0706.3368/ms.tex
1: %\documentclass[12pt,article]{aastex}
2: %\documentclass[manuscript]{aastex}
3: \documentclass[12pt,preprint]{aastex}
4: %\textwidth 17cm
5: %\textheight 21.75cm
6: %\oddsidemargin -0.5cm
7: \tolerance=1000
8: %\topmargin 0cm
9: %\parskip 10pt
10: %\baselineskip= 10pt
11: 
12: \newcommand \kms{km~$\rm{s}^{-1}$}
13: \newcommand \cc{$\rm{cm}^{-3}$}
14: \newcommand \lam{$\lambda$}
15: \newcommand \Ha{H$\alpha$}
16: \newcommand \Hb{H$\beta$}
17: \newcommand \Hg{H$\gamma$}
18: \newcommand \Hd{H$\delta$}
19: \newcommand \He{H$\epsilon$}
20: \newcommand \mum{$\mu$m}
21: \newcommand \lsol{L$_{\odot}$}
22: \newcommand \msol{M$_{\odot}$}
23: \newcommand \rsol{R$_{\odot}$}
24: \newcommand \mdot{M$_{\odot}$ yr$^{-1}$}
25: \newcommand \fdens{erg s$^{-1}$ cm$^{-2}$ arcsec$^{-2}$}
26: \newcommand \flux{erg s$^{-1}$ cm$^{-2}$}
27: \newfont{\rten}{cmr10} 
28: \def \arcdeg{\hbox{$^\circ$}}
29: \def \arcmin{\hbox{$^\prime$}}
30: \def \arcsec{\hbox{$^{\prime\prime}$}}
31: 
32: \begin{document}
33: 
34: %\normalsize
35: 
36: \title{The dust, planetesimals and planets of HD 38529}
37: 
38: \author{Amaya Moro-Mart\'{\i}n\altaffilmark{1}, 
39: Renu Malhotra\altaffilmark{2}, 
40: John M. Carpenter\altaffilmark{3}, 
41: Lynne A. Hillenbrand\altaffilmark{3}, 
42: Sebastian Wolf\altaffilmark{4}, 
43: Michael R. Meyer\altaffilmark{5}, 
44: David Hollenbach\altaffilmark{6},
45: Joan Najita\altaffilmark{7} and
46: Thomas Henning\altaffilmark{4}}
47: 
48: \altaffiltext{1}{Department of Astrophysical Sciences, Peyton Hall, Ivy Lane, 
49: Princeton University, Princeton, NJ 08544, USA, amaya@astro.princeton.edu}
50: 
51: \altaffiltext{2}{Department of Planetary Sciences, University of Arizona, 1629 E. 
52: University Boulevard, Tucson, AZ 85721, USA}
53: 
54: \altaffiltext{3}{Department of Astronomy, California Institute of Technology, 
55: Pasadena, CA 91125, USA}
56: 
57: \altaffiltext{4}{Max-Planck-Institut f\"ur Astronomie, K\"onigstuhl 17, 69117 
58: Heidelberg, Germany} 
59: 
60: \altaffiltext{5}{Steward Observatory, University of Arizona, 933 North Cherry Ave., 
61: Tucson, AZ 85721, USA}
62: 
63: \altaffiltext{6}{NASA Ames, Moffet Field, CA 94035, USA}
64: 
65: \altaffiltext{7}{National Optical Astronomy Observatory, 950 North Cherry Ave.,  
66: Tucson AZ 85721, USA} 
67: 
68: \begin{abstract}
69: HD~38529 is a post-main sequence G8III/IV star (3.5 Gyr old) with a planetary system 
70: consisting of at least two planets having $\it{M}$sin$\it{i}$ of 
71: 0.8~M$_{Jup}$ and 12.2~M$_{Jup}$, semimajor axes of 
72: 0.13 AU and 3.74 AU, and eccentricities of 0.25 and 0.35, respectively.
73: $\it{Spitzer}$ observations show that HD~38529 has an excess emission above 
74: the stellar photosphere, with a signal-to-noise ratio (S/N) at 70 $\mu$m of 4.7, 
75: a small excess at 33 $\mu$m (S/N=2.6) and no excess $<$30 $\mu$m. 
76: We discuss the distribution of the potential dust-producing planetesimals
77: from the study of the dynamical perturbations of the two known planets, considering
78: in particular the effect of secular resonances. We identify three dynamically stable 
79: niches at 0.4--0.8 AU, 20--50 AU and beyond 60 AU. We model the spectral energy 
80: distribution of HD~38529 to find out which of these niches show signs 
81: of harboring dust-producing plantesimals. 
82: The secular analysis, together with the SED modeling resuls, 
83: suggest that the planetesimals responsible for most of the dust emission 
84: are likely located within 20--50 AU, a configuration that resembles that of the Jovian planets + 
85: Kuiper Belt in our Solar System. Finally, we place upper limits 
86: (8$\times$10$^{-6}$~lunar masses~of 10 $\mu$m particles) to the amount of dust 
87: that could be located in the dynamically stable region that exists between the 
88: two planets (0.25--0.75 AU).
89: \end{abstract}
90: 
91: \keywords{circumstellar matter --- Kuiper Belt
92: --- infrared: stars
93: --- planetary systems
94: --- stars: HD 38529
95: } 
96: 
97: \section{Introduction}
98:  
99: HD~38529 is a post-main sequence star (G8 III/IV) with an estimated age
100: of 3.5$\pm$1 Gyr (Valenti \& Fischer~\citeyear{vale05} and Gonzalez et al.~\citeyear{gonz01}), 
101: a distance of 42$\pm$2 pc (Perryman et al.~\citeyear{perr97}), and 
102: T$_{eff}$ = 5697 K, L$_{*}$ = 6.31~\lsol, M$_{*}$ = 1.47~\msol~
103: [Fe/H] = 0.445 (Valenti \& Fischer~\citeyear{vale05}). 
104: High precision radial velocity monitoring of HD~38529 has led to the discovery 
105: of a close-in Jupiter-mass planet (HD~38529b -- Fischer et al.~\citeyear{fisc01}) 
106: and a second more massive and more distant planet (HD~38529c -- 
107: Fischer et al.~\citeyear{fisc03}). Butler et al. (\citeyear{butl06})
108: report the current estimates of the parameters of the two planets: 
109: masses ($\it{M}$sin$\it{i}$) of 0.8~M$_{Jup}$ and 12.2~M$_{Jup}$, 
110: semimajor axes of 0.13 AU and 3.74 AU, 
111: and eccentricities of 0.25 and 0.35, for HD~38529b and  HD~38529c respectively. 
112: 
113: As part of the $\it{Spitzer}$ Legacy Program FEPS (Meyer et al.~\citeyear{meye06}), 
114: we searched for debris disks around the nine stars in the FEPS sample known from 
115: radial velocity (RV) studies to have one or more massive planets
116: (Moro-Mart\'{\i}n et al.~\citeyear{ama07}). 
117: HD~38529 was found to be the only star in that sub-sample to have an excess emission above the stellar 
118: photosphere, with a signal-to-noise ratio (S/N) at 70 $\mu$m of 4.7, a 
119: small excess at 33 $\mu$m (S/N = 2.6) 
120: and no excess at $\lambda$ $<$ 30 $\mu$m (Moro-Mart\'{\i}n et al.~\citeyear{ama07}). 
121: HD~38529 therefore joined the small group of stars known to date that have both IR 
122: excess and one or more known planetary companions. Table 1 summarizes the properties of these systems, 
123: showing a wide diversity of planetary architectures. 
124: Six of these nine sources, HD 33636, HD 50554, HD 52265, HD 82943, HD 117176 
125: and HD 128311, are similar to HD~38529 in that their
126: $\it{Spitzer}$ observations also show an excess at 70 $\mu$m~but 
127: no excess at 24 $\mu$m, 
128: implying that the bulk of the excess emission is arising from cool material (T$<$100 K) 
129: located mainly beyond 10 AU (Beichman et al.~\citeyear{beic05a}). This means that
130: these stars not only harbor planets (as inferred from their radial velocity observations) 
131: but also harbor an outer belt of dust-producing planetesimals (responsible for 
132: their IR excess) and in this regard they resemble the Solar System in its
133: Jovian planets +  Kuiper Belt configuration. The other two sources in Table 1 (HD 69830
134: and $\epsilon$-Eri) have warm dust and are therefore less relevant to the present discussion. 
135: 
136: In this paper we constrain the distribution of the potential dust-producing planetesimals
137: from the study of the dynamical perturbations of the two known planets, considering
138: in particular the effect of secular resonances. This allows us to identify three regions
139: where planetesimals could be dynamically stable ($\S$2). After examining the lifetimes 
140: of the dust grains and concluding that the dust traces the location of the parent 
141: planetesimals ($\S$3), we show how we can further constrain the location of the dust (and the 
142: parent planetesimals) by modeling HD 38529 spectral energy distribution (SED -- $\S$4). 
143: In $\S$5 we discuss together the results from the dynamical and SED analysis, and finally $\S$6
144: summarizes our results. 
145: 
146: \section{Possible Location of the Dust-Producing Planetesimals: Effect of Gravitational Perturbations by the Planets}
147: 
148: We can identify the possible location of the dust-producing planetesimals by studying
149: the effect of the planetary perturbations on the stability of the planetesimals' orbits. 
150: Consider the orbital parameter space of semimajor axis and eccentricity, $(a,e)$. 
151: First, we can eliminate planetesimals with orbits that would 
152: cross the orbits of the planets, i.e., orbits that have 
153: either periastron or apoastron within the periastron-apoastron range
154: of each of the planets.  Second, we note that planetesimals in initially circular 
155: orbits would be strongly unstable in the vicinity of the orbits 
156: of each of the two known planets, in a range of semimajor axis, 
157: $\Delta a \simeq \pm 1.5 (m_{\hbox{planet}}/m_\star)^{2/7}$ 
158: (Duncan et al.~\citeyear{dunc89}).
159: These two considerations identify several regions where planetesimals could not
160: be stable, shown as the grey and red shaded zones in Fig.~1. In addition to these
161: perturbations that operate over short timescales, we need to consider the effect of 
162: perturbations operating over much longer timescales. These are the secular perturbations, 
163: and as we will see below, they can cause a strong eccentricity excitation of the planetesimals, 
164: particularly at secular resonance locations, that can significantly shorten their lifetime.
165: 
166: We now calculate the effect of secular perturbations of the two planets on the 
167: planetesimals (taken as test particles on initially circular orbits). 
168: For the two planets, we assume co-planar orbits and minimum masses, taking
169: their orbital parameters from Butler et al. (\citeyear{butl06}).
170: For a two planet system, there are two linear modes that excite eccentricities
171: of test particles; we follow the Laplace-Lagrange secular perturbation analysis 
172: (see Murray and Dermott~\citeyear{murr99}).  The frequencies, $g_i$, phases, $\beta_i$, and 
173: amplitudes, $E^{(i)}$, of the two secular modes for the HD 38529 system are:
174: \begin{equation}
175: {g_1 = 0.106 \hbox{arcsec/yr}, E^{(1)} = (6.697\times10^{-3}, 0.3300), \beta_1=0.2271}
176: \end{equation}
177: \begin{equation}
178: {g_2 = 19.7 \hbox{arcsec/yr}, E^{(2)} = (0.2787, -6.411\times10^{-5}), \beta_2=1.594}
179: \end{equation}
180: 
181: The secular variations of the eccentricity vectors of planets b,c are 
182: given by
183: 
184: \begin{equation}
185: {e_b\cos\varpi_b = E^{(1)}_1\cos(g_1t+\beta_1)+E^{(2)}_1\cos(g_2t+\beta_2)}
186: \end{equation}
187: \begin{equation}
188: {e_b\sin\varpi_b = E^{(1)}_1\sin(g_1t+\beta_1)+E^{(2)}_1\sin(g_2t+\beta_2)}
189: \end{equation}
190: \begin{equation}
191: {e_c\cos\varpi_c = E^{(1)}_2\cos(g_1t+\beta_1)+E^{(2)}_2\cos(g_2t+\beta_2)}
192: \end{equation}
193: \begin{equation}
194: {e_c\sin\varpi_c = E^{(1)}_2\sin(g_1t+\beta_1)+E^{(2)}_2\sin(g_2t+\beta_2)}
195: \end{equation}
196: 
197: The fast mode dominates in planet b's eccentricity vector,
198: whereas the slow mode dominates planet c's eccentricity vector.
199: 
200: Now consider a test particle (i.e. a planetesimal) in this system.
201: Its secular perturbations have two components: 
202: a ``free'' apsidal precession due to the overall quadrupole potential 
203: (arising from the combined effects of the post-Newtonian stellar 
204: gravitational field, and the gravitational effects of planetary masses 
205: approximated by uniform density circular rings of radius equal to their 
206: semimajor axes), and a ``forced'' eccentricity-apsidal perturbation due to 
207: the planetary secular modes (i.e., due to the secular variations of the 
208: elliptical planetary orbits).  The amplitude of the latter perturbation 
209: is a function of the semimajor axis of the test particle; at some values 
210: of semimajor axis, where the free precession frequency is close to a 
211: planetary secular frequency, the test particle is subjected to a large 
212: amplitude resonant eccentricity excitation. To determine the maximum 
213: forced eccentricity obtained by a test particle, we use the secular 
214: resonance analysis given in Malhotra (\citeyear{malh98}) which includes the effects 
215: of non-linear saturation of the amplitude of the resonantly forced 
216: eccentricity from each secular mode.
217: The results are shown in Fig.~1: a test particle initially in a circular 
218: orbit would have its eccentricity excited by the slow mode to the values
219: indicated by the blue curve; the green curve is for the eccentricity
220: excited by the fast mode.  
221: 
222: Clearly, very significant secular eccentricity excitation occurs over 
223: a wide zone that extends to distances much larger than the unstable zones 
224: identified above. Remarkably, even though the strongly unstable zone of the outer 
225: planet extends outward to only about 5.5 AU, the eccentricity excitation exceeds 
226: 0.3 to more than 10 AU.  In general, we see that the forced eccentricities exceed 0.1 
227: everywhere up to $\sim57$ AU.  A strong secular resonance with 
228: the slow mode occurs at a semimajor axis value of $\sim55$ AU, where the 
229: particles become planet-crossing or can even collide with the star;
230: from the secular analysis (Malhotra 1998), we calculate that the timescale 
231: to excite the eccentricity from $\sim0$ to $\sim1$ is $\sim1.9\times10^8$ years.
232: (An analogous phenomenon occurs at the inner edge of the asteroid belt 
233: in our Solar System, at the location of the so-called $\nu_6$ secular 
234: resonance where asteroidal apsidal precession rates are resonant with the 
235: 6th secular mode of the Solar System.)  Thus, tentatively, we can
236: identify three regions that could harbor planetesimal populations
237: (or perhaps small planets) in low eccentricity orbits: 
238: an inner region 0.4--0.8 AU in-between the two planets' orbits, and two outer 
239: regions, 20--50 AU and beyond 60 AU. This conclusion is confirmed by numerical 
240: integrations of test particles as demonstrated in Figures 2 and 3. 
241: 
242: What does the high eccentricity excitation of test particles imply
243: for dust-producing planetesimal populations? 
244: Because these are forced eccentricities, neighboring particle orbits 
245: could have a high degree of apsidal alignment, so that the relative
246: velocities of planetesimals would be determined by Kepler shear, and
247: therefore would be small. However, it is likely that, in the formation
248: and evolution process, free eccentricities of planetesimal populations 
249: become comparable to the forced eccentricities, as observed in the solar 
250: system's asteroid belt and Kuiper Belt.  In that case, the forced
251: eccentricities would indeed be a measure of the relative velocities 
252: for collisions amongst planetesimals in these regions.
253: Thus, if the zones identified above are populated by planetesimals,
254: the planetary perturbations would cause mutual planetesimal collisions 
255: that would result in the production of dust. 
256: 
257: Alternatively, because the primordial disk of HD 38529 was massive enough to form 
258: at least two giant planets, arguably, these zones could harbor one or more large 
259: planetesimals, or even sub-Jovian mass planets so far undetected: 
260: sufficiently massive planets would tend to suppress the secularly forced eccentricities, and 
261: clear out planetesimal populations from their vicinities, 
262: thereby removing sources of dust; while large 1000 km-size planetesimals 
263: could have stirred up and ground away the planetesimal disk at early times. 
264: 
265: \section{A Collision-Dominated System: the Dust Traces the Planetesimals}
266: 
267: To evaluate whether the HD~38529 debris disk is radiation-dominated or collision-dominated we
268: consider the lifetimes of the dust grains: 
269: \begin{itemize}
270: 
271: \item The collisional lifetime at a distance $\it{R}$ from the star can be 
272: estimated as t$_{col}$ = P$_{orb}$/8$\Sigma$$\sigma_{geo}$ (Backman \& Paresce~\citeyear{back93}), 
273: where P$_{orb}$ is the orbital period, $\Sigma$ is the surface density of the 
274: grains and $\sigma_{geo}$ is the grain geometric cross section. 
275:  $\Sigma$$\sigma_{geo}$ is the dimensionless fractional surface area of the disk, of
276: the order of L$_{dust}$/L$_*$ $\sim$ 10$^{-5}$(5600/T$_*$)$^3$(F$_{70,dust}$/F$_{70,*}$)
277: (Backman \& Paresce~\citeyear{back93}). 
278: For HD~38529,  F$_{70,*}$ = 17.4 mJy, F$_{70,dust}$ = 66.9 mJy, T$_*$ = 5697 K 
279: and M$_{*}$ = 1.47 \msol~ (Moro-Mart\'{\i}n et al.~\citeyear{ama07}), so that 
280: L$_{dust}$/L$_{*}$ = 3.6 $\times$10$^{-5}$ and 
281: \begin{equation}
282: {t_{col}  = 3.4 \times 10^3 P_{orb} = 2.8  \times 10^3 ({R \over AU})^{3/2} yr.}
283: \end{equation}
284: 
285: \item The  Poynting-Robertson (P-R) lifetime at distance $\it{R}$, i.e. the time it takes for 
286: the dust grain to migrate from $\it{R}$ to the star, is given by 
287: \begin{equation}
288: {t_{PR} = {4 \pi b \rho \over 3} {c^2 R^2 \over L_*} = 710 
289: ({b \over \mu m}) ({\rho \over g/cm^3}) ({R \over AU})^2 
290: ({L_\odot \over L_*}) {1 \over 1+albedo} ~yr}
291: \end{equation}
292: (Burns, Lamy \& Soter,~\citeyear{burn79}; Backman \& Paresce~\citeyear{back93}), 
293: where $\it{b}$ and $\rho$ are the grain radius and density, respectively.  
294: More relevant is the time it takes for the particle to drift under
295: P-R drag from a populated to a relatively unpopulated region, i.e. the time it would
296: take to fill out the gap under P-R drag ($\it{t_{PR}^{fill}}$). 
297: If the scale over which the dust density 
298: significantly decreases is $\it{x}$\% of $\it{R}$, this time is
299: \begin{equation}
300: {t_{PR}^{fill} =  (1-(1-{x \over 100})^{2}) ~t_{PR} = 2.0 \times 10^4 
301: ({R \over AU})^2 yr,}
302: \end{equation}
303: where we have assumed that x = 10\%, L$_{*}$ = 6.3 \lsol, $\rho$ = 2.5 g/cm$^3$, 
304: $\it{b}$ = 10 $\mu$m and $\it{albedo}$ = 0.1. 
305: 
306: \end{itemize}
307: 
308: If the sources of dust are outside the orbit of the planet, and if P-R drag dominates
309: the dynamics, as the dust particles drift inward due to P-R drag they 
310: are likely to be scattered out of the system when crossing the orbit of a planet, 
311: creating a dust depleted region inside the planet's orbit
312: (Roques et al.~\citeyear{roqu94}; Liou \& Zook~\citeyear{liou99}; 
313: Moro-Mart\'{\i}n \& Malhotra~\citeyear{ama02},~\citeyear{ama03}). 
314: The ejection is very efficient for planets in circular orbits: 
315: planets with masses of 3--10 M$_{Jup}$ located 
316: between 1--30 AU eject $>$90\% of the particles that go past their orbits, while a 
317: 1 M$_{Jup}$ planet at 30 AU ejects $>$ 80\% of the particles, and about 50\%--90\%
318: if located at 1 AU (Moro-Mart\'{\i}n \& Malhotra~\citeyear{ama05b}); 
319: these results are valid for dust particle sizes in the range
320: 0.7--135~$\mu$m (assuming astronomical silicate composition), and when the central
321: star is solar-type. 
322: 
323: On the other hand, if the system is collision-dominated (Krivov et al.~\citeyear{kriv00}; 
324: Dominik \& Decin~\citeyear{domi03}; Wyatt~\citeyear{wyat05}), 
325: mutual collisions can fragment the larger particles to smaller and smaller 
326: sizes, until they are blown out from the system by radiation pressure. This means 
327: that the dust particles would be destroyed before they migrate far from the 
328: location of their parent bodies under P-R drag, i.e. the dust traces the location of 
329: the parent bodies. In this case, an inner cavity in 
330: the dust density distribution could only arise if the planetesimals themselves 
331: are confined to a belt, i.e. if there is an inner edge to their spatial distribution. 
332: This scenario may suggest a massive planet confining the inner edge of the 
333: dust-producing planetesimals. 
334: 
335: For the HD~38529 system, the estimates in equations [7] 
336: and [9] suggest that $\it{t_{col}}$ $<<$ $\it{t_{PR}^{fill}}$ at all relevant radii, 
337: i.e., we are in the collision-dominated regime. This means that the dust 
338: emission traces the location of the dust-producing planetesimals, with 
339: dynamically stable niches at 0.4--0.8 AU, 20--50 AU and beyond 60 AU. To find out  
340: which of these niches do actually show signs of harboring dust-producing plantesimals 
341: we turn now to the study of the IR excess emission detected by $\it{Spitzer}$.
342: 
343: \section{SED Modeling}
344: 
345: \subsection{Single Temperature Models are Insufficient}
346: 
347: The 70 $\mu$m observations of HD~38529 are not spatially resolved and therefore
348: it is not possible to know unambiguously the spatial distribution of the dust.
349: We can learn about the characteristic temperature of the dust from the spectral 
350: energy distribution (Hillenbrand et al. in preparation). 
351: From the ratio of the excess emission at 70 $\mu$m to that at 33 $\mu$m, and 
352: assuming these wavelengths are at the Wien's tail of the dust blackbody emission, 
353: the ratio of the excess fluxes yield a characteristic temperature of of 43$\pm$4 K.
354: If the blackbody grains are in thermal equilibrium, we can calculate the 
355: location of the dust from
356: \begin{equation}
357: {{R_{dust} \over AU} = ({L_* \over L_{\sun}})^{1/2}({278 K \over T_{dust}})^2}
358: \end{equation}
359: (Backman et al.~\citeyear{back93}); for  HD~38529, R$_{dust}$ = 106$\pm$18 AU. The 
360: dust mass can then be estimated by assuming the dust is in a thin shell at 
361: a distance R$_{dust}$, and that the grains are spherical with cross section
362: equal to their geometric cross section, so that the total number of dust particles is 
363: $\sim$(4$\pi$R$_{dust}^2$/$\pi$$\it{b}^2$)$\times$$\tau$ and 
364: M$_{dust}$ $\sim$ 16$\pi$R$_{dust}$$^2$$\tau$$\it{b}$$\rho$/3. The optical 
365: depth, $\tau$, can be approximated as L$_{dust}$/L$_{*}$ so that, 
366: \begin{equation}
367: {{M_{dust} \over M_{\earth}} \sim 6.28\times10^{-5}({L_{dust} \over L_*})({\rho \over g cm^{-3}})({<b> \over \mu m})({R_{dust} \over AU})^2}
368: \end{equation}
369: (Jura et al.~\citeyear{jura95}), where $\it{b}$ and $\rho$ are the particle's 
370: radius and density. 
371: It is important to note that the above estimate is a lower limit: in the absence of 
372: sensitive (sub)millimeter detections, no realistic constraints 
373: can be made to the dust mass, a significant fraction of which could be locked in grains with
374: sizes of $\sim$1 mm that emit efficiently in the (sub)millimeter but contribute little to 
375: the infrared emission. For HD~38529, L$_{dust}$/L$_{*}$ = 3.6$\times$10$^{-5}$, and
376: using  $\it{b}$ = 10~$\mu$m and $\rho$ = 2.5 g/cm$^{-3}$ we obtain 
377: M$_{dust}$ $>$ 1.9$\times$10$^{-9}$~\msol\footnote{Dust mass estimates for the KB dust disk range 
378: from a total dust 
379: mass $<$ 3$\times$10$^{-10}$~\msol~(Backman et al.~\citeyear{back95}) to 
380: $\sim$ 4$\times$10$^{-11}$~\msol~for dust particles $<$ 150 $\mu$m 
381: (Moro-Mart\'{\i}n \& Malhotra~\citeyear{ama03}); with a 
382: fractional luminosity of L$_{dust}$/L$_*$$\sim$10$^{-7}$--10$^{-6}$ (Stern~\citeyear{ster96}). 
383: The fractional luminosity of the asteroid belt dust (a.k.a zodiacal cloud) is estimated to 
384: be L$_{dust}$/L$_*$$\sim$10$^{-8}$--10$^{-7}$ (Dermott et al.~\citeyear{derm02}).}.
385: 
386: Another estimate of the dust temperature can be made from fitting a photosphere plus a single 
387: temperature blackbody to the IRS spectrum only; this results in a dust temperature of 79 K, which
388: would corresponds to a distance of 31 AU. The discrepancy with the characteristic temperature 
389: derived from the ratio of the excess emission at 70 $\mu$m to that at 33 $\mu$m indicates 
390: that it is not possible to conclude that the dust is confined to a narrow ring as implied by 
391: the assumption of a single grain temperature. Similarly, Hillenbrand et al. (in preparation) 
392: find that the SEDs of about 1/3 of the FEPS targets with 70~$\mu$m excess emission are better 
393: fit by multi-temperature rather than single-temperature blackbody models.
394: The need for a multi-temperature grain distribution has previously been 
395: found in systems like $\beta$-Pic (e.g. Li \& Greenberg~\citeyear{li98} and Telesco 
396: et al.~\citeyear{tele05}) and HR4796A (e.g. Li \& Lunine~\citeyear{li03a}), 
397: and has been unambiguously confirmed by the 
398: eight spatially resolved observations in scattered light of debris disks known to date, 
399: which led Kalas et al. (\citeyear{kala06}) to conclude that 
400: debris disks show two basic architectures: 1) belts about 20--30 AU 
401: wide and with well-defined outer boundaries (HR 4796A, Fomalhaut and HD 139664); 
402: and 2) wider belts with sensitivity-limited edges implying widths $>$50 AU 
403: (HD 32297, $\beta$-Pic, AU-Mic, HD 107146 and HD 53143).
404: In the absence of spatially resolved observations, the SED modeling of HD~38529 
405: (or any other source under consideration) should enable to explore a range of 
406: disk architectures.
407: 
408: \subsection{Multi Temperature (SED) Models}
409: \subsubsection{SED Modeling Assumptions} 
410: 
411: For the modeling of the SED, we use the radiative transfer code developed by 
412: Wolf \& Hillenbrand (\citeyear{wolf03}). Because of the above considerations, 
413: we model the dust disk as an annulus of inner radius R$_{in}$, outer
414: radius R$_{out}$, total dust mass M$_{dust}$,
415: and a constant surface density ($\Sigma \propto r^0$, so that the number density, 
416: n(r)~$\propto$~r$^{-1}$).
417: We assume that the dust grains are composed of silicates\footnote{For a study on how the SEDs depend on the grain composition we refer to Wolf \& Hillenbrand 
418: (\citeyear{wolf03}) and Moro-Mart\'{\i}n, Wolf \& Malhotra (\citeyear{ama05a}).} with 
419: optical constants from Weingartner \& Draine (\citeyear{wein01}). 
420: For the particle sizes we consider two options:
421: 1) a single grain size of 10 $\mu$m in radius, and 2) a particle size distribution following
422: a power law, $\it{n(b)}$ $\propto$ $\it{b^{-q}}$, where $\it{b}$ is the particle radius, 
423: $\it{q}$ = 3.5 (for grains in collisional equilibrium), 
424: $\it{b_{min}}$ = 2 $\mu$m and 
425: $\it{b_{max}}$ = 10 $\mu$m.  In both cases, the radius of 10 $\mu$m was chosen 
426: because such a grain radiates efficiently at 70 $\mu$m.
427: Because of the last property, this choice of grain size provides a
428: lower limit for M$_{dust}$. The value selected for $\it{b_{min}}$ corresponds to the 
429: ``blow-out'' size, i.e. the minimum size of the grains that can remain bound in the 
430: system, given by
431: \begin{equation}
432: { {b_{min} \over \mu m} = 0.52 \times {2.5g/cm^{-3} \over \rho} \times {1 + albedo \over 1.1}
433: \times {L_*/L_{\odot} \over (M_*/M_{\odot})(T_*/5780K)}}
434: \end{equation}
435: (Artymowicz et al.~\citeyear{arty89}). For HD~38529, L$_{*}$ = 6.31 \lsol, M$_{*}$ = 1.47 \msol~and  
436: T$_{*}$ = 5697 K; using a grain density $\rho$ = 2.5 g/cm$^{-3}$ 
437: and albedo = 0.1, we get $\it{b_{min}}$ = 2 $\mu$m . 
438: 
439: We assume that the dust disk is optically thin with the dust being in thermal 
440: equilibrium with the stellar radiation field. 
441: Only scattering, absorption and reemission of stellar radiation by dust 
442: grains are taken into account, neglecting scattering and dust heating 
443: from the infrared radiation produced by the optically thin dust disk. 
444: With these assumptions, the temperature of the dust grain,
445: for a given dust size and a chemical composition, depends only on the distance to the 
446: central star
447: \footnote{Our SED modeling assumes that the observed dust excess arises
448: from a cicumstellar disk. However, the outermost planet has a $\it{M}$sin$\it{i}$ of 12.2~M$_{Jup}$, 
449: placing it in the boundary between planets and brown dwarfs. Even though the latter 
450: could potentially harbor a disk, the prevalence of cold dust in the HD 38529 system indicates
451: that this could not be the dominant source of the observed dust because this disk 
452: would be located near the star (the outermost planet's semimajor axis is 3.74 AU)
453: implying the presence of warm dust that is not observed.}.
454: 
455: The outer radius of the disk, R$_{out}$, cannot be constrained with data currently available. 
456: %without multiwavelength sub-mm and mm photometry or spatially resolved observations.
457: Based on scattered-light observations from 
458: nearby debris disks, disk sizes of about 100 AU 
459: to several hundred AU are inferred (Kalas et al.~\citeyear{kala05}; Dent et al.~\citeyear{dent00}; 
460: Greaves et al.~\citeyear{grea00}; Wilner et al.~\citeyear{wiln02}; 
461: Holland et al.~\citeyear{holl03}; Liu~\citeyear{liu05}; Metchev et al.~\citeyear{metc05}; 
462: Ardila et al.~\citeyear{ardi04}). 
463: %However, submillimeter observations of the 
464: %center of the Trapezium (Williams, Andrews \& Wilner~\citeyear{will05}) 
465: %and of the Taurus region indicate 
466: %that the majority of primordial disks may be smaller (in particular if they
467: %are born within a cluster). 
468: For this reason we consider disk sizes of R$_{out}$ = 50 AU (Solar System size), 
469: R$_{out}$ = 100 AU and R$_{out}$ = 500 AU.
470: 
471: With the above three values for R$_{out}$, assuming a uniform density distribution and
472: with the grain size and composition fixed, we then vary R$_{in}$ and
473: M$_{dust}$ (our only two free parameters) to create a grid of models where
474: we allow R$_{in}$ to vary from the silicate sublimation radius (R$_{sub}$, 
475: where T$_{sub}$ = 1550 K) to R$_{out}$. 
476: This accounts for the possibility of having either a dust disk of wide 
477: radial extent or a narrow ring of dust. 
478: 
479: The observations to be modeled are the $\it{Spitzer}$-IRS\footnote{The IRS spectrum for $\lambda$ $<$ 14.21 $\mu$m 
480: (short-low module) shows a small offset with respect to the Kurucz model. Because the emission at those
481: wavelengths is clearly photospheric, we have multiplied the spectrum for 
482: $\lambda$ $<$ 14.21 $\mu$m by 1.045 to make it 
483: coincide with the stellar photosphere (so these wavelengths do not dominate the $\chi$$^2$ 
484: statistics). In addition, the IRS spectrum shows a small discontinuity at 14.21 $\mu$m, 
485: possibly because the star was not centered in the slit of the long-low module. To correct
486: for this discontinuity, we have multiplied the IRS spectrum for $\lambda$ $>$ 14.21 $\mu$m 
487: by 1.108. Accounting for this, the corrected 33 $\mu$m flux is 95$\pm$2 mJy, where the 
488: error is the internal uncertainty. For the calculation of the S/N of the excess 
489: we used the total uncertainty, 
490: obtained from adding in quadrature the internal uncertainty and the calibration uncertainty, 
491: the latter taken to be 6\%. This is an underestimate of the significance of the departure from 
492: a pure photosphere because this calibration uncertainty is for the overall spectrum, not 
493: for individual wavelengths relative to each other.} synthetic 
494: photometric points at 13 $\mu$m, 24 $\mu$m  and 33 $\mu$m and the $\it{Spitzer}$-MIPS
495: photometric points at 24 $\mu$m and 70 $\mu$m (from Moro-Mart\'{\i}n et al.~\citeyear{ama07}). 
496: To evaluate whether a particular model is a valid fit to these data points, or if not, to what 
497: degree of certainty the model can be excluded, we calculate its $\chi$$^{2}$ probability, 
498: P($\chi$$^{2}$ $\mid$ $\nu$), where $\nu$ is the number of degrees of freedom. In 
499: our case $\nu$ = 2, as the only two free parameters are R$_{in}$ and M$_{dust}$, and
500: all the other disk parameters and dust properties are fixed to the values given above. 
501: %This is the probability that the observed $\chi$$^2$ for the correct model is less than 
502: %the value of $\chi$$^{2}$ for the model under consideration, with P(0 $\mid$ $\nu$) = 0 
503: %and  P(1 $\mid$ $\nu$) = 1. 
504: The probability is defined so that P(0 $\mid$ $\nu$) = 0 and  P(1 $\mid$ $\nu$) = 1. 
505: Models with P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.9973 can be 
506: excluded with a 3-$\sigma$ certainty, while models with P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.683 
507: could be excluded with a 1-$\sigma$ certainty. 
508: 
509: \subsubsection{SED Modeling Results}
510: 
511: Figure 4 shows the two-dimensional grids of SED models described above: 
512: in $\it{green}$ are the models with 
513: P($\chi$$^{2}$ $\mid$ $\nu$) $<$ 0.683, i.e. models that are consistent with
514: the observations; $\it{red}$ represents models with P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.683,
515: i.e. models that could be excluded with a 1-$\sigma$ certainty; and $\it{black}$
516: corresponds to models with  P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.9973, i.e. models
517: that can be excluded with a 3-$\sigma$ certainty. 
518: Some of these SEDs are shown in Fig. 5 together with the observations. 
519: As can be seen in both figures, the models are degenerate and for a given R$_{out}$, 
520: there are many pairs of R$_{in}$ and M$_{dust}$ that could fit the observed SED. 
521: Because of this degeneracy, our main interest lies in identifying the models that can 
522: be excluded. These are of particular interest because they allow us to identify dust depleted regions
523: whose origin can be studied in terms of the overall dynamics of the planetary system (see $\S$2).
524: 
525: The main results from Fig. 4(a--c) regarding the models that assume 
526: grains 10 $\mu$m in radius are the following
527: \begin{itemize}
528: \item For R$_{out}$ = 50 AU, we can exclude models 
529: with R$_{in}$ $\la$ 5 AU with a certainty of 3-$\sigma$, or
530: we could exclude models with R$_{in}$ $\la$ 14 AU with a certainty of 
531: 1-$\sigma$; 
532: the data is consistent with models having 15 AU $\la$ R$_{in}$ $\la$ 50 AU,
533: i.e., models ranging from a 35 AU wide disk (extending from 15 AU--50 AU) 
534: to a very  narrow ring of dust located at 50 AU. 
535: In other words, the observations require the presence of an inner cavity 
536: in the dust density distribution of at least 5 AU in radius, with larger 
537: cavities consistent with the data (the latter implying narrower rings
538: of larger dust mass). 
539: \item For R$_{out}$ = 100 AU, none of the R$_{in}$ considered 
540: (ranging from the dust sublimation radius, R$_{sub}$, to 100 AU) can be
541: excluded with a 3-$\sigma$ certainty; while models with R$_{in}$ $\la$ 7 AU
542: and those with R$_{in}$ $\ga$ 40 AU could be excluded to 1-$\sigma$. Possible
543: fits include models with 8 AU $\la$ R$_{in}$ $\la$ 35 AU.
544: \item For R$_{out}$ = 500 AU, again none of the R$_{in}$ considered 
545: (from R$_{sub}$ to 400 AU) can be excluded with a 3-$\sigma$ certainty; 
546: while models with R$_{in}$ $\ga$ 14 AU
547: could be excluded to 1-$\sigma$. Possible models 
548: include those with  R$_{sub}$ $\la$ R$_{in}$ $\la$ 12 AU. 
549: \end{itemize}
550: 
551: We can understand these results as follows. For a constant surface density,
552: $\Sigma \propto r^0$, there is 100 times more dust mass between 50 and 100 AU 
553: than there is between 5 and 10 AU, while the dust temperature is proportional 
554: to $\it{r}$$^{-1/2}$. This results in that there is a significant contribution
555: to the 70 $\mu$m emission from the largest radii, R$_{out}$. To keep the 
556: 70 $\mu$m emission constant, the dust surface area or mass at R$_{out}$ must be similar 
557: as R$_{out}$ varies, i.e. $\Sigma$R$_{out}$ $\sim$ constant, this means that: 
558: 1) larger R$_{out}$ requires
559: a smaller $\Sigma$, and smaller $\Sigma$ means that we can tolerate smaller inner 
560: radii since there is now less dust mass placed there; this explains 
561: why the SEDs can be fitted with a smaller inner cavity for 
562: R$_{out}$ = 100 AU, and with a constant surface density all the way to the sublimation 
563: radius in the 500 AU case; and 2) smaller R$_{out}$ requires a larger $\Sigma$,
564: and in order to keep the 24 $\mu$m and 33 $\mu$m fluxes below their upper limits, this
565: requires to increase the inner cavity size. A larger inner cavity would also be needed if 
566: we had assumed $\Sigma \propto r^{-1}$ instead of constant. 
567: 
568: In the models shown in Fig. 4(d) we relax the assumption that 
569: all the grains are 10 $\mu$m in radius and allow for the presence of 
570: smaller grains, with the particle size distribution following
571: $\it{n(b)}$ $\propto$ $\it{b^{-3.5}}$, and with $\it{b_{min}}$ = 2 $\mu$m and 
572: $\it{b_{max}}$ = 10 $\mu$m. We find that most of the models can be 
573: excluded with a 3-$\sigma$ certainty, i.e. we can exclude the presence of 
574: a significant population of small grains inside 100 AU based on the lack of a significant 
575: continuum emission at $\lambda$ $<$ 30 $\mu$m. A similar conclusion has been obtained from 
576: several other debris disks whose spectroscopy show little or no solid state
577: features, indicating that the dust grains have sizes $\ga$ 10 $\mu$m
578: (e.g. Jura et al.~\citeyear{jura04}; Stapelfeldt et al.~\citeyear{stap04}).
579: 
580: \section{Discussion of the Dynamical and the SED Analysis}
581: 
582: \subsection{Location of Cold Dust}
583: 
584: The two outermost regions identified in $\S$2 where planetesimals could 
585: survive for extended periods of time in the presence of the two known radial 
586: velocity-detected planets, namely 20--50 AU and beyond 60 AU, are in broad agreement 
587: with the allowed dust locations that result from the modeling of the SED.
588: In particular, we saw that the models with R$_{out}$ = 50 AU predicted 
589: an inner cavity of radius $>$ 5 AU;  while a range of dust disks with R$_{out}$ = 50 AU 
590: and R$_{in}$ $>$ 20 AU could fit the observed SED. From the models with 
591: R$_{out}$ = 100 AU, we can conclude that it is very unlikely that 
592: the observed dust emission arises only from planetesimals located beyond 60 AU because
593: the models with R$_{out}$ = 100 AU and R$_{in}$ = 60 AU, 70 AU, 80 AU and 90 AU 
594: have P($\chi$$^{2}$ $\mid$ $\nu$) larger than 0.83, 0.87, 0.89 and 0.91, respectively 
595: A similar conclusion can be drawn from the models with R$_{out}$ = 500 AU, 
596: and R$_{in}$ $>$ 60 AU, for which P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.94.
597: Therefore, from the dynamical and SED modeling we conclude that the planetesimals
598: responsible for most of the dust emission are likely located within the 
599: 20--50 AU region. 
600: 
601: Given that the maximum eccentricities in the 20--50 AU region are moderately high, 
602: 0.25--0.3 (see Figures 1 and 2) -- implying collisional velocities at 30 AU of approximately 
603: 0.3$\times$$\it{v_{circ}}$ = 2.8 km/s -- 
604: could a moderate rate of dust production persist in this relatively old system?
605: (A ``moderate rate" is important in order for the dust production to be long-lived, 
606: instead of being depleted rapidly in the early history of the system). 
607: Following Wyatt et al. (\citeyear{wyat07}), we can estimate 
608: the maximum fractional luminosity of the excess, $\it{f_{max}}$, that could originate from 
609: a planetesimal belt of a given age that is evolving in quasi-steady state. 
610: Wyatt's analytical model assumes that the 
611: planetesimals and the dust are in collisional equilibrium, and that their size distribution follows
612: a continuous power law of index -3.5. For HD 38529, and given the results above, we assume that 
613: the planetesimal belt extends from 20--50 AU, with a mean planetesimal eccentricity 
614: of 0.3. Other parameters in the model are the diameter of the largest planetesimal in the 
615: cascade, 2000 km, and the specific incident energy required to catastrophically
616: destroy a planetesimal, 200 J/kg. At the age of the HD 38529 system, 3.5 Gyr, the model predicts
617: $\it{f_{max}}$ = 2.26$\times$10$^{-6}$ (see eq. [20] in Wyatt et al.~\citeyear{wyat07}). 
618: This value is smaller than the observed fractional 
619: luminosity of the excess, $\it{f_{obs}}$ = 3.6$\times$10$^{-5}$~(see $\S$3). However, because
620: there are two orders of magnitude uncertainty in the estimate of $\it{f_{max}}$, 
621: we cannot reject a scenario in which the dust observed is the result of the steady grinding 
622: down of planetesimals. If we were to assume that the above value of $\it{f_{max}}$ is correct, 
623: an estimate of the timescale over which the fractional luminosity can be mantained at the value of
624: $\it{f_{obs}}$ is given by $\it{t_{age}}$$\times$$\it{f_{max}}$/$\it{f_{obs}}$ = 220 Myr. 
625: In this case, a possible scenario could be that the stable region beyond 60 AU supplies some 
626: planetesimals that drift into the 20--50 AU region by non-gravitational effects. 
627: 
628: \subsection{Upper Limits on Warm Dust}
629: 
630: In $\S$2 we identified a  modestly stable small zone in-between the two known planets
631: (0.4--0.8 AU), that has also been identified in numerical
632: simulations by Barnes \& Raymond (\citeyear{barn04}); our secular perturbation
633: analysis provides a theoretical explanation for those numerical
634: stability results.  Raymond \& Barnes (\citeyear{raym06}) consider terrestrial planet 
635: accretion in this zone.  They conclude that HD 38529 is likely to support 
636: an asteroid belt and perhaps Mars-sized planets, but not larger planets, 
637: because the potential feeding zone for the accretion of a terrestrial 
638: planet would be limited by the high eccentricities of the planetesimals 
639: in this region. 
640: The lack of an IR excess at wavelengths shorter than 30 $\mu$m allows
641: us to place an upper limit on the amount of warm (asteroidal) dust that
642: could be located in this inner region.  We use the IRAC 5$\mu$m and 8$\mu$m,
643: IRS 13$\mu$m and MIPS 24$\mu$m photometric measurements, and we assume
644: 10$\mu$m size silicate grains (optical constants taken from
645: Weingartner \& Draine~\citeyear{wein01}).  We find that a 3-$\sigma$ upper limit to the
646: dust mass in this potential asteroid belt is $3\times$10$^{-13}$ M$_\odot$ or 
647: 10$^{-7}$~M$_{\earth}$. For comparison, the mass estimate for the zodiacal cloud in the 
648: terrestrial planet region of the Solar System is 3$\times$10$^{-10}$~M$_{\earth}$
649: (Hahn et al.~\citeyear{hahn02}), i.e. 330 times smaller than the estimated upper limit 
650: of warm dust in HD~38529.
651: 
652: \section{Conclusions}
653: HD~38529 harbors a planetary system consisting of at least two planets (with 
654: $\it{M}$sin$\it{i}$ of 0.8~M$_{Jup}$ and 12.2~M$_{Jup}$, semimajor axes of 0.13 AU 
655: and 3.74 AU, and eccentricities of 0.25 and 0.35) and a likely population of dust-producing 
656: planetesimals that are responsible for the 70 $\mu$m excess emission detected 
657: by $\it{Spitzer}$. Using analytical and numerical dynamical analysis, in this paper 
658: we have constrained the distribution of the potential dust-producing planetesimals 
659: from the study of the dynamical perturbations of the two known planets, considering 
660: in particular the effect of secular resonances. A dust disk inner edge at 5.5 AU 
661: would naturally arise from the gravitational scattering of planetesimals and dust 
662: grains by the outermost planet. We show that larger inner cavities in the dust disk, 
663: that would be consistent with the observed SED, can be created due to the secular effects 
664: that arise from the interaction between the two massive planets. From the analysis of 
665: the secular perturbations we identify three regions that could harbor planetesimal 
666: populations in low eccentricity orbits (where the planetesimals could be long-lived): 
667: 0.4--0.8 AU, 20--50 AU and beyond 60 AU.  
668: From the modeling of the SED we conclude that the planetesimals responsible for most 
669: of the dust emission observed by $\it{Spitzer}$ are likely located within the 20--50 AU region.
670: In this regard, HD 38529 resembles the configuration of the Solar System's Jovian 
671: planets + Kuiper Belt (KB). 
672: The SED models give a dust mass estimate of 1--5$\times$10$^{-10}$~\msol~of 10 $\mu$m particles.
673: The presence of a significant population of small grains inside 100 AU
674: (with the particle size distribution following
675: $\it{n(b)}$ $\propto$ $\it{b^{-3.5}}$, and with $\it{b_{min}}$ = 2 $\mu$m and 
676: $\it{b_{max}}$ = 10 $\mu$m) is excluded to a 3-$\sigma$ certainty level based on the lack
677: of a significant continuum emission at $\lambda$ $<$ 30 $\mu$m. 
678: We do not find evidence of dust emission within the innermost region, 
679: with a 3-$\sigma$ dust mass upper limit of 10$^{-7}$~M$_{\earth}$ (in 10 $\mu$m
680: grains), suggesting any remnant dust belt would have a mass smaller than 330 times that in
681: the Solar System's zodiacal cloud. 
682: 
683: The SED models are degenerate. We need to break this degeneracy to get a better 
684: understanding of this planetary system, in particular, of how the spatial distribution
685: of the dust and the planetesimals are affected by the gravitational perturbations
686: of the two planets. This requires spatially resolved images to constrain the disk sizes, 
687: and/or high resolution spectroscopy observations to look for spectral features that 
688: could constrain the grain size and composition, and/or accurate photometric points 
689: in the 33 $\mu$m--70 $\mu$m range and in the sub-mm to better determine the shape of the SED. 
690: 
691: \begin{center} {\it Acknowledgments} \end{center}
692: We thank the rest of the FEPS team members, colleagues at the Spitzer Science Center, 
693: and members of all the Spitzer instrument teams for advice and support. 
694: This work is based [in part] on observations made with the Spitzer Space Telescope,
695: which is operated by the Jet Propulsion Laboratory, California Institute of Technology 
696: unders NASA contrast 1407. 
697: A.M.M. is under contract with the Jet Propulsion Laboratory (JPL) funded by NASA through
698: the Michelson Fellowship Program. JPL is managed for NASA by the California Institute of
699: Technology. A.M.M. is also supported by the Lyman Spitzer Fellowship at Princeton University.
700: R.M. and M.R.M. are supported in part through the LAPLACE node of NASA's Astrobiology Institute.
701: R.M. also acknowledges support from NASA-Origins of Solar Systems research program.
702: S.W. was supported through the DFG Emmy Noether grant WO 875/2-1 and WO875/2-2.
703: FEPS is pleased to acknowledge support from NASA contracts 1224768 and 1224566
704: administered through JPL.  
705: 
706: \begin{thebibliography}{}
707: \bibitem[Ardila et al. (2004)]{ardi04}
708: Ardila, D R., Golimowski, D.A., Krist, J.E., Clampin, M., Williams, J.P. 2004, ApJ, 617, L147
709: \bibitem[Artymowicz, Burrows \& Paresce (1989)]{arty89}
710: Artymowicz, P., Burrows, C. \& Paresce, F. 1989, 337, 494
711: %\bibitem[Allende Prieto \& Lambert (1999)]{alle99}
712: %Allende Prieto, C. \& Lambert, D.L., 1999, A\&A, 352, 555
713: \bibitem[Backman \& Paresce~(1993)]{back93}
714: Backman, D.E. \& Paresce, F. 1993, Protostars and Planets III 
715: (ed. E.H. Levy \& J.I Lunine, Tucson: Univ. of Arizona Press), 1253
716: \bibitem[Backman, Dasgupta \& Stencel (1995)]{back95}
717: Backman, D.E., Dasgupta, A. \& Stencel, R.E. 1995, ApJ, 450, L35 
718: %\bibitem[Backman (2004)]{back04}
719: %Backman, D.E. 2004, ASP Conf. Ser. 324, Debris Disks and the Formation of Planets, 
720: %ed. L. Caroff, L.J. Moon, D.E. Backman \& E. Praton (San Francisco: ASP), 9
721: %\bibitem[Barbieri \& Gratton (2002)]{barb02}
722: %Barbieri, M. \& Gratton, R.G., 2002, A\&A, 384, 879
723: \bibitem[Barnes \& Raymond (2004)]{barn04}
724: Barnes, R. \& Raymond, S.N. 2004, ApJ, 617, 569 
725: \bibitem[Beichman et al. (2005a)]{beic05a}
726: Beichman, C.A., Bryden, G., Rieke, G.H. et al., 2005, ApJ, 622, 1160
727: \bibitem[Beichman et al. (2005b)]{beic05b}
728: Beichman, C.A., Bryden, G., Gautier, T.N., Stapelfeldt, K.R., Werner, M.W. et al. 2005, ApJ, 626, 1061
729: \bibitem[Bryden et al. (2006)]{bryd06}
730: Bryden, G. Beichman, C.A., Trilling, D.E. et al. 2006, ApJ, 636, 1098
731: \bibitem[Burns, Lamy \& Soter (1979)]{burn79}
732: Burns, J.A., Lamy, P.L., Soter, S., 1979, Icarus, 40, 1
733: \bibitem[Butler (2006)]{butl06}
734: Butler, R.P., Wright, J.T., Marcy, G.W. et al. 2006, ApJ, 646, 505
735: \bibitem[Carpenter \& Stauffer (2003)]{carp03}
736: Carpenter, J. M. \& Stauffer, J. 2003, FEPS Memo No.4
737: \bibitem[Carpenter et al. (2005)]{carp05}
738: Carpenter, J. M., Wolf, S., Schreyer, K., Launhardt, R., \& Henning, Th. 2005, ApJ, 129, 1049
739: \bibitem[Dent et al. (2000)]{dent00}
740: Dent, W.R.F., Walker, H.J., Holland, W.S. \& Greaves, J.S. 2000, MNRAS, 314, 702
741: \bibitem[Dermott et al. (2002)]{derm02}
742: Dermott, S.F., Kehoe, T.J.J., Durda, D.D. et al. 2002, in Proc. Asteroids, Comets, Meteors, ed. B. Warmbein (ESA SP-500; Noordwijk; ESA), 319
743: \bibitem[Dominik \& Decin (2003)]{domi03}
744: Dominik, C. \& Decin, G. 2003, ApJ, 598, 626 
745: \bibitem[Duncan et al. (1989)]{dunc89}
746: Duncan, M., Quinn, T. \& Tremaine, S. 1989, Iracus, 82, 402
747: \bibitem[Duncan, Levison \& Lee (1998)]{dunc98}
748: Duncan, M.J., Levison, H.F. \& Lee, M.H. 1998, AJ, 116, 2067
749: \bibitem[Fischer et al. (2001)]{fisc01}
750: Fischer, D.A., Marcy, G.W., Butler, R.P. et al. 2001, ApJ, 551, 1107\bibitem[Fischer et al. (2003)]{fisc03}
751: Fischer, D.A., Marcy, G.W., Butler, R.P. et al. 2003, ApJ, 586, 1394
752: \bibitem[Fischer et al. (2005)]{fisc05}
753: Fischer, D.A. \& Valenti, J.A. 2005, ApJ, 622, 1102
754: \bibitem[Gonzalez et al. (2001)]{gonz01}
755: Gonzalez, G., Laws, C., Tyagi, S. et al. 2001, ApJ, 121, 432
756: \bibitem[Greaves et al. (1998)]{grea98}
757: Greaves J. S., Holland, W. S., Moriarty-Schieven G., Jenness, T., Dent, W. R. F.et al. 1998, ApJ, 506, L133
758: \bibitem[Greaves, Mannings \& Holland (2000)]{grea00}
759: Greaves, J.S., Mannings, V. \& Holland, W.S. 2000, Icarus, 143, 155
760: \bibitem[Greaves et al. (2005)]{grea05}
761: Greaves J. S., Holland W. S., Wyatt M. C., Dent W. R. F. \& Robson E. I. 2005, ApJ, 619, L187
762: \bibitem[Greaves, Fischer \& Wyatt~(2006)]{grea06}
763: Greaves, J.S., Fischer, D.A. \& Wyatt, M.C. 2006, MNRAS, 366, 283
764: \bibitem[Hahn et al. (2002)]{hahn02}
765: Hahn, J.M., Zook, H.A., Cooper, B. \& Sunkara, B. 2002, Icarus, 158, 360
766: \bibitem[Hatzes et al. (2000)]{hatz00}
767: Hatzes A. P., Cochran W. D., McArthur, B., Baliunas S. L., Walker G. A. H. et al. 2000, ApJ, 544, L145
768: \bibitem[Holland (2003)]{holl03}
769: Holland, W.S. et al. 2003, ApJ, 582, 1141
770: %\bibitem[Houk (1975)]{houk75}
771: %Houk \& Cowley, 1975, Michigan Catalog of HD stars, Vol 1
772: \bibitem[Houk (1980)]{houk80}
773: Houk 1980, Michigan Catalog of HD stars, Vol 3
774: %\bibitem[Houk \& Smoth-Moore (1988)]{houk88}
775: %Houk \& Smoth-Moore 1988, Michigan Catalog Vol 4
776: %\bibitem[Houk \& Smoth-Moore (1999)]{houk99}
777: %Houk \& Smoth-Moore 1999, Michigan Catalog Vol 5
778: %\bibitem[Jaschek (1978)]{Jasc78}
779: %Jaschek, M., 1978, BICDS, 15, 121
780: \bibitem[Jura et al. (1995)]{jura95}
781: Jura, M., Ghez, A.M., White, R.J., McCarthy, D.W., Smith, R.C. \& Martin, P.G.
782: 1995, ApJ, 445, 451	
783: \bibitem[Jura et al. (2004)]{jura04}
784: Jura M., Chen C.H., Furlan, E., Green, J., Sargent B. et al. 2004, ApJS, 154, 453
785: \bibitem[Kalas, Graham \& Clampin (2005)]{kala05}
786: Kalas, P., Graham, J.R. \& Clampin, M. 2005, Nature, 435, 1067
787: \bibitem[Kalas et al. (2006)]{kala06}
788: Kalas, P., Graham, J.R., Clampin, M. \& Fitzgerald, M.P. 2006, ApJ, 637, L57
789: \bibitem[Krivov, Mann \& Krivova(2000)]{kriv00}
790: Krivov, A.V., Mann, I. \& Krivova, N.A., 2000, A\&A, 362, 1127
791: \bibitem[Li \& Greenberg (1998)]{li98}
792: Li, A. \& Greenberg, J. M., 1998, A\&A, 331, 291
793: \bibitem[Li \& Lunine (2003)]{li03a}
794: Li, A. \& Lunine, J., 2003, ApJ, 590, 368
795: %\bibitem[Li, Lunine \& Bendo (2003)]{li03b}
796: %Li, A., Lunine, J. \& Bendo, G.J., 2003, ApJ, 598, L51
797: \bibitem[Liou \& Zook (1999)]{liou99}
798: Liou, J.C. \& Zook, H.A., 1999, AJ, 118, 580
799: \bibitem[Liu (2005)]{liu05}
800: Liu, M.C. 2004, Sci, 305, 144
801: \bibitem[Malhotra (1998)]{malh98}
802: Malhotra, R. 1998,in ASP Conf. Ser. 149, 
803: Solar System Formation and Evolution, 
804: ed. D. Lazzaro, R. Vieira Martins, S. Ferraz-Mello, J. Fernandez and Beauge, 
805: (San Francisco: ASP), 37
806: %\bibitem[Mayor et al. (2004)]{mayo04}
807: %Mayor, M., Udry, S., Naef, D. et al 2004, A\&A, 415, 391
808: \bibitem[Menou \& Tabachnik (2003)]{meno03}
809: Menou, K. \& Tabachnik, S. 2003, ApJ, 583, 473
810: \bibitem[Metchev et al. (2005)]{metc05}
811: Metchev, S.A., Eisner, J.A., Hillenbrand, L.A. \& Wolf, S. 2005, ApJ, 622, 451
812: \bibitem[Meyer et al. (2004)]{meye04}
813: Meyer, M.R., Hillenbrand, L.A., Backman, D.E. et al. 2004, ApJS, 154, 422
814: \bibitem[Meyer et al. (2006)]{meye06}
815: Meyer, M.R. et al. 2006, PASP, 118, 1690
816: \bibitem[Moro-Mart\'{\i}n \& Malhotra (2002)]{ama02}
817: Moro-Mart\'{\i}n, A. \& Malhotra, R. 2002, AJ, 124, 2305
818: \bibitem[Moro-Mart\'{\i}n \& Malhotra (2003)]{ama03}
819: Moro-Mart\'{\i}n, A. \& Malhotra, R. 2003, AJ, 125, 2255
820: \bibitem[Moro-Mart\'{\i}n, Wolf \& Malhotra (2005)]{ama05a}
821: Moro-Mart\'{\i}n, A., Wolf, S. \& Malhotra, R. 2005, ApJ, 621, 1079
822: \bibitem[Moro-Mart\'{\i}n \& Malhotra (2005)]{ama05b}
823: Moro-Mart\'{\i}n, A. \& Malhotra, R. 2005, ApJ, 633, 1150
824: \bibitem[Moro-Mart\'{\i}n et al. (2007)]{ama07}
825: Moro-Mart\'{\i}n, A., Carpenter, J.M., Meyer, M.R., Hillenbrand, L.A., Malhotra, R. et al.
826: 2007, ApJ, 658, in press
827: \bibitem[Murray \& Dermott (1999)]{murr99}
828: Murray, N. \& Dermott, S., 1999, $\it{Solar~System~Dynamics}$, Cambridge University Press.  
829: %\bibitem[Naef et al. (2001)]{naef01}
830: %Nordstrom, B., Mayor, M., Andersen, J. et al., 2004, A\&A, 418, 989
831: %\bibitem[Perrier et al. (2005)]{peri05}
832: %Perrier, C., Sivan, J.-P., Naef, D. et al. 2003, A\&A, 410, 1039
833: \bibitem[Perryman et al. (1997)]{perr97}
834: Perryman et al. 1997, A\&A, 323, L49
835: \bibitem[Raymond, Barnes \& Kaib (2006)]{raym06}
836: Raymond, S.N., Barnes, R. \& Kaib, N.A.	2006, ApJ, 644, 1223
837: \bibitem[Roques et al. (1994)]{roqu94}
838: Roques, F., Scholl, H., Sicardy, B. \& Smith, B.A. 1994, Icarus, 108, 37
839: %\bibitem[Santos et al. (2004)]{sant04}
840: %Santos, N.C., Israelian, G. \& Mayor, M.,  2004, A\&A, 415, 1153
841: \bibitem[Stapelfeldt et al. (2004)]{stap04}
842: Stapelfeldt, K.R., Holmes, E.K., Chen, C., Rieke, G.H., Su, K.Y.L. et al. 2004, ApJS, 154, 458
843: \bibitem[Stern (1996)]{ster96}
844: Stern, S. A. 1996, AJ, 112, 1203
845: \bibitem[Telesco et al. (2005)]{tele05}
846: Telesco, C. M., Fisher, R.S., Wyatt, M.C., Dermott, S.F., Kehoe, T.J. et al., Nature, 433, 133
847: %\bibitem[Udry, Mayor \& Queloz. (2003)]{udry03}
848: %Udry, S., Mayor, M. \& Queloz, D., 2003, ASP Conf. Ser. 294, 17-26.
849: \bibitem[Valenti \& Fischer (2005)]{vale05}
850: Valenti, J.A. \& Fischer, D.A. 2005, ApJS, 159, 141
851: \bibitem[Weingartner \& Draine (2001)]{wein01}
852: Weingartner, J.C. \& Draine, B.T., 2001, ApJ, 548, 296 
853: %\bibitem[Williams, Andrews \& Wilner (2005)]{will05}
854: %Williams, J.P., Andrews, S.M., \& Wilner, D.J., 2005, ApJ, 634, 495
855: \bibitem[Wilner et al. (2002)]{wiln02}
856: Wilner, D.J., Holman, M.J, Kuchner, M.J. \& Ho, P.T.P., 2002, ApJ, 569, 115
857: \bibitem[Wisdom \& Holman (1991)]{wisd91}
858: Wisdom, J., \& Holman, M. 1991, AJ, 102, 1528
859: \bibitem[Wolf \& Hillenbrand (2003)]{wolf03}
860: Wolf, S. \& Hillenbrand, L.A. 2003, ApJ, 596, 603
861: \bibitem[Wyatt (2005)]{wyat05}
862: Wyatt, M.C. 2005, A\&A, 433, 1007
863: \bibitem[Wyatt et al. (2007)]{wyat07}
864: Wyatt, M.C., Smith, R., Greaves, J.S., Beichman, C.A., Bryden, G. \& Lisse, C.M., 
865: 2007, ApJ, in press
866: \end{thebibliography}
867: 
868: 
869: \clearpage
870: 
871: %\begin{deluxetable}{lcccccc}
872: \begin{deluxetable}{llllllllllc}
873: \rotate
874: %\tablewidth{80pc}
875: \tablewidth{0pc}
876: \tablecaption{Stars with debris disks and planets\tablenotemark{a}}
877: \tablehead{
878: \colhead{Source} & 
879: %\colhead{Planet} &
880: \colhead{} &
881: \colhead{$\it{M}$sin$\it{i}$} &
882: \colhead{Period} &
883: \colhead{$\it{a}$} &
884: \colhead{$\it{e}$} &
885: \colhead{$\it{T_p}$} &
886: \colhead{$\omega$} &
887: \colhead{N$_{obs}$} &
888: \colhead{Excess\tablenotemark{b}} &
889: \colhead{Ref.}\\
890: \colhead{(HD $\#$)} & 
891: %\colhead{} &
892: \colhead{} &
893: \colhead{(M$_{Jup}$)} &
894: \colhead{(days)} &
895: \colhead{(AU)} &
896: \colhead{} &
897: \colhead{(JD-2440000)} &
898: \colhead{(deg)} &
899: \colhead{} &
900: \colhead{$\lambda$ [L$_{dust}$/L$_{*}$]} &
901: \colhead{}}
902: \startdata
903: 
904:  38529   & b    & 0.852           & 14.3093(13)   & 0.131      & 0.248(23)   & 9991.59(23) & 91.2(6.2) & 162 & 70$\mu$m [3.6x10$^{-5}$] & (1)\\
905:          & c    & 13.2            & 2165(14)      & 3.74       & 0.3506(85)  & 10085(15)   & 15.7(1.9) & 162 &  & \\
906:  33636   & b    & 9.28            & 2127.7(8.2)   & 3.27       & 0.4805(60)  & 11205.8(6.4)& 339.5(1.4)& 38  & 70$\mu$m [4.9$\times$10$^{-5}$] & (2)\\
907:  50554   & b    & 4.38            & 1223(12)      & 2.28       & 0.437(38)   & 10649(16)   & 7.9(4.3)  & 47  & 70$\mu$m [4.4$\times$10$^{-5}$] & (2)\\
908:  52265   & b    & 1.09            & 119.290(86)   & 0.504      & 0.325(65)   & 10833.7(4.2)& 243(15)   & 28  & 70$\mu$m [2.9$\times$10$^{-5}$] & (2)\\
909:  82943   & b    & 1.81            & 219.50(13)    & 0.752      & 0.39(26)    & \nodata     & 121.0(3.1)& 165 & 70$\mu$m [1.2$\times$10$^{-4}$] & (2)\\ 
910:          & c    & 1.74            & 439.2(1.8)    & 1.19       & 0.020(98)   & \nodata     & 260.0(10) & 165 & & \\
911:  117176  & b    & 7.49            & 116.6884(44)  & 0.484      & 0.4007(35)  & 7239.82(21) & 358.71(54)& 74  & 70$\mu$m [1.0$\times$10$^{-5}$] & (2)\\
912:  128311  & b    & 2.19            & 458.6(6.8)    & 1.10       & 0.25(10)    & 10211(76)   & 111(36)   & \nodata& 70$\mu$m [3.0$\times$10$^{-5}$] & (2)\\
913:          & c    & 3.22	          & 928(18)       & 1.76       & 0.170(90)   & 10010(400)  & 200(150)  & \nodata & & \\
914:  69830   & b    & 0.0322          & 8.6670(30)	  & 0.0789     & 0.100(40)   & 13496.800(60)&340(26)   & 74  & 24$\mu$m [2$\times$10$^{-4}$] & (3)\\
915:          & c    & 0.0374          & 31.560(40)	  & 0.187      & 0.130(60)   & 13469.6(2.8)& 221(35)   & 74 & & \\
916:          & d    & 0.0573          & 197.0(3.0)    & 0.633      & 0.070(70)   & 13358(34)   & 224(61)   & 74 & & \\
917:  $\epsilon$-Eri & b & 1.06        & 2500(350)     &  3.38      & 0.25(23)    & 8940(520)   & 6.15      & 120 & IR to sub-mm   &(4)\\
918: \tablenotetext{a}{Orbital parameters from Butler et al. (2006). $\it{a}$ and $\it{e}$ are the 
919: semimajor axis and eccentricity of the planet;  $\it{T_p}$ is the time of periastron passage (as a 
920: Julian day) and $\omega$ is the longitude of periastron. N$_{obs}$ is the number of radial velocity 
921: observations.  The number in parenthesis indicate the uncertainty in the last significant figures.}
922: \tablenotetext{b} {Wavelength at which the excess is detected, with the fractional luminosity of 
923: the excess in brakets. For stars with 70 $\mu$m excess, and assuming that the peak of the emission
924: is at 70 $\mu$m (T$_{dust}$ = 52.7 K), L$_{dust}$/L$_{\star}$ $\sim$ 10$^{-5}$(5600/T$_*$)$^3$(F$_{70,dust}$/F$_{70,*}$), 
925: where F$_{70,dust}$ and F$_{70,*}$ are the dust excess and photospheric flux at 
926: 70~$\mu$m and T$_*$ is the 
927: stellar temperature (Bryden et al. (\citeyear{bryd06}). 
928: For HD 69830, the fractional luminosity of the excess is calculated by integrating the excess and photospheric emission
929: beteween 7 and 35 $\mu$m.
930: References are: (1)  Moro-Mart\'{\i}n et al. (2007); 
931: (2) Beichman et al. (2005a); 
932: (3) Beichman et al. (2005b) and
933: (4) Greaves et al. (1998, 2005).}
934: \enddata
935: \end{deluxetable}
936: 
937: \clearpage
938: 
939: \begin{figure}
940: \epsscale{1.0}
941: \plotone{f1.eps}
942: \caption{Test particle orbits are planet-crossing (hence
943: unstable) in the red-shaded zone, and strongly unstable in the grey-shaded
944: zone (owing to overlapping first order mean motion resonances).  The secular
945: modes of the two planets, HD38529b and HD38529c, excite the eccentricities
946: of circular test particle orbits: the maximum eccentricity due to the fast
947: mode is shown by the green curve, while that due to the slow mode is shown
948: by the blue curve.  The inner region, interior to the outer planet,
949: is shown in the upper panel. The lower panel shows that the planetary
950: perturbations are very wide ranging: secular eccentricity excitation
951: exceeds 0.1 to nearly 60 AU; the sharp peak at 55 AU is due to a resonance
952: with the slow mode.}
953: \end{figure}
954: 
955: \clearpage
956: 
957: \begin{figure}
958: \epsscale{0.8}
959: \plotone{f2.eps}
960: \caption{Numerical simulations of 300 test particles in the HD 38529 system of two planets. The 
961: simulations include 100 particles uniformly spaced between 0.01--5 AU, 
962: and 200 particles uniformly spaced between 
963: 5--75 AU, all in initially circular orbits coplanar with the planets.
964: The angular elements 
965: were chosen randomly between 0 and $2\pi$. 
966: Particles were removed if they approached the star closer
967: than 0.005 AU or approached a planet closer than the Hill
968: radius of the planet.
969: The orbits were integrated for 200 Myr using a symplectic 
970: integrator (Wisdom \& Holman~\citeyear{wisd91}). Overall, the simulations confirm the results from the secular analysis. 
971: They differ in that the maximum eccentricity in the 20--50 
972: AU region calculated from the secular analysis is smaller than that found in the numerical
973: integrations; this is because the latter include secular and non-secular perturbations (e.g. mean motion 
974: resonances), and the test particles have non-zero initial eccentricities and inclinations.}
975: \end{figure}
976: 
977: \clearpage
978: 
979: \begin{figure}
980: \epsscale{0.8}
981: \plotone{f3.eps}
982: \caption{Orbital evolution over the 200 Myr integration period of a test particle
983: near the secular resonance. The quantity  $\it{dw}$ is the difference between the
984: longitude of periastron of the particle and that of the outermost planet.
985: The timescale to excite the eccentricity is about 50 Myr.}
986: \end{figure}
987: 
988: \clearpage
989: 
990: \begin{figure}
991: \epsscale{0.8}
992: \plotone{f4.epsi}
993: \caption{Each point of these two dimensional grids represents a modeled SED, 
994: where R$_{in}$ and M$_{dust}$ are the two free parameters. Panels (a), (b) and (c) correspond to models 
995: with a single grain size of 10 $\mu$m and with R$_{out}$ = 50 AU, 100 AU 
996: and 500 AU, respectively. Panel (d) corresponds to models with a distribution of grain
997: sizes given by $\it{n(b)}$ $\propto$ $\it{b^{-3.5}}$, with $\it{b_{min}}$ = 2 $\mu$m and 
998: $\it{b_{max}}$ = 10 $\mu$m.  The models in $\it{green}$ are those with
999: P($\chi$$^{2}$ $\mid$ $\nu$) $<$ 0.683; $\it{red}$ in panels 
1000: (a)--(c) represents models with 
1001: P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.683,
1002: i.e. models that can be excluded with 1-$\sigma$ certainty; and $\it{black}$
1003: corresponds to models with  P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.9973, i.e. models
1004: that are excluded with 3-$\sigma$ certainty. The red dots in panel (d) have 
1005: P($\chi$$^{2}$ $\mid$ $\nu$) $>$ 0.988, i.e., they are all close to be excluded
1006: with a 3-$\sigma$ certainty.}
1007: \end{figure}
1008: 
1009: \clearpage
1010: 
1011: \begin{figure}
1012: \epsscale{0.7}
1013: \plotone{f5.epsi}
1014: \caption{Observed and modeled SEDs for HD~38529. 
1015: The $\it{dotted~line}$ is the Kurucz model. The $\it{black~thick}$ 
1016: line is the IRS low-resolution spectrum. The 
1017: photometric points are identified as follows: $\it{black~circles}$ are the new Spitzer observations 
1018: (IRAC, MIPS and synthetic photometry from IRS); $\it{red~diamonds}$ are IRAS observations. In all cases, 
1019: the error bars correspond to 1-$\sigma$ uncertainties.  Upper limits are represented by $\it{triangles}$ 
1020: and are given when F/$\Delta$F $<$ 3 and placed at F + 3$\times$$\Delta$F if F $>$ 0, or 3$\times$$\Delta$F 
1021: if F  $<$ 0. The colored continuous lines in the main panel 
1022: show three sets of models that fit the observations with a 
1023: $\chi$$^{2}$ probability $<$ 0.68 (corresponding to the models represented in $\it{green}$ in
1024: Fig. 4). The models include the emission from the photosphere and from a dust 
1025: disk composed of astronomical silicate grains 10 $\mu$m in radius. 
1026: We assume the dust disk extends from R$_{in}$ to R$_{out}$ with a constant surface density. 
1027: We consider three values for R$_{out}$: 50 AU ($\it{red}$), 100 AU ($\it{blue}$) and 500 AU 
1028: ($\it{green}$). 
1029: R$_{in}$ and M$_{dust}$ are allowed to vary.  The insert at the lower left shows 
1030: the most relevant excluded models. The solid line represents models excluded
1031: with a certainty of 3-$\sigma$, while the dashed line corresponds to 1-$\sigma$. The model
1032: shown in $\it{light~blue}$ gives an upper limit to the amount of warm dust located between 0.25--0.75 AU.}
1033: 
1034: \end{figure}
1035: 
1036: \end{document}
1037: