1: \documentclass{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: %\documentclass[12pt,preprint]{/astro/krivov/SMALL_PROJECTS/PLANETESIMAL_BELTS/aastex52/aastex}
4: %\documentclass[10pt,preprint2]{/astro/krivov/SMALL_PROJECTS/PLANETESIMAL_BELTS/aastex52/aastex}
5:
6: %-----------------------------------------------------------------------
7: % Authors' Macros
8: %-----------------------------------------------------------------------
9:
10: %\newcommand{\rev}[1]{{\bf [#1}]}
11: %\newcommand{\corrected}[1]{{\bf #1}}
12: \newcommand{\rev}[1]{}
13: \newcommand{\corrected}[1]{#1}
14:
15: \newcommand{\ph}{\phantom{0}}
16:
17: \newcommand{\AAp}{\aap}
18: \newcommand{\ApJ}{\apj}
19: \newcommand{\ApJS}{\apjs}
20: \newcommand{\AJ}{\aj}
21: \newcommand{\ARevEPS}{Ann. Rev. Earth Planet. Sci.}
22: \newcommand{\JGR}{\jgr}
23: \newcommand{\MNRAS}{\mnras}
24: \newcommand{\AdvSR}{\apspr}
25: \newcommand{\PSS}{\planss}
26: \newcommand{\PASP}{\pasp}
27:
28: \newcommand{\be}{\begin{equation}}
29: \newcommand{\ee}{\end{equation}}
30: \newcommand{\bd}{\begin{displaymath}}
31: \newcommand{\ed}{\end{displaymath}}
32: \newcommand{\bea}{\begin{eqnarray}}
33: \newcommand{\eea}{\end{eqnarray}}
34: \newcommand{\mum}{\,\mu\hbox{m}}
35: \newcommand{\m}{\,\mbox{m}}
36: \newcommand{\mm}{\,\hbox{mm}}
37: \newcommand{\cm}{\,\mbox{cm}}
38: \newcommand{\nm}{\,\mbox{nm}}
39: \newcommand{\km}{\,\mbox{km}}
40: \newcommand{\AU}{\,\mbox{AU}}
41: \newcommand{\g}{\,\mbox{g}}
42: \newcommand{\s}{\,\mbox{s}}
43: \newcommand{\pc}{\,\mbox{pc}}
44: \newcommand{\yr}{\,\mbox{yr}}
45: \newcommand{\Jy}{\,\mbox{Jy}}
46: \newcommand{\erg}{\,\mbox{erg}}
47: \newcommand{\K}{\,\mbox{K}}
48: \newcommand{\D}{\,\mbox{d}}
49:
50: \shortauthors{Krivov et al.}
51: \shorttitle{Debris disk models and planetesimal properties}
52:
53: \begin{document}
54:
55: %===========================================================================
56: % "Title page"
57: %===========================================================================
58:
59: \title{Collisional and Thermal Emission Models of Debris Disks:\\
60: Towards Planetesimal Population Properties
61: %Seeing Dust, Thinking of Planetesimals:
62: %A Link through Debris Disk Models
63: }
64: \author{Alexander V. Krivov,
65: Sebastian M{\"u}ller,
66: Torsten L{\"o}hne, and
67: Harald Mutschke}
68: \affil{Astrophysikalisches Institut und Universit{\"a}tssternwarte,
69: Friedrich Schiller University Jena,
70: Schillerg{\"a}{\ss}chen~ 2--3, 07745 Jena, Germany;
71: krivov@astro.uni-jena.de}
72:
73: \begin{abstract}
74: Debris disks around main-sequence stars are believed to derive
75: from planetesimal populations that have accreted at
76: early epochs and survived possible planet formation processes.
77: While debris disks must contain solids in a broad range of sizes~---
78: from big planetesimals down to tiny dust grains~--- debris disk observations
79: are only sensitive to the dust end of the size distribution.
80: Collisional models of debris disks are needed
81: to ``climb up'' the ladder of the collisional cascade, from dust
82: towards parent bodies, representing the main mass reservoir of the disks.
83: We have used our collisional code to
84: generate five disks around a sun-like star, assuming
85: planetesimal belts at 3, 10, 30, 100, and $200\AU$ with 10 times the Edgeworth-Kuiper-belt mass
86: density, and to evolve them for 10~Gyr.
87: Along with an appropriate scaling rule, this effectively yields a
88: three-parametric set of reference disks (initial mass, location of
89: planetesimal belt, age).
90: For all the disks, we have generated
91: spectral energy distributions (SEDs),
92: assuming homogeneous spherical astrosilicate dust grains.
93: A comparison between generated and actually observed SEDs
94: yields estimates of planetesimal properties (location, total mass etc.).
95: As a test and a first application of this approach,
96: we have selected five disks around sun-like stars with well-known SEDs.
97: In four cases, we have reproduced the data
98: with a linear combination of two disks from the grid
99: (an ``asteroid belt'' at $3\AU$ and an outer ``Kuiper belt'');
100: in one case a single, outer component was sufficient.
101: The outer components are compatible with ``large Kuiper belts''
102: of 0.2--50 earth masses (in the bodies up to $100\km$ in size)
103: with radii of $100$--$200\AU$.
104: \end{abstract}
105:
106: \keywords{circumstellar matter --- planetary systems: formation ---
107: Kuiper belt --- stars: individual (HD~377, HD~70573, HD~72905, HD~107146,
108: HD~141943)
109: }
110:
111: %===========================================================================
112: \section{Introduction}
113: %===========================================================================
114:
115: Since the IRAS discovery of the excess infrared emission around Vega by
116: \citet{aumann-et-al-1984}, infrared surveys with IRAS, ISO, Spitzer, and
117: other space-based and ground-based telescopes
118: have shown the Vega phenomenon to be common for main-sequence stars
119: \cite[e.g.][]{meyer-et-al-2004,beichman-et-al-2005,najita-williams-2005,rieke-et-al-2005,
120: bryden-et-al-2006,siegler-et-al-2006, su-et-al-2006,
121: trilling-et-al-2007,hillenbrand-et-al-2008,trilling-et-al-2008}.
122: The observed excesses are attributed to
123: circumstellar disks of second-generation dust,
124: sustained by numerous planetesimals in orbit around the stars.
125: Jostling collisions between planetesimals grind them all the way down to
126: smallest dust grains which are then blown away by stellar radiation.
127: While the bulk of such a
128: debris disk's mass is hidden in invisible parent bodies,
129: the observed luminosity is dominated by small particles at dust sizes.
130: Hence the studies of dust emission have a potential to shed light onto
131: the properties of parent planetesimal populations as well
132: as planets that may shape them and,
133: ultimately, onto the evolutionary history of circumstellar planetary systems.
134:
135: However, there is no direct way to infer
136: the properties of invisible planetesimal populations
137: from the observed dust emission. Dust and planetesimals can only be linked
138: through models.
139: First, dynamical models can be used
140: to predict, for a given planetesimal family (mass, location, age, etc.), the distribution of dust.
141: Such models have become available in recent years
142: \citep[e.g.][]{thebault-et-al-2003,krivov-et-al-2006,thebault-augereau-2007,wyatt-et-al-2007,loehne-et-al-2007}.
143: After that, standard thermal emission models will describe the resulting dust emission.
144: Comparison of that emission to the one actually observed would then reveal the probable properties of
145: underlying, dust-producing planetesimal families.
146:
147: %To facilitate interpretation of SEDs of debris disks, we have compiled
148: In this paper, we follow this approach
149: %a catalog of SEDs for a set of hypothetical debris disks around G2 dwarfs
150: %of different spectral classes (F, G, and K)
151: and generate a set of hypothetical debris disks around G2 dwarfs
152: with different ages (10~Myr -- 10~Gyr), assuming debris dust to stem from planetesimal
153: belts with different initial masses at different distances from the
154: central star.
155: For every set of these parameters, we simulate steady-state dust distributions
156: with our collisional code
157: \citep{krivov-et-al-2005,krivov-et-al-2006,loehne-et-al-2007}.
158: This is different from a traditional, ``empirical'' approach,
159: in which dust distributions are postulated, usually in form of power laws,
160: parameterized by ranges and exponents that play a role of fitting parameters
161: \citep[e.g.][]{wolf-hillenbrand-2003}.
162: Interestingly, replacing formal dust distributions with those coming out
163: of dynamical modeling does not increase the number of fitting parameters.
164: Just the opposite: the number of parameters reduces and those parameters that
165: we keep free all have clear astrophysical meaning.
166: The most important are
167: location of a parent planetesimal belt and its current mass \citep{wyatt-et-al-2007}.
168: %its initial mass (at the onset of
169: %the collisional evolution), and system's age
170:
171: Having produced a set of model debris disks, we compute thermal emission fluxes in a wide
172: range of wavelengths from mid-infrared to millimeter. In so doing, we completely abandon simple blackbody
173: or modified blackbody calculations and solve a thermal balance equation
174: instead. At this stage, we assume compact spherical grains composed of astronomical silicate
175: \corrected{\citep{laor-draine-1993}}
176: and employ standard Mie calculations to compute dust opacities.
177: Although this is still a noticeable simplification, it represents a natural
178: step towards considering realistic materials and using more involved methods of light
179: scattering theory that we leave for subsequent papers.
180:
181: As a test and a first application of the results,
182: we re-interprete available observational data on a selection of disks
183: around sun-like stars with well known spectral energy distributions (SEDs).
184:
185: This paper is organized as follows.
186: Section 2 describes the dynamical and thermal emission models.
187: In section~3, a set of reference disks is introduced and the model parameters
188: are specified.
189: Section~4 presents the modeling results for this set of disks: size and spatial
190: distribution of dust, dust temperatures, and the generated SEDs.
191: Application to selected observed disks is made in section 5.
192: Section~6 summarizes the paper.
193: %our conclusions and a discussion.
194:
195:
196: %===========================================================================
197: \section{Model}
198: %===========================================================================
199: \subsection{Dynamical model}
200:
201: To simulate the dust production by the planetesimal belt and
202: the dynamical evolution of a disk, we use our collisional
203: code (ACE, Analysis of Collisional Evolution).
204: The code numerically solves the Boltzmann-Smoluchowski kinetic equation
205: to evolve a disk of solids in a broad range of sizes
206: (from smallest dust grains to planetesimals),
207: orbiting a primary in nearly Keplerian orbits
208: (gravity + direct radiation pressure + drag forces) and
209: experiencing disruptive and erosive (cratering) collisions.
210: Collision outcomes are simulated with available material- and size-dependent scaling
211: laws for fragmentation and dispersal in both strength and gravity regime.
212: The current version implements a 3-dimensional kinetic model, with masses,
213: semimajor axes, and eccentricities as phase space variables.
214: This approach automatically enables a study of the simultaneous evolution of
215: mass, spatial, and velocity distribution of particles.
216: The code is fast enough to easily follow the evolution of a debris disk over
217: Gyr timescales.
218: A detailed description of our approach, its numerical implementation, and astrophysical
219: applications can be found in our previous papers
220: \citep{krivov-et-al-2000b,krivov-et-al-2005,krivov-et-al-2006,loehne-et-al-2007}.
221:
222: \subsection{Thermal emission model}
223:
224: For spherical dust grains with radius $s$ and temperature $T_{\mathrm{g}}$
225: we can calculate their
226: distance $r$ to the star under the assumption of thermal equilibrium as
227: \bea
228: r = \frac{R_*}{2} \sqrt{ \frac{\int_0^{\infty} \D\lambda \; Q_{\lambda}^{\mathrm{abs}}(s)
229: F_{\lambda, *}(T_*)}
230: {\int_0^{\infty} \D\lambda \; Q_{\lambda}^{\mathrm{abs}}(s)
231: B_{\lambda}(T_{\mathrm{g}})} }
232: \label{equ:r} .
233: \eea
234: Here, $R_*$ denotes the radius and $F_{\lambda, *}(T_*)$ the flux
235: of the star with an effective temperature $T_*$
236: and $B_{\lambda}(T_{\mathrm{g}})$ the Planck function.
237: The absorption efficiency $Q_{\lambda}^{\mathrm{abs}}(s)$
238: is a function of wavelength $\lambda$ and particle size.
239:
240: We now consider a rotationally symmetric dust disk at a distance $D$ from the observer.
241: Denote by $ N(r,s)$
242: the surface number density of grains with radius $s$ at a distance $r$ from the star,
243: so that $N(r,s)ds$ is the number of grains with radii $[s,s+ds]$ in a
244: narrow annulus of radius $r$, divided by the surface area of that annulus.
245: Then the specific flux emitted from the entire disk at a given wavelength can be calculated as
246: \bea
247: F_{\lambda, \mathrm{disk}}^{\mathrm{tot}}
248: & = & \int \D r \int \D s \; F_{\lambda, \mathrm{disk}}(r,s) \label{equ:F1}\\
249: & = & \frac{2\pi^2}{D^2} \int \D T_g \, r(T_g) \, \frac{\D r(T_g)}{\D T_g} \int \D s \; s^2
250: \;\times\nonumber\\
251: & & \times \; N(r,s) \,Q_{\lambda}^{abs}(s) \, B_{\lambda}(T_g)
252: \label{equ:F} .
253: \eea
254:
255: %===========================================================================
256: \section{Reference disks}
257: %===========================================================================
258:
259: \subsection{Central star}
260:
261: The parameters of the central star (mass and photospheric spectrum)
262: affect both the dynamics of solids
263: (by setting the scale of orbital velocities and
264: determining the radiation pressure strength)
265: and their thermal emission
266: (by setting the dust grain temperatures).
267: We take the Sun (a G2V dwarf with a solar metallicity)
268: as a central star and calculate its photospheric spectrum with the NextGen
269: grid of models \citep{hauschildt-et-al-1999}.
270:
271: \subsection{Forces}
272:
273: In the dynamical model, we include central star's gravity and
274: direct radiation pressure. We switch off the drag forces (both the
275: Poynting-Robertson and stellar wind drag), which are of little
276: importance for the optical depths in the range from $\sim 10^{-5}$ to $\sim 10^{-3}$)
277: considered here \citep{artymowicz-1997,krivov-et-al-2000b,wyatt-2005}.
278:
279: \subsection{Collisions}
280:
281: The radii of solids in every modeled disk cover the interval from $0.1\mum$ to $100\km$.
282: The upper limit of $100\km$ is justified by the fact that planetesimal accretion
283: models predict larger objects to have a steeper size distribution and thus to contribute
284: less to the mass budget of a debris disk \citep[e.g.][]{kenyon-luu-1999b}.
285: To describe the collisional outcomes, we make the same assumptions
286: as in \citet{loehne-et-al-2007}.
287: This applies, in particular, to the critical energy for disruption and
288: dispersal, $Q_D^*(s)$, as well as to the size distribution of fragments of
289: an individual collision. However, in contrast to \citet{loehne-et-al-2007}
290: where only catastrophic collisions were taken into account, we include here
291: cratering collisions as well. This is necessary, as cratering collisions
292: alter the size distribution of dust in the disk markedly, which shows up in
293: the SEDs \citep{thebault-et-al-2003,thebault-augereau-2007}.
294: The actual model of cratering collisions used here is close to that
295: by \citet{thebault-augereau-2007}.
296: An essential difference is our assumption of a single power law for
297: the size distribution of the fragments of an individual collision instead
298: of the broken power law proposed originally in \citet{thebault-et-al-2003}.
299: However, this difference has little effect on the resulting size distribution in
300: collisional equilibrium.
301:
302: \subsection{Optical properties of dust}
303:
304: An important issue is a choice of grain composition and morphology.
305: These affect both the dynamical model (through radiation pressure
306: efficiency
307: % and mechanical properties determining collisional outcomes,
308: as well as bulk density)
309: and thermal emission model (through absorption efficiency).
310: Here we assume compact spherical grains composed of astronomical silicate
311: \citep[a.k.a. astrosilicate or astrosil,][]{laor-draine-1993}, similar to the MgFeSiO$_4$ olivine,
312: with density of $3.3\g\cm^{-3}$.
313: Taking optical constants from \citet{laor-draine-1993}, we calculated
314: radiation pressure efficiency $Q_{pr}$ and absorption efficiency $Q_{abs}$ with a standard
315: Mie routine \citep{bohren-huffman-1983}.
316:
317: To characterize the radiation pressure strength, it is customary to
318: use the radiation pressure to gravity ratio $\beta$ \citep{burns-et-al-1979},
319: which is independent of distance from the star and, for a given star,
320: only depends on $Q_{pr}$ and particle size. If grains that are small enough
321: to respond to radiation pressure derive from collisions of larger objects
322: in nearly circular orbits, they will get in orbits with eccentricities
323: $e \sim \beta / (1 - \beta)$. This implies that grains with $\beta < 0.5$
324: remain orbiting the star, whereas those with $\beta > 0.5$
325: leave the system in hyperbolic orbits.
326: The $\beta$ ratio for compact astrosil grains, computed from $Q_{pr}$,
327: is shown in Fig.~\ref{fig_beta}.
328: \corrected{The blowout limit, $\beta = 0.5$, corresponds to the grain
329: radius of $s = 0.4\mum$.}
330: Note that the tiniest astrosil grains ($ \la 0.1\mum$) would have $\beta < 0.5$ again
331: and thus could orbit the star in bound orbits.
332: However, the dynamics of these small motes would be
333: subject to a variety of effects (e.g. the Lorentz force) not included in our model,
334: and their lifetimes may be shortened by erosion processes (e.g. stellar wind sputtering).
335: Altogether, we expect them to make little contribution to the thermal emission in the mid-IR to
336: sub-mm. By setting the minimum radius of grains to $0.1\mum$, we
337: therefore do not take into account these grains here.
338:
339: \begin{figure}[h!]
340: \begin{center}
341: \includegraphics[scale=0.6]
342: {f1.eps}
343: %{/astro/krivov/ACE/ACE_2008/FGK_CATALOG/beta_of_s.eps}
344: \caption
345: {
346: Radiation pressure to gravity ratio $\beta$ for astrosilicate grains
347: as a function of their size.
348: \corrected{Horizontal lines at $\beta = 0.5$ and $\beta = 1.0$ show,
349: which particles typically move in bound elliptic orbits, in hyperbolas,
350: as well as in anomalous hyperbolas (open outward from the star).}
351: \label{fig_beta}
352: }
353: \end{center}
354: \end{figure}
355:
356: The spectral dependence of the absorption efficiency $Q_{abs}$ of different-sized
357: astrosil spheres is depicted in Fig.~\ref{fig_Qabs}.
358: %Further on, the stellar spectrum of a G2 V star and the Planck curves for
359: %150 and 20 K are added to indicate the spectral ranges of absorption and emission.
360:
361: \begin{figure}[h!]
362: \begin{center}
363: \includegraphics[scale=0.6]
364: {f2.eps}
365: %{/astro/sebastian/Desktop/Programs/DustTemperature/Qabs/Figures/qabs_G2-Draine93.eps}
366: \caption
367: {
368: Top: absorption efficiency of astrosilicate compact spherical grains as a function of
369: wavelength for different grain sizes.
370: Bottom: the spectrum of a G2 V star and
371: the Planck curves for 150 and 20 K (in arbitrary vertical scale)
372: to indicate the spectral ranges
373: most important for absorption and emission.
374: \label{fig_Qabs}
375: }
376: \end{center}
377: \end{figure}
378:
379: \subsection{Parent planetesimal belts}
380:
381: To have a representative set of ``reference'' debris disks
382: around sun-like stars, we consider possible planetesimal rings centered
383: at the semimajor axes of $a = 3$, $10$, $30$, $100$, and $200\AU$ from the primary.
384: All five rings are assumed to have the same relative width initially
385: (again, in terms of semimajor axis) of $\Delta a /a = 0.2 $ ($\pm 0.1$)
386: and share the same semi-opening angle (the same as the maximum orbital inclination of the objects) of
387: $\varepsilon = 0.1$~rad. The orbital eccentricities of planetesimals
388: are then distributed uniformly between 0.0 and 0.2, in accordance with
389: the standard equipartition condition.
390: \corrected{The initial (differential) mass distribution of all solids is given
391: by a power law with the index $1.87$, a value that accounts for the
392: modification of the classical Dohnanyi's (\citeyear{dohnanyi-1969})
393: $1.833$ through the size dependence of material strength
394: \citep[see, e.g.,][]{durda-dermott-1997}.
395: }
396:
397: \begin{deluxetable*}{rcccc}
398: \tablecaption{Description of reference disks
399: \label{tab_runs}
400: }
401: \tablewidth{0pt}
402: \tablehead{
403: \colhead{Disk} &
404: \colhead{Belt} &
405: \colhead{Initial} &
406: \colhead{$a$ range} &
407: \colhead{$r$ range}\\
408: \colhead{identifier} &
409: \colhead{location [AU]} &
410: \colhead{disk mass [$M_\oplus$]} &
411: \colhead{[AU]} &
412: \colhead{[AU]}
413: }
414: \startdata
415: 10EKBD \@@ \ph\ph 3AU & 3 & 0.001 & 0.3 -- 30 & 0.5 -- 20\\
416: 10EKBD \@@ \ph 10AU & 10 & 0.03 & 1 -- 100 & 2 -- 50\\
417: 10EKBD \@@ \ph 30AU & 30 & 1 & 3 -- 300 & 5 -- 200\\
418: 10EKBD \@@ 100AU & 100 & 30 & 10 -- 1000 & 20 -- 500\\
419: 10EKBD \@@ 200AU & 200 & 200 & 20 -- 2000 & 30 -- 1000\\
420: \enddata
421: \end{deluxetable*}
422:
423: The initial disk mass
424: is taken to be $1 M_\oplus$ (earth mass) for
425: a $30\AU$ ring, roughly corresponding to ten (or slightly more) times
426: the Edgeworth-Kuiper belt (EKB) mass \citep[e.g.][]{gladman-et-al-2001b,hahn-malhotra-2005}.
427: For other parent ring locations, the initial mass is taken in such a way
428: as to provide approximately the same spatial {\em density} of material.
429: Since the circumference of a ring $2 \pi a$,
430: its absolute width $\Delta a$,
431: and vertical thickness $2 a \varepsilon$
432: are all proportional to $a$,
433: the condition of a constant density requires the mass scaling $\propto a^3$.
434: This corresponds to the initial mass ranging from $\approx 0.001 M_\oplus$ in the $3\AU$ case
435: to $\approx 200 M_\oplus$ in the $200\AU$ case.
436: With these values, all reference disks have about ten times the
437: EKB density (10~EKBD).
438:
439: \corrected{
440: That all the belts share the same volume density of material is
441: purely a matter of convention. Instead, we could choose them to have the same
442: surface density or the same total mass. Given the scaling rules, as discussed in the
443: text and Appendix~\ref{app_scaling}, none of these choices would have strong
444: advantages or disadvantages.
445: }
446:
447: All five reference disks are listed in Table~\ref{tab_runs}.
448: We evolved them with the collisional code, ACE,
449: and stored all results between the ages of 10~Myr and 10~Gyr
450: at reasonable time steps.
451: In what follows, we use self-explanatory identifiers like
452: \mbox{10EKBD \@@ 10AU \@@ 300Myr} to refer to
453: a particular disk of a particular age.
454:
455: Importantly, the same runs of the collisional code automatically
456: provide the results for disks of any other initial density (or mass).
457: This is possible due to the mass-time scaling of \citet{loehne-et-al-2007},
458: which can be formulated as follows.
459: Denote by $M(M_0,t)$ the mass that a disk with initial mass $M_0$ has
460: at time $t$. Then, the mass of another disk with $x$ times larger initial
461: mass at time instant $t / x$ is simply
462: \be
463: M(x M_0, t / x) = x M(M_0,t) ,
464: \label{mass_scaling}
465: \ee
466: For instance, the mass of the 1EKBD \@@ 10AU \@@ 10Gyr disk is one-tenth
467: of the 10EKBD \@@ 10AU \@@ 1Gyr disk mass.
468: Note that the same scaling applies to any other quantity directly proportional
469: to the amount of disk material.
470: In other words, $M$ may equally stand for the mass of dust,
471: its total cross section, thermal radiation flux, etc.
472: See Appendix~\ref{app_scaling} for additional explanations.
473:
474: %===========================================================================
475: \section{Results}
476: %===========================================================================
477:
478: \subsection{Size and spatial distributions of dust}
479:
480: As noted above, the collisional code ACE uses masses and orbital elements
481: of disk particles as phase space variables. At any time instant, their
482: phase space distribution is transformed to usual mass/size and spatial distributions.
483: It is important to understand that mass/size distributions and spatial distributions
484: cannot, generally, be decoupled from each other. Grains of different sizes
485: have different radial distributions and conversely, the size distribution of
486: material is different at different distances from the star.
487:
488: A typical size distribution of solids is shown in Fig.~\ref{fig_size_dist}
489: for one of the disks, namely for 10EKBD \@@ 30AU \@@ 100Myr.
490: Different lines correspond to different distances from the primary.
491: As expected, the size distribution is the broadest within the parent ring
492: of planetesimals. Farther out, it only contains grains which are small enough
493: to develop orbits with sufficiently large apocentric distances due to radiation
494: pressure.
495:
496: \begin{figure}[h!]
497: \begin{center}
498: \includegraphics[scale=0.6]
499: {f3.eps}
500: %{/astro/krivov/ACE/ACE_2008/FGK_CATALOG/size_dist.eps}
501: \caption
502: {
503: Size distribution in the 10EKBD \@@ 30AU \@@ 100Myr disk at three
504: different distances from the star.
505: \label{fig_size_dist}
506: }
507: \end{center}
508: \end{figure}
509:
510: The spatial distribution of material in the same
511: disk is shown in Fig.~\ref{fig_spat_dist}.
512: Here, different lines refer to different particle sizes.
513: The ring of the biggest particles shown ($100\mum$), for which radiation
514: pressure is negligible, nearly coincides with the initial ring
515: of planetesimals (semimajor axes: from $27$ to $33\AU$, eccentricities:
516: from 0.0 to 0.2, hence radial distances from $22$ to $40\AU$).
517: The larger the particles, the more confined their rings.
518: The rings are more extended outward with respect to the parent planetesimal ring
519: than inward.
520:
521: \begin{figure}[h!]
522: \begin{center}
523: \includegraphics[scale=0.6]
524: {f4.eps}
525: %{/astro/krivov/ACE/ACE_2008/FGK_CATALOG/spat_dist.eps}
526: \caption
527: {
528: Spatial distribution of grains with three characteristic radii
529: for the 10EKBD \@@ 30AU \@@ 100Myr disk. The ring of the biggest
530: particles shown ($100\mum$, hatched) is the narrowest.
531: Its radial extension is nearly the same as that of the initial
532: planetesimal ring; vertical ``walls'' are artifacts due
533: to a discrete distance binning.
534: \label{fig_spat_dist}
535: }
536: \end{center}
537: \end{figure}
538:
539:
540: Radial profiles of the normal geometrical optical depth for
541: three reference disks (planetesimal rings at $10$, $30$, and $100\AU$) are depicted
542: in Fig.~\ref{fig_tau_dist}. Initially, the peak optical depth of the
543: disks is proportional to the distance of the parent ring, making
544: the $100\AU$ disk ten times optically thicker than the $10\AU$ one.
545: The subsequent collisional evolution of \corrected{the disks depends
546: on their initial mass and distance from the star, as explained
547: in detail in \citet{loehne-et-al-2007} and Appendix~\ref{app_scaling}.
548: Once a collisional steady state is reached (which is the case after 10~Myr
549: for all three disks),
550: the optical depth decays with time approximately as $t^\xi$,
551: where $\xi \approx {-0.3 \ldots -0.4}$,
552: i.e. roughly by one order of magnitude from 10~Myr to 10~Gyr.
553: In a steady-state regime, the optical depth is proportional
554: to $r^{1 + 1.3\xi} \sim r^{1.5}$. This explains why, at any age between
555: 10~Myr and 10~Gyr, the $100\AU$ ring is $\approx 30$ times optically thicker
556: than the $10\AU$ one.
557: }
558:
559: \begin{figure}[h!]
560: \begin{center}
561: \includegraphics[scale=0.6]
562: {f5.eps}
563: %{/astro/krivov/ACE/ACE_2008/FGK_CATALOG/tau_of_r.eps}
564: \caption
565: {
566: Radial profiles of the normal geometrical optical depth
567: for three out of five basic runs
568: (10EKBD \@@ 10AU, solid lines;
569: 10EKBD \@@ 30AU, dashed;
570: 10EKBD \@@ 100AU, dotted)
571: at different ages.
572: The thinner the line, the older the disk,
573: as marked in the legend.
574: The dashed-dotted lines are initial optical depths,
575: artificially enhanced by a factor of ten for a better visibility.
576: \label{fig_tau_dist}
577: }
578: \end{center}
579: \end{figure}
580:
581:
582: \subsection{Dust temperatures}
583:
584: Figure~\ref{fig_temp} shows the dust temperatures as a function of two variables:
585: grain distances from the star and their radii.
586: In a parallel scale on the right,
587: we show typical size distributions (cf. Fig.~\ref{fig_size_dist}).
588: Similarly, under the temperature plot, typical radial profiles of the disk
589: are drawn (cf. Fig.~\ref{fig_spat_dist}). This enables a direct ``read-out''
590: of the typical\footnote{``Typical'' in the sense that it is the temperature
591: of cross-section dominating grains in the densest part of the disk.}
592: temperature in one or another disk. We find, for example,
593: $130\K$ at $10\AU$, $90\K$ at $30\AU$, and $50\K$ at $100\AU$.
594:
595: \begin{figure*}[hb!]
596: \begin{center}
597: \includegraphics[scale=0.6]
598: {f6.eps}
599: %{/astro/sebastian/Desktop/Programs/DustTemperature/Temperatures/Figures/T_G2_Draine93.eps}
600: \caption
601: {
602: The left upper plot shows the equilibrium temperature of dust particles as a function
603: of their distance from the star (horizontal axis) and size (vertical axis).
604: Contours are isotherms.
605: The blackbody dust temperatures are given along the upper edge of the plot for comparison.
606: In the right-hand plot the size distribution at the ``central'' distance of
607: the systems ($10\AU$, solid; $30\AU$, dashed; and $100\AU$, dotted)
608: at 100~Myr is given.
609: The lowest left plot gives the normal optical depth for the same three disks
610: as a function of distance to the star.
611: An intersection of a horizontal straight line going through the maximum
612: of the size distribution in a disk (right) with a vertical line through the peak
613: of its radial profile (bottom) provides the typical dust temperature in that disk.
614: \label{fig_temp}
615: }
616: \end{center}
617: \end{figure*}
618:
619: \begin{figure}[h!]
620: \begin{center}
621: \includegraphics[scale=0.6]
622: {f7.eps}
623: %{/astro/sebastian/Desktop/Programs/SED/SEDs/G2-Variation/Figures/G2_Black-SED.eps}
624: \caption
625: {
626: \rev{Figure changed.}
627: The emission from one and the same, 1EKBD \@@ 30AU \@@ 100Myr, disk, calculated under
628: different assumptions about absorbing and emitting properties of dust grains:
629: blackbody, astrosil (our nominal case), and amorphous carbon particles.
630: Thin solid line: photosphere of a G2V star.
631: \label{fig_bb}
632: }
633: \end{center}
634: \end{figure}
635:
636: \begin{figure*}
637: \begin{center}
638: \includegraphics[scale=0.6]
639: {f8a.eps}
640: %{/astro/sebastian/Desktop/Programs/SED/SEDs/G2-Variation/Figures/G2-Variation.eps}
641: \includegraphics[scale=0.6]
642: {f8b.eps}
643: %{/astro/sebastian/Desktop/Programs/SED/SEDs/G2-Variation/Figures/G2_KB-Variation.eps}
644: \caption
645: {
646: \rev{Some labels changed.}
647: Spectral energy distributions of disks
648: stemming from planetesimal rings with different masses at different locations
649: and at different time steps.
650: To obtain the absolute values of fluxes, a distance of 10~pc was assumed.
651: Left: reference disks (10EKBD),
652: right: less massive disks (1EKBD).
653: The results for the latter have been obtained with the aid of
654: Eq.~(\ref{mass_scaling}).
655: From top to bottom: the SEDs of the simulated planetesimal rings at 10, 30 and $100\AU$.
656: In each panel, lines of decreasing thickness correspond
657: to the ages of 10~Myr, 100~Myr, 1~Gyr, and 10~Gyr.
658: Note that the evolution of the 1EKBD \@@ 100AU disk at the beginning is
659: very slow, so that the SEDs at 10 and 100~Myr are indistinguishable.
660: Vertical lines indicate centers of observational bands of several instruments (in $\mum$):
661: Spitzer MIPS (24, 70, 160),
662: Herschel PACS (100, 160),
663: \corrected{Sofia HAWC} (200),
664: CSO Sharc (350),
665: JCMT SCUBA/SCUBA 2 (450, 850),
666: \corrected{MPIfR IRAM} (1300).
667: A thin line from top left to bottom right is the stellar photosphere.
668: \label{fig_variation}
669: }
670: \end{center}
671: \end{figure*}
672:
673: \begin{figure*}
674: \begin{center}
675: \includegraphics[scale=0.6]
676: {f9a.eps}
677: %{/astro/sebastian/Desktop/Programs/SED/SEDs/G2-Variation/Figures/G2-Contribution-s.eps}
678: \includegraphics[scale=0.6]
679: {f9b.eps}
680: %{/astro/sebastian/Desktop/Programs/SED/SEDs/G2-Variation/Figures/G2-Contribution-r.eps}
681: \caption
682: {
683: \rev{Figure changed.}
684: Contribution of individual grain size decades (shown with different linestyles in the
685: left panel) and individual radial annuli of the disks (different linestyles, right) to the SED.
686: \corrected{As the grain blowout radius is $\approx 0.4\mum$, see Fig.~\ref{fig_beta},
687: in the left panels we split the lowest
688: size decade into blowout grains with $s \in [0.1\mum, 0.4\mum]$
689: and bound ones with $s \in [0.4\mum, 1.0\mum]$.}
690: Panels from top to bottom correspond to planetesimal rings at 10, 30 and $100\AU$.
691: The initial density of all disks is 10EKBD and their age is 1~Gyr.
692: \label{fig_contribution}
693: }
694: \end{center}
695: \end{figure*}
696:
697: These values are noticeably higher than the blackbody values
698: of $88\K$, $51\K$, and $28\K$, respectively.
699: The reason for these big deviations and for the S-shaped isotherms
700: in Fig.~\ref{fig_temp}
701: is the astronomical silicate's
702: spectroscopic properties with relatively high absorption at
703: visible wavelengths and steeply decreasing absorption coefficient
704: at \corrected{longer wavelengths} (see Fig.~\ref{fig_Qabs}).
705: The cross-section dominating astrosil grains are in a size range where the
706: absorption efficiency for visible and near-infrared wavelengths
707: (around $1\mum$) has already reached the blackbody value while
708: emission is still rather inefficient. With the enhancement of the
709: emission efficiencies relative to the ``saturated'' absorption,
710: temperatures drop drastically for somewhat larger grains.
711: The larger the distance from the star (yielding lower average temperature
712: and lower emission efficiency), the wider the size range
713: over which the temperature decreases, and the stronger the
714: temperature difference between small and large grains.
715: This explains why the S-shape of the isotherms gets more pronounced
716: from the left to the right in Fig.~\ref{fig_temp}.
717:
718: Further, we note that Mie resonances
719: can increase the absorption/emission efficiencies even beyond unity
720: for wavelengths somewhat longer than the grain size (see $1\mum$, $10\mum$,
721: and $100\mum$ curves in Fig.~\ref{fig_Qabs}). This explains
722: the temperature maximum for grains of about $0.3\mum$ radius
723: (``resonance'' with the stellar radiation maximum) and the minimum
724: with temperatures even below the blackbody values for 10 to $50\mum$
725: grain radius (``resonance'' with the blackbody emission peak).
726:
727:
728: \subsection{Spectral energy distributions}
729:
730: We start with a single, ``typical'' SED for one of the disks.
731: Such an SED for the 1EKBD \@@ 30AU \@@ 100Myr disk is shown in
732: Fig.~\ref{fig_bb} with a thick solid line.
733: It peaks at about $50\mum$, which is consistent with
734: the dust temperatures (Fig.~\ref{fig_temp}).
735: The hump at $\approx 10\mum$ is due to a classical silicate feature,
736: as discussed below.
737:
738: For comparison, we have overplotted the SEDs calculated for the same disk,
739: but under different assumptions about the absorbing and emitting properties of grains:
740: in a black-body approximation \corrected{(grey line)}
741: and for amorphous carbon \corrected{(dashed line)}.
742: Note that the difference applies only to the calculation of thermal emission.
743: In other words, the dynamical modeling was still done by assuming the radiation pressure of astrosil
744: and not of perfectly absorbing or carbon particles, but we assumed the grains to absorb
745: and emit like a blackbody or carbon when calculating the thermal emission.
746: There is a striking difference between the curves, especially the blackbody SED
747: deviates from the others dramatically.
748: The blackbody assumption leads to a strong increase of the total flux
749: as well as to a shift of the
750: maximum in the SED from 50 to $130\mum$!
751: In addition the excess drops towards longer wavelengths much slower
752: than in the case of the astronomical silicate. In fact,
753: it will never intersect the stellar photospheric flux.
754:
755: We now proceed with a set of SEDs for our grid of reference disks.
756: Some of them are shown in Fig.~\ref{fig_variation}.
757: The main features of these plots reveal no surprises.
758: The absolute level of excess emission is higher for more massive disks,
759: as well as for distant ones (which is just the consequence of the
760: assumed ``same-density'' scaling, as described in Sect. 3.5, see also
761: Fig.~\ref{fig_tau_dist}). The amount of dust emission is roughly comparable
762: with the photospheric emission for the mid-aged 1EKBD \@@ 30AU disk.
763: This is consistent with the known fact that a several Gyr-old EKB counterpart
764: would only slightly enhance the photospheric
765: emission even at the ``best'' wavelengths.
766: The position of the maximum emission ranges from $\approx 30\mum$ for the $10\AU$ disk
767: to $\approx 70\mum$ for the $100\AU$ disk. Note that blackbody calculation would
768: predict the emission to peak at longer wavelengths; beyond $100\mum$ for a $100\AU$ disk.
769:
770: Again, the hump seen in all SEDs slightly below $10\mum$ is due to a silicate feature in $Q_{abs}$;
771: furthermore, some traces of the second feature at $20\mum$ are barely visible.
772: This explanation is supported by Fig.~\ref{fig_Qabs}
773: that shows the absorption efficiency feature in this spectral range for
774: small particles. This becomes even more obvious by comparing the contribution of the
775: different grain size decades. For $0.1$ to $1\mum$ particles the hump is more pronounced
776: than for larger ones (see left panels in Fig.~\ref{fig_contribution} below), as it is
777: the case for the absorption efficiency. Further on,
778: the $10\mum$ ``excess'' becomes less visible for most distant disks (from top to bottom
779: panels in Fig.~\ref{fig_variation}), where
780: the average temperatures are lower, the maximum emission shifts to longer wavelengths,
781: and therefore the Planck curve at $\lambda \sim 10\mum$--$20\mum$ is steeper.
782:
783: The left panels in Fig.~\ref{fig_contribution} illustrate relative contributions of different-sized
784: particles to the full SEDs.
785: This is useful to get an idea which
786: instrument is sensitive to which grain sizes.
787: \corrected{The blowout grains with radii less than $0.4\mum$ make only modest contribution
788: to the flux even at $10\mum$.
789: The mid-IR fluxes are always dominated by bound grains with
790: $0.4\mum$ to $1\mum$ radii (for the $10\AU$ and $30\AU$ rings)
791: or those with $0.4\mum$ to $10\mum$ (for the $100\AU$ ring).
792: In the far-IR, particles up to $100\mum$ in size play a role.
793: The greatest effect on the sub-mm fluxes is that of $100\mum$ to $1\mm$ grains.
794: }
795:
796: The position of the different maxima \corrected{in Fig.~\ref{fig_contribution}}
797: can be understood by comparing the size decades to the dust temperature plot, Fig.~\ref{fig_temp}.
798: Particles of $0.1\mum$ to $1\mum$ are on the average a bit warmer than particles of $1$ to $10\mum$.
799: However, the size distribution shows that the second decade is dominated by particles only slightly
800: larger than $1\mum$, which are still nearly as warm as the particles in the decade below.
801: Thus, the maxima of the corresponding SED contributions are shifted only slightly.
802: It is the step to the next decade where the decrease of temperature becomes very obvious
803: by a large shift of the maximum. From that size on, the maxima stay nearly at the same
804: position (in fact the maxima are shifted again to smaller wavelengths) as the
805: temperature changes only marginally.
806:
807: Similar to the contribution of the different size decades in the left panel,
808: the right panels in Fig.~\ref{fig_contribution} demonstrate
809: the contribution of the different radial parts of the disk to the total SED.
810: As expected, most of the flux comes from the medium distances
811: as this is the location of the birth ring. The second largest contribution
812: is made by the outer part of the ring.
813:
814:
815: %===========================================================================
816: \section{Application to selected debris disks}
817: %===========================================================================
818:
819: \rev{This section was further split into several additional subsections}
820:
821: \subsection{Measured fluxes}
822:
823: To test the plausibility of our models, we have selected several nearby
824: sun-like stars known to possess debris dust. We used published datasets
825: to search for stars with (i) spectral classes most likely G2V (or very close),
826: and (ii) unambiguous excesses probed in a wide range of wavelengths
827: from near-IR to far-IR or sub-mm.
828: The resulting five stars and their properties are
829: listed in Table~\ref{tab_stars}, a summary of observational data on them
830: is given in Table~\ref{tab_obs},
831: and the disk properties as derived in original papers are collected in Table~\ref{tab_props_previous}.
832: The data include those from various surveys
833: with IRAS, ISO, Spitzer, Keck II, and JCMT
834: (Table~\ref{tab_obs}).
835: The estimated ages of the systems
836: range from 30 to 400~Myr
837: (Table~\ref{tab_stars})
838: and the fractional luminosities from $\sim 10^{-5}$ to $\sim 10^{-3}$
839: (Table~\ref{tab_props_previous}).
840: \corrected{The collected data points for our sample stars (photosphere $+$ dust)
841: are plotted in Fig.~\ref{fig_obs}.}
842:
843: \begin{deluxetable}{l c c c c}
844: \tablecaption{Stellar parameters \label{tab_stars}}
845: \tablewidth{0pt}
846: \tablehead{
847: \colhead{Star} &
848: \colhead{$T_{\mathrm{eff}}$ [K]}&
849: \colhead{$\log{L_*/L_\odot}$} &
850: \colhead{D [pc]} &
851: \colhead{age [Myr]}
852: }
853: \startdata
854: HD 377 & 5852 $^{a)}$ & 0.09 $^{a)}$ & 40 $^{a)}$ & 32 $^{a)}$ \\
855: HD 70573 & 5841 $^{a)}$ & -0.23 $^{a)}$ & 46 $^{a)}$ & 100 $^{a)}$ \\
856: HD 72905$^1$ & 5831 $^{a)}$ & -0.04 $^{a)}$ & 13.85 $^{d)}$ & 420 $^{d)}$ \\
857: HD 107146 & 5859 $^{a)}$ & 0.04 $^{a)}$ & 29 $^{a)}$ & $100^{+100}_{-20}$ $^{c)}$ \\
858: HD 141943 & 5805 $^{a)}$ & 0.43 $^{a)}$ & 67 $^{a)}$ & 32 $^{a)}$ \\
859: \enddata
860: \tablerefs{
861: $^{a)}$ \citet{hillenbrand-et-al-2008};
862: $^{b)}$ \citet{rhee-et-al-2007};
863: $^{c)}$ \citet{moor-et-al-2006}, \citet{trilling-et-al-2008}
864: }
865: \tablecomments{$^1$A G1.5 star.}
866: \end{deluxetable}
867:
868: \begin{deluxetable*}{l l l l}
869: \tablecaption{Observational data for the five G2 stars and their disks \label{tab_obs}}
870: \tablewidth{0pt}
871: \tablehead{
872: \colhead{Star} &
873: \colhead{Instrument, $\lambda$ ($\mum$)} &
874: \colhead{Reference} &
875: \colhead{Notes}
876: }
877: \startdata
878: HD 377 & {\it IRAC} 3.6/4.5/8.0 & \citet{hillenbrand-et-al-2008} & \\
879: & {\it IRAS} 13/33 & \citet{hillenbrand-et-al-2008} & \\
880: & {\it IRAS} 60 & \citet{moor-et-al-2006} & \\
881: & {\it MIPS} 24/70/160 & \citet{hillenbrand-et-al-2008} & \\ \hline
882: HD 70573 & {\it IRAC} 3.6/4.5/8.0 & \citet{hillenbrand-et-al-2008} & A planet host star \\
883: & {\it IRS} 13/33 & \citet{hillenbrand-et-al-2008} & \citep{setiawan-et-al-2007} \\
884: & {\it MIPS} 24/70/160 & \citet{hillenbrand-et-al-2008} & \\ \hline
885: HD 72905 & {\it IRAC} 3.6/4.5/8.0 & \citet{hillenbrand-et-al-2008} & \\
886: & {\it IRS} 13/33 & \citet{beichman-et-al-2006} & \\
887: & {\it IRAS} 12/25 & \citet{spangler-et-al-2001} & \\
888: & {\it ISOPHOT} 60/90 & \citet{spangler-et-al-2001} & \\
889: & {\it MIPS} 24 & \citet{bryden-et-al-2006} & \\
890: & {\it MIPS} 70 & \citet{hillenbrand-et-al-2008} & \\ \hline
891: HD 107146 & {\it IRAC} 3.6/4.5/8.0 & \citet{hillenbrand-et-al-2008} & Resolved in V and I\\
892: & {\it LWS} 11.7/17.8 & \citet{williams-et-al-2004} & bands \citep{ardila-et-al-2004},\\
893: & {\it IRS} 13/33 & \citet{hillenbrand-et-al-2008} & at 350 and $450\mum$ \\
894: & {\it IRAS} 60/100 & \citet{moor-et-al-2006} & \citep{williams-et-al-2004},\\
895: & {\it MIPS} 24/70 & \citet{hillenbrand-et-al-2008} & and at $3\mm$\\
896: & {\it SCUBA} 450/850 & \citet{williams-et-al-2004} & \citep{carpenter-et-al-2005}\\ \hline
897: HD 141943 & {\it IRAC} 3.6/4.5/8.0 & \citet{hillenbrand-et-al-2008} & \\
898: & {\it IRS} 13/33 & \citet{hillenbrand-et-al-2008} & \\
899: & {\it MIPS} 24/70 & \citet{hillenbrand-et-al-2008} & \\ \hline
900: \enddata
901: \end{deluxetable*}
902:
903: \begin{deluxetable*}{l c c c c}
904: \tablecaption{Previously derived disk properties \label{tab_props_previous}}
905: \tablewidth{0pt}
906: \tablehead{
907: \colhead{Star} &
908: \colhead{$T_{\mathrm{dust}}$ [K]} &
909: \colhead{$R_{\mathrm{dust}}$ [AU]} &
910: \colhead{$M_{\mathrm{dust}}$ [$M_\oplus$]} &
911: \colhead{$L_{\mathrm{dust}}/L_*$}
912: }
913: \startdata
914: HD 377 & 58 $^{a),1}$ & 23 $^{a),5}$ & $3.98\times 10^{-4}$ $^{a),8}$ & $3.98\times 10^{-4}$ $^{a),11}$ \\
915: & & & & $(4.0\pm0.3)\times 10^{-4}$ $^{f),12}$ \\ \hline
916: HD 70573 & 41 $^{a),1}$ & 35 $^{a),5}$ & $2.0\times 10^{-5}$ $^{a),8}$ & $1.0\times 10^{-4}$ $^{a),11}$ \\ \hline
917: HD 72905 & 103 $^{a),1}$ & 7 $^{a),5}$ & $1.58\times 10^{-6}$ $^{a),8}$ & $2.0\times 10^{-5}$ $^{a),11}$ \\
918: & $63 - 67$ $^{b),3}$ & $12.2 - 15.9$ $^{b),3}$ & $3.3\times 10^{-6}$ $^{b),3}$ & $2.9\times 10^{-5}$ $^{b),13}$ \\
919: & 123 $^{g),2}$ & 6.2 $^{g),5}$ & & $(0.6-1.5)\times 10^{-5}$ $^{g),14}$ \\
920: & & & & $1.6\times 10^{-5}$ $^{e),15}$ \\
921: & & & & $2.8\times 10^{-4}$ $^{g),16}$ \\ \hline
922: HD 107146 & 52 $^{a),1}$ & 30 $^{a),5}$ & $1.26\times 10^{-3}$ $^{a),8}$ & $4.94\times 10^{-4}$ $^{a),11}$ \\
923: & & $13.6 - >200$ $^{a),6}$ & & \\
924: & & & $3.2\times 10^{-7}$ $^{c),9}$ & ($9.2\pm0.9)\times 10^{-4}$ $^{f),12}$ \\
925: & 55 $^{d),2}$ & 29 $^{d),5}$ & $8.99\times 10^{-2}$ $^{d),10}$ & $9.5\times 10^{-4}$ $^{d),12}$ \\
926: & 51 $^{h),4}$ & $>31 - 150$ $^{h),7}$ & $0.1$ $^{h),4}$ & $1.2\times 10^{-3}$ $^{h)}$ \\ \hline
927: HD 141943 & 85 $^{a),1}$ & 18 $^{a),5}$ & $7.94\times 10^{-5}$ $^{a),8}$ & $1.58\times 10^{-4}$ $^{a),11}$ \\
928: & & $8.6 - 40$ $^{a),6}$ & & \\ \hline
929: \enddata
930: \tablerefs{
931: a) \citet{hillenbrand-et-al-2008},
932: b) \citet{beichman-et-al-2006},
933: c) \citet{carpenter-et-al-2005},
934: d) \citet{rhee-et-al-2007},
935: e) \citet{bryden-et-al-2006},
936: f) \citet{moor-et-al-2006},
937: g) \citet{spangler-et-al-2001},
938: h) \citet{williams-et-al-2004}
939: }
940: \tablecomments{\scriptsize
941: $^{1}$ Color temperature ($33 - 70 \mu\mbox{m}$) from blackbody SED fitting.
942: $^{2}$ From SED fitting using a single temperature blackbody.
943: $^{3}$ From SED fitting using $10\mu\mathrm{m}$ silicate grains with a temperature
944: profile following a power law (favored model in \citet{beichman-et-al-2006}).
945: $^{4}$ From single temperature SED fitting using a modified blackbody and a mass absorption
946: coefficient $\kappa_{850} = 1.7~\mathrm{cm}^2/\mathrm{g}$.
947: $^{5}$ Derived from $T_{\mathrm{dust}}$ assuming blackbody (lower limit).
948: $^{6}$ Extended ring derived from blackbody SED fitting assuming a constant surface density.
949: $^{7}$ Inner border derived from SED fitting, outer border taken from resolved image.
950: $^{8}$ Derived from fractional luminosity for an average grain size of $<a> = 10\mu\mathrm{m}$
951: and a density of $\rho = 2.5\mathrm{g}/\mathrm{cm}^3$.
952: $^{9}$ Derived for $T_{\mathrm{dust}} = 40~\mathrm{K}$ using a frequency dependent mass
953: absorption coefficient.
954: $^{10}$ Derived from submillimeter observations using a dust opacity of $1.7~\mathrm{cm}^2/\mathrm{g}$ at $850~\mum$.
955: $^{11}$ Derived from $T_{\mathrm{dust}}$ and $R_{\mathrm{dust}}$ using Stefan-Boltzmann relation.
956: $^{12}$ $L_{\mathrm{dust}}/L_* = L_{\mathrm{IR}}/L_{\mathrm{bol}}$.
957: $^{13}$ $L_{\mathrm{dust}}$ obtained by integrating {\it IRS} spectrum ($10 - 34~\mu\mathrm{m}$) after
958: extrapolation to $70~\mu\mathrm{m}$.
959: $^{14}$ $L_{\mathrm{dust}}$ is derived from the SED fitting and $L_*$ is obtained by integrating
960: the corresponding Kurucz model.
961: $^{15}$ Minimum value, derived from the $70~\mu\mathrm{m}$ measurement.
962: $^{16}$ $L_*$ is the stellar bolometric luminosity and $L_{\mathrm{dust}}$ is the sum of the
963: luminosities in each ({\it IRAS}) wavelength band with a correction (for longer wavelengths).
964: }
965: \end{deluxetable*}
966:
967: \begin{figure*}[b!]
968: \begin{center}
969: \includegraphics[scale=0.6]
970: {f10.eps}
971: %{/astro/sebastian/Desktop/Programs/SED/SEDs/G2-Variation/Figures/G2-Observations.eps}
972: \caption
973: {
974: \rev{Figure changed.}
975: Observational data for five selected G2V stars.
976: Note that all fluxes have been scaled to the same standard distance of 10pc.
977: \corrected{Symbols} in the left-hand, \corrected{grey-shaded} part of each panel ($\lambda < 10\mum$)
978: are IRAC observations.
979: They are used to find an appropriate Hauschildt model to the photosphere
980: (thin solid line), assuming that no excess is already present in the near infrared.
981: \rev{Text moved to the caption of the next figure.}
982: Vertical error bars are $1\sigma$ observational uncertainties, taken from the source papers.
983: Horizontal bars indicate the band width of the
984: respective detector.
985: \label{fig_obs}
986: }
987: \end{center}
988: \end{figure*}
989:
990:
991: \subsection{Observed excesses}
992:
993: \corrected{Symbols in Fig.~\ref{fig_comp} represent}
994: the observed excess emission for our sample stars.
995: In the cases where the photospheric subtraction was done in the source papers, we just used the
996: published data points. In the cases where only the total measured flux (star + dust) was given,
997: we proceeded as follows.
998: Three IRAC points ($3.6$, $4.5$, and $8.0\mum$) were fitted by
999: an appropriate NextGen model \citep{hauschildt-et-al-1999}, and the resulting photospheric
1000: spectrum was subtracted from the fluxes measured at longer wavelengths.
1001: As far as the data quality is concerned,
1002: the best case is clearly HD~107146, where
1003: the data points cover a broad range between $10\mum$ and $1\mm$.
1004: In other cases, the longest wavelengths probed lay at $70$--$160\mum$.
1005: As a result, it \corrected{is} sometimes unclear where exactly the excess peaks.
1006: This is exemplified by HD~70573 where the $160\mum$ point has a huge error bar.
1007:
1008: \begin{figure*}[b!]
1009: \begin{center}
1010: \includegraphics[scale=0.6]
1011: {f11.eps}
1012: %{/astro/sebastian/Desktop/Programs/SED/SEDs/G2-Variation/Figures/G2_SED-Comp.eps}
1013: \caption
1014: {
1015: \corrected{Observed (symbols) and modeled (lines) excess emission,
1016: scaled to the distance of 10pc.
1017: The wavelength range matches the unshaded part of Fig.~\ref{fig_obs}.
1018: Here, in contrast to Fig.~\ref{fig_obs}, symbols represent the {\em excess} emission.
1019: Squares mark the cases where the scaled NextGen model shown on that figure was used to
1020: subtract the photosphere.
1021: Circles indicate that for these observations the stellar emission was subtracted
1022: using photospheric fluxes as given in the respective papers.
1023: }
1024: Dashed lines: two ``underlying'' SEDs of reference disks (unscaled, i.e. with 10EKBD),
1025: one for ``cold'' excess and one for ``warm'' excess (except for HD~377 where only cold
1026: component is observed).
1027: Solid line: a linear combination of two scaled reference SEDs that provides
1028: a reasonable fit to the data points (except for HD~377 where a single scaled reference
1029: SED is sufficient).
1030: \label{fig_comp}
1031: }
1032: \end{center}
1033: \end{figure*}
1034:
1035: Yet before any comparison with the modeled SEDs,
1036: the resulting points in \corrected{Fig.~\ref{fig_comp}} allow several quick conclusions.
1037: Notwithstanding the paucity of long-wavelength data just discussed,
1038: in all five systems the excess seems to peak at or slightly beyond $100\mum$,
1039: suggesting a ``cold EKB'' as a source of dust. Additionally, in all systems
1040: except for HD~377, a warm emission at $\lambda <20\mum$ \corrected{seems to be} present,
1041: implying a closer-in ``asteroid belt''.
1042:
1043: \subsection{Comparison of measured and modeled SEDs}
1044:
1045: We now proceed with a comparison between the observed dust emission
1046: and the modeled emission.
1047: We stress that our goal here is {\em not} to provide
1048: the {\em best} fit to the observations possible with our approach, but rather to demonstrate
1049: that a set of reference disks modeled in the previous sections can be used to make
1050: rough preliminary conclusions about the planetesimal families.
1051:
1052: To make such a comparison, we employ the following procedure:
1053:
1054: 1. For each star, we first look whether only cold or cold + warm excess emission
1055: is present. In the former case (HD~377), we fit the data points with a single
1056: ``cold'' reference disk. In the latter case (all other systems), we invoke
1057: a two-component model:
1058: a close-in $3\AU$ disk and an appropriate ``cold'' disk.
1059:
1060: 2. The location of the ``cold'' planetesimal belt is chosen according
1061: to the peak wavelength of the measured excess:
1062: $100\AU$ (HD~72905 and HD~141943) or
1063: $200\AU$ (HD~377, HD~70573, and HD~107146).
1064:
1065: 3. We then scale each of the two reference SEDs, ``warm'' and ``cold''
1066: (or only one for HD~377) vertically to come to the observed absolute flux.
1067: %a reasonable fit to all data points.
1068: Physically, it necessitates a change in the initial disk mass.
1069: However, it is {\em not} sufficient to
1070: change the initial disk mass by the ratio of the observed flux and the flux from
1071: a reference disk.
1072: The reason is that a change in the initial mass also alters the rate
1073: of the collisional evolution, whereas we need the ``right'' flux at a fixed
1074: time instant, namely the actual age of the system (Tab.~\ref{tab_stars}).
1075: Therefore, to find the mass modification factor we apply
1076: scaling rules, as explained in Appendix~\ref{app_scaling}.
1077: Specifically, we solve Eq.~({\ref{o-c}}).
1078: In the systems that reveal both warm and cold emission, this is done separately
1079: for the inner and outer disk.
1080:
1081: The results presented in Fig.~\ref{fig_comp} \corrected{with lines}
1082: show that the modeled SEDs can, generally, reproduce
1083: the data points within their error bars.
1084: Again, the judgment should take into account the fact that
1085: we are just using one or two pre-generated SEDs for a rather coarse grid of
1086: reference disks.
1087: Much better fits would certainly be possible if we allowed a more exact positioning
1088: of parent belts and let additional model parameters vary.
1089: Dust opacities, initial distributions of planetesimals' \corrected{sizes and} orbital elements,
1090: \corrected{as well as}
1091: their mechanical properties that were fixed in modeling of the collisional outcomes
1092: would all be at our disposal for this purpose. Further, more than two-component
1093: planetesimal belts could be astrophysically relevant as well,
1094: as is the case \corrected{in} our solar system
1095: (asteroid belt, different cometary families, various populations in the EKB).
1096:
1097: We now come to the interpretation of the fitting results, trying to recover
1098: the properties of dust-producing planetesimal belts.
1099: Table~\ref{tab_props} lists them for all systems.
1100: The most important information is the deduced mass and location of the belts.
1101:
1102: \begin{deluxetable*}{l l c c c c}
1103: \tablecaption{Disk properties derived in this study \label{tab_props}}
1104: \tablewidth{0pt}
1105: \tablehead{
1106: \colhead{Star} &
1107: \colhead{Component} &
1108: \colhead{$M_{\mathrm{disk}}$ [$M_\oplus$] $^{1)}$} &
1109: \colhead{$R_{\mathrm{belt}}$ [AU] $^{2)}$} &
1110: \colhead{$M_{\mathrm{dust}}$ [$M_\oplus$] $^{3)}$} &
1111: \colhead{$T_{\mathrm{dust}}$ [K] $^{4)}$}
1112: }
1113: \startdata
1114: HD \ph\ph\ph 377 & Outer & $(32)$ \ph $32$ & $200$ & $3.1\times 10^{-2}$ & $40$ \\
1115: \hline
1116: HD \ph 70573 & Inner & $(0.0063)$ \ph $0.0046$ & $3$ & $1.4\times 10^{-7}$ & $200$ \\
1117: & Outer & $(2.6)$ \ph $2.5$ & $200$ & $2.0\times 10^{-3}$ & $40$ \\
1118: \hline
1119: HD \ph 72905 & Inner & $(0.054)$ \ph $0.019$ & $3$ & $3.4\times 10^{-8}$ & $200$ \\
1120: & Outer & $(0.23)$ \ph $0.23$ & $100$ & $2.1\times 10^{-4}$ & $50$ \\
1121: \hline
1122: HD 107146 & Inner & $(0.039)$ \ph $0.023$ & $3$ & $4.9\times 10^{-7}$ & $200$ \\
1123: & Outer & $(47)$ \ph $47$ & $200$ & $4.8\times 10^{-2}$ & $40$ \\
1124: \hline
1125: HD 141943 & Inner & $(0.039)$ \ph $0.027$ & $3$ & $8.0\times 10^{-7}$ & $200$ \\
1126: & Outer & $(6.1)$ \ph $6.1$ & $100$ & $5.5\times 10^{-3}$ & $50$ \\
1127: \enddata
1128: \tablecomments{
1129: $^{1)}$ {\em Initial} mass (in parentheses) and the current mass of the whole planetesimal disk
1130: (bodies up to $100\km$ in radius).\\
1131: $^{2)}$ Location of the parent planetesimal belt.\\
1132: $^{3)}$ Current mass of ``visible'' dust (grains up to $1\mm$ in radius).\\
1133: $^{4)}$ Temperature of cross-section dominating astrosil grains at the location of the parent planetesimal belt,
1134: see explanation at Fig.~\ref{fig_temp}.
1135: }
1136: \end{deluxetable*}
1137:
1138: \subsection{Results for hot dust}
1139:
1140: As far as the hot dust components in four out of five systems are concerned,
1141: our results show that these can be explained by ``massive asteroid belts''
1142: with roughly the lunar mass in bodies up to $\sim 100\km$ in size, located
1143: at $3\AU$, with a width of $\sim 1\AU$.
1144: However, the quoted distance of inner components~--- $3\AU$~--- is only due to
1145: the fact that this is the smallest disk in our grid.
1146: This distance can only be considered as an upper limit: the SEDs seem perfectly
1147: compatible with disks as far in as $0.3\AU$, as suggested for the case of
1148: HD~72905 \citep{wyatt-et-al-2007}.
1149:
1150: What is more, even the very fact that hot excess is real can sometimes be questioned,
1151: since it can be mimicked by photospheric emission slightly larger than the assumed values.
1152: Indeed, the excess for HD~70573 and HD~72905 at wavelengths around and
1153: below $25\mum$ does not exceed 10\%,
1154: which is comparable with the average calibration uncertainty
1155: and therefore has to be considered marginal \citep{bryden-et-al-2006,hillenbrand-et-al-2008}.
1156: Only in the case of HD~72905, the Spitzer/IRS detection of the $10\mum$ emission from hot silicates
1157: provides an independent confirmation that the hot excess is real \citep{beichman-et-al-2006}.
1158: However, the HD~72905 plot in Fig.~\ref{fig_comp} makes it
1159: obvious that some problems occurred in terms of the photosphere fitting.
1160: All data points that we obtained by subtracting the IRAC photospheric
1161: fluxes (squares) systematically lie above the data points where a photosphere
1162: from the literature was subtracted (circles).
1163: The origin of the difference is unclear; on any account, the problem cannot
1164: be mitigated by the assumption that an excess is already present at IRAC
1165: wavelengths, since this would shift the squares further upwards.
1166: Considering the circles to be more trustworthy, the shape of the SED to
1167: fit changes. Then a closer-in disk at $\sim 0.3\AU$ could better reproduce
1168: the fluxes in the near and mid infrared, while
1169: the outer ring would have to be shifted to a distance somewhat larger than
1170: $100\AU$ in order not to surpass the measured flux at $33~\mum$.
1171: A problem would arise with the inner disk: at $\sim 0.3\AU$, the collisional
1172: evolution is so rapid that an unrealistically large initial belt mass would
1173: be necessary.
1174: Similar arguments have led \citet{wyatt-et-al-2007} to a conclusion that
1175: HD~72905 must be a system at a transient phase rather than a system
1176: collisionally evolving in a steady state.
1177:
1178: Still, treating the derived sizes and masses of the inner disks as upper limits
1179: yields physical implications.
1180: Because the collisional evolution close to the star is rapid,
1181: such belts must have lost up to two-thirds of their initial mass
1182: before they have reached their present age
1183: (cf. initial and current mass in Table~\ref{tab_props}).
1184: In the case of HD~70573, the known giant planet with $a=1.76\AU$ and $e=0.4$
1185: \citep{setiawan-et-al-2007}
1186: does not seem to exclude the existence of a dynamically stable planetesimal
1187: belt \corrected{either inside $\sim 1\AU$
1188: or outside $\sim 3\AU$.}
1189:
1190: \subsection{Results for cold dust}
1191:
1192: The estimated parameters of the outer components of the disks suggest
1193: ``massive and large Kuiper belts''.
1194: The radii of the outer rings are larger
1195: than the radii derived in previous studies
1196: (cf. Table~\ref{tab_props_previous} and Table~\ref{tab_props}).
1197: This traces back to our using astrosilicate instead of blackbody when calculating
1198: the dust emission, so that the same dust temperatures are attained
1199: at larger distances (see Fig.~\ref{fig_bb}).
1200:
1201: Since one disk in our sample, that of HD 107146, has been resolved,
1202: it is natural to compare our derived disk radius with the one obtained from the images.
1203: \citet{williams-et-al-2004} report an outer border of the system of $150\AU$ based on submillimeter
1204: images. In contrast, \citet{ardila-et-al-2004} detected an $85\AU$-wide ring peaking in
1205: density at about $130\AU$. This is comparable to, although somewhat smaller than,
1206: our $200\AU$ radius. However, moving the outer ring to smaller
1207: distances would increase the fluxes in the mid infrared where the SED already surpasses the
1208: observations and the other way round in the sub-mm region.
1209: The resulting deficiency of sub-mm fluxes, though,
1210: could be due to roughness of Mie calculations.
1211: As pointed out by \citet{stognienko-et-al-1995},
1212: an assumption of homogeneous particles typically leads to
1213: underestimation of the amount of thermal radiation in the sub-mm region.
1214:
1215: Large belt radii imply large masses.
1216: Dust masses derived here are by two orders of magnitude larger
1217: than previous estimates
1218: (cf. Table~\ref{tab_props} and Table~\ref{tab_props_previous}).
1219: The total masses of the belts we derive range
1220: from several to several tens earth masses, to be compared
1221: with $\sim 0.1 M_\oplus$ in the present-day EKB
1222: (although there is no unanimity on that point~---
1223: cf. \citeauthor{stern-colwell-1997} \citeyear{stern-colwell-1997}).
1224: Note that, as the collisional evolution at
1225: $100$--$200\AU$ is quite slow, whereas the oldest system in our sample is
1226: only 420~Myr old, the difference between the initial disk mass and the current
1227: disk mass is negligible.
1228: Assuming several times the minimum mass solar nebula with a standard
1229: surface density of solids $\Sigma \sim 50\g\cm^2 (r/1\AU)^{-3/2}$ \citep[e.g.][]{hayashi-et-al-1985},
1230: the mass of solids in the EKB region would be a few tens $M_\oplus$;
1231: and current models \citep[e.g.][]{kenyon-luu-1999b} successfully accumulate 100\km-sized
1232: EKB objects in tens of Myr. However, it is questionable whether
1233: the assumed radial surface density profile could extend much farther out
1234: from the star.
1235: As a result, it is difficult to say, whether a progenitor disk could contain enough
1236: solids as far as at $200\AU$ from the star to form a belt of $30$ to $50 M_\oplus$.
1237:
1238: However, such questions may be somewhat premature.
1239: On the observational side, more data are needed,
1240: especially at longer wavelengths; for instance,
1241: the anticipated Herschel data
1242: (PACS at $100/160\mum$ and SPIRE at $250$ to $500\mum$)
1243: would help a lot.
1244: On the modeling side, a more systematic study is needed
1245: to clarify, how strongly various assumptions of the current model
1246: \corrected{(especially the collisional outcome prescription and
1247: the material choices)} may affect the calculated size distributions
1248: of dust, the dust grain temperatures, and the amount of their thermal emission.
1249:
1250: At this point, we can only state that in the five systems
1251: analyzed (with a possible exception of HD 72905)
1252: and with the caveat that available data are quite scarce,
1253: the observations are not incompatible with a standard steady-state
1254: scenario of collisional evolution and dust production.
1255: Of course, other possibilities, such as major collisional breakups
1256: \corrected{\citep{kenyon-bromley-2005,grigorieva-et-al-2006}}
1257: or events
1258: similar to the Late Heavy Bombardment (as suggested, for instance for HD 72905,
1259: \citeauthor{wyatt-et-al-2007} \citeyear{wyatt-et-al-2007})
1260: cannot be ruled out for the inner disks.
1261:
1262: %===========================================================================
1263: \section{Summary}
1264: %===========================================================================
1265:
1266: Debris disks around main-sequence stars may serve as
1267: tracers of planetesimal populations that have accumulated at earlier,
1268: protoplanetary and transitional, phases of systems' evolution, and have not been
1269: used up to form planets.
1270: However, observations of debris disks are only sensitive to the lowest end of the size distribution.
1271: Using dynamical and collisional models of debris disks is the only way
1272: to ``climb up'' the ladder of the collisional cascade, past the ubiquitous $\mu$m-sized
1273: grains towards parent bodies and towards the main mass reservoir of the disks.
1274:
1275: The main idea of this paper has been to take a grid of planetesimal families
1276: (with different initial masses, distances from a central star etc.),
1277: to collisionally ``generate'' debris disks from these families
1278: and evolve them with the aid of an elaborated collisional code,
1279: and finally, to calculate SEDs for these disks.
1280: A comparison/fit of the observed SEDs with the pre-generated
1281: SEDs is meant to allow quick conclusions about the properties of
1282: the planetesimal belt(s) that maintain one or another observed disk.
1283:
1284: Our specific results are as follows:
1285:
1286: 1. We have produced five reference disks around a G2V star from planetesimal belts at 3, 10, 30,
1287: 100, and $200\AU$ with 10 times the EKB mass density and evolved them for
1288: 10~Gyr. With an appropriate scaling rule (Eq.~\ref{scaling1}),
1289: we can translate these results to an arbitrary initial disk mass
1290: and any age between 10~Myr and 10~Gyr. Thus, effectively we have a
1291: three-parametric set of reference disks (initial mass, location of
1292: planetesimal belt, age). For all the disks, we have generated SEDs,
1293: assuming astrosilicate (with tests made also for blackbody and amorphous carbon).
1294:
1295: 2. We have selected five G2V stars with good data (IRAS, ISO/ISOPHOT,
1296: Spitzer/IRAC, /IRS, /MIPS, Keck II/LWS, and JCMT/SCUBA) and tested our grid against these
1297: data. For all five systems, we have reproduced the data points within
1298: the error bars with a linear combination of two disks from the grid
1299: (an ``asteroid belt'' at $3\AU$ and an outer ``Kuiper belt'').
1300: This automatically gives us the desired estimates of planetesimals
1301: (location, total mass etc.).
1302:
1303: 3. A comparison of the observational data on the five stars with
1304: the grid of models leads us to a conclusion that the cold emission
1305: (with a maximum at the far-IR) is compatible with ``large Kuiper belts'',
1306: with masses in the range 3--50 earth masses and radii of $100$--$200\AU$.
1307: These large sizes trace back to the facts that the collisional model
1308: predicts the observed emission to stem from micron-sized dust grains,
1309: whose temperatures are well in excess of a blackbody temperature
1310: at a given distance from the star
1311: \citep[as discussed, e.g., in][]{hillenbrand-et-al-2008}.
1312: This conclusion is rather robust against variation in parameters
1313: of collisional and thermal emission models, and is roughly consistent
1314: with disk radii revealed in scattered light images (e.g. HD~107146).
1315: Still, quantitative conclusions about the
1316: mass and location of the planetesimal belts would significantly
1317: depend on (i) the adopted model of collision outcomes
1318: \corrected{(which, in turn, depend on the dynamical excitation of
1319: the belts, i.e. on orbital eccentricities and inclinations of planetesimals)}
1320: and (ii) the assumed grains' absorption and emission efficiencies.
1321: For example, a less efficient cratering
1322: (retaining more grains with radii $\sim10\mum$ in the disk)
1323: and/or more ``transparent'' materials
1324: (making dust grains of the same sizes at the same locations colder)
1325: would result in ``shifting'' the parent belts closer to the star.
1326:
1327: In future, we plan to extend this study in two directions.
1328: First, we will investigate more systematically the influence of
1329: the dust composition by trying relevant materials with available optical data
1330: rather than astrosilicate; this should be done consistently in the
1331: dynamical/collisional and thermal emission models.
1332: Second, it is planned to extend this study to stars
1333: with a range of spectral classes. This will result in a catalog of disk colors
1334: that should be helpful for interpretation of data expected to come, most notably
1335: from the Herschel Space Observatory.
1336:
1337: %===========================================================================
1338: %\newpage
1339: % Acknowledgments
1340: %===========================================================================
1341:
1342: \acknowledgements
1343: This work has been particularly motivated by the Herschel Open Time Key Program
1344: ``DUNES'' (DUst disks around NEarby Stars, PI: C.Eiroa) and we wish to thank
1345: many colleagues involved in DUNES (in particular, Jean-Charles Augereau,
1346: Jens Rodmann, and Philippe Th\'ebault) for encouragement and numerous discussions.
1347: \corrected{A speedy and constructive review of an anonymous reviewer helped to improve the paper.}
1348: This research has been funded by the Deutsche Forschungsgemeinschaft (DFG),
1349: projects Kr 2164/5-1 and Mu 1164/6-1, \corrected{by} the Deutscher Akademischer Austauschdienst
1350: (DAAD), project D/0707543,
1351: \corrected{and by the International Space Science Institute (Bern)}.
1352:
1353: %===========================================================================
1354: %\clearpage
1355: % Appendix
1356: %===========================================================================
1357:
1358: \appendix
1359:
1360: \rev{The Appendix has been reworked and extended.}
1361:
1362: \section{Scaling rules}
1363: \label{app_scaling}
1364:
1365: {\em 1. Dependence of evolution on initial disk mass.}
1366: %Denote by $F(M_0,r,t)$ the radiation flux, emitted at a certain wavelength
1367: Consider a disk with initial mass $M(t=0) \equiv M_0$ at a distance $r$ from the star
1368: with age $t$.
1369: Denote by $F(M_0,r,t)$ any quantity directly proportional
1370: to the amount of disk material in any size regime, from dust grains to
1371: planetesimals.
1372: In other words, $F$ may equally stand for the total disk mass,
1373: the mass of dust, its total cross section, etc.
1374: As found by \citet{loehne-et-al-2007}, there is a scaling rule:
1375: \be
1376: F(x M_0, r, t) = x F(M_0, r, x t) ,
1377: \label{scaling1}
1378: \ee
1379: valid for any factor $x > 0$. This scaling is an {\em exact} property of
1380: every disk of particles, provided these are produced, modified and lost in
1381: binary collisions and not in any other physical processes.
1382:
1383: \bigskip
1384: {\em 2. Dependence of evolution on distance.}
1385: Another scaling rule is the dependence of the evolution timescale on
1386: the distance from the star \citep{wyatt-et-al-2007,loehne-et-al-2007}.
1387: Then
1388: \be
1389: F(M_0, x r, t) \approx F(M_0, r, t^{-4.3}) .
1390: \label{scaling2}
1391: \ee
1392: Unlike Eq.~(\ref{scaling1}), this scaling is approximate.
1393:
1394: \bigskip
1395: {\em 3. Dust mass as a function of time.}
1396: Finally, the third scaling rule found in
1397: \citet{loehne-et-al-2007}
1398: is the power-law decay of the dust mass
1399: \be
1400: F(M_0, r, x t) \approx x^{-\xi} F(M_0, r, t) ,
1401: \label{scaling3}
1402: \ee
1403: where $\xi \approx 0.3 \ldots 0.4$
1404: (Fig.~\ref{fig_xi}).
1405: This scaling is also approximate and,
1406: unlike Eq.~(\ref{scaling1}) and Eq.~(\ref{scaling2}), only applies
1407: %Let now $F(M_0,r,t)$ be
1408: %to any quantity proportional to the amount of those
1409: %disk objects that have reached {\em collisional equilibrium}.
1410: %These are objects whose collisional lifetime is shorter than the current system's age,
1411: %typically smaller than $1$--$100\km$.
1412: %Now, $F$ may stand for the mass or cross section of dust or small planetesimals,
1413: %etc.~--- but not for the total disk mass.
1414: to every quantity directly proportional to the amount of {\em dust}.
1415: In this context, ``dust'' refers to all objects in the strength rather than gravity
1416: regime, implying radii less than about 100 meters.
1417: The scaling is sufficiently accurate for disks that are much older than
1418: the collisional lifetime of these $100\m$-sized bodies.
1419: This is also seen in Fig.~\ref{fig_xi}: while for the $3\AU$ disk the power law (\ref{scaling3})
1420: sets in after $\ll 1$~Myr, the $200\AU$ disk needs $\sim 100$~Myr to reach this regime.
1421:
1422: \begin{figure}[h!]
1423: \begin{center}
1424: \includegraphics[scale=0.6]
1425: {f12.eps}
1426: %{/astro/krivov/ACE/ACE_2008/FGK_CATALOG/xi.eps}
1427: \caption
1428: {
1429: The time evolution of dust mass ($s< 1\mm$) for our five reference disks (thick lines).
1430: For comparison, power laws $t^{-\xi}$ with $\xi = 0.3$ and $\xi = 0.4$ are shown with thin dashed
1431: lines.
1432: \label{fig_xi}
1433: }
1434: \end{center}
1435: \end{figure}
1436:
1437: Note that the ``pre-steady-state'' phase of collisional evolution may actually
1438: require a more sophisticated treatment. Our runs assume initially a power-law size distribution
1439: of planetesimals, and an instantaneous start of the collisional cascade at $t=0$.
1440: In reality, an initial size distribution is set up by the accretion history of
1441: planetesimals and will surely deviate from a single power law.
1442: Moreover, at a certain phase cratering and destruction of objects may increasingly come into play
1443: simultaneously with ceasing, yet ongoing accretion; the efficiencies and timescales of these
1444: processes will be different for different size ranges and different spatial locales in the disk
1445: \citep[e.g.][]{davis-farinella-1997,kenyon-luu-1998,kenyon-luu-1999a,kenyon-luu-1999b}.
1446:
1447: The usefulness of these scaling rules can be illustrated with the
1448: following examples.
1449:
1450: \bigskip
1451: {\em Example~1.}
1452: %\subsection{Example~1}
1453: Assuming now $F$ to be the total amount of dust,
1454: from Eqs.~(\ref{scaling1})--(\ref{scaling3}) one finds
1455: \be
1456: F(xM_0, yr, t) \approx x^{1-\xi} y^{4.3\xi}F(M_0, r, t) .
1457: \ee
1458: Our choice of reference disks (different distances, but the same volume density)
1459: implies $x = y^3$.
1460: The normal optical optical depth $\tau \propto F/r^2$ scales as
1461: \be
1462: \tau(y^3 M_0, y r, t) \approx y^{1+1.3\xi} \tau (M_0, r, t) .
1463: \ee
1464: Therefore, once a steady-state is reached ($\xi \approx 0.3...0.4$),
1465: a $y$ times more distant planetesimal belt gives rise to a
1466: $y^{1+1.3\xi}$ times optically thicker disk.
1467: This explains, in particular, why in Fig.~\ref{fig_spat_dist}
1468: any $100\AU$ ring is $\approx 30$ times optically thicker
1469: than the co-eval $10\AU$ one.
1470:
1471: \bigskip
1472: {\em Example~2.}
1473: %\subsection{Example~2}
1474: Since the distance $r$ in Eqs.~(\ref{scaling1}) and (\ref{scaling3})
1475: is kept fixed, $F$ in these equations can also denote the radiation flux,
1476: emitted by a disk at a certain wavelength.
1477: Let $F_o(t)$ be the observed flux from a disk of age $t$.
1478: Imagine a model of a disk of the same age with an initial mass $M_0$
1479: predicts a flux $F(M_0,t)$ which is by a factor $A$ lower than the observed one:
1480: \be
1481: F_o(t) = A F(M_0, r, t).
1482: \ee
1483: Our goal is to find the ``right'' initial mass, i.e. a factor $B$ such that
1484: \be
1485: F_o(t) = F(B M_0, r, t).
1486: \ee
1487: With the aid of Eq.~(\ref{scaling1}), this can be rewritten as
1488: \be
1489: F_o(t) = B F(M_0, r, B t) .
1490: \label{o-c}
1491: \ee
1492: Eq.~(\ref{scaling3}) gives now
1493: \be
1494: F_o(t)
1495: \approx B F(M_0, r, t) B^{-\xi}
1496: = B ^{1-\xi} F(M_0, r, t) ,
1497: \ee
1498: whence
1499: \be
1500: B \approx A^{1/(1-\xi)} .
1501: \label{B(A)}
1502: \ee
1503: For instance, a 10 times higher flux at a certain age requires a 27--46 times larger
1504: initial disk mass if $\xi = 0.3 \ldots 0.4$.
1505:
1506: Although this rule is convenient for quick estimates, it should be used with caution.
1507: As described above, the value of $\xi$ at the beginning of collisional evolution
1508: (which lasts up to 100~Myr for the $200\AU$ belt) can be much smaller~--- close to zero or even
1509: negative~--- than the ``normal'' $\xi = 0.3 \ldots 0.4$. For this reason, we prefer to use only
1510: the first scaling rule, Eq.~(\ref{scaling1}). Therefore, instead of
1511: applying Eq.~(\ref{B(A)}), we find $B$ by solving Eq.~(\ref{o-c}) numerically with a simple iterative
1512: routine.
1513: It is this way Fig.~\ref{fig_comp} was constructed.
1514:
1515: %===========================================================================
1516: %\clearpage
1517: % References
1518: %===========================================================================
1519:
1520: %\bibliography{/astro/krivov/LIBRARY/BIB/english}
1521: %\bibliographystyle{/astro/krivov/SMALL_PROJECTS/PLANETESIMAL_BELTS/aastex52/apj}
1522: %\input ms.bbl
1523:
1524: \begin{thebibliography}{46}
1525: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
1526:
1527: \bibitem[{{Ardila} {et~al.}(2004){Ardila}, {Golimowski}, {Krist}, {Clampin},
1528: {Williams}, {Blakeslee}, {Ford}, {Hartig}, \&
1529: {Illingworth}}]{ardila-et-al-2004}
1530: {Ardila}, D.~R.,
1531: et al.
1532: %{Golimowski}, D.~A., {Krist}, J.~E., {Clampin}, M.,
1533: %{Williams}, J.~P., {Blakeslee}, J.~P., {Ford}, H.~C., {Hartig}, G.~F., \&
1534: %{Illingworth}, G.~D.
1535: 2004, \ApJ, 617, L147
1536:
1537: \bibitem[{{Artymowicz}(1997)}]{artymowicz-1997}
1538: {Artymowicz}, P. 1997, \ARevEPS, 25, 175
1539:
1540: \bibitem[{{Aumann} {et~al.}(1984){Aumann}, {Beichman}, {Gillett}, {de Jong},
1541: {Houck}, {Low}, {Neugebauer}, {Walker}, \& {Wesselius}}]{aumann-et-al-1984}
1542: {Aumann}, H.~H.,
1543: %{Beichman}, C.~A., {Gillett}, F.~C., {de Jong}, T., {Houck},
1544: %J.~R., {Low}, F.~J., {Neugebauer}, G., {Walker}, R.~G., \& {Wesselius}, P.~R.
1545: et al.
1546: 1984, \ApJ, 278, L23
1547:
1548: \bibitem[{Beichman {et~al.}(2005)Beichman, Bryden, Rieke, Stansberry, Trilling,
1549: Stapelfeldt, Werner, Engelbracht, Blaylock, Gordon, Chen, Su, \&
1550: Hines}]{beichman-et-al-2005}
1551: Beichman, C.~A.,
1552: et al.
1553: %Bryden, G., Rieke, G.~H., Stansberry, J.~A., Trilling, D.~E.,
1554: %Stapelfeldt, K.~R., Werner, M.~W., Engelbracht, C.~W., Blaylock, M., Gordon,
1555: %K.~D., Chen, C.~H., Su, K.~Y.~L., \& Hines, D.~C.
1556: 2005, \ApJ, 622, 1160
1557:
1558: \bibitem[{{Beichman} {et~al.}(2006){Beichman}, {Tanner}, {Bryden},
1559: {Stapelfeldt}, {Werner}, {Rieke}, {Trilling}, {Lawler}, \&
1560: {Gautier}}]{beichman-et-al-2006}
1561: {Beichman}, C.~A.,
1562: et al.
1563: %{Tanner}, A., {Bryden}, G., {Stapelfeldt}, K.~R., {Werner},
1564: %M.~W., {Rieke}, G.~H., {Trilling}, D.~E., {Lawler}, S., \& {Gautier}, T.~N.
1565: 2006, \ApJ, 639, 1166
1566:
1567: \bibitem[{Bohren \& Huffman(1983)}]{bohren-huffman-1983}
1568: Bohren, C.~F., \& Huffman, D.~R. 1983, Absorption and Scattering of Light by
1569: Small Particles (Wiley and Sons: New York -- Chichester -- Brisbane --
1570: Toronto -- Singapore)
1571:
1572: \bibitem[{{Bryden} {et~al.}(2006){Bryden}, {Beichman}, {Trilling}, {Rieke},
1573: {Holmes}, {Lawler}, {Stapelfeldt}, {Werner}, {Gautier}, {Blaylock}, {Gordon},
1574: {Stansberry}, \& {Su}}]{bryden-et-al-2006}
1575: {Bryden}, G.,
1576: et al.
1577: %{Beichman}, C.~A., {Trilling}, D.~E., {Rieke}, G.~H., {Holmes},
1578: %E.~K., {Lawler}, S.~M., {Stapelfeldt}, K.~R., {Werner}, M.~W., {Gautier},
1579: %T.~N., {Blaylock}, M., {Gordon}, K.~D., {Stansberry}, J.~A., \& {Su},
1580: %K.~Y.~L.
1581: 2006, \ApJ, 636, 1098
1582:
1583: \bibitem[{Burns {et~al.}(1979)Burns, Lamy, \& Soter}]{burns-et-al-1979}
1584: Burns, J.~A., Lamy, P.~L., \& Soter, S. 1979, Icarus, 40, 1
1585:
1586: \bibitem[{Carpenter {et~al.}(2005)Carpenter, Wolf, Schreyer, Launhardt, \&
1587: Henning}]{carpenter-et-al-2005}
1588: Carpenter, J.~M., Wolf, S., Schreyer, K., Launhardt, R., \& Henning, T. 2005,
1589: \AJ, 129, 1049
1590:
1591: \bibitem[{Davis \& Farinella(1997)}]{davis-farinella-1997}
1592: Davis, D.~R., \& Farinella, P. 1997, Icarus, 125, 50
1593:
1594: \corrected{
1595: \bibitem[{Dohnanyi(1969)}]{dohnanyi-1969}
1596: Dohnanyi, J.~S. 1969, \JGR, 74, 2531
1597: }
1598:
1599: \corrected{
1600: \bibitem[{Durda \& Dermott(1997)}]{durda-dermott-1997}
1601: Durda, D.~D., \& Dermott, S.~F. 1997, Icarus, 130, 140
1602: }
1603:
1604: \bibitem[{{Gladman} {et~al.}(2001){Gladman}, {Kavelaars}, {Petit},
1605: {Morbidelli}, {Holman}, \& {Loredo}}]{gladman-et-al-2001b}
1606: {Gladman}, B.,
1607: et al.
1608: %{Kavelaars}, J.~J., {Petit}, J., {Morbidelli}, A., {Holman},
1609: %M.~J., \& {Loredo}, T.
1610: 2001, \AJ, 122, 1051
1611:
1612: \corrected{
1613: \bibitem[{Grigorieva {et~al.}(2007)Grigorieva, Artymowicz, \&
1614: Th\'ebault}]{grigorieva-et-al-2006}
1615: Grigorieva, A., Artymowicz, P., \& Th\'ebault, P. 2007, \AAp, 461, 537
1616: }
1617:
1618: \bibitem[{{Hahn} \& {Malhotra}(2005)}]{hahn-malhotra-2005}
1619: {Hahn}, J.~M., \& {Malhotra}, R. 2005, \AJ, 130, 2392
1620:
1621: \bibitem[{Hauschildt {et~al.}(1999)Hauschildt, Allard, \&
1622: Baron}]{hauschildt-et-al-1999}
1623: Hauschildt, P., Allard, F., \& Baron, E. 1999, \ApJ, 512, 377
1624:
1625: \bibitem[{{Hayashi} {et~al.}(1985){Hayashi}, {Nakazawa}, \&
1626: {Nakagawa}}]{hayashi-et-al-1985}
1627: {Hayashi}, C., {Nakazawa}, K., \& {Nakagawa}, Y. 1985, in Protostars and
1628: Planets II, ed. D.~C. {Black} \& M.~S. {Matthews}, 1100--1153
1629:
1630: \bibitem[{{Hillenbrand} {et~al.}(2008){Hillenbrand}, {Carpenter}, {Kim},
1631: {Meyer}, {Backman}, {Moro-Martin}, {Hollenbach}, {Hines}, {Pascucci}, \&
1632: {Bouwman}}]{hillenbrand-et-al-2008}
1633: {Hillenbrand}, L.~A.,
1634: et al.
1635: %{Carpenter}, J.~M., {Kim}, J.~S., {Meyer}, M.~R.,
1636: %{Backman}, D.~E., {Moro-Martin}, A., {Hollenbach}, D.~J., {Hines}, D.~C.,
1637: %{Pascucci}, I., \& {Bouwman}, J.
1638: 2008, \corrected{\ApJ, 677, 630}
1639:
1640: \corrected{
1641: \bibitem[{{Kenyon} \& {Bromley}(2005)}]{kenyon-bromley-2005}
1642: {Kenyon}, S.~J., \& {Bromley}, B.~C. 2005, \AJ, 130, 269
1643: }
1644:
1645: \bibitem[{{Kenyon} \& {Luu}(1998)}]{kenyon-luu-1998}
1646: {Kenyon}, S.~J., \& {Luu}, J.~X. 1998, \AJ, 115, 2136
1647:
1648: \bibitem[{{Kenyon} \& {Luu}(1999{\natexlab{a}})}]{kenyon-luu-1999a}
1649: ---. 1999{\natexlab{a}}, \AJ, 118, 1101
1650:
1651: \bibitem[{{Kenyon} \& {Luu}(1999{\natexlab{b}})}]{kenyon-luu-1999b}
1652: ---. 1999{\natexlab{b}}, \ApJ, 526, 465
1653:
1654: \bibitem[{Krivov {et~al.}(2006)Krivov, L\"ohne, \&
1655: Srem\v{c}evi\'c}]{krivov-et-al-2006}
1656: Krivov, A.~V., L\"ohne, T., \& Srem\v{c}evi\'c, M. 2006, \AAp, 455, 509
1657:
1658: \bibitem[{Krivov {et~al.}(2000)Krivov, Mann, \& Krivova}]{krivov-et-al-2000b}
1659: Krivov, A.~V., Mann, I., \& Krivova, N.~A. 2000, \AAp, 362, 1127
1660:
1661: \bibitem[{Krivov {et~al.}(2005)Krivov, Srem\v{c}evi\'c, \&
1662: Spahn}]{krivov-et-al-2005}
1663: Krivov, A.~V., Srem\v{c}evi\'c, M., \& Spahn, F. 2005, Icarus, 174, 105
1664:
1665: \bibitem[{{Laor} \& {Draine}(1993)}]{laor-draine-1993}
1666: {Laor}, A., \& {Draine}, B.~T. 1993, \ApJ, 402, 441
1667:
1668: \bibitem[{L\"ohne {et~al.}(2008)L\"ohne, Krivov, \&
1669: Rodmann}]{loehne-et-al-2007}
1670: L\"ohne, T., Krivov, A.~V., \& Rodmann, J. 2008, \ApJ, 673, 1123
1671:
1672: \bibitem[{{Meyer} {et~al.}(2004){Meyer}, {Hillenbrand}, {Backman}, {Beckwith},
1673: {Bouwman}, {Brooke}, {Carpenter}, {Cohen}, {Gorti}, {Henning}, {Hines},
1674: {Hollenbach}, {Kim}, {Lunine}, {Malhotra}, {Mamajek}, {Metchev},
1675: {Moro-Martin}, {Morris}, {Najita}, {Padgett}, {Rodmann}, {Silverstone},
1676: {Soderblom}, {Stauffer}, {Stobie}, {Strom}, {Watson}, {Weidenschilling},
1677: {Wolf}, {Young}, {Engelbracht}, {Gordon}, {Misselt}, {Morrison}, {Muzerolle},
1678: \& {Su}}]{meyer-et-al-2004}
1679: {Meyer}, M.~R.,
1680: et al.
1681: %{Hillenbrand}, L.~A., {Backman}, D.~E., {Beckwith}, S.~V.~W.,
1682: %{Bouwman}, J., {Brooke}, T.~Y., {Carpenter}, J.~M., {Cohen}, M., {Gorti}, U.,
1683: %{Henning}, T., {Hines}, D.~C., {Hollenbach}, D., {Kim}, J.~S., {Lunine}, J.,
1684: %{Malhotra}, R., {Mamajek}, E.~E., {Metchev}, S., {Moro-Martin}, A., {Morris},
1685: %P., {Najita}, J., {Padgett}, D.~L., {Rodmann}, J., {Silverstone}, M.~D.,
1686: %{Soderblom}, D.~R., {Stauffer}, J.~R., {Stobie}, E.~B., {Strom}, S.~E.,
1687: %{Watson}, D.~M., {Weidenschilling}, S.~J., {Wolf}, S., {Young}, E.,
1688: %{Engelbracht}, C.~W., {Gordon}, K.~D., {Misselt}, K., {Morrison}, J.,
1689: %{Muzerolle}, J., \& {Su}, K.
1690: 2004, \ApJS, 154, 422
1691:
1692: \bibitem[{{Mo{\'o}r} {et~al.}(2006){Mo{\'o}r}, {{\'A}brah{\'a}m}, {Derekas},
1693: {Kiss}, {Kiss}, {Apai}, {Grady}, \& {Henning}}]{moor-et-al-2006}
1694: {Mo{\'o}r}, A.,
1695: et al.
1696: %{{\'A}brah{\'a}m}, P., {Derekas}, A., {Kiss}, C., {Kiss},
1697: %L.~L., {Apai}, D., {Grady}, C., \& {Henning}, T.
1698: 2006, \ApJ, 644, 525
1699:
1700: \bibitem[{{Najita} \& {Williams}(2005)}]{najita-williams-2005}
1701: {Najita}, J., \& {Williams}, J.~P. 2005, \ApJ, 635, 625
1702:
1703: \bibitem[{{Rhee} {et~al.}(2007){Rhee}, {Song}, {Zuckerman}, \&
1704: {McElwain}}]{rhee-et-al-2007}
1705: {Rhee}, J.~H., {Song}, I., {Zuckerman}, B., \& {McElwain}, M. 2007, \ApJ, 660,
1706: 1556
1707:
1708: \bibitem[{Rieke {et~al.}(2005)Rieke, Su, Stansberry, Trilling, Bryden,
1709: Muzerolle, White, Gorlova, Young, Beichman, Stapelfeldt, \&
1710: Hines}]{rieke-et-al-2005}
1711: Rieke, G.~H.,
1712: et al.
1713: %Su, K.~Y.~L., Stansberry, J.~A., Trilling, D., Bryden, G.,
1714: %Muzerolle, J., White, B., Gorlova, N., Young, E.~T., Beichman, C.~A.,
1715: %Stapelfeldt, K.~R., \& Hines, D.~C.
1716: 2005, \ApJ, 620, 1010
1717:
1718: \bibitem[{{Setiawan} {et~al.}(2007){Setiawan}, {Weise}, {Henning}, {Launhardt},
1719: {M{\"u}ller}, \& {Rodmann}}]{setiawan-et-al-2007}
1720: {Setiawan}, J.,
1721: et al.
1722: %{Weise}, P., {Henning}, T., {Launhardt}, R., {M{\"u}ller}, A.,
1723: %\& {Rodmann}, J.
1724: 2007, \ApJ, 660, L145
1725:
1726: \bibitem[{{Siegler} {et~al.}(2007){Siegler}, {Muzerolle}, {Young}, {Rieke},
1727: {Mamajek}, {Trilling}, {Gorlova}, \& {Su}}]{siegler-et-al-2006}
1728: {Siegler}, N.,
1729: et al.
1730: %{Muzerolle}, J., {Young}, E.~T., {Rieke}, G.~H., {Mamajek},
1731: %E.~E., {Trilling}, D.~E., {Gorlova}, N., \& {Su}, K.~Y.~L.
1732: 2007, \ApJ, 654, 580
1733:
1734: \bibitem[{{Spangler} {et~al.}(2001){Spangler}, {Sargent}, {Silverstone},
1735: {Becklin}, \& {Zuckerman}}]{spangler-et-al-2001}
1736: {Spangler}, C., {Sargent}, A.~I., {Silverstone}, M.~D., {Becklin}, E.~E., \&
1737: {Zuckerman}, B. 2001, \ApJ, 555, 932
1738:
1739: \bibitem[{{Stern} \& {Colwell}(1997)}]{stern-colwell-1997}
1740: {Stern}, S.~A., \& {Colwell}, J.~E. 1997, \ApJ, 490, 879
1741:
1742: \bibitem[{{Stognienko} {et~al.}(1995){Stognienko}, {Henning}, \&
1743: {Ossenkopf}}]{stognienko-et-al-1995}
1744: {Stognienko}, R., {Henning}, T., \& {Ossenkopf}, V. 1995, \AAp, 296, 797
1745:
1746: \bibitem[{{Su} {et~al.}(2006){Su}, {Rieke}, {Stansberry}, {Bryden},
1747: {Stapelfeldt}, {Trilling}, {Muzerolle}, {Beichman}, {Moro-Martin}, {Hines},
1748: \& {Werner}}]{su-et-al-2006}
1749: {Su}, K.~Y.~L.,
1750: et al.
1751: %{Rieke}, G.~H., {Stansberry}, J.~A., {Bryden}, G.,
1752: %{Stapelfeldt}, K.~R., {Trilling}, D.~E., {Muzerolle}, J., {Beichman}, C.~A.,
1753: %{Moro-Martin}, A., {Hines}, D.~C., \& {Werner}, M.~W.
1754: 2006, \ApJ, 653, 675
1755:
1756: \bibitem[{{Th{\' e}bault} \& {Augereau}(2007)}]{thebault-augereau-2007}
1757: {Th{\' e}bault}, P., \& {Augereau}, J.-C. 2007, \AAp, 472, 169
1758:
1759: \bibitem[{Th\'ebault {et~al.}(2003)Th\'ebault, Augereau, \&
1760: Beust}]{thebault-et-al-2003}
1761: Th\'ebault, P., Augereau, J.-C., \& Beust, H. 2003, \AAp, 408, 775
1762:
1763: \bibitem[{{Trilling} {et~al.}(2008){Trilling}, {Bryden}, {Beichman}, {Rieke},
1764: {Su}, {Stansberry}, {Blaylock}, {Stapelfeldt}, {Beeman}, \&
1765: {Haller}}]{trilling-et-al-2008}
1766: {Trilling}, D.~E.,
1767: et al.
1768: %{Bryden}, G., {Beichman}, C.~A., {Rieke}, G.~H., {Su},
1769: %K.~Y.~L., {Stansberry}, J.~A., {Blaylock}, M., {Stapelfeldt}, K.~R.,
1770: %{Beeman}, J.~W., \& {Haller}, E.~E.
1771: 2008, \ApJ, 674, 1086
1772:
1773: \bibitem[{{Trilling} {et~al.}(2007){Trilling}, {Stansberry}, {Stapelfeldt},
1774: {Rieke}, {Su}, {Gray}, {Corbally}, {Bryden}, {Chen}, {Boden}, \&
1775: {Beichman}}]{trilling-et-al-2007}
1776: {Trilling}, D.~E.,
1777: et al.
1778: %{Stansberry}, J.~A., {Stapelfeldt}, K.~R., {Rieke}, G.~H.,
1779: %{Su}, K.~Y.~L., {Gray}, R.~O., {Corbally}, C.~J., {Bryden}, G., {Chen},
1780: %C.~H., {Boden}, A., \& {Beichman}, C.~A.
1781: 2007, \ApJ, 658, 1289
1782:
1783: \bibitem[{Williams {et~al.}(2004)Williams, Najita, Liu, Bottinelli, Carpenter,
1784: Hillenbrand, Meyer, \& Soderblom}]{williams-et-al-2004}
1785: Williams, J.~P.,
1786: et al.
1787: %Najita, J., Liu, M.~C., Bottinelli, S., Carpenter, J.~M.,
1788: %Hillenbrand, L.~A., Meyer, M.~R., \& Soderblom, D.~R.
1789: 2004, \ApJ, 604, 414
1790:
1791: \bibitem[{{Wolf} \& {Hillenbrand}(2003)}]{wolf-hillenbrand-2003}
1792: {Wolf}, S., \& {Hillenbrand}, L.~A. 2003, \ApJ, 596, 603
1793:
1794: \bibitem[{Wyatt(2005)}]{wyatt-2005}
1795: Wyatt, M.~C. 2005, \AAp, 433, 1007
1796:
1797: \bibitem[{{Wyatt} {et~al.}(2007){Wyatt}, {Smith}, {Greaves}, {Beichman},
1798: {Bryden}, \& {Lisse}}]{wyatt-et-al-2007}
1799: {Wyatt}, M.~C.,
1800: et al.
1801: %{Smith}, R., {Greaves}, J.~S., {Beichman}, C.~A., {Bryden}, G.,
1802: %\& {Lisse}, C.~M.
1803: 2007, \ApJ, 658, 569
1804:
1805: \end{thebibliography}
1806:
1807:
1808: \end{document}
1809:
1810: