$ in accreting systems is not strong. A KS test gives

500: only marginally significant difference ($P_{KS} = 0.1$) in

501: $L_{X,char}$ between high-accretors and diskfree stars. Median

502: values of $\log (L_{X,char})$ are 30.16 erg~s$^{-1}$  for diskfree

503: stars, 30.10 erg~s$^{-1}$  for MIR disk stars,  and 30.10

504: erg~s$^{-1}$ for high-accretors.}. The latter effect is the

505: well-established suppression of time-integrated X-ray emission in

506: accreting vs. non-accreting PMS systems \citep[][and references

507: therein]{Gregory07}. Due to shorter flare timescales, $E_{flare}$

508: in high-accretors shows a somewhat narrower distribution than that

509: of other stars but the difference is not statistically significant

510: (panel $e$).

511: 

512: 

513: A more interesting effect is seen in the ratio

514: $E_{flare}/E_{char}$ which is systematically larger in

515: high-accretors that other stars. The median value

516: $E_{flare}/E_{char} \simeq 10$ compared to 5 in diskfree stars

517: ($P_{KS}=0.008$, Figure \ref{fig_energy_contrib}$f$). In the

518: 840~ks of the COUP observation, a typical single bright flare with

519: a duration of 90~ks (median of ${\rm t}_{flare2}-{\rm t}_{flare1}$

520: for all 216 flares) may increase the time-integrated source X-ray

521: luminosity $1.5-2$ times if $E_{flare}/E_{char} = 5-10$. For a

522: shorter more typical $Chandra$ exposure,  a 50~ks bright flare

523: ($<22\%$ of flares analyzed here have durations $<50$~ks) will

524: change the time-integrated source X-ray luminosity $3-5.5$ times

525: if $E_{flare}/E_{char} = 5-10$.

526: 

527: 

528: The major result of the section is that no difference is found in

529: flare duration, peak luminosity and total energy between flares

530: occurring in disky and diskfree systems. However, as a tentative

531: finding flares from high-accreting disky stars seem to be somewhat

532: shorter and thus have weaker total X-ray energies than the rest of

533: the analyzed flares. This tentative finding is supported by two

534: statistically significant findings: 1. super-hot COUP flares are

535: found to be shorter than cooler COUP flares (Paper~I) and 2.

536: super-hot flares preferentially present in accreting systems (\S

537: \ref{superhot_section}).

538: 

539: 

540: 

541: 

542: \section{DISKS MAY TRUNCATE PRE-MAIN SEQUENCE MAGNETOSPHERES

543: \label{truncation_section}}

544: 

545: 

546: 

547: Figures \ref{fig_L_vs_dHKs},  \ref{fig_L_vs_ch12} and

548: \ref{fig_L_vs_ewca} compare the flare loop lengths inferred from

549: analysis of the X-ray spectral evolution of the decay phases

550: (Paper I) with our disk and accretion indicators. After careful

551: investigation of various measures of magnetospheric size, we

552: choose to examine the ratio of the coronal extent of the loop as

553: measured from the star center, $L+R_{\star}$, to the Keplerian

554: corotation radius $R_{cor}$ determined by the stellar rotation

555: rate.  We are thus less interested in the loop size measured in

556: meters than the relative sizes of the loop to the likely location

557: of the inner edge of the disk.  This measure reduces variations

558: associated with star mass, age and rotation and focuses on the

559: question of the relationship between PMS disks and magnetospheres.

560: 

561: 

562: 

563: Figure \ref{fig_L_vs_dHKs} plots $(L+R_{\star})/R_{cor}$ against

564: the NIR $\Delta(H-K_s)$ disk indicator. The vertical lines for

565: each flare do not represent error bars, but are the low and upper

566: boundary of the inferred loop size ranges (Paper I) with symbols

567: positioned at the mean of those ranges. We find that, except for

568: COUP \#~1608\footnote{This outlier is a visual double unresolved

569: by 2MASS, and we suspect the NIR photometry is unreliable. It lies

570: outside the field of the $Spitzer$ IRAC images, so its MIR

571: properties are unavailable.}, coronal structures responsible for

572: flares detected from a dozen sources with NIR inner disks do not

573: exceed $R_{cor}$, while coronal structures responsible for $\sim

574: 40\%$ of flares from 43 sources without NIR inner disks exceed

575: $R_{cor}$. Some of these flares in diskfree systems arise from

576: loops reaching $\ga 2 \times R_{cor}$. This pattern is present in

577: each morphological flare type (typical, step, double, slow

578: rise/flat top) as indicated by different symbols.

579: 

580: 

581: 

582: The same pattern is seen when  $(L+R_{\star})/R_{cor}$ is plotted

583: against the MIR disk indicator $[3.6]-[4.5]$ (Figure

584: \ref{fig_L_vs_ch12}) and the accretion indicator ${\rm EW(Ca~II)}$

585: (Figure \ref{fig_L_vs_ewca}).  Virtually all of the flares whose

586: inferred sizes exceed the host star's corotation radius are

587: diskfree and non-accreting systems, while disky and accreting

588: flare loops all lie within the corotation radius.  The MIR plot

589: adds nine flares from five sources which were not available in the

590: NIR plot; all of these follow the NIR trend.  Two outliers, COUP

591: \#205 and 485, are those for which inner part of the disk is

592: likely cleared of circumstellar material (judging from their NIR

593: colors). Again, the trend is strong and applicable to each

594: morphological type of flares. The sample of stars with strong

595: ${\rm EW(Ca~II)}$ emission indicating active accretion is smaller

596: than those with infrared photometric excesses, but the trend of

597: smaller loop sizes in accreting systems is still clearly seen.

598: 

599: 

600: These three figures provide strong and consistent support for a

601: model where protoplanetary disks truncate PMS magnetospheres. We

602: do not know of any selection effect that would have produced this

603: pattern in a spurious fashion.  The plots in Figure

604: \ref{fig_corona_extent_suppl} help elucidate this trend. Here the

605: symbol colors represent the classification of disky (blue) and

606: diskfree (red) stars based on MIR colors, and only the mean value

607: of inferred flare loop size is shown. Recall from Figure

608: \ref{fig_rotation} that, due to the well-established connection

609: between disks and rotation, corotation radii scaled to stellar

610: radii are systematically larger for disky compared to diskfree

611: stars (the difference is roughly a factor of two). This difference

612: in $R_{cor}/R_{\star}$ is the major contributor to the difference

613: in $(L + R_{\star})/R_{cor}$ between the disky and diskfree stars

614: seen in Figures \ref{fig_L_vs_dHKs}-\ref{fig_L_vs_ewca}.   If the

615: loop sizes $L/R_\star$ were to be considered without normalization

616: to the corotation radii, the difference between the two classes

617: becomes marginal\footnote{To avoid observational bias of Class~III

618: stars towards higher $R_{star}$ and thus restricting the flare

619: sample to stars with stellar radius in the range of

620: $R_{\star}=[1-3]$~R$_{\odot}$ gives the following results: KS test

621: shows no statistical difference ($P_{KS} = 0.2$) in $L/R_{\star}$

622: between Class~II and Class~III; median values of

623: $L/R_{\star}=3.2(2.4)$ for Class~II(III) suggest that

624: $L/R_{\star}$ in Class~II stars is 1.3 times larger than that of

625: Class~III.}.

626: 

627: We thus find that X-ray coronal extents are somewhat similar in

628: diskfree and disky systems but, due to the well-established fact

629: that diskfree stars are faster rotators, the X-ray flares often

630: exceed the corotation radius in these systems. In contrast, X-ray

631: flare loops on disky stars never exceed the corotation radius,

632: although in some cases they reach the corotation radius which is

633: also the likely truncation radius for the circumstellar disk. The

634: fact that X-ray loops of disky stars are close to but never exceed

635: {\bf the one corotation radius} supports long-standing models of

636: star-disk magnetic interaction at this inner edge involving

637: accretion, outflow ejection, and regulation of the stellar angular

638: momentum. The very large loop sizes seen in all types of PMS stars

639: point to confinement by strong magnetic fields of T-Tauri stars

640: which, particularly in rapidly rotating diskfree systems, are

641: capable of withstanding the effects of centrifugal forces

642: \citep{Jardine99}.

643: 

644: 

645: 

646: 

647: 

648: \section{Super-hot flares and disks \label{superhot_section}}

649: 

650: Paper I describes the selection of 73 of the 216 COUP flares as

651: `super-hot' with peak plasma temperatures $T_{pk} \ga 100$~MK.

652: Similar flares have been occasionally reported in other PMS

653: systems such as the diskfree binary PMS star system V773 Tau

654: observed with ASCA \citep{Tsuboi98}, two embedded young systems in

655: the NGC~2264 star forming region \citep{Simon05}, and about half

656: of the COUP flares studied by \citet{Favata05}.  However, here we

657: have a sufficiently rich sample to study super-hot PMS flares as a

658: class. This is made possible by the large Orion population, the

659: unusually long COUP  exposure, and our new highly-sensitive flare

660: analysis techniques.

661: 

662: There is some concern that the $Chandra$ telescope cannot discern

663: differences in plasma temperatures above $\sim 100$~MK due to the

664: rapid decline in mirror reflectivity above $\sim 8$~keV.  However,

665: we explain in detail in Appendix B of Paper I and the Appendix

666: below that, when high-signal flares are considered, that

667: discriminations between 100~MK and $\ga 200$~MK peak temperatures

668: as well as between $\la 100$~MK and $\ga 100$~MK are possible

669: using the median energy as a temperature indicator (our MASME

670: technique described in Paper I). We argue that $\ga 200$~MK

671: super-hot peak temperatures were missed by the more traditional

672: time-resolved spectroscopy techniques.

673: 

674: To check on the applicability of the Reale model to unusually large and hot 

675: loops the detailed time-dependent hydrodynamic simulation was applied to a

676: typical super-hot flare of the COUP source \#1343 with observed peak flare 

677: temperature of a few-to-several hundred MK and derived sizes of associated coronal 

678: flaring structures of $\sim 10^{12}$~cm \citep[section 4.1 in][]{Favata05}. 

679: The model achieved peak temperature of $\sim 200$~MK and matched both, the observed flare spectrum 

680: and lightcurve. The loop plasma was heated rapidly to $\sim 200$~MK on time-scale of

681: 1 hour following by explosive chromospheric evaporation with chromospheric 

682: plasma reaching the loop apex on similar time-scale of 1 hour. This was followed

683: by the decay phase from the nearly equilibrium state governed by conduction

684: and radiation cooling processes on a time-scale of a few to several hours.

685: 

686: 

687: Table \ref{tbl_correl_new} shows that the flare peak temperatures of stars with

688: MIR disks are systematically higher than peak temperatures for stars without MIR

689: disks at high statistical significance ($P_{KS} = 0.9\%$). It is not clear why

690: the effect is not seen in $K$-band excess systems.  Figure \ref{fig.peakTdisk}$a$

691: shows that this effect is even more prominent when disk-free stars are compared

692: to high accreting stars based on the Ca~II emission line indicator for accretion.

693: Flares from MIR-excess disk stars (solid black distribution) are on average

694: hotter than those from non-disk stars (dotted) and, among flares from disk stars,

695: those from high-accretors (dashed) are the hottest. As a consequence,

696: high-accretors have the largest fraction of super-hot ($T_{obs,pk} > 100$~MK)

697: flares (53\%) compared to that of stars with MIR disks (40\%)

698: and diskfree stars (27\%). The effect of systematically hotter flares in disk

699: and highly accreting stars becomes even more prominent (at a significance level

700: of $P_{KS} = 0.02\%$) when only flares from $M<2$~M$_{\odot}$ stars are

701: considered (Figure \ref{fig.peakTdisk}$b$). Even when uncertainties on

702: individual values of $T_{obs,pk}$ (which can be large for very high

703: temperatures; Appendix B in Paper~I) are taken into account through Monte-Carlo

704: simulations, this significance level does not exceed $P_{KS}=1.5\%$\footnote{

705: We simulated and compared 10000 temperature distributions for each of the sample

706: (diskless, disk and high-accretors), with individual temperature values randomly

707: drawn from Gaussian distributions with mean equal to the measured $T_{obs,pk}$ 

708: and variance as an average error reported in Appendix B of Paper~I. We find 

709: that $(50\%,68\%,90\%)$ of 10000 resulting significance levels from both, the 

710: K-S tests between diskless vs. disk and diskless vs. high-accretors comparisons,

711: do not exceed values of $P_{KS} = (0.3\%,0.5\%,1.5\%)$, respectively.}. We thus

712: find strong evidence that super-hot flares are preferentially associated with 

713: PMS stars undergoing substantial accretion.

714: 

715: Disky stars which possess super-hot flares, including accretors,

716: tend to have spectral types late-K and M corresponding to masses

717: $M \la 1 - 2$~M$_\odot$, while the majority of diskfree stars with

718: super-hot flares are more massive with spectral types early-K

719: through F (thin dashed line in Figure \ref{fig.peakTdisk}$b$). This 

720: latter group is not large: for sources with known

721: masses, 66\% of the super-hot flares are produced by stars with

722: $M<1$~M$_{\odot}$ while only 15\% are produced by stars with

723: $M>2$~M$_{\odot}$. This can be attributed to the rarity of

724: intermediate-mass stars compared to lower mass stars in the

725: Initial Mass Function\footnote{Out of 161 COUP stars analyzed here

726: 83 have known masses of $M<1$~M$_{\odot}$ and 23 have masses of

727: $M>2$~M$_{\odot}$. Out of 57 stars producing super-hot flares 27

728: have known masses of $M<1$~M$_{\odot}$ and 7 have masses of

729: $M>2$~M$_{\odot}$. The fraction of low-mass super-hot stars

730: ($27/83$) is comparable to that of higher-mass stars ($7/23$).}.

731: The bottom line seems to be that in low-mass stars

732: ($M<1$~M$_{\odot}$) the appearance of super-hot flares is

733: connected to the presence of active (accreting) disks, however

734: super-hot flares may also arise in massive ($M>2$~M$_{\odot}$)

735: diskfree stars.

736: 

737: 

738: 

739: 

740: \section{Magnetic fields geometries \label{magnetic_fields.section}}

741: 

742: Our analysis can provide indirect access to the geometry and

743: strength of the large-scale magnetic fields responsible for

744: confining the X-ray emitting plasma during the bright PMS flares

745: studied here. As described in Paper I, the flare decay model gives

746: estimates of the flare plasma peak emission measure ($EM_{pk}$),

747: coronal loop size (we use here the mean value of the size range,

748: $L$), and plasma temperature ($T_{obs,pk}$). If we add an

749: assumption concerning the ratio of the cylindrical loop

750: cross-sectional radius to the loop length $\beta$, we can derive

751: the plasma electron density $n_e$ from the emission measure and

752: loop length. The magnetic field confining the plasma can then be

753: estimated assuming pressure equilibrium,

754: \begin{equation}

755: B_{eq} \simeq (8 \pi \times 2 n_e k T^\prime_{pk})^{1/2} ~ {\rm where} ~ \\

756: n_e \simeq (EM_{pk}/(2 \pi \beta^{2} L^{3}))^{1/2}.

757: \label{Bfield_eqn}

758: \end{equation}

759: \noindent Following past flare models, we adopt a 10:1 cylindrical

760: geometry, $\beta = 0.1$.  $T^\prime_{pk}$ used here is the plasma

761: temperature at the loop apex, which is hotter than the observed

762: X-ray temperature integrated over the entire loop according to

763: $T^\prime_{pk} = 0.068 \times T_{obs,pk}^{1.2}$ where both

764: temperatures are in units of K \citep{Reale02,Favata05}.

765: 

766: 

767: 

768: Figure \ref{fig_ultrahot_5} shows the resulting inferred plasma

769: densities (panel $a$) and magnetic field strengths (panel $b$) for

770: all (147 of the 162) flares with known loop sizes and stellar

771: radii of their host stars. Plasma density $n_e$ is anticorrelated

772: with the loop size $L$ as expected from equation (1); the grey

773: line shows the relation $n_e \propto L^{-3/2}$ expected for

774: constant emission measure and stellar radius.   Estimated plasma

775: densities\footnote{ We note that the highest loop plasma densities

776: derived here from modeling flare events are comparable to the high

777: densities inferred from high-resolution X-ray spectra which are

778: usually attributed to shocks at the base of accretion streams

779: \citep[e.g.,][]{Kastner02, Schmitt05, Drake05, Argiroffi07,

780: Gudel07b, Huenemoerder07}.  This indicates that plasma density

781: alone may not be a reliable discriminant between flare and

782: accretional X-rays.} range  from $4 \times 10^9$~cm$^{-3}$ for

783: very large loop sizes ($L \sim 10$~$R_{\star}$) to $1 \times

784: 10^{12}$~ cm$^{-3}$ for small loop sizes of $L \la

785: 0.2$~$R_{\star}$.

786: 

787: 

788: 

789: Figure \ref{fig_ultrahot_5}$b$ plots the equilibrium magnetic

790: field strengths obtained from equation (1) of individual COUP

791: flares against the inferred flare loop lengths.  But the

792: relationship between magnetic field strength and loop length is

793: also a function of field geometry.  The simplest geometry often

794: assumed for PMS stars is a dipole,

795: \begin{equation}

796: B = \frac{B_{ph}}{(L/R_{\star}+1)^3}

797: \end{equation}

798: where $B_{ph}$ is the photospheric magnetic field which can be

799: directly measured. Figure \ref{fig_ultrahot_5}$b$ shows these

800: relations for several photospheric field strengths in the range

801: $1-6$~kG consistent with measurements of Zeeman broadening and

802: circular polarization of PMS photospheric lines

803: \citep{Johns-Krull99, Symington05, Johns-Krull07, Donati07}.

804: 

805: Figure \ref{fig_ultrahot_5}$b$ shows that for most of the flares

806: the magnetic fields estimated assuming pressure equilibrium are

807: consistent with the assumption of dipole geometry with

808: photospheric magnetic strengths comparable to the observed values

809: of $1-6$~kG. These include all types of systems: diskfree stars

810: (red circles), MIR disk stars (blue circles), active accretors

811: (green boxes), and higher mass stars (black triangles). Thus the

812: majority of the flares from our flare sample are associated with

813: the extended, dipole-like loops, and as already discussed in

814: Paper~I are different from the previously reported smaller stellar

815: flares which take place in the complex surface field regions

816: likely making up the ``characteristic'' level of X-ray emission.

817: 

818: 

819: However about 40 flares have inferred field strengths too strong

820: for their inferred loop lengths and/or have loop lengths too long

821: for any realistic dipole model.  Nearly all of these discrepant

822: flares have super-hot peak temperatures (magenta circles in Figure

823: \ref{fig_ultrahot_5}(b)).  If a dipolar topology and pressure

824: equilibrium holds, the implied surface magnetic field is tens of

825: kilogauss, far above reasonable values.

826: 

827: 

828: 

829: This discrepancy can not be fully explained by uncertainties in

830: inferred peak flare temperatures. Both Appendix B of Paper I and

831: the Appendix below provide detailed evidence that the existence of

832: super-hot peak temperatures with $T_{obs,pk} > 100$~MK, and often

833: $> 200$~MK, is reliable. The uncertainties of these super-hot

834: temperatures shown in Figure~15 of Paper~I propagate into loop

835: size uncertainties of 40\% and magnetic strength uncertainties of

836: 10\%. The confidence regions of most super-hot flares in Figure

837: \ref{fig_ultrahot_5}$b$ thus lie above the locus of dipolar fields

838: with photospheric strength of $B_{ph} \sim 10$~kG.

839: 

840: 

841: As discussed in \S \ref{superhot_section}, super-hot flares appear

842: to be associated with the presence of a disk and high accretion.

843: We thus emerge with evidence for anomalous magnetic fields

844: associated with accreting PMS stars. Super-hot flares typically

845: arise in large coronal loops $L/R_{\star} \ga 2$ but still inside

846: the corotation radius. In \S \ref{disc_stardisk.section}, we

847: speculate that such anomalous fields may result from the

848: distortion of magnetic topologies or the thickening of magnetic

849: loops by the process of accretion.

850: 

851: 

852: 

853: \section{Discussion \label{discussion_section}}

854: 

855: 

856: \subsection{Does flaring occur in star-disk magnetic fields?

857: \label{disc_stardisk.section}}

858: 

859: 

860: 

861: The first result we encounter in this study is negative: disks

862: appear to have no effects on flare morphology, timescales, or

863: energetics (\S \ref{morphology_section}-\ref{energetics_section}).

864: This is not a trivial finding; there is no reason to believe that

865: flares arising in field lines attached to the inner rim of the

866: circumstellar disk would have the same power, plasma properties

867: and temporal evolution as flares from magnetic loops anchored in

868: the stellar surface.   Reconnection in star-disk loops can occur

869: in several ways: by magnetic interactions between star and disk

870: fields near the corotation radius \citep{Hayashi96}, by stochastic

871: fluctuations in the accretion rate near the corotation radius

872: \citep{Shu97}, or by twisting of star-disk loops from stars

873: slightly out of corotation with their disks \citep{Birk00,

874: Montmerle00}.  \citet{Isobe03} calculated the hydrodynamical

875: response of cool plasma to a sudden reconnection event  in the

876: middle of $1 \times 10^{12}$~cm star-disk loop.  They find that

877: the shock reaches both the stellar and disk surfaces in

878: $1-2$~hours with bulk plasma motions sometimes exceeding 1000~km

879: s$^{-1}$. Temperatures quickly reach $50-100$~MK but the rise in

880: emission measure sometimes appears slower and less regular than in

881: typical solar-type flares.  These might be classified as

882: slow-rise-flat-top or double flares.  Remarkably, Isobe et al.

883: find that  the entire inner disk gas can be vaporized by powerful

884: reconnection events.

885: 

886: 

887: Thus, although SRTF and double flares conceivably could arise

888: exclusively in star-disk loops, our results indicate this is not

889: the case.  Instead, we find that all flare morphological types

890: occur equally in both diskfree and disky systems with similar

891: luminosity and duration distributions.  We therefore conclude that

892: all flare types likely arise in traditional solar-type magnetic

893: loops where both footprints are rooted in the stellar surface.

894: This idea is supported by the similarity of COUP flaring

895: statistics and plasma elemental abundance anomalies to those seen

896: in older magnetically active stars \citep{Wolk05, Maggio07,

897: Stelzer07}.   Figure 11 in Paper I shows that step and double

898: flares are commonly seen in the contemporary Sun, often because of

899: a triggered reconnection or a reheating event. Slow-rise-flat-top

900: flares may be similar to other flares with reconnection

901: sequentially progressing along multiple magnetic arcades. Thus,

902: all types of COUP flares may arise from solar-type magnetic

903: morphologies.

904: 

905: 

906: However, the solar-type flare analogy to powerful, long-lasting

907: COUP flares faces a challenge. \citet{Jardine99} has raised the

908: issue that very large magnetic loops anchored to rapidly rotating

909: stars may be destroyed when the centrifugal force of their

910: confined plasma exceeds the weakening magnetic tension in the

911: outer part of the loop. \citet{Favata05} considered this

912: instability to be a strong argument in favor of attaching long

913: loops to the circumstellar disks.  We tentatively conclude from

914: our findings that, while this centrifugal force might break some

915: magnetic loops, others survive to produce the observed powerful

916: flares in diskfree high rotating COUP stars.

917: 

918: 

919: There is one flare parameter which is linked to the presence of a

920: disk: flares on Class~II and high-accretion stars have

921: systematically hotter peak flare temperatures (\S

922: \ref{superhot_section} and Table \ref{tbl_correl_new}). This may

923: point to star-disk reconnection as modeled by \citet{Isobe03}

924: where temperatures of 100~MK are easily achieved. Additional

925: calculations should be made to establish the conditions where peak

926: temperatures of 200~MK or more occur.  However, it is also

927: possible that these higher temperatures are a byproduct of

928: systematically higher surface magnetic fields in Class II systems

929: which would be capable of confining hotter plasmas.  Younger

930: systems might have stronger fields due to vigorous dynamo

931: precesses \citep{Browning07}, a primordial field component

932: \citep{Johns-Krull07}, and/or accretional processes

933: \citep{Bessolaz08}.  Hotter peak temperatures could be a byproduct

934: of the magnetosphere compression by disks discussed in \S

935: \ref{disk_truncation.section}, or atypical coronal loop geometries

936: with larger aspect ratios than typical of solar coronal loops,

937: $\beta \gg 0.1$. Rare examples of such ``thick'' stellar coronal

938: loops are reported in the literature; for example, the AB~Dor~29

939: Nov 1997 flare has estimated $\beta \sim 1$

940: \citep{Maggio00,Reale07}.  We conclude that the correlation

941: between peak flare temperature and disks is an interesting new

942: result, but does not appear to have a unique explanation and does

943: not clarify whether or not reconnection occurs in star-disk field

944: lines.

945: 

946: 

947: 

948: 

949: \subsection{Magnetospheric truncation by a disk \label{disk_truncation.section}}

950: 

951: 

952: Our second observational result is positive: the COUP flares

953: provide the first direct evidence that the magnetic loops of

954: powerful X-ray PMS flares can exceed the stellar corotation radius

955: in diskfree stars, but do not exceed the corotation radius in

956: disky stars (Figures \ref{fig_L_vs_dHKs}-\ref{fig_L_vs_ewca}).

957: This finding is presented visually in Figure

958: \ref{fig_corona_extent_sketch} which shows with realistic relative

959: scaling the largest inferred X-ray coronal extents for Class~II

960: and Class~III stars.  The high-order multi-polar component of the

961: magnetic field is expected to dominate at the stellar surface but

962: fall off rapidly with height, leaving the dipolar component

963: dominant at distances above a couple of stellar radii

964: \citep[Figure 15 in][]{Donati07,Donati08}. Figure

965: \ref{fig_corona_extent_sketch} also shows schematically a

966: compression and distortion of magnetic field lines by the

967: accretion disk.  More realistic field configurations with both

968: accreting field lines and coronal loops are calculated by

969: \citet{Jardine06, Long07}.

970: 

971: 

972: For a T-Tauri star with radius of $\sim 2$~R$_{\odot}$ and mass of

973: $\sim 0.5$~M$_{\odot}$, the typical large flaring magnetic loops

974: are roughly the same for the disk classes with $L/R_{\star} \sim

975: 5$.  But, due to the strong rotational acceleration after disks

976: are gone, the corotation radii shrink so that loops formerly

977: $(L+R_{\star})/R_{cor} \la 1$ during their disky phase are often

978: $1 \la (L+R_{\star})/R_{cor} \la 2$ during their diskfree phase.

979: Table \ref{tbl_correl_known} shows that, as stars evolved from

980: Class~II to Class~III phases, the average corotation radii

981: decrease from $R_{cor} \sim 7$~$R_{\star}$ to $R_{cor} \sim

982: 3$~$R_{\star}$ due to a shortening of the average rotation periods

983: from $P \sim 9$~day to $P \sim 3$~day.

984: 

985: 

986: 

987: There are three independent supporting lines of evidence

988: suggesting that long loop structures can be present in

989: magnetically active stars when disks are not present.

990: \begin{enumerate}

991: 

992: 

993: \item Even the relatively inactive Sun has helmet-like X-ray

994: streamers that reach up to $\ga 0.5$~R$_{\odot}$ above the

995: photosphere.  These occur in the subclass of the solar long decay

996: events (LDEs).  The analogies between solar LDEs and COUP bright

997: X-ray flares are discussed in Paper I and in

998: \S~\ref{disc_LDE.section} below.

999: 

1000: 

1001: \item \citet{Mullan06} have carefully compared the loop sizes

1002: inferred from 106 flares studied with the $EUVE$ satellite from 33

1003: magnetically active main sequence stars, including members of RS

1004: CVn binaries and dMe stars.  They find that stars with $B-V<1.4$

1005: (hotter than M0) generally have loop lengths $L/R_\star < 0.5$

1006: while stars with $B-V>1.4$ (cooler than M0) often exhibit loop

1007: lengths $L/R_\star \sim 1.0-1.5$.

1008: 

1009: 

1010: \item Radio VLBI studies have shown that magnetospheres likely

1011: extend several stellar radii in very active stars, filling with

1012: trans-relativistic electrons emitting gyrosynchrotron radiation.

1013: This is indicated by VLBI imaging of the RS CVn binary systems UX

1014: Ari HR 1099, and HR 5110 and the dMe stars YY Gem and UV Cet

1015: \citep{Lang94, Alef97, Benz98, Franciosini99, Ransom02, Ransom03}.

1016: While none of these systems showed loops larger than $\sim 2$

1017: times the photospheric radius, recent VLBI mapping of the nearby

1018: diskfree PMS binary system V773 Tau reveals radio structures

1019: extraordinarily far from the component stars.  Their elongated

1020: magnetospheres resembling huge solar helmet streamers extending

1021: $\simeq 20$~R$_\star$ yet are still anchored to the stellar

1022: surface \citep{Massi07}.

1023: 

1024: 

1025: \end{enumerate}

1026: As PMS stars lie at the extreme high-luminosity end of the

1027: $10^{10}$ range of correlated X-ray and radio luminosities

1028: \citep[the Benz-G\"udel relation; see Figure 6 in][]{Guedel02}, it

1029: is reasonable that they also have the largest flare loop sizes.

1030: While the typical large coronal structures we find from COUP X-ray

1031: flares are $\sim 5$ times the stellar radius, even larger

1032: structures are found for some cases. These magnetic structures

1033: must extend significantly beyond the corotation surface, implying

1034: that the loop magnetic fields are, at least temporarily, capable

1035: of withstanding both the thermal pressure of the confined hot gas

1036: and the effect of centrifugal forces \citep{Jardine99}.

1037: 

1038: 

1039: 

1040: Our finding that the coronal structures in slower rotating

1041: Class~II stars do not exceed their corotation surface agrees well

1042: with the T-Tauri coronal models of \citet{Jardine06}.  Combining

1043: average COUP levels of X-ray emission with optical band

1044: measurements of multipolar surface magnetic fields and magnetic

1045: circumstellar disks,  they calculate self-consistent 3-dimensional

1046: force-free field configurations.  Their model emerges with a

1047: complicated combination of closed loops confining X-ray emitting

1048: plasma and open field lines available for accreting cool disk

1049: material or releasing a high velocity wind \citep[see

1050: also][]{Lovelace95, Long07}.  In the model of Jardine et al.,

1051: X-ray coronal extents for diskfree PMS  stars reach

1052: $1.5-2.5$~R$_{cor}$, precisely the range we find in the current

1053: work (Figures \ref{fig_L_vs_dHKs}-\ref{fig_L_vs_ewca}). The

1054: centrifugal stripping of the coronal outer parts proposed for

1055: rapidly rotating main-sequence stars, such as VXR 45 with a

1056: rotation period around $0.2$~days \citep{Jardine04}, is apparently

1057: unimportant for relatively slow rotating T-Tauri stars.

1058: 

1059: 

1060: 

1061: We thus find that, while flare properties are mostly similar in

1062: Class~II and Class~III stars (\S \ref{disc_stardisk.section}),

1063: some differences are found. X-ray coronal structures on Class~II

1064: stars are on average somewhat larger and are more likely to

1065: produce super-hot flares.  This suggests the magnetic fields on

1066: accreting Class~II stars may have distorted magnetic topologies as

1067: \citet{Jardine06} predict.  We discuss this further in \S

1068: \ref{disc_superhot.section}.

1069: 

1070: 

1071: 

1072: 

1073: \subsection{Comparison with solar long decay events

1074: \label{disc_LDE.section}}

1075: 

1076: 

1077: 

1078: It is reasonable to propose that the majority of the bright long

1079: X-ray flares detected from Orion T-Tauri stars are enhanced

1080: analogs of eruptive solar flare events classified as long decay

1081: events \citep[LDEs,][]{Kahler77}. The {\it Skylab}, SMM and {\it

1082: Yohkoh} space observatories show these flares produce X-ray

1083: emitting arches and streamers with altitudes reaching up to

1084: several $\times 10^{10}$~cm \citep[$\ga 0.2$~R$_\odot$,][]

1085: {Svestka95, Farnik96, Svestka97}. The 24 Jan 1992 X-ray streamers,

1086: which reached up to $L \ga 0.5$~ R$_{\odot}$ in altitude

1087: \citep{Hiei94,Hiei97}, are representative of magnetic structures

1088: emerging from solar LDEs (see inset in Figure

1089: \ref{fig_corona_extent_sketch}). The X-ray lightcurves of LDE

1090: flares last from a few hours to a day, similar to the duration of

1091: the COUP flares studied here (see Figure 14 in Paper I for two

1092: examples).

1093: 

1094: 

1095: The origin of giant solar X-ray emitting arches and streamers is

1096: not well-understood.  The most widely accepted model is that the

1097: impulsive flare near the solar surface ($L \la 10^{-2}$~R$_\odot$)

1098: blows open the overlying large-scale magnetic field with

1099: subsequent reconnection of magnetic lines through a vertical

1100: current sheet. This model was developed over many years

1101: \citep[e.g.][] {Sturrock66, Kopp76, Forbes96} and is reviewed by

1102: \citet{Priest02}. Peak observed temperatures in giant solar arches

1103: and streamers are typically of several to $\ga 10$~MK with a wide

1104: range of plasma densities.  For example, the 24 Jan 1992 event

1105: showed a relatively low density $\la 10^{8}$~cm$^{-3}$ while the

1106: 15 Mar 1993 event showed densities $\ga 10^{10}$~cm$^{-3}$

1107: \citep{Getman99, Getman00}.  Recall, however, that the peak

1108: luminosities and total energies of these solar flares are far

1109: below those we are studying in COUP stars, roughly $E \sim

1110: 10^{30-31}$~ergs $vs.$ $E \sim 10^{35-37}$~ergs in the X-ray band

1111: integrated over the flare.

1112: 

1113: 

1114: Confining a solar flare plasma with $n_e \la 10^8$~cm$^{-3}$ and

1115: peak temperatures $T \la 10$~MK requires a local field strength

1116: exceeding 1~G. In the solar corona,  the dipole component appears

1117: to become dominant at 2.5~$R_{\odot}$ \citep{Luhmann98} and the

1118: Sun's magnetic dipole field, only $\sim 1$~G at the surface,

1119: likely falls $\la 0.1$~G at this distance. Small-scale multipolar

1120: fields in active regions have surface strengths around 1000~G, but

1121: these quickly decay at large distances. Thus, the solar fields are

1122: too weak to confine X-ray emitting plasma at distances comparable

1123: to or exceeding the solar radius. The large-scale magnetic field

1124: of T-Tauri stars must thus be far stronger than in the Sun, as

1125: argued by \citet{Jardine06} and others (\S

1126: \ref{disk_truncation.section}, \ref{disc_superhot.section}), and

1127: can sustain giant X-ray arches and streamers with sizes

1128: $L/R_{\star} \sim 1-10$. The recent reported discovery of the two

1129: solar-like radio-flaring streamers in the young binary system

1130: V773~Tau~A reaching up the altitudes of $>20$~R$_{\star}$, but

1131: apparently anchored at the surface of its host star

1132: \citep{Massi07}, further strengthens the idea of the strong

1133: large-scale magnetic fields and giant X-ray flaring structures

1134: anchored at the stellar surface of T-Tauri stars.

1135: 

1136: 

1137: 

1138: We recall in this context our tentative finding (limited by small

1139: samples to a low statistical significance) that flares in

1140: accreting systems are systematically shorter than in other systems

1141: (\S \ref{energetics_section}). If this is true, it might be that

1142: high accretion may prevent very long lasting flares.  In order to

1143: have relatively long-lasting flare events, magnetic stresses must

1144: build up over an extended time, storing magnetic energy

1145: accumulated from shearing and twisting of magnetic field lines.  A

1146: similar process occurs in two-ribbon solar flares

1147: \citep{Priest02}, although with much lower total energies. The

1148: magnetic field must be stable in order to store large stresses. It

1149: is possible that the process of accretion destabilizes the field,

1150: preventing the accumulation of magnetic stresses and forcing

1151: shorter flares.

1152: 

1153: 

1154: We thus suggest that it is systematically shorter flare durations

1155: due to disrupted magnetic field configurations that may be

1156: responsible for the well-established reduction in time-integrated

1157: X-ray luminosities of accreting PMS stars compared to

1158: non-accreting stars.  This interpretation is different from past

1159: explanations. \citet{Preibisch05b} suggested that X-ray emission

1160: from accretors is suppressed because it cannot arise in magnetic

1161: field lines which are mass-loaded with disk material.

1162: \citet{Jardine06} argued that the outer magnetosphere of accretors

1163: is stripped by interaction with the disk.  \citet{Gregory07}

1164: proposed that soft X-ray emission is attenuated by dense material

1165: in accretion columns. The true cause of the reduced X-ray emission

1166: of Class~II systems is thus still uncertain.

1167: 

1168: 

1169: 

1170: 

1171: \subsection{Anomalous super-hot flares in accreting stars

1172: \label{disc_superhot.section}}

1173: 

1174: 

1175: 

1176: We described in \S \ref{superhot_section} and in Paper I a

1177: significant subset of COUP flares exhibiting peak temperatures

1178: exceeding 100~MK, some apparently exceeding 200~MK.   These

1179: temperatures are hotter than any solar flare plasma, and hotter

1180: than nearly all reported stellar flares.  The only comparable

1181: event we have identified is the $T>180$~MK plasma temperature in

1182: the 16 Dec 2005 flare in the RS CVn system II Peg using the hard

1183: X-ray detector (designed for gamma ray burst discovery) on board

1184: the $Swift$ satellite \citep{Osten07}.  While the calibration of

1185: $Chandra$ ACIS median energies to plasma temperatures above $\sim

1186: 100$~MK is not precise, we argue in Appendix B of Paper I and the

1187: Appendix below that these events are indeed hotter than other

1188: flares in the sample.

1189: 

1190: 

1191: 

1192: While the super-hot flare phenomenon may have more than one cause,

1193: the clearest relation is to active accretion in lower-mass PMS

1194: stars. These are systems where the stellar radius is small and the

1195: stellar magnetosphere appears truncated by the inner disk at the

1196: corotation radius (\S \ref{disk_truncation.section}).  We show in

1197: \S \ref{magnetic_fields.section} that these temperatures are too

1198: hot to be explained by standard dipolar fields rooted in the

1199: stellar surface. We speculate in \S~\ref{disc_stardisk.section}

1200: that these conditions lead to a compression and intensification of

1201: field strength in the outer regions of the loop where the observed

1202: X-ray emission originates. This might lead to more violent

1203: reconnection and successful confinement of higher pressure

1204: plasmas.  This interpretation qualitatively agrees with recent MHD

1205: computations of accretion funnels through PMS magnetospheres where

1206: distortion of initially dipolar field lines are predicted

1207: \citep{Jardine06, Bessolaz08, Romanova08}.  Other explanations

1208: such as reconnection in disk-derived fields

1209: \citep[e.g.,][]{Hayashi96} seem less attractive given the

1210: similarity of superhot and normal temperature flares in other

1211: properties (flare morphology and energetics).

1212: 

1213: 

1214: 

1215: 

1216: \section{Conclusions \label{conclusion_section}}

1217: 

1218: 

1219: 

1220: We examine here empirical relationships between magnetic

1221: reconnection flares, protoplanetary disks and accretion. The

1222: current work provides detailed flare modeling of a much larger

1223: dataset than previously available: we consider 216 bright X-ray

1224: flares from 161 PMS stars observed during the 13-day COUP exposure

1225: of the Orion Nebula.  Our sample is larger because, as described

1226: in Paper I, we use data analysis techniques that permit modeling

1227: of fainter flares than feasible with traditional parametric

1228: modeling of variable $Chandra$ ACIS spectra.  We thus have

1229: opportunity to uncover more subtle relationships between flaring,

1230: disks and accretion than were previously possible.  The main

1231: results of our study are as follows:

1232: 

1233: 

1234: 

1235: 1. Perhaps with the exception of the $Chandra$ study of NGC~2264,

1236: where CTTS are seen to be more variable than WTTS

1237: \citep{Flaccomio06}, past studies using smaller samples have not

1238: found differences in flare statistics or properties as a function

1239: of PMS evolutionary state or mass: accreting Class II systems

1240: appeared to flare like non-accreting Class III systems

1241: \citep[e.g.][]{Stelzer00,  Wolk05, Favata05, Stelzer07}. We

1242: confirm here that Class II and Class III systems produce X-ray

1243: flares with indistinguishable morphologies. Typical fast-rise

1244: slow-decay flares, composite step and double flares, and the

1245: unusual slow-rise flat-top flares are present in both classes. We

1246: have no clear evidence for distinctive flare types emerging from

1247: star-disk magnetic field lines. Accretion variations producing

1248: optical band variations are not associated with X-ray flares

1249: \citep{Stassun06}. Accretion and magnetic flaring thus appear to

1250: be unrelated, even though the theory of PMS accretion involves

1251: magnetic truncation of disks and magnetic funneling of disk

1252: material onto the stellar surface.  The COUP PMS flares are

1253: consistent with solar-type magnetic structures with both

1254: footpoints anchored in the stellar surface.  Many may represent

1255: analogies of a much less powerful class of solar flares known as

1256: long duration events which produce giant coronal arches and X-ray

1257: streamers.

1258: 

1259: 

1260: 

1261: 2. A distinct difference is found in the distribution of loop

1262: sizes. We find that X-ray coronal extents in fast rotating

1263: Class~III stars sometimes exceeds the Keplerian corotation radius,

1264: whereas X-ray magnetospheres in Class~II stars appear to be

1265: truncated at the inner edge of accreting protoplanetary disks.

1266: This directly supports theoretical models of magnetically mediated

1267: accretion, magnetic star-disk rotational coupling, and  disk

1268: confinement of PMS magnetospheres.

1269: 

1270: 

1271: 

1272: 3. A related, but statistically less secure, result is that flares

1273: from highly accreting Class~II stars have somewhat shorter

1274: durations and weaker total X-ray energies than Class~III flares.

1275: This might reflect the destabilization of magnetic arcades in

1276: accreting systems, and could account for the reduction of

1277: time-averaged X-ray luminosities in Class II compared to Class III

1278: populations noted in previous studies.

1279: 

1280: 

1281: 

1282: 4. A subclass of super-hot flares with peak plasma temperatures

1283: greater than 100~MK is noted for the first time.  These are

1284: inconsistent with formation in normal dipolar magnetic loops

1285: attached to the stellar surface, and appear preferentially in

1286: accreting Class~II systems.  They may reflect compression and

1287: distortion of the large scale magnetospheric topology by star-disk

1288: magnetic interactions, as predicted by recent theoretical

1289: calculations.

1290: 

1291: 

1292: 

1293: 

1294: 

1295: \acknowledgements  We thank the anonymous referee for his time and 

1296: useful comments that improved this work. The work was supported by the $Chandra$ ACIS

1297: Team (G. Garmire, PI) through the SAO grant SV4-74018. G.M.

1298: acknowledges contribution from contract ASI-INAF I/088/06/0. This

1299: publication makes use of data products from the Two Micron All Sky

1300: Survey (a joint project of the University of Massachusetts and the

1301: Infrared Processing and Analysis Center/California Institute of

1302: Technology, funded by NASA and NSF), and archival data obtained with

1303: the {\it Spitzer Space Telescope} (operated by the Jet Propulsion

1304: Laboratory, California Institute of Technology under a contract

1305: with NASA).

1306: 

1307: 

1308: 

1309: \appendix

1310: 

1311: 

1312: \section{SUPER-HOT FLARES  \label{appendix_section}}

1313: 

1314: 

1315: It is difficult to accurately determine temperatures of thermal

1316: components $\ga 100$~MK from $Chandra$-ACIS spectra. To insure

1317: that the peak temperatures of $T_{obs,pk}>100$~MK do indeed

1318: describe the hottest flare plasmas observed in our sample, we

1319: describe here checks that have been performed beyond the analysis

1320: of the Appendix B of Paper I.

1321: 

1322: 

1323: For each of the 216 COUP flares treated here, a spectrum has been

1324: extracted within the total time range $[{\rm t}_{flare1}-{\rm

1325: t}_{flare2}]$. Spectra were fitted with WABS~$\times$~MEKAL model

1326: \citep[for compatibility with][]{Getman05} with 0.3 times solar

1327: elemental abundances allowing both temperature and column density

1328: to be free parameters. The following confirmatory results are

1329: obtained. First, for the majority of the flares, the X-ray column densities

1330: obtained from the integrated flare spectral fits are in excellent

1331: agreement with the source column densities obtained from the full 

1332: COUP observation by \citet{Getman05}

1333: that we used as fixed parameters during our flare analysis

1334: (see also Figure~16 in Paper~I). Second, out of the 73 flares with peak flare

1335: temperatures $T_{obs,pk}>100$~MK, only 8 (11\%) have their

1336: integrated flare spectral fits $T_{flare}<40$~MK, while 80\%  of

1337: the flares with $T_{obs,pk}<100$~MK have $T_{flare}<40$~MK (Figure

1338: \ref{fig_super_hot_analysis1}). This confirms a distinct

1339: difference between super-hot and ordinary flares.  Third, high

1340: accreting stars\footnote{The twelve highly accreting starswith

1341: $EW(CaII) < -2$~\AA\ are COUP  \#11, 66, 141, 567, 579, 1044,

1342: 1045, 1080, 1096, 1335, 1409, and 1444.} have the largest fraction

1343: of reported super-hot flares (9 out of 17 flares or 53\%), three

1344: more flares from high accretors  have their

1345: $80<T_{obs,pk}<100$~MK. This confirms the discussion in \S

1346: \ref{superhot_section}  that high accretors preferentially exhibit

1347: super-hot flares.

1348: 

1349: 

1350: Figure \ref{fig_ultrahot_4} further illustrates spectral

1351: differences between super-hot ($T_{obs,pk}>100$~MK) and ordinary

1352: flares.  To avoid possible confounding effects of differing

1353: absorptions, flare histories and disk properties, we extract the

1354: spectrum within a brief 6~ks interval around the flare peak for

1355: all MIR disk sources in the narrow interval

1356: $21.5<\log(N_H)<22$~cm$^{-2}$. The figure compares combined flare

1357: peak spectra for sources with ordinary (panel $a$) and super-hot

1358: flares (panel $b$). Merging these spectra to improve the

1359: statistical accuracy, the best-fit model parameters of the

1360: composite 6~ks spectra are $N_H = 4.8_{-0.5}^{+0.5} \times

1361: 10^{21}$~cm$^{-2}$ and  $kT = 5.5_{-0.8}^{+1.0}$~keV for the

1362: ordinary flares, and $N_H = 6.6_{-0.7}^{+0.6} \times

1363: 10^{21}$~cm$^{-2}$ and $kT = 15.2_{-3.6}^{+12.1}$~keV for the

1364: super-hot flares. The composite spectrum of super-hot flares can

1365: not be described by ordinary flare models (compare black and green

1366: lines in Figure~\ref{fig_ultrahot_4}$b$). This again demonstrates

1367: that super-hot flares are the hottest flares of our sample.

1368: 

1369: 

1370: 

1371: 

1372: \begin{thebibliography}

1373: 

1374: 

1375: \bibitem[Alef et al.(1997)]{Alef97} Alef, W., Benz, A.~O., \&

1376: Guedel, M.\ 1997, \aap, 317, 707

1377: 

1378: 

1379: \bibitem[Argiroffi et al.(2007)]{Argiroffi07} Argiroffi, C.,

1380: Maggio, A., \& Peres, G.\ 2007, \aap, 465, L5

1381: 

1382: 

1383: \bibitem[Arzner et al.(2007)]{Arzner07} Arzner, K., G{\"u}del,

1384: M., Briggs, K., Telleschi, A., \& Audard, M.\ 2007, \aap, 468, 477

1385: 

1386: 

1387: \bibitem[Bally et al.(1987)]{Bally87} Bally, J., Stark, A.~A.,

1388: Wilson, R.~W., \& Langer, W.~D.\ 1987, \apjl, 312, L45

1389: 

1390: 

1391: \bibitem[Benz et al.(1998)]{Benz98} Benz, A.~O., Conway, J.,

1392: \& Gudel, M.\ 1998, \aap, 331, 596

1393: 

1394: 

1395: \bibitem[Bessolaz et al.(2008)]{Bessolaz08} Bessolaz, N., Zanni,

1396: C., Ferreira, J., Keppens, R., \& Bouvier, J.\ 2008, \aap, 478,

1397: 155

1398: 

1399: 

1400: \bibitem[Birk et al.(2000)]{Birk00} Birk, G.~T., Schwab, D.,

1401: Wiechen, H., \& Lesch, H.\ 2000, \aap, 358, 1027

1402: 

1403: 

1404: \bibitem[Bouvier et al.(2007)]{Bouvier07a} Bouvier, J., Alencar,

1405: S.~H.~P., Harries, T.~J., Johns-Krull, C.~M., \& Romanova, M.~M.\

1406: 2007, Protostars and Planets V, 479

1407: 

1408: 

1409: 

1410: \bibitem[Bouvier \& Appenzeller(2007)]{Bouvier07b} Bouvier, J., \&

1411: Appenzeller, I.\ 2007, IAU Symposium, 243,

1412: 

1413: 

1414: 

1415: \bibitem[Browning(2007)]{Browning07} Browning, M.\ 2007, ArXiv

1416: e-prints, 712, arXiv:0712.1603

1417: 

1418: 

1419: 

1420: \bibitem[Caramazza et al.(2007)]{Caramazza07} Caramazza, M.,

1421: Flaccomio, E., Micela, G., Reale, F., Wolk, S.~J., \& Feigelson,

1422: E.~D.\ 2007, \aap, 471, 645

1423: 

1424: 

1425: 

1426: \bibitem[Daou et al.(2006)]{Daou06} Daou, A.~G., Johns-Krull,

1427: C.~M., \& Valenti, J.~A.\ 2006, \aj, 131, 520

1428: 

1429: 

1430: 

1431: \bibitem[Donati et al.(2007)]{Donati07} Donati, J.-F., et al.\

1432: 2007, \mnras, 380, 1297

1433: 

1434: 

1435: 

1436: \bibitem[Donati et al.(2008)]{Donati08} Donati, J.~F., et al.\

1437: 2008, ArXiv e-prints, 802, arXiv:0802.2052

1438: 

1439: 

1440: 

1441: \bibitem[Drake et al.(2005)]{Drake05} Drake, J.~J., Testa, P.,

1442: \& Hartmann, L.\ 2005, \apjl, 627, L149

1443: 

1444: 

1445: 

1446: \bibitem[Farnik et al.(1996)]{Farnik96} Farnik et al.\ 1996, \solphys,

1447: 168, 331

1448: 

1449: 

1450: 

1451: \bibitem[Favata et al.(2005)]{Favata05} Favata, F., Flaccomio,

1452: E., Reale, F., Micela, G., Sciortino, S., Shang, H., Stassun,

1453: K.~G., \& Feigelson, E.~D.\ 2005, \apjs, 160, 469

1454: 

1455: 

1456: 

1457: \bibitem[Feigelson \& Montmerle(1999)]{Feigelson99} Feigelson,

1458: E.~D., \& Montmerle, T.\ 1999, \araa, 37, 363

1459: 

1460: 

1461: \bibitem[Feigelson et al.(2005)]{Feigelson05} Feigelson, E.~D., et

1462: al.\ 2005, \apjs, 160, 379

1463: 

1464: 

1465: \bibitem[Feigelson et al.(2007)]{Feigelson07} Feigelson, E.,

1466: Townsley, L., G{\"u}del, M., \& Stassun, K.\ 2007, Protostars and

1467: Planets V, 313

1468: 

1469: 

1470: 

1471: \bibitem[Flaccomio et al.(2003)]{Flaccomio03} Flaccomio, E.,

1472: Damiani, F., Micela, G., Sciortino, S., Harnden, F.~R., Jr.,

1473: Murray, S.~S., \& Wolk, S.~J.\ 2003, \apj, 582, 398

1474: 

1475: 

1476: 

1477: \bibitem[Flaccomio et al.(2005)]{Flaccomio05} Flaccomio, E.,

1478: Micela, G., Sciortino, S., Feigelson, E.~D., Herbst, W., Favata,

1479: F., Harnden, F.~R., Jr., \& Vrtilek, S.~D.\ 2005, \apjs, 160, 450

1480: 

1481: 

1482: 

1483: \bibitem[Flaccomio et al.(2006)]{Flaccomio06} Flaccomio, E., Micela, G., \&

1484: Sciortino, S.\ 2006, \aap, 455, 903

1485: 

1486: 

1487: 

1488: \bibitem[Forbes \& Acton(1996)]{Forbes96} Forbes, T.~G., \&

1489: Acton, L.~W.\ 1996, \apj, 459, 330

1490: 

1491: 

1492: \bibitem[Franciosini et al.(1999)]{Franciosini99} Franciosini, E.,

1493: Massi, M., Paredes, J.~M., \& Estalella, R.\ 1999, \aap, 341, 595

1494: 

1495: 

1496: 

1497: \bibitem[Franciosini et al.(2007)]{Franciosini07} Franciosini, E., et

1498: al.\ 2007, \aap, 468, 485

1499: 

1500: 

1501: 

1502: \bibitem[Getman \& Livshitz(1999)]{Getman99} Getman, K.~V., \&

1503: Livshitz, M.~A.\ 1999, Astronomy Reports, 43, 615

1504: 

1505: 

1506: 

1507: \bibitem[Getman \& Livshits(2000)]{Getman00} Getman, K.~V., \&

1508: Livshits, M.~A.\ 2000, Astronomy Reports, 44, 255

1509: 

1510: 

1511: 

1512: \bibitem[Getman et al.(2005a)]{Getman05} Getman, K.~V., Flaccomio,

1513: E., Broos, P.~S. and 21 co-authours\ 2005, \apjs, 160, 319

1514: 

1515: 

1516: 

1517: \bibitem[Getman et al.(2006)]{Getman06} Getman, K.~V.,

1518: Feigelson, E.~D., Townsley, L., Broos, P., Garmire, G., \&

1519: Tsujimoto, M.\ 2006, \apjs, 163, 306

1520: 

1521: 

1522: 

1523: \bibitem[Getman et al.(2008)]{Getman08} Getman, K.~V.,

1524: Feigelson, E.~D., Broos, P.~S., Micela, G., \& Garmire, G.~P.\

1525: 2008, \apj, in press (Paper I)

1526: 

1527: 

1528: 

1529: \bibitem[Getman et al.(2007b)]{Getman07b} Getman, K.~V.,

1530: Feigelson, E.~D., Garmire, G., Broos, P., \& Wang, J.\ 2007, \apj,

1531: 654, 316

1532: 

1533: 

1534: \bibitem[Gregory et al.(2006a)]{Gregory06a} Gregory, S.~G.,

1535: Jardine, M., Simpson, I., \& Donati, J.-F.\ 2006a, \mnras, 371,

1536: 999

1537: 

1538: 

1539: 

1540: \bibitem[Gregory et al.(2006b)]{Gregory06} Gregory, S.~G.,

1541: Jardine, M., Cameron, A.~C., \& Donati, J.-F.\ 2006b, \mnras, 373,

1542: 827

1543: 

1544: 

1545: 

1546: \bibitem[Gregory et al.(2007)]{Gregory07} Gregory, S.~G., Wood,

1547: K., \& Jardine, M.\ 2007, \mnras, 379, L35

1548: 

1549: 

1550: \bibitem[Grosso et al.(2005)]{Grosso05} Grosso, N., et al.\

1551: 2005, \apjs, 160, 530

1552: 

1553: \bibitem[G{\"u}del(2002)]{Guedel02} G{\"u}del, M.\ 2002, \araa,

1554: 40, 217

1555: 

1556: 

1557: \bibitem[G{\"u}del(2004)]{Gudel04} G{\"u}del, M.\ 2004, \aapr,

1558: 12, 71

1559: 

1560: 

1561: 

1562: \bibitem[G{\"u}del et al.(2007)]{Gudel07} G{\"u}del, M., et

1563: al.\ 2007, \aap, 468, 353

1564: 

1565: 

1566: 

1567: \bibitem[G{\"u}del \& Telleschi(2007)]{Gudel07b} G{\"u}del, M.,

1568: \& Telleschi, A.\ 2007, \aap, 474, L25

1569: 

1570: 

1571: 

1572: \bibitem[Hartmann(1998)]{Hartmann98} Hartmann, L.\ 1998, Accretion

1573: processes in star formation / Lee Hartmann.~Cambridge, UK ; New

1574: York : Cambridge University Press

1575: 

1576: 

1577: 

1578: \bibitem[Hartmann et al.(2005)]{Hartmann05} Hartmann, L., Megeath,

1579: S.~T., Allen, L., Luhman, K., Calvet, N., D'Alessio, P.,

1580: Franco-Hernandez, R., \& Fazio, G.\ 2005, \apj, 629, 881

1581: 

1582: 

1583: \bibitem[Hayashi et al.(1996)]{Hayashi96} Hayashi, M.~R.,

1584: Shibata, K., \& Matsumoto, R.\ 1996, \apjl, 468, L37

1585: 

1586: 

1587: \bibitem[Herbst et al.(2002)]{Herbst02} Herbst, W.,

1588: Bailer-Jones, C.~A.~L., Mundt, R., Meisenheimer, K., \&

1589: Wackermann, R.\ 2002, \aap, 396, 513

1590: 

1591: 

1592: 

1593: \bibitem[Hiei(1994)]{Hiei94} Hiei, E.\ 1994, IAU Colloq.~144:

1594: Solar Coronal Structures, 163

1595: 

1596: 

1597: 

1598: \bibitem[Hiei(1997)]{Hiei97} Hiei, E.\ 1997, Memorie della

1599: Societa Astronomica Italiana, 68, 491

1600: 

1601: 

1602: 

1603: \bibitem[Hillenbrand(1997)]{Hillenbrand97} Hillenbrand, L.~A.\ 1997,

1604: \aj, 113, 1733

1605: 

1606: 

1607: 

1608: \bibitem[Huenemoerder et al.(2007)]{Huenemoerder07} Huenemoerder,

1609: D.~P., Kastner, J.~H., Testa, P., Schulz, N.~S., \& Weintraub,

1610: D.~A.\ 2007, \apj, 671, 592

1611: 

1612: 

1613: 

1614: \bibitem[Imanishi et al.(2003)]{Imanishi03} Imanishi, K.,

1615: Nakajima, H., Tsujimoto, M., Koyama, K., \& Tsuboi, Y.\ 2003,

1616: \pasj, 55, 653

1617: 

1618: 

1619: \bibitem[Isobe et al.(2003)]{Isobe03} Isobe, H., Shibata, K.,

1620: Yokoyama, T., \& Imanishi, K.\ 2003, \pasj, 55, 967

1621: 

1622: 

1623: \bibitem[Jardine \& Unruh(1999)]{Jardine99} Jardine, M., \&

1624: Unruh, Y.~C.\ 1999, \aap, 346, 883

1625: 

1626: 

1627: \bibitem[Jardine(2004)]{Jardine04} Jardine, M.\ 2004, \aap, 414,

1628: L5

1629: 

1630: 

1631: \bibitem[Jardine et al.(2006)]{Jardine06} Jardine, M., Cameron,

1632: A.~C., Donati, J.-F., Gregory, S.~G., \& Wood, K.\ 2006, \mnras,

1633: 367, 917

1634: 

1635: 

1636: 

1637: \bibitem[Johns-Krull et al.(1999)]{Johns-Krull99} Johns-Krull, C.~M.,

1638: Valenti, J.~A., Hatzes, A.~P., \& Kanaan, A.\ 1999, \apjl, 510,

1639: L41

1640: 

1641: 

1642: 

1643: \bibitem[Johns-Krull(2007)]{Johns-Krull07} Johns-Krull, C.~M.\ 2007,

1644: \apj, 664, 975

1645: 

1646: 

1647: 

1648: \bibitem[Kahler(1977)]{Kahler77} Kahler, S.\ 1977, \apj, 214,

1649: 891

1650: 

1651: 

1652: 

1653: \bibitem[Kastner et al.(2002)]{Kastner02} Kastner, J.~H.,

1654: Huenemoerder, D.~P., Schulz, N.~S., Canizares, C.~R., \&

1655: Weintraub, D.~A.\ 2002, \apj, 567, 434

1656: 

1657: \bibitem[Kelly(2007)]{Kelly07} Kelly, B.~C.\ 2007, \apj, 665, 

1658: 1489 

1659: 

1660: 

1661: \bibitem[Kopp \& Pneuman(1976)]{Kopp76} Kopp, R.~A., \&

1662: Pneuman, G.~W.\ 1976, \solphys, 50, 85

1663: 

1664: 

1665: \bibitem[Koyama et al.(1996)]{Koyama96} Koyama, K., Hamaguchi,

1666: K., Ueno, S., Kobayashi, N., \& Feigelson, E.~D.\ 1996, \pasj, 48,

1667: L87

1668: 

1669: 

1670: \bibitem[Lang(1994)]{Lang94} Lang, K.~R.\ 1994, \apjs, 90, 753

1671: 

1672: 

1673: 

1674: \bibitem[Long et al.(2005)]{Long05} Long, M., Romanova, M.~M.,

1675: \& Lovelace, R.~V.~E.\ 2005, \apj, 634, 1214

1676: 

1677: 

1678: 

1679: \bibitem[Long et al.(2007)]{Long07} Long, M., Romanova, M.~M.,

1680: \& Lovelace, R.~V.~E.\ 2007, \mnras, 374, 436

1681: 

1682: 

1683: 

1684: \bibitem[Lovelace et al.(1995)]{Lovelace95} Lovelace, R.~V.~E.,

1685: Romanova, M.~M., \& Bisnovatyi-Kogan, G.~S.\ 1995, \mnras, 275,

1686: 244

1687: 

1688: 

1689: \bibitem[Luhmann et al.(1998)]{Luhmann98} Luhmann, J.~G.,

1690: Gosling, J.~T., Hoeksema, J.~T., \& Zhao, X.\ 1998, \jgr, 103,

1691: 6585

1692: 

1693: 

1694: \bibitem[Maggio et al.(2000)]{Maggio00} Maggio, A., Pallavicini, R.,

1695: Reale, F., \& Tagliaferri, G.\ 2000, \aap, 356, 627

1696: 

1697: 

1698: 

1699: \bibitem[Maggio et al.(2007)]{Maggio07} Maggio, A., Flaccomio,

1700: E., Favata, F., Micela, G., Sciortino, S., Feigelson, E.~D., \&

1701: Getman, K.~V.\ 2007, \apj, 660, 1462

1702: 

1703: 

1704: 

1705: \bibitem[Mathis(1990)]{Mathis90} Mathis, J.~S.\ 1990, The

1706: Evolution of the Interstellar Medium, 12, 63

1707: 

1708: 

1709: 

1710: \bibitem[Massi et al.(2007)]{Massi07} Massi, M., et al.\ 2007,

1711: ArXiv e-prints, 712, arXiv:0712.0718

1712: 

1713: 

1714: 

1715: \bibitem[Montmerle et al.(2000)]{Montmerle00} Montmerle, T.,

1716: Grosso, N., Tsuboi, Y., \& Koyama, K.\ 2000, \apj, 532, 1097

1717: 

1718: 

1719: 

1720: \bibitem[Mullan et al.(2006)]{Mullan06} Mullan, D.~J., Mathioudakis,

1721: M., Bloomfield, D.~S., \& Christian, D.~J.\ 2006, \apjs, 164, 173

1722: 

1723: 

1724: \bibitem[Osten et al.(2007)]{Osten07} Osten, R.~A., Drake, S.,

1725: Tueller, J., Cummings, J., Perri, M., Moretti, A., \& Covino, S.\

1726: 2007, \apj, 654, 1052

1727: 

1728: 

1729: \bibitem[Preibisch et al.(2005)]{Preibisch05b} Preibisch, T., et

1730: al.\ 2005b, \apjs, 160, 401

1731: 

1732: 

1733: 

1734: \bibitem[Priest \& Forbes(2002)]{Priest02} Priest, E.~R., \&

1735: Forbes, T.~G.\ 2002, \aapr, 10, 313

1736: 

1737: 

1738: \bibitem[Prisinzano et al.(2007)]{Prisinzano07} Prisinzano, L., et

1739: al.\ 2007, ArXiv e-prints, 712, arXiv:0712.2975

1740: 

1741: 

1742: \bibitem[Ransom et al.(2002)]{Ransom02} Ransom, R.~R., Bartel,

1743: N., Bietenholz, M.~F., Lebach, D.~E., Ratner, M.~I., Shapiro,

1744: I.~I., \& Lestrade, J.-F.\ 2002, \apj, 572, 487

1745: 

1746: 

1747: \bibitem[Ransom et al.(2003)]{Ransom03} Ransom, R.~R., Bartel,

1748: N., Bietenholz, M.~F., Ratner, M.~I., Lebach, D.~E., Shapiro,

1749: I.~I., \& Lestrade, J.-F.\ 2003, \apj, 587, 390

1750: 

1751: 

1752: 

1753: \bibitem[Reale et al.(1997)]{Reale97} Reale, F., Betta, R.,

1754: Peres, G., Serio, S., \& McTiernan, J.\ 1997, \aap, 325, 782

1755: 

1756: 

1757: 

1758: \bibitem[Reale(2002)]{Reale02} Reale, F.\ 2002, Stellar Coronae

1759: in the Chandra and XMM-NEWTON Era, 277, 103

1760: 

1761: 

1762: \bibitem[Reale(2007)]{Reale07} Reale, F.\ 2007, \aap, 471, 271

1763: 

1764: 

1765: \bibitem[Rebull et al.(2006)]{Rebull06} Rebull, L.~M., Stauffer,

1766: J.~R., Megeath, S.~T., Hora, J.~L., \& Hartmann, L.\ 2006, \apj,

1767: 646, 297

1768: 

1769: 

1770: \bibitem[Romanova et al.(2008)]{Romanova08} Romanova, M.~M.,

1771: Kulkarni, A.~K., \& Lovelace, R.~V.~E.\ 2008, \apjl, 673, L171

1772: 

1773: 

1774: 

1775: \bibitem[Schmitt et al.(2005)]{Schmitt05} Schmitt, J.~H.~M.~M.,

1776: Robrade, J., Ness, J.-U., Favata, F., \& Stelzer, B.\ 2005, \aap,

1777: 432, L35

1778: 

1779: 

1780: 

1781: \bibitem[Shu et al.(1994)]{Shu94} Shu, F., Najita, J.,

1782: Ostriker, E., Wilkin, F., Ruden, S., \& Lizano, S.\ 1994, \apj,

1783: 429, 781

1784: 

1785: 

1786: 

1787: \bibitem[Shu et al.(1997)]{Shu97} Shu, F.~H., Shang, H.,

1788: Glassgold, A.~E., \& Lee, T.\ 1997, Science, 277, 1475

1789: 

1790: 

1791: 

1792: \bibitem[Siess et al.(2000)]{Siess00} Siess, L., Dufour, E., \&

1793: Forestini, M.\ 2000, \aap, 358, 593

1794: 

1795: 

1796: 

1797: \bibitem[Simon \& Dahm(2005)]{Simon05} Simon, T., \& Dahm,

1798: S.~E.\ 2005, \apj, 618, 795

1799: 

1800: 

1801: \bibitem[Stassun et al.(2006)]{Stassun06} Stassun, K.~G., van den

1802: Berg, M., Feigelson, E., \& Flaccomio, E.\ 2006, \apj, 649, 914

1803: 

1804: 

1805: \bibitem[Stelzer et al.(2000)]{Stelzer00} Stelzer, B.,

1806: Neuh{\"a}user, R., \& Hambaryan, V.\ 2000, \aap, 356, 949

1807: 

1808: 

1809: 

1810: \bibitem[Stelzer et al.(2007)]{Stelzer07} Stelzer, B., Flaccomio,

1811: E., Briggs, K., Micela, G., Scelsi, L., Audard, M., Pillitteri,

1812: I., G{\"u}del, M.\ 2007, \aap, 468, 463

1813: 

1814: 

1815: \bibitem[Sturrock(1966)]{Sturrock66} Sturrock, P.~A.\ 1966, \nat,

1816: 211, 695

1817: 

1818: 

1819: 

1820: \bibitem[{\v S}vestka et al.(1995)]{Svestka95} {\v S}vestka, Z.,

1821: F{\'a}rn{\'{\i}}k, F., Hudson, H.~S., Uchida, Y., Hick, P., \&

1822: Lemen, J.~R.\ 1995, \solphys, 161, 331

1823: 

1824: 

1825: 

1826: \bibitem[Svestka et al.(1997)]{Svestka97} Svestka, Z., Farnik,

1827: F., Hick, P., Hudson, H.~S., \& Uchida, Y.\ 1997, \solphys, 176,

1828: 355

1829: 

1830: 

1831: 

1832: \bibitem[Symington et al.(2005)]{Symington05} Symington, N.~H.,

1833: Harries, T.~J., Kurosawa, R., \& Naylor, T.\ 2005, \mnras, 358,

1834: 977

1835: 

1836: 

1837: 

1838: \bibitem[Tsuboi et al.(1998)]{Tsuboi98} Tsuboi, Y., Koyama, K.,

1839: Murakami, H., Hayashi, M., Skinner, S., \& Ueno, S.\ 1998, \apj,

1840: 503, 894

1841: 

1842: 

1843: 

1844: \bibitem[Vuong et al.(2003)]{Vuong03} Vuong, M.~H., Montmerle,

1845: T., Grosso, N., Feigelson, E.~D., Verstraete, L., \& Ozawa, H.\

1846: 2003, \aap, 408, 581

1847: 

1848: 

1849: \bibitem[Wolk et al.(2005)]{Wolk05} Wolk, S.~J., Harnden,

1850: F.~R., Jr., Flaccomio, E., Micela, G., Favata, F., Shang, H., \&

1851: Feigelson, E.~D.\ 2005, \apjs, 160, 423

1852: 

1853: 

1854: 

1855: \bibitem[Yang et al.(2007)]{Yang07} Yang, H., Johns-Krull,

1856: C.~M., \& Valenti, J.~A.\ 2007, \aj, 133, 73

1857: 

1858: 

1859: 

1860: \end{thebibliography}

1861: 

1862: 

1863: 

1864: 

1865: 

1866: 

1867: 

1868: \clearpage

1869: 

1870: 

1871: 

1872: \begin{deluxetable}{cccccccc}

1873: \centering \rotate \tabletypesize{\small} \tablewidth{0pt}

1874: \tablecolumns{8}

1875: 

1876: 

1877: 

1878: \tablecaption{Summary of Stellar Properties

1879: \label{tbl_correl_known}}

1880: 

1881: 

1882: 

1883: \tablehead{

1884: 

1885: 

1886: 

1887: \colhead{Quantity} & \colhead{N} & \colhead{Flag} & \colhead{Min}

1888: & \colhead{Max} &

1889: \colhead{Median} & \colhead{Mean$\pm$SD} & \colhead{$P_{KS}$}\\

1890: 

1891: 

1892: 

1893: (1)&(2)&(3)&(4)&(5)&(6)&(7)&(8)}

1894: 

1895: 

1896: 

1897: \startdata

1898: X-ray Absorbing Column Density  (cm$^{-2}$) & & & & & & & \nodata\\

1899: $\log(NH)$ (all available) & 161 & s & 20.0 & 22.8 & 21.6 & $21.5\pm0.6$ & \nodata\\

1900: $\log(NH_{NIRdisk})$ (NIR inner disk) & 56 & s & 20.0 & 22.6 & 21.7 & $21.6\pm0.6$ & \nodata\\

1901: $\log(NH_{noNIRdisk})$ (no NIR inner disk) & 84 & s & 20.0 & 22.3 & 21.5 & $21.4\pm0.6$ & 0.09\\

1902: $\log(NH_{MIRdisk})$ (MIR disk) & 62 & s & 20.0 & 22.8 & 21.8 & $21.7\pm0.6$ &\nodata \\

1903: $\log(NH_{noMIRdisk})$ (no MIR disk) & 50 & s & 20.0 & 22.5 & 21.3 & $21.3\pm0.6$ & 0.02\\

1904: \cline{1-8}

1905: Stellar Mass (M$_{\odot}$) & & & & & & & \\

1906: $M$ (all available) & 128 & s & 0.2 & 5.0 & 0.7 & $1.0\pm0.9$ & \nodata\\

1907: $M_{NIRdisk}$ (NIR inner disk) & 42 & s & 0.2 & 5.0 & 0.7 & $1.1\pm1.1$ & \nodata\\

1908: $M_{noNIRdisk}$ (no NIR inner disk) & 72 & s & 0.2 & 2.3 & 0.6 & $1.0\pm0.8$ & 0.1\\

1909: $M_{MIRdisk}$ (MIR disk) & 44 & s & 0.2 & 5.0 & 0.7 & $1.1\pm0.9$ &\nodata\\

1910: $M_{noMIRdisk}$ (no MIR disk) & 42 & s & 0.2 & 2.3 & 0.6 & $1.0\pm0.9$ & 0.4\\

1911: \cline{1-8}

1912: Stellar Radius ($10^{10}$~cm) & & & & & & & \\

1913: $R_{\star}$ (all available) & 130 & s & 4 & 90 & 18 & $20\pm12$ & \nodata\\

1914: $R_{\star,NIRdisk}$ (NIR inner disk) & 44 & s & 5 & 90 & 16 & $19\pm15$ & \nodata\\

1915: $R_{\star,noNIRdisk}$ (no NIR inner disk) & 72 & s & 4 & 58 & 18 & $21\pm11$ & 0.5\\

1916: $R_{\star,MIRdisk}$ (MIR disk) & 45 & s & 4 & 32 & 14 & $17\pm7$ & \nodata\\

1917: $R_{\star,noMIRdisk}$ (no MIR disk) & 43 & s & 10 & 58 & 17 & $22\pm12$ & 0.3\\

1918: \cline{1-8}

1919: Stellar Rotational Period (day) & & & & & & & \\

1920: $P$ (all available) & 79 & s & 1.1 & 19.5 & 6.5 & $6.7\pm3.9$ & \nodata\\

1921: $P_{NIRdisk}$ (NIR inner disk) & 20 & s & 1.4 & 14.4 & 8.5 & $7.7\pm2.8$ & \nodata\\

1922: $P_{noNIRdisk}$ (no NIR inner disk) & 48 & s & 1.1 & 19.5 & 6.3 & $6.3\pm4.2$ & 0.06\\

1923: $P_{MIRdisk}$ (MIR disk) & 24 & s & 5.2 & 14.4 & 9.0 & $8.8\pm1.9$ & \nodata\\

1924: $P_{noMIRdisk}$ (no MIR disk) & 35 & s & 1.1 & 19.5 & 3.5 & $5.5\pm4.6$ & $<$0.0001\\

1925: \cline{1-8}

1926: Keplerian Co-rotation Radius ($10^{10}$~cm) & & & & & & & \\

1927: $R_{cor}$ (all available) & 67 & s & 22 & 270 & 85 & $90\pm47$ & \nodata\\

1928: $R_{cor,NIRdisk}$ (NIR inner disk) & 14 & s & 39 & 180 & 120 & $107\pm39$ & \nodata\\

1929: $R_{cor,noNIRdisk}$ (no NIR inner disk) & 44 & s & 22 & 270 & 84 & $87\pm50$ & 0.03\\

1930: $R_{cor,MIRdisk}$ (MIR disk) & 21 & s & 54 & 180 & 98 & $108\pm33$ & \nodata\\

1931: $R_{cor,noMIRdisk}$ (no MIR disk) & 31 & s & 22 & 270 & 76 & $83\pm55$ & 0.007\\

1932: \cline{1-8}

1933: Ratio of Co-rotation Radius to Stellar Radius & & & & & & & \\

1934: $R_{cor}/R_{\star}$ (all available) & 67 & s & 1.4 & 20.9 & 5.1 & $5.7\pm3.9$ & \nodata\\

1935: $(R_{cor}/R_{\star})_{NIRdisk}$ (NIR inner disk) & 14 & s & 3.0 & 16.3 & 7.4 & $8.1\pm3.6$ &\nodata \\

1936: $(R_{cor}/R_{\star})_{noNIRdisk}$ (no NIR inner disk) & 44 & s & 1.5 & 20.9 & 4.2 & $5.3\pm4.0$ & 0.01\\

1937: $(R_{cor}/R_{\star})_{MIRdisk}$ (MIR disk) & 21 & s & 3.0 & 17.2 & 7.2 & $8.1\pm3.6$ & \nodata\\

1938: $(R_{cor}/R_{\star})_{noMIRdisk}$ (no MIR disk) & 31 & s & 1.4 & 13.6 & 3.2 & $4.0\pm2.7$ & $<$0.0001\\

1939: 

1940: 

1941: 

1942: \enddata

1943: 

1944: 

1945: 

1946: \tablecomments{Columns 1: Quantity name. Column 2: Number of

1947: flares (Flag$=$f, see Column 3) or sources (Flag$=$s) in a sample.

1948: Column 3: Flag indicates if this is a flare (f) or source (s)

1949: sample. Columns 4-6: Min, max, median values for a considered

1950: quantity's distribution. Column 7: Mean and standard deviation for

1951: a considered quantity's distribution. Column 8: K-S probability

1952: for the assumption that distributions of a considered quantity in

1953: disky and diskfree stars are drawn from the same underlying

1954: distribution.}

1955: 

1956: 

1957: 

1958: \end{deluxetable}

1959: 

1960: 

1961: 

1962: \clearpage

1963: 

1964: 

1965: 

1966: \begin{deluxetable}{cccccccc}

1967: \centering \rotate \tabletypesize{\small} \tablewidth{0pt}

1968: \tablecolumns{8}

1969: 

1970: 

1971: 

1972: \tablecaption{Summary of Flare Properties \label{tbl_correl_new}}

1973: 

1974: 

1975: 

1976: \tablehead{

1977: 

1978: 

1979: 

1980: \colhead{Quantity} & \colhead{N} & \colhead{Flag} & \colhead{Min}

1981: & \colhead{Max} &

1982: \colhead{Median} & \colhead{Mean$\pm$SD} & \colhead{$P_{KS}$}\\

1983: 

1984: 

1985: 

1986: (1)&(2)&(3)&(4)&(5)&(6)&(7)&(8)}

1987: 

1988: 

1989: 

1990: \startdata

1991: 

1992: 

1993: 

1994: \cline{1-8}

1995: Peak Flare Luminosity (erg s$^{-1}$) & & & & & & & \\

1996: $\log(L_{X,pk})$ (all available)  & 216 & f & 30.45 & 32.92 & 31.33 & $31.38\pm0.51$ & \nodata\\

1997: $\log(L_{X,pk,NIRdisk})$ (NIR inner disk)  & 78 & f & 30.44 & 32.51 & 31.35 & $31.36\pm0.46$ & \nodata\\

1998: $\log(L_{X,pk,noNIRdisk})$ (no NIR inner disk)  & 106 & f & 30.53 & 32.92 & 31.38 & $31.41\pm0.52$ & 0.9\\

1999: $\log(L_{X,pk,MIRdisk})$ (MIR disk)  & 89 & f & 30.49 & 32.51 & 31.27 & $31.32\pm0.44$ & \nodata\\

2000: $\log(L_{X,pk,noMIRdisk})$ (no MIR disk)  & 60 & f & 30.57 & 32.77 & 31.38 & $31.40\pm0.53$ & 0.4\\

2001: \cline{1-8}

2002: Peak Flare Temperature (MK) & & & & & & & \\

2003: $T_{pk}$ (all available)  & 216 & f & 19 & 700 & 63 & $141\pm178$ & \nodata\\

2004: $T_{pk,NIRdisk}$ (NIR inner disk)  & 78 & f & 23 & 700 & 73 & $170\pm209$ & \nodata\\

2005: $T_{pk,noNIRdisk}$ (no NIR inner disk)  & 106 & f & 19 & 700 & 59 & $123\pm153$ & 0.5\\

2006: $T_{pk,MIRdisk}$ (MIR disk)  & 89 & f & 19 & 700 & 84 & $181\pm209$ & \nodata\\

2007: $T_{pk,noMIRdisk}$ (no MIR disk)  & 60 & f & 20 & 600 & 57 & $102\pm114$ & 0.009\\

2008: \cline{1-8}

2009: Inferred Loop Size ($10^{10}$~cm) & & & & & & & \\

2010: $L$ (all available)  & 175 & f & 0.4 & 510 & 43 & $64\pm70$ & \nodata\\

2011: $L_{NIRdisk}$ (NIR inner disk)  & 59 & f & 4 & 420 & 43 & $60\pm63$ & \nodata\\

2012: $L_{noNIRdisk}$ (no NIR inner disk)  & 89 & f & 6 & 510 & 43 & $67\pm71$ & 0.8\\

2013: $L_{MIRdisk}$ (MIR disk)  & 68 & f & 4 & 510 & 56 & $79\pm86$ & \nodata\\

2014: $L_{noMIRdisk}$ (no MIR disk)  & 54 & f & 0.4 & 300 & 37 & $54\pm55$ & 0.12\\

2015: \cline{1-8}

2016: Ratio of Loop Size to Stellar Radius& & & & & & & \\

2017: $L/R_{\star}$ (all available) & 147 & f & 0.03 & 45 & 2.4 & $4.4\pm7.3$ & \nodata\\

2018: $(L/R_{\star})_{NIRdisk}$ (NIR inner disk)  & 48 & f & 0.14 & 33 & 2.5 & $4.0\pm5.1$ & \nodata\\

2019: $(L/R_{\star})_{noNIRdisk}$ (no NIR inner disk)  & 81 & f & 0.2 & 46 & 2.4 & $4.6\pm7.9$ & 0.5\\

2020: $(L/R_{\star})_{MIRdisk}$ (MIR disk)  & 52 & f & 0.4 & 46 & 3.0 & $5.9\pm8.2$ & \nodata\\

2021: $(L/R_{\star})_{noMIRdisk}$ (no MIR disk)  & 47 & f & 0.03 & 29 & 1.8 & $2.9\pm4.4$ & 0.009\\

2022: \cline{1-8}

2023: Ratio of Loop Size to Co-rotation Radius & & & & & & & \\

2024: $L/R_{cor}$ (all available) & 81 & f & 0.007 & 5.4 & 0.51 & $0.73\pm0.76$ & \nodata\\

2025: $(L/R_{cor})_{NIRdisk}$ (NIR inner disk) & 18 & f & 0.08 & 1.3 & 0.51 & $0.54\pm0.30$ & \nodata\\

2026: $(L/R_{cor})_{noNIRdisk}$ (no NIR inner disk) & 51 & f & 0.05 & 5.4 & 0.51 & $0.82\pm0.88$ & 0.15\\

2027: $(L/R_{cor})_{MIRdisk}$ (MIR disk) & 30 & f & 0.07 & 5.4 & 0.45 & $0.66\pm0.94$ & \nodata\\

2028: $(L/R_{cor})_{noMIRdisk}$ (no MIR disk) & 33 & f & 0.007 & 2.4 & 0.69 & $0.78\pm0.66$ & 0.19\\

2029: \cline{1-8}

2030: Ratio of Loop Size plus Stellar Radius & & & & & & & \\

2031: to Co-rotation Radius &&&&&&& \\

2032: $(L+R_{\star})/R_{cor}$ (all available) & 81 & f & 0.16 & 5.5 & 0.72 & $0.98\pm0.80$ & \nodata\\

2033: $((L+R_{\star})/R_{cor})_{NIRdisk}$ (NIR inner disk) & 18 & f & 0.28 & 1.5 & 0.66 & $0.69\pm0.32$ &\nodata\\

2034: $((L+R_{\star})/R_{cor})_{noNIRdisk}$ (no NIR inner disk) & 51 & f & 0.16 & 5.5 & 0.88 & $1.11\pm0.90$ & 0.08\\

2035: $((L+R_{\star})/R_{cor})_{MIRdisk}$ (MIR disk) & 30 & f & 0.24 & 5.49 & 0.62 & $0.81\pm0.94$ & \nodata\\

2036: $((L+R_{\star})/R_{cor})_{noMIRdisk}$ (no MIR disk) & 33 & f & 0.18 & 3.11 & 0.98 & $1.14\pm0.75$ & 0.01\\

2037: \cline{1-8}

2038: 

2039: 

2040: 

2041: \enddata

2042: 

2043: 

2044: \tablecomments{Columns 1: Quantity name. In case of loop sizes,

2045: summary is provided for a distribution of a mean value of the

2046: inferred loop size ranges. Column 2: Number of flares (Flag$=$f,

2047: see Column 3) or sources (Flag$=$s) in a sample. Column 3: Flag

2048: indicates if this is a flare (f) or source (s) sample. Columns

2049: 4-6: Min, max, median values for a considered quantity's

2050: distribution. Column 7: Mean and standard deviation for a

2051: considered quantity's distribution. Column 8: K-S probability for

2052: the assumption that distributions of a considered quantity in

2053: disky and diskfree stars are drawn from the same underlying

2054: distribution.}

2055: 

2056: 

2057: 

2058: \end{deluxetable}

2059: 

2060: 

2061: 

2062: \clearpage

2063: 

2064: 

2065: 

2066: 

2067: \begin{figure}

2068: \centering

2069: \includegraphics[angle=0.,width=6.5in]{f1.eps}

2070: \caption{Disk indicators $K_s - [3.6]$ $vs.$ $[3.6]-[4.5]$

2071: color-color diagram for 98 COUP sources with available infrared

2072: photometry. For 59 sources $\Delta(H-K_s) < -0.06$~mag indicating

2073: no inner disk, while for 39 sources (marked by $\times$)

2074: $\Delta(H-K_s) > -0.06$~mag indicating an inner disk. Reddening

2075: vector of $A_V \sim 30$~mag is shown as the grey arrow. The grey

2076: serpent line roughly discriminates between stars with and without

2077: a MIR excess from a circumstellar disk.  Typical errors on $K_s -

2078: [3.6]$ are 0.02~mag, while errors on $[3.6]-[4.5]$ are shown as

2079: grey bars. \label{fig_Ksch1_vs_ch12}}

2080: \end{figure}

2081: 

2082: 

2083: 

2084: \clearpage

2085: 

2086: 

2087: 

2088: \begin{figure}

2089: \centering

2090: \includegraphics[angle=0.,width=6.5in]{f2.eps}

2091: \caption{Inner disk indicator $\Delta(H-K_s)$ $vs.$ accretion

2092: indicator from the 8542~\AA\ Ca~II line for 128 flares on 96 COUP

2093: stars with available data. Symbols represent flare morphologies:

2094: 57 ``typical'' flares (black circles); 21 ``step'' flares (red

2095: boxes); 12 ``slow-rise-and/or-top-flat'' flares (blue diamonds);

2096: 23 ``incomplete'' flares (brown triangles); 5 ``double'' flares

2097: (green $\times$); and 10 ``other'' flares (magenta triangles). The

2098: grey serpent line roughly discriminates between stars with and

2099: without inner disk. The orange serpent line very roughly

2100: discriminates between stars with and without accretion.

2101: \label{fig_dhks_vs_ewca}}

2102: \end{figure}

2103: 

2104: 

2105: 

2106: \clearpage

2107: 

2108: 

2109: \begin{figure}

2110: \centering

2111: \includegraphics[angle=0.,width=6.5in]{f3.eps}

2112: \caption{Confirmation of PMS disk-rotation relationship: (a)

2113: Keplerian corotation radius scaled to stellar radius $vs.$ stellar

2114: rotational period; and (b) Keplerian corotation radius scaled to

2115: stellar radius versus stellar radius. Panels present results for

2116: all stars with available stellar properties (67 out of 161).

2117: Symbols indicate subsamples: stars with MIR excess disk stars

2118: (blue circles), without MIR disks (red crosses), uncertain disks

2119: (grey circles), and highly accreting stars with $EW(CaII) <-2$~\AA

2120: (green boxes). \label{fig_rotation}}

2121: \end{figure}

2122: 

2123: 

2124: 

2125: \clearpage

2126: 

2127: 

2128: 

2129: \begin{figure}

2130: \centering

2131: \includegraphics[angle=0.,width=6.5in]{f4.eps}

2132: \caption{Locations of subsamples of X-ray bright flaring COUP

2133: stars plotted on the map of the OMC derived from the

2134: velocity-integrated intensity of $^{13}$CO \citep{Bally87}.

2135: Symbols are coded by their X-ray absorption:  $\log(N_H)<21$~

2136: cm$^{-2}$ (red), $21<\log(N_H)<21.5$~cm$^{-2}$ (green),

2137: $21.5<\log(N_H)<22$~cm$^{-2}$ (magenta), and $22<\log(N_H)<

2138: 22.8$~cm$^{-2}$ (blue). Stars with masses $M>2$~M$_{\odot}$ are

2139: outlined by yellow diamonds; stars harboring super-hot

2140: ($T_{obs,pk} > 100$~MK) flares are labelled by yellow $\times$.

2141: Two small cyan boxes indicate BN/KL and OMC1-S regions

2142: \citep{Grosso05}, and the cyan circle marks $\theta^{1}$~Ori~C at

2143: the center of the Orion Nebula Cluster. \label{fig_ultrahot_3}}

2144: \end{figure}

2145: 

2146: 

2147: 

2148: 

2149: \clearpage

2150: 

2151: 

2152: \begin{figure}

2153: \centering

2154: \includegraphics[angle=0.,width=6.0in]{f5.eps}

2155: \caption{\footnotesize Cumulative distributions of flare rise

2156: ($a$) and decay ($b$) timescales; X-ray flare peak luminosity

2157: ($c$); energy from the ``characteristic'' state within flare

2158: duration time; ($d$); flare energy ($e$); and ratio of energies in

2159: flare and ``characteristic'' states ($f$).  Line types indicate

2160: samples: all 216 flares (grey), 89 flares from MIR disk stars

2161: (solid black), 60 diskfree stars (dotted black), and 17

2162: high-accretor stars (dashed black). Legends indicate median

2163: values, and KS test probability compare diskfree and high-accretor

2164: stars. \label{fig_energy_contrib}}

2165: \end{figure}

2166: 

2167: 

2168: 

2169: 

2170: 

2171: \clearpage

2172: 

2173: 

2174: \begin{figure}

2175: \centering

2176: \includegraphics[angle=0.,width=6.5in]{f6.eps}

2177: \caption{Inferred sizes of coronal flaring structures scaled to

2178: stellar corotation radii $vs.$ $K_s$-excess inner disk indicator.

2179: Results are shown for 69 flares from 56 COUP stars with available

2180: data.  Flares are labelled by corresponding COUP source numbers.

2181: Symbols distinguish flare morphologies: 32 ``typical'' flares

2182: (black circles); 14 ``step'' flares (red boxes); 7

2183: ``slow-rise-and/or-top-flat'' flares (blue diamonds); 11

2184: ``incomplete'' flares (brown triangles); and 5 ``other'' flares

2185: (magenta triangles). The grey serpent line roughly discriminates

2186: between stars with and without inner disks. The vertical bars

2187: represent boundaries of the inferred loop size ranges (Paper~I)

2188: with symbols positioned at the mean of those ranges.

2189: \label{fig_L_vs_dHKs}}

2190: \end{figure}

2191: 

2192: 

2193: \clearpage

2194: 

2195: 

2196: 

2197: \begin{figure}

2198: \centering

2199: \includegraphics[angle=0.,width=6.5in]{f7.eps}

2200: \caption{Inferred sizes of coronal flaring structures scaled to

2201: stellar corotation radii $vs.$ [3.6] $-$ [4.5] MIR disk indicator.

2202: Results are shown for 63 flares from 47 COUP stars with available

2203: data.  See Figure~\ref{fig_L_vs_dHKs} for symbols.

2204: \label{fig_L_vs_ch12}}

2205: \end{figure}

2206: 

2207: 

2208: 

2209: \clearpage

2210: 

2211: 

2212: 

2213: \begin{figure}

2214: \centering

2215: \includegraphics[angle=0.,width=6.5in]{f8.eps}

2216: \caption{Inferred sizes of coronal flaring structures scaled to

2217: stellar corotation radii $vs.$ 8542~\AA\ Ca~II line accretion

2218: indicator. Results are shown for 66 flares from 53 COUP stars with

2219: available information data.  The grey serpent line roughly

2220: discriminates between stars with and without accretion. See

2221: Figure~ \ref{fig_L_vs_dHKs} for symbols.  \label{fig_L_vs_ewca}}

2222: \end{figure}

2223: 

2224: 

2225: 

2226: \clearpage

2227: 

2228: 

2229: 

2230: \begin{figure}

2231: \centering

2232: \includegraphics[angle=0.,width=6.0in]{f9.eps}

2233: \caption{Coronal extents of analyzed flaring structures $vs.$: (a)

2234: stellar rotational period; (b) loop size normalized to stellar

2235: radius; and (c) stellar mass.  See Figure~\ref{fig_rotation} for

2236: symbols. \label{fig_corona_extent_suppl}}

2237: \end{figure}

2238: 

2239: 

2240: 

2241: \clearpage

2242: 

2243: 

2244: \begin{figure}

2245: \centering

2246: \includegraphics[angle=0.,width=7.0in]{f10.eps}

2247: \caption{(a) Cumulative distributions of flare peak temperature for

2248: all 216 flares (grey line), 89 flares from MIR disk stars (solid black), 60

2249: from diskfree stars (dotted), and 17 from high-accretors (dashed). (b) Mass-stratified cumulative

2250: distributions of flare peak temperature for 57 flares from MIR disk stars with

2251: $M<2$~M$_{\odot}$ (thick solid), 38 from diskfree stars with $M<2$~M$_{\odot}$

2252: (thick dotted), 15 from high-accretors with $M<2$~M$_{\odot}$ (thick dashed),

2253: 8 from MIR disk stars with $M>2$~M$_{\odot}$ (thin solid), and 12 from diskfree

2254: stars with $M>2$~M$_{\odot}$ (thin dotted). Legends indicate median values. KS 

2255: test probabilities compare temperature distributions between non-disk and disk 

2256: or high-accretor stars.

2257: \label{fig.peakTdisk}}

2258: \end{figure}

2259: 

2260: 

2261: 

2262: \clearpage

2263: 

2264: 

2265: 

2266: \begin{figure}

2267: \centering

2268: \includegraphics[angle=0.,width=6.5in]{f11.eps}

2269: \caption{Estimated plasma density at flare peak (panel $a$) and

2270: magnetic field from pressure equipartition (panel $b$) are plotted

2271: against inferred loop sizes scaled to stellar radii. Symbols

2272: discriminate subsamples: MIR disk stars (blue circles); no MIR

2273: disk stars (red crosses); uncertain MIR disk stars (grey circles);

2274: highly accreting stars with $EW(CaII) <-2$~\AA\ (green boxes);

2275: masses $M>2$~M$_{\odot}$ (black triangles); and super-hot flares

2276: with $T_{obs,pk}>100$~MK (magenta circles in panel $b$). In panel

2277: $a$, the dashed grey line indicates $n_e \propto L^{-3/2}$. In

2278: panel $b$, the cyan curves represent loci of photospheric dipolar

2279: magnetic fields $B_{ph}=(1, 3, 6)$~kG (solid) and $(10, 50)$~kG

2280: (dashed). \label{fig_ultrahot_5}}

2281: \end{figure}

2282: 

2283: 

2284: \clearpage

2285: 

2286: 

2287: 

2288: \begin{figure}

2289: \centering

2290: \includegraphics[angle=0.,width=6.5in]{f12.eps}

2291: \caption{Sketch showing typical sizes of the largest X-ray

2292: emitting magnetic structures assuming a dipolar geometry (hatched)

2293: for Class~II and Class~III T-Tauri stars, respectively.   Relative

2294: sizes are shown to scale for typical stars.   The inset shows the

2295: X-ray image of the Sun, both to scale and expanded by a factor of

2296: 4 for clarity, showing a large helmet-like X-ray streamers  (24

2297: Jan 1992 flare, $Yohkoh$~SXT).   \label{fig_corona_extent_sketch}}

2298: \end{figure}

2299: 

2300: 

2301: 

2302: \clearpage

2303: 

2304: 

2305: 

2306: \begin{figure}

2307: \centering

2308: \includegraphics[angle=0.,width=6.5in]{f13.eps}

2309: \caption{For all 216 COUP flares, comparison of the flare peak

2310: temperature from MASME analysis (Paper~I) with the temperature

2311: derived from the spectral fit of the integrated flare spectrum.

2312: Dashed lines show 100~MK and 40~MK, respectively.

2313: \label{fig_super_hot_analysis1}}

2314: \end{figure}

2315: 

2316: 

2317: 

2318: \clearpage

2319: 

2320: 

2321: \begin{figure}

2322: \centering

2323: \includegraphics[angle=0.,width=5.5in]{f14.eps}

2324: \caption{\footnotesize Composite flare peak spectra from MIR disk

2325: stars with source X-ray column densities in the narrow range of

2326: absorption.  See Appendix for sample details. (a) Composite of 14

2327: ordinary flares from 12 stars with $T_{obs,pk} < 100$~MK. (b)

2328: Composite of 21 super-hot flares from 14 stars with $T_{obs,pk}

2329: > 100$~MK.  Three spectral models and residuals are:

2330: ordinary flare model (black), super-hot flare model (red),

2331: ordinary flare model with $N_H$ fixed to that of super-hot model

2332: (green). \label{fig_ultrahot_4}}

2333: \end{figure}

2334: 

2335: 

2336: \end{document}

2337: