1: \documentclass[12pt,preprint]{aastex}
2:
3: %\received{}
4: %\revised{}
5: %\accepted{}
6:
7: \shortauthors{Harker et al.}
8: \shorttitle{Disks of Three Herbig Stars}
9:
10: %\received{2003 November 13}
11: \begin{document}
12:
13: \title{The Disk Atmospheres of Three Herbig Ae/Be Stars}
14:
15: \author{David E.\ Harker \altaffilmark{1}}
16: \affil{Center for Astrophysics and Space Sciences, University of California,
17: San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0424}
18:
19: \author{Charles E.\ Woodward \altaffilmark{1}} \affil{Astronomy Department,
20: School of Physics and Astronomy, 116 Church Street, S.E., University of
21: Minnesota, Minneapolis, MN 55455}
22:
23: \author{Diane H.\ Wooden \altaffilmark{1}, Pasquale Temi}
24: \affil{NASA Ames Research Center, Space Science Division, MS245-1, Moffet
25: Field, CA 94035-1000}
26:
27: \altaffiltext{1}{Visiting Astronomer, Cerro Tololo Interamerican
28: Observatory, National Optical Astronomy Observatories, operated by the
29: Association of Universities for Research in Astronomy, Inc., under
30: cooperative agreement with the National Science Foundation.}
31:
32: %%Insert a gray "draft" (with date) across page.
33: %\special{!userdict begin /bop-hook{gsave 340 50 translate
34: %20 rotate /Times-Roman findfont 30 scalefont setfont
35: %0 0 moveto 0.9 setgray (V6/02.Nov.2003) show grestore}def end}
36:
37:
38: \begin{abstract}
39:
40: We present infrared (IR) spectrophotometry ($R \simeq 180$) of three
41: Herbig Ae/Be stars surrounded by possible protoplanetary disks: HD~150193,
42: HD100546 and HD~179218. We construct a mid-IR spectral energy
43: distributions (SED) for each object by using $7.6 - 13.2$~\micron\ HIFOGS
44: spectra, 2.4 -- 45~\micron\ spectrophotometry from the {\it ISO} SWS, the
45: 12, 25, 60, and 100~\micron\ photometric points from IRAS, and for
46: HD~179218, photometric bolometric data points from the Mt.\ Lemmon
47: Observing Facility. The SEDs are modeled by using an expanded version of
48: the \citet{chigol97} two-layer, radiative and hydrostatic equilibrium, passive
49: disk. This expanded version includes the emission from Mg-pure crystalline
50: olivine (forsterite) grains in the disk surface layer. Each of the three
51: objects studied vary in the amount of crystals evident from their
52: spectrophotometry. HD~150193 contains no crystals while HD~100546 and
53: HD~179218 respectively show evidence of having crystalline silicates
54: in the surface layers of their disks. We find that the inner region of
55: HD100546 has a 37\% higher crystalline-to-amorphous silicate ratio in its
56: inner disk region ($\leq 5$~AU) compared to the outer disk region, while the
57: inner disk region of HD 179218 has a 84\% higher crystalline-to-amorphous
58: silicate ratio in its inner disk region ($\leq 5$~AU) compared to the outer
59: region. All three objects are best-fit using a grain size distribution power
60: law which falls as $a^{-3.5}$. HD~150193 is best-fit by a small disk
61: ($\sim 5$~AU in radius) while HD~100546 and HD~179218 are best fit by larger
62: disks ($\sim 150$~AU in radius). Furthermore, HD~100546's disk flares larger
63: than compared to HD~150193 (25\% more at 5~AU) and HD~179218 (80\% more at
64: 5~AU). We discuss the implications of our results and compare them with
65: other modeling efforts.
66:
67: \end{abstract}
68:
69: \keywords{stars: general}
70:
71: \section{INTRODUCTION}
72: \label{intro}
73:
74: The ubiquity of accretion disks around pre-main sequence and young main
75: sequence stars having the potential to form planetary systems is now well
76: established \citep{Koerner01}. However, unknown is an accurate estimate of
77: the fraction of single stars with disks that have produced planetary
78: systems. If not all such systems have planets, why then do only some
79: accretion disks form planets and others do not? Theoretical models of
80: particle aggregation show that if particles can grow from submicron to mm
81: to cm in size, then the formation of planetesimals is possible in the time
82: before the disk dissipates \citep{Habi99}. The problem
83: remains to understand how
84: grains condense from nebular gasses, and how relic interstellar grains
85: survive and are modified by their transport in the disk. These pristine
86: submicron size grains, be they nebular condensates or relic interstellar
87: grains, then grow into mm size particles. However, investigating how
88: grains grow is complicated by the fact that most of the grain processing
89: probably occurs in the hot mid-planes of these disks \citep{Bell2000},
90: hidden from view by the optically thick disk photospheres. If grains are
91: lofted above the photosphere by disk processes such as winds, turbulent
92: convection, or changes in vertical structure, the evolution of dust can be
93: investigated by observing the properties of the small ($\leq 1$~\micron)
94: grains in the disk surface layer or atmosphere.
95:
96: In particular, observing and modeling the 10~\micron \ silicate resonances
97: in nearby (d~$\leq 500$~pc), young (age~$\leq 5$~Myr) Herbig Ae/Be (HAEBE)
98: stars, the sites of possible on-going planetary formation \citep{WW98},
99: establishes the silicate mineralogy, Mg content of the grains, the degree
100: of crystallinity vs.\ amorphousness, grain porosity, and grain size
101: distributions. These grain properties can be linked to conditions in
102: protoplanetary disks and constrain theoretical models of protoplanetary
103: disk evolution. For example, amorphous silicates commonly detected in
104: solar system comets are considered to be interstellar relic grains
105: \citep{LiGre97} and probably are represented by Glasses Embedded with
106: Metals and Sulfides (GEMS) in interplanetary dust particles, IDPs
107: \citep{Brownl2000}. Conversely, Mg-rich crystalline silicates are thought
108: to be either pristine solar nebula condensates that condensed at $\sim
109: 1450$~K or amorphous silicates annealed into crystals at temperatures
110: $\gtrsim 1000$~K in the inner hot zones of the accretion disk. If these
111: silicate crystals are formed in the hot inner zones of the disk, then the
112: crystals must migrate to large radial distances in order to be
113: incorporated into comets and outer nebula protoplanetesimals. However,
114: the hot temperatures required to form the silicate crystals could have
115: occurred in the early phases ($\sim 300,000$~yrs) of the solar nebula in
116: the disk \citep{BM02} or in nebula shocks in the $5 - 10$~AU region,
117: obviating the need for radial transport of the crystals to comet forming
118: zones. Indeed, Mg-rich silicate crystals within IDPs are detected through
119: cometary IR spectra \citep{Wooden2000a,Wooden2000b}. Crystalline silicates
120: also have been spectroscopically identified in pre-main sequence HAEBE
121: stars including HD~100546 \citep{Malf98} and in $\beta$~Pic
122: \citep{Pant97}, a debris
123: disk system continually repopulated with dust grains via comet-comet
124: collisions and cometary accretion events. The HAEBE stars (the high mass
125: [$2 - 10$~M$_{\odot}$] analogs of T Tauri stars [$\sim 1$~M$_{\odot}$])
126: are the probable progenitors of the $\beta$-Pic stars since both have a
127: common occurrence of `exocometary activity,' are similar in spectral type,
128: and have an apparent overlap in Hipparcos ages \citep{Crifo97}.
129:
130: In this work, we examine the thermal emission from three Herbig Ae/Be
131: stars of similar spectral type: HD~150193 (A1V), HD~100546 (B9V), and
132: HD~179218 (B9). The stellar ages of the objects are: $2 \times 10^6$~yrs,
133: $10 \times 10^6$~yrs, and $0.5 \times 10^6$~yrs for HD~150193, HD~100546
134: and HD~179218, respectively \citep{vda98}. We chose these objects for study
135: because they are isolated (not belonging to any know star forming region),
136: have little to no active accretion, possess possible processing
137: circumstellar disks, and have varying degrees of silicate crystallinity
138: \citep{Meeus01}. We assemble and model SEDs for each system using a passive
139: reprocessing circumstellar disk to constrain the amount of crystalline
140: silicates in the inner and outer parts of the disk, and flaring of the disk
141: in each of the systems.
142:
143: \section{OBSERVATIONS}
144: \label{observations}
145:
146: Spectrophotometry of HD~150193, HD~100546, and HD~179218 was obtained
147: using the NASA Ames Research Center {\it HI\/}gh efficiency {\it F\/}aint
148: {\it O\/}bject {\it G\/}rating {\it S\/}pectrometer (HIFOGS:
149: \citet{Witteborn91}). The spectrum of HD~179218 was obtained on the 2.34-m
150: telescope at the Wyoming Infrared Observatory (WIRO) using 32-bit FORTH
151: telescope software. The spectra of HD~150193 and HD~100546 were obtained
152: on the 4-m Blanco telescope at the National Optical Astronomical
153: Observatories Cerro Tololo Inter-American Observatory (CTIO) in Chile.
154: Table~\ref{tab:obshf} summarizes the HIFOGS observations used in this
155: work.
156:
157: The HIFOGS is a grating spectrometer with a resolution of $\Delta \lambda
158: \simeq 0.045$~\micron\ per detector, yielding a resolving power of $R =
159: \lambda /\Delta\lambda \simeq 350 - 180$ covering the $7.5 -
160: 13.4$~\micron\ spectral region. A $3^{\prime\prime}$ diameter circular
161: entrance aperture was used for each observation at each observing
162: facility. This aperture corresponds to a radial size of 448~AU, 309~AU,
163: and 732~AU for HD~150193, HD~100546 and HD~179218, respectively. Standard
164: IR observing techniques of chopping and nodding were employed, with a chop
165: frequency of 7~Hz and with a chop throw of $30^{\prime\prime}$ in the E-W
166: direction. Absolute calibrated flux spectra of the standard stars used in
167: the data reduction were taken from the UKIRT CGS 3 and HIFOGS measurements
168: compiled by \citet{Coh96}. Ratios of two or more measurements for each
169: standard at different air masses were fitted to ratios of atmospheric
170: transmissions (using ATRAN software, \citet{Lord93}), for a precipitable
171: water vapor of 4.5~mm for both WIRO and CTIO to verify the wavelength
172: calibration and the atmospheric transmission corrections as functions of
173: airmass. Each HIFOGS spectrum was flux calibrated by dividing the
174: measured spectrum by the standard star, multiplying by the corresponding
175: star's flux spectrum \citep{Coh96}, and multiplying by the inverse ratio of
176: computed atmospheric transmission spectra. All observations presented
177: here were obtained under photometric conditions. The spectra are shown in
178: Fig.~\ref{fig:hifogs}.
179:
180: Complete Nyquist sampling of the 10~\micron\ spectral signatures requires
181: two HIFOGS grating settings. We were able to achieve Nyquist sampling for
182: the object HD~179218. The other two objects, HD~150193 and HD~100546,
183: were observed at a single grating setting. Observations of HD~179218 on
184: 1998 June 23~UT, June 27~UT, and June 30~UT were obtained at the first
185: grating setting that spanned $7.811 - 13.356$~\micron. The 1998 June
186: 24~UT and June 25~UT observations were obtained at a second grating
187: setting that was shifted such that the wavelength interval of the
188: detectors was half way between the wavelength interval of the detectors in
189: the first grating setting. The second grating setting spanned a
190: wavelength range of $7.519 - 13.122$~\micron. The spectra from each
191: observational epoch are shown in Fig.~\ref{fig:hifogs}. Spectra for each
192: grating setting were scaled to the spectrum obtained during the best
193: photometric conditions (1998 June 23~UT and 1998 June 25~UT for the first
194: and second grating settings, respectively) and combined by computing a
195: statistically weighted average. The scaling factors for each spectrum are
196: listed in Table~\ref{tab:obshf}. The resultant HIFOGS spectra for
197: HD~179218 at each grating setting were subsequently interlaced to produce
198: the final spectrum (spanning $7.528 - 13.356$~\micron). Therefore, for
199: HD~179218, there are approximately twice as many data points between
200: 7.5~\micron\ and 13.4~\micron\ than for HD~100546 and HD~150193. No
201: scaling factor was used to interlace the spectra from the two grating
202: settings.
203:
204: We also obtained 1.2 -- 23.0 \micron\ IR broad- and narrow-band
205: photometric observations of HD~179218 on 2001 April 30.42~UT using the
206: University of Minnesota (UM) Mount Lemmon Observing Facility (MLOF) 1.52~m
207: telescope and the UM multi-filter bolometer \citep{Hanner90}. Observations
208: where conducted using 9.33\arcsec\ circular aperture, a 29\arcsec\ N-S
209: nod, and a 7.5~Hz chop-frequency. In addition to standard near-IR
210: broad-band filters, the bolometer filter set also included the six
211: Infrared Telescope Facility (IRTF) narrow-band 7--13~\micron\ ``silicate''
212: filters \citep{Tokunaga86}. Multiple observations of the photometric
213: standard star $\alpha$~Boo (= HD~124897) at a variety of air masses during
214: the night established the photometric zero point over the entire spectral
215: range of the bolometer. The photometric systems, magnitude scales, and
216: absolute flux calibrations for the bolometer are given by \citet{Mason01}
217: and \citet{rdg97a,rdg97b}. Standard extinction values of 0.2
218: magnitudes per air mass (mag AM$^{-1}$) were used for filters $[J] - [L]$
219: and 7.9--12.6 \micron, 0.3~mag~AM$^{-1}$ for filter $[M]$, and
220: 0.5~mag~am$^{-1}$ at 18.3~\micron. The use of standard extinction values
221: introduces no more than a few percent uncertainty in the derived
222: photometry, as observations of HD 179218 where obtained at $\leq 1.2$ air
223: masses. Observed magnitudes corrected for atmospheric extinction, are
224: summarized in Tables~\ref{tab:obsbolo}.
225:
226: \subsection{Assembly of Data Sets for SEDs}
227: \label{sec:assembly}
228:
229: Data from the European Space Agency Infrared Space Observatory ({\it ISO})
230: Short-Wavelength Spectrometer (SWS) and the Infrared Astronomical Observatory
231: (IRAS) are used to expand the wavelength coverage of the spectral energy
232: distribution (SED) of each of our three HAEBEs.
233: The {\it ISO} SWS observations were conducted using the
234: SWS01 scanning mode (Astronomical Observing Template, referred to as AOTs)
235: covering the entire 2.4 -- 45~\micron\ SWS wavelength range.
236: Each object has been observed at a different scanning speed; in particular,
237: data have been taken at ``speed 4'' (HD100546), ``speed 3'' (HD179218),
238: and ``speed 1'' (HD150193), corresponding to raw resolution of
239: approximately $\lambda/\Delta\lambda \simeq $ 1400, 800, and 400,
240: respectively.
241:
242: Since the HAEBEs were faint and the standard pipeline process did not
243: adequately apply a dark subtraction to the data points, we have
244: reprocessed the data from a raw level to a fully processed stage using the
245: Observer's Spectral Interactive Analysis Package (OSIA v2.0) and the {\it ISO}
246: Spectral Analysis Package (ISAP v2.1). Each of the 12 bands into which the
247: SWS01 is divided have been processed separately; dark current subtraction,
248: scan direction matching and flat-fielding have been applied interactively
249: to each band. Using ISAP, we sigma clipped outliers and averaged the data
250: of all 12 detectors for each AOT band, retaining the instrument
251: resolution. For those bands affected by fringes, we applied the defringe
252: routines incorporated in ISAP.
253:
254: For extended sources a flux discontinuity may occur between band limits as
255: a consequence of a change in aperture size at some band edges. Our three
256: objects show very little flux jumps at the detector band limits, and in
257: most cases, we did not have to normalize the segments to their neighbors
258: when we combined the 12 spectral segments into a single spectrum. For
259: HD~179218 we use an additive term to normalize the last two segments
260: ($19.5 - 45$~\micron) to the flux in the previous band. Since we have {\it ISO}
261: SWS data from objects that have been taken at different scanning speeds and,
262: consequently, have different resolutions, we have produced the final
263: spectrum for each object rebinning the data to a bin size that corresponds
264: to a spectral resolution of $\lambda/\Delta\lambda \simeq 400$.
265:
266: The {\it ISO} SWS spectra are scaled to the HIFOGS spectrophotometry using
267: the following procedure: 1)~the resolution of the {\it ISO} SWS spectrum is
268: degraded to match the wavelength scale of the HIFOGS spectrum; 2)~a scaling
269: factor is calculated for each data point (approximately 30 points for
270: HD~100546 and HD~150193, and 60 points for HD~179218) in the high S/N region
271: between 8 and 10~\micron; 3)~a statistically weighted average of the
272: scaling factors is computed; and 4)~the averaged scaling factor is applied to
273: the entire original, undegraded {\it ISO} SWS spectrum. The scaling factor
274: of the {\it ISO} SWS spectra to the HIFOGS spectrophotometry for each object
275: is listed in Table~\ref{tab:sedscale}.
276:
277: In addition to spectra from {\it ISO} SWS, we use IRAS flux values
278: from \citet{Oudm92} for each of the three objects. We apply
279: the method outlined in the IRAS Explanatory Supplement (1988) to
280: produce color corrected, monochromatic flux points. We assume the IRAS fluxes
281: are larger due to the larger IRAS aperture. First, the 12~\micron\ IRAS
282: data point is scaled to the HIFOGS spectrum. Then, the resulting
283: scaling factor is applied to each of the other three IRAS data points
284: (Table~\ref{tab:sedscale}).
285:
286: Finally, we added the IR photometric points from MLOF to the SED of
287: HD~179218. The near-IR points were corrected for interstellar reddening by
288: assuming the extinction law of \citet{Mathis00} and the published
289: extinction for HD~179218 of $A_v = 1.6$. The scaling factor applied to
290: the MLOF data to match the HIFOGS spectrum of HD~179218 is listed in
291: Table~\ref{tab:sedscale}. The final assembled SEDs for all three objects
292: are shown in Fig.~\ref{fig:sed}.
293:
294: The factors needed to scale the {\it ISO} SWS data to the HIFOGS data
295: arise due to a difference in aperture size and distance to the object.
296: The effective aperture size of {\it ISO} SWS is 9.44\arcsec, and for
297: HIFOGS is 3\arcsec. Based on millimeter studies, the diameter of the disk
298: around HD~179218 is $230 - 450$~AU \citep{ms2000} and for HD~150193 is
299: $\sim 250$~AU \citep{ms1997}. The distance to HD~179218 (based on Hipparcos
300: data) is 244~pc and for HD~150193 is 149~pc. This results in projected
301: diameters of the {\it ISO} SWS and HIFOGS apertures on HD~179218 of
302: 2302~AU and 732~AU. Similarly, the projected diameters on HD~150193 are
303: 1409~AU and 448~AU. Therefore, the apertures contain the millimeter
304: measured disk sizes resulting in {\it ISO} SWS to HIFOGS scaling factors
305: close to 1 for these two objects. Conversely, the distance of HD~100546
306: is 103~pc yielding a projected aperture diameter from {\it ISO} SWS and
307: HIFOGS as 973~AU and 309~AU, respectively. Although there are no
308: millimeter measurements of HD100546, HST Chronographic Imaging reveals a disk
309: approximately 760 -- 1030~AU in diameter \citep{Auge2001}, \citep{Grad2001}.
310: Therefore, significant mid-IR flux may be contained in the {\it ISO} SWS
311: aperture which would account for the comparatively large scaling factor needed
312: to scale {\it ISO} SWS to HIFOGS for HD~100546.
313:
314: \section{Passive Disk Model}
315:
316: We use a modified version of the radiative, hydrostatic models of
317: passively irradiated circumstellar disks developed \citet{chigol97} (hereafter
318: CG97) and updated by \citet{chiang01} (hereafter C01). The CG97 disk is
319: composed of two parts: an optically thin surface layer and an optically
320: thick interior (Fig.~\ref{fig:sketch}). The surface layer is directly heated
321: by light from the
322: central star. Half of the reprocessed emission from the surface layer
323: escapes into space while the other half radiates onto the disk interior.
324: The emission from the disk surface layer heats the cooler disk interior.
325: Since the disk is in hydrostatic equilibrium, the disk flares with
326: increasing distance from the central star. The amount of disk flaring is
327: influenced by the temperature of the interior. Flaring of the disk
328: results in more starlight hitting the disk at larger distances from the
329: central star as compared to an otherwise geometrically flat disk. C01
330: updated the CG97 model by adding a range of grain sizes and incorporating
331: a simple mineralogy through the use of laboratory determined optical
332: constants, including water ice, amorphous olivine, and metallic Fe.
333: Although the two-layered disk approach is simple, when compared to more
334: detailed vertical structure models, the CG97 model was found to be a
335: robust model for computing SEDs from HAEBEs \citep{Dull02}.
336:
337: In this work, we use the updated model of C01 with our own enhanced
338: feature of adding Mg-pure crystalline olivine grains (forsterite) into the
339: optically thin surface layer. We also are able to vary the
340: crystalline-to-amorphous silicate ratio in the inner and outer radial regions
341: of the disk. The details of the standard disk model are explained in CG97 and
342: C01; however, we will highlight the disk and dust parameters here.
343:
344: Disk parameters include the disk surface density at 1~AU
345: ($\Sigma_{\circ}$), the outer disk radius ($r_{\circ}$), and the height of
346: the disk photosphere in units of the gas scale height ($H/h$). Dust grain
347: parameters include grain mineralogy, slope of the grain size distribution
348: ($q$), maximum grain size ($a_{max}$), and sublimation temperature
349: ($T_{sub}$). The shape of the SED in the region covered by our data sets
350: (2 -- 45~\micron) are largely unaffected by changes in the input
351: parameters. The dust parameter which most affects the shape of the resonance
352: features is the slope of the grain size distribution, $q$ (C01).
353: Changes in disk radius, disk surface density, and maximum
354: grain size for the interior and the surface layer mostly affect the shape
355: of the SED at wavelengths longer than 50~\micron\ (C01; Creech-Eakman et
356: al.\ 2002). The one disk parameter that produces the most change at
357: wavelengths less than 50~\micron\ is $H/h$, which can be used as a measure
358: of dust settling in the disk (C01). However, as noted in C01,
359: $H/h$ is not a fixed parameter and should be calculated self-consistently at
360: each radius in the disk. We follow the procedure of C01 by keeping $H/h$
361: constant throughout the disk and using it as an indicator of dust settling
362: in the disk. Another unknown parameter expressed in terms
363: of $H/h$ may actually represent the settling of dust in the disk.
364: Finally, the relative mass fraction of the minerals in the disk and surface
365: layer will influence the SED in the mid-IR.
366:
367: To model the emission from dust particles, we use the method of C01 and
368: assume that the disk is composed of a size distribution of isolated
369: spherical grains of discrete mineralogy; i.e., we ignore gas as a source
370: of opacity. The specific mineralogy is addressed in the next section
371: (\S\ref{sec:mins}). The opacity of the grains is computed by using Mie
372: Theory. For grains coated with a water ice mantle, Mie-G\"uttler Theory
373: is used to compute the opacity. The grain size distribution in the disk
374: surface layer and in the disk interior are equal, and unchanged in each of
375: our models. The central stars are modeled as simple blackbodies. The
376: radius, temperature, and luminosity of the central stars of our three
377: objects are taken from the compilation by \citet{vda98}. We use the
378: luminosity to calculate the radius of each of the stars using the
379: relation: $L_{\star} (\rm{ergs~cm}^{-2}~\rm{s}^{-1}) = 4\pi \sigma R_{\star}^2
380: T_{\star}^4$.
381:
382: Two final assumptions are made when using the C01 model. The first is
383: that we are viewing the disk face-on. The stars discussed here have low A$_v$
384: ($A_v = 0.28$ for HD~100546; $A_v = 1.27$ for HD~179218; and $A_v = 1.6$ for
385: HD~150193), therefore, the disk is not inclined at such an angle as to
386: significantly block the starlight. CG99 showed that until the disk is
387: inclined at such an angle as to block the starlight ($i \sim 45^{\circ}$),
388: the SED is mostly unaffected by disk inclination. Therefore,
389: we use the approximation that the disks are face on for all three of our
390: objects. The second assumption is that any accretion in these systems
391: is negligible and does not significantly contribute to the measured
392: emission from these systems. By making such an assumption, we are placing an
393: upper limit on the contribution of flux from the disk alone on the measured
394: emission from each of these systems.
395:
396: Flux from the central star, disk interior and disk surface are
397: co-added to produce the model SED. The relative amount of amorphous
398: to crystalline silicates in the disk surface layer is adjusted to produce the
399: best fit within a $2\sigma$ (95\% confidence) level \citep{Press92}.
400:
401: \subsection{Dust Mineralogy}
402: \label{sec:mins}
403:
404: The mineralogy used by C01 is a good representation of the basic
405: minerals extant in young disk systems \citep{Malf98,Bouwm2000}. We
406: maintain the mineralogy of C01, using amorphous olivine, metallic Fe, and
407: water ice. However, we have expanded the mineralogy by including Mg-pure
408: crystalline olivine grains in the disk surface layer. Crystalline olivine
409: has been discovered to be a very important component of the dust in
410: circumstellar disks and solar system comets. In an attempt to quantify
411: the amount of crystalline olivine grains, we make this important step of
412: including them in the C01 passive disk model. Table~\ref{tab:mins} lists
413: the minerals used in the modeling and the reference for the indices of
414: refraction.
415:
416: It is difficult to model the thermal emission from crystalline silicate
417: grains \citep{yfh99}. Many authors choose to use a form of a continuous
418: distribution of ellipsoids while others use direct comparisons from
419: laboratory transmission experiments. In this work, we choose to use the
420: methods outlined by \citet{Fabi01} who calculate the emission from
421: ellipsoids by elongating the crystals along one of the three crystalline
422: axes. Based on the location of the resonance peaks in the HIFOGS and {\it
423: ISO} SWS spectra of HD~150193 and HD~179218, we use an axis ratio of
424: 10:1:1 to compute the optical efficiencies (Q$_s$) of the crystals. The
425: crystalline olivines are not coated with ice since to do so requires
426: mixing theory which eliminates the distinctive crystalline resonances. We
427: compute the thermal emission from the crystals in the disk surface layer
428: of each object. We make the simplified assumption that the crystals are at
429: all radii in the disk surface layer. This assumption may not be valid
430: since ice should coat the crystals out at larger disk radii. However, the
431: measured crystalline resonances seem to originate from bare crystalline
432: grains. Finally, we vary the ratio of amorphous to crystalline silicates
433: to find the best model fit to the SEDs.
434:
435: We note that in addition to solid-state emission features from
436: silicates, HD~100546 and HD~179218 exhibit emission from polycyclic
437: aromatic hydrocarbons (PAHs) \citep{Meeus01}. We do not attempt to model
438: the emission from these molecules here, but will examine them in a future
439: paper.
440:
441: \section{RESULTS}
442: \label{sec:results}
443:
444: Our modeling results are displayed in Figs.~\ref{fig:model} and
445: \ref{fig:10model} and summarized in Table~\ref{tab:fitparams}. To produce the
446: best-fit model to the SEDs, HD~100546 and HD~179218 are both modeled with a
447: disk 150~AU in radius, while HD~150193 is modeled with a disk 5~AU in radius.
448: The best fit SEDs for
449: HD~100546 uses a scale height of $H/h = 4$, a scale height of $H/h = 3$ for
450: HD~150193, and a scale height of $H/h = 1$ for HD~179218. This means that
451: at a radius of 5~AU the disk around HD~100546 flares about 80\% more than the
452: disk around HD~179218, and about 25\% more than the disk around HD~150193.
453: The current model fits suggests that the measured emission from
454: HD~150193 is dominated by warm dust in a flaring disk close to the
455: central star, and that the emission from grains in an extended disk ($> 5$~AU)
456: are relatively small compared to the other two objects. We should note that
457: for HD~150193, a larger disk of radius 50~AU with a scale height of $H/h = 1$
458: can be used to match the contrast of the measured silicate resonance features
459: in HD~150193, but the calculated flux at longer wavelengths exceeds the
460: measured IRAS data point at 60~\micron\ by a factor of 4. Therefore, we are
461: able to match the long wavelength photometry and the contrast of the silicate
462: resonance features by reducing the radius of the disk and increasing the scale
463: height.
464:
465: The qualitative crystalline content of these three HAEBEs has previously been
466: reported by \citet{Meeus01} with HD~100546 and HD~179218 showing evidence of
467: crystalline silicates, while the emission from HD~150193 is dominated by
468: amorphous silicates. For the two
469: objects with crystalline silicates, we find that a better fit is produced to
470: the observed SEDs of the objects if we use a model in which a higher ratio of
471: crystalline-to-amorphous silicates is located in the inner regions of the disk
472: ($\leq 5$~AU) compared to the outer regions of the disk (5 -- 150 AU). The
473: inner region of HD~100546 has 37\% higher crystalline-to-amorphous silicate
474: ratio compared to the outer region. This is contrary to the findings of
475: \citep{Bouwm2003} (hereafter, BdKDW) who used a spherical shell model to
476: calculate a factor of almost 10 higher fraction of crystalline silicates in
477: regions greater than 10 AU. The inner region of HD~179218 has 84\% higher
478: crystalline-to-amorphous silicate ratio compared to the outer region.
479: Our model results are consistent with the recent findings of \citet{vanb2004}.
480: \citet{vanb2004} observed three Herbig Ae stars (HD~142527, HD~163296, and
481: HD~144432) using the Mid-Infrared Interferometric Instrument on the Very
482: Large Telescope Interferometer. In all three of their objects, they found
483: evidence for a larger fraction of crystalline silicates in closer to the star
484: (1 -- 2~AU) compared to the outer region of the disk (2 -- 20~AU).
485:
486: Finally, all three of our objects are best fit using a grain size distribution
487: with a slope of $q = 3.5$ for both the optically thick inner disk and for the
488: optically thin surface layer.
489:
490: \section{DISCUSSION}
491: \label{sec:discussion}
492:
493: It is difficult to make any statistically significant conclusions about disk
494: evolution from modeling only three objects. However, we can make some
495: interesting observations based on our results. \citet{Meeus01} defined
496: two groups of Herbig Ae/Be stars based on the shape of their SEDs. Group I
497: objects exhibit near-IR and far-infrared (far-IR) emission, with and without
498: silicate emission features (denoted Groups Ia and Ib, respectively). Group II
499: objects exhibit near-IR emission and silicate emission features, but much less
500: far-IR emission compared to the objects in Group~I. HD~179218 and HD~100546
501: are considered Group Ia objects and HD~150193 is considered a Group II object
502: under this scenario \citep{Meeus01}.
503:
504: \citet{Meeus01} suggest a disk geometry interpretation of their groupings.
505: Group I objects have a geometrically thin, optically thin
506: inner region with a flaring outer region. Group II objects contain an
507: optically thick inner region which puffs up, shielding the optically thin, low
508: mass, outer disk
509: region from stellar flux, thereby preventing it from flaring.
510: Qualitatively our modeling results match these groupings. HD~150193 is
511: modeled with a relatively small (5~AU), low mass disk, which suppresses
512: the amount of emission from ice coated silicate grains contributing to the
513: emission around 45~\micron. HD~179218 and HD~100546 are modeled with
514: relatively large disks. HD~100546 is modeled with a disk that flares more
515: than that of HD~150193 and HD~179218. We find it interesting that the
516: disk of HD~179218 flares less than the disk of HD~100546, even though the
517: two are both Group I sources. At least on the quantitative level, the two
518: disks differ.
519:
520: Although our model fits show no evidence for grain growth (all three SEDs are
521: modeled with the same grain size distribution) there is a difference in the
522: radial distribution of crystalline silicates. A possible evolutionary
523: scenario is one in which when circumstellar disks form around young stars,
524: they are thought to be primarly composed of amorphous ISM grains
525: \citep{Wood2004}. As the disk evolves, crystalline silicates are condensed
526: \citep{Gros1972} or annealed \citep{Riet2002} in the inner radial regions of
527: the disk, either through heating \citep{BM02} or through shocks in the
528: disk \citep{Hark2002}. The crystals are then transported to the outer
529: regions of the disk \citep{Cuzzi93}. Therefore, from this scenario, we
530: can conclude that the disk around HD150193 is the least evolved since it does
531: not contain any crystals. This is followed by HD179218 which has a large
532: crystalline-to-amorphous in the inner radial region compared to the outer
533: radial region. Finally, this is followed by HD100546 which has a slightly
534: lower crystalline-to-amorphous silicate ratio at inner disk radii and a
535: larger ratio at large disk radii compared to
536: HD179218. Such a scenario is supported by the findings of \citet{vanB2003}
537: who find that the shape and strength (band over continuum) of the 10~\micron\
538: silicate feature are correlated. Strong features (from small grains)
539: show mostly amorphous type silicate grains, while weaker, flatter features
540: show evidence of crystalline silicates (grain processing).
541:
542: It should be noted that there may be evidence of grain growth in both HD~179218
543: and HD~100546 as evidenced by the flatness of their 10~\micron\ silicate
544: features. We have not been able to adequately model ``by eye'' the shape of
545: the feature with the current model. Attempts to use a more shallow grain
546: size distribution and/or truncate the lower end of the grain sizes did not
547: result in better fits (either by eye, or statistically) to the 10~\micron\
548: feature. A more diverse mineralogy combined with a radially varying grain
549: size distribution may improve the fits to the 10~\micron\ spectral feature.
550: However, such models are beyond the scope of this paper.
551:
552: \subsection{Comparison with Other Modeling Efforts}
553:
554: Other authors have modeled the three objects presented here.
555: As stated earlier, BdKDW modeled the SED of HD~100546 using
556: a optically thin spherically symmetric dust distribution (i.e., no optically
557: thick disk component). BdKDW find that
558: the the amount of crystals used to fit the SED increased with radial
559: distance from the star. This contrasts with our model which shows that
560: there needs to be a higher crystalline-to-amorphous silicate ratio in
561: closer to the star, in agreement with radial mixing models. BdKDW
562: also finds that HD~100546 has a large, flaring disk, consistent with our
563: model results.
564:
565: \citet{Domi2003}, using another variation of the CG97 model \citep{Dull01}
566: (hereafter DDN), used a simple mineralogy of grains 0.1~\micron\ in size.
567: They modeled HD~150193 with a disk smaller than that of the other two and
568: found that the emission feature in HD~100546 is best fit with larger sized
569: grains than the other two objects. While we can not conclusively make a
570: quantitative statement about the grain size in HD~100546, our results are
571: qualitatively similar with \citet{Domi2003} for HD~150193. However, there are
572: some important differences between our model results and those of
573: \citet{Domi2003}. They find that their fits for these three objects are
574: best when the surface density increases with distance from the star
575: (instead of decreasing as we have assumed here). They interpret these
576: results as evidence for a ``gap'' in their disks, possibly produced by a
577: planet or other large body. We do not find such evidence for an
578: increasing surface density in our model calculations. Furthermore, based
579: on DDN, it is extremely difficult to disentangle the degeneracy of the DDN
580: model when simultaneously varying the surface density, the slope of the
581: surface density with stellar distance, and the radial size of the disk.
582: Therefore, while both of our results are intriguing, it underlines the
583: need (as pointed out in Dominik) for high spatial resolution imaging of
584: these objects in the mid-IR to resolve these modeling differences.
585:
586: \section{CONCLUSIONS}
587: \label{sec:conclusions}
588:
589: In this paper we constructed mid-IR SEDs of three Herbig Ae/Be stars using
590: data from the HIFOGS, {\it ISO} SWS, IRAS, and MLOF photometry. We modeled
591: the SEDs using the simple two-layer radiative transfer disk model of CG97
592: and C01 with the added feature of calculating the emission from Mg-pure
593: crystalline olivine grains. Our findings are highlighted as follows:
594:
595: 1)~HD~150193 has thermal emission arising from a relatively small
596: ($\simeq 5$~AU in radius) flared ($H/h = 3$) disk.
597:
598: 2) HD~179218 has thermal emission arising from a large ($\simeq 150$~AU
599: in radius) less flared ($H/h = 1$) disk.
600:
601: 3) HD~100546 has thermal emission arising from a large ($\simeq 150$~AU
602: in radius) more flared ($H/h = 4$) disk. At 5~AU, the disk around HD~100546
603: flares 80\% more than the disk around HD~179218 and 25\% more than HD~150193.
604:
605: 4) All three objects were modeled using a power law grain size distribution
606: with a slope of $a^{-3.5}$ for both the optically thick inner region and
607: the optically thin surface layer.
608:
609: 5) HD~150193 shows no evidence of emission from crystalline silicates
610: while HD~179218 and HD~100546 both show emission from Mg-rich crystalline
611: olivine grains.
612:
613: 6) The SED of HD~179218 was best fit using a crystalline-to-amorphous
614: ratio 84\% larger in the inner radial regions of the disk ($\leq 5$~AU)
615: compared to the outer regions.
616:
617: 7) The SED of HD~100546 was best fit using a crystalline-to-amorphous
618: ratio 37\% larger in the inner radial regions of the disk ($\leq 5$~AU)
619: compared to the outer regions.
620:
621: In circumstellar disks, grain growth and crystallization may require both
622: sufficient disk mass (i.e., size) and scale height. Scale height may be
623: a signature of turbulent movement in the disk which could aid grain growth
624: and transport \citep{Cuzzi93}. More disks need to be fitted with similar
625: model structures to increase the sample to make more sense of the links
626: between grain properties and disk structure.
627:
628: \acknowledgements
629: DEH and CEW acknowledge partial support for this work from from NSF Grant
630: AST-0205814. In addition, DEH and DW acknowledge support from the NASA Ames
631: Research Center Director's Discretionary Fund. The authors thank
632: the day-crew and staff of NOAO CTIO for their assistance at the Blanco 4-m
633: (in particular Ron Probst) and James E. Lyke for his assistance with the
634: UM MLOF observations. Finally, the authors wish to thank an anonymous
635: referee for providing useful comments on the paper.
636:
637:
638: \newpage
639: %% cew rearranges citations for e-citing 03.Nov.2003
640:
641: \begin{thebibliography}{}
642:
643: \bibitem[Augereau, Lagrange, Mouillet, \& M{\'e}nard(2001)]{Auge2001}
644: Augereau, J.~C., Lagrange, A.~M., Mouillet, D., \& M{\' e}nard, F.\ 2001,
645: \aap, 365, 78
646:
647: \bibitem[Bell et al.(2000)]{Bell2000}
648: Bell, K.R., Cassen, P.M., Wasson, J.T., Woolum, D.S.\ 2000, in
649: Protostars and Planets IV, ed. V. Mannings, A.P. Boss, \& S.S. Russell,
650: University of Arizona Press, Tucson, 897
651:
652: \bibitem[Bouwman, de Koter, Dominik, \& Waters(2003)]{Bouwm2003}
653: Bouwman, J., de Koter, A., Dominik, C., \& Waters, L.~B.~F.~M.\ 2003, \aap,
654: 401, 577
655:
656: \bibitem[Bouwman et al.(2000)]{Bouwm2000}
657: Bouwman, J., de Koter, A., van den Ancker, M.~E., \& Waters,
658: L.~B.~F.~M.\ 2000, \aap, 360, 213
659:
660: \bibitem[Brockl\'ee-Morvan et al.(2002)]{BM02} Bockel{\'
661: e}e-Morvan, D., Gautier, D., Hersant, F., Hur{\' e}, J.-M., \& Robert, F.\
662: 2002, \aap, 384, 1107
663:
664: \bibitem[Brownlee et al.(2000)]{Brownl2000}
665: Brownlee, D.E., Joswiak, D.J., Bradley, J.P., Gezo, J.C., Hill,
666: H.G.M.\ 2000, Lunar Planet.\ Inst.\ 31, Abs.\ No.\ 1921
667:
668: \bibitem[Chiang \& Goldreich(1997)]{chigol97}
669: Chiang, E.I.\ \& Goldreich, P.\ 1997, \apj, 490, 368 (CG97)
670:
671: \bibitem[Chiang et al.(2001)]{chiang01} Chiang, E.I.,
672: Joung, M.K., Creech-Eakman, M.J., Qi, C., Kessler, J.E.,
673: Blake, G.A., van Dishoeck, E.F.\ 2001, \apj, 547, 1077 (C01)
674:
675: \bibitem[Cohen et al.(1996)]{Coh96}
676: Cohen, M., Witteborn, F.C., Carbon, D.F., Davies, J.K., Wooden, D.H.,
677: \& Bregman, J.D. 1996, \aj, 112, 2274
678:
679: \bibitem[Crifo et al.(1997)]{Crifo97}
680: Crifo, F., Vidal-Madjar, A., Lallement, R., Ferlet, R., \& Gerbaldi, M.\
681: 1997, \aap, 320, L29
682:
683: \bibitem[Cuzzi, Dobrovolskis, \& Champney(1993)]{Cuzzi93}
684: Cuzzi, J.~N., Dobrovolskis, A.~R., \& Champney, J.~M.\ 1993, Icarus, 106,
685: 102
686:
687: \bibitem[Dominik et al.(2003)]{Domi2003}
688: Dominik, C., Dullemond, C.~P., Waters, L.~B.~F.~M., \& Walch, S.\ 2003,
689: \aap, 398, 607
690:
691: \bibitem[Dorschner et al.(1995)]{Dorsc95}
692: Dorschner, J., Begemann, B., Henning, T., Jaeger, C., \& Mutschke, H.\ 1995,
693: \aap, 300, 503
694:
695: \bibitem[Dullemond, Dominik, \& Natta(2001)]{Dull01}
696: Dullemond, C.~P., Dominik, C., \& Natta, A.\ 2001, \apj, 560, 957
697:
698: \bibitem[Dullemond, van Zadelhoff, \& Natta(2002)]{Dull02}
699: Dullemond, C.P., van Zadelhoff, G.J., \& Natta, A.\ 2002, \aap, 389, 464
700:
701: \bibitem[Fabian et al.(2001)]{Fabi01}
702: Fabian, D., Henning, T., J{\" a}ger, C., Mutschke, H., Dorschner, J.,
703: Wehrhan, O.\ 2001, \aap, 378, 228
704:
705: \bibitem[Gehrz(1997a)]{rdg97a} Gehrz, R. D. 1997a, in The International
706: Comet Quarterly, ed. D. W. E. Green, April 1997
707:
708: \bibitem[Gehrz(1997b)]{rdg97b} Gehrz, R. D. 1997b, in The International
709: Comet Quarterly Guide to Observing Comets, ed. D. W. E. Green,
710: (Cambridge: Smithsonian Astrophysical Observatory), p117
711:
712: \bibitem[Grady et al.(2001)]{Grad2001}
713: Grady, C.~A., et al.\ 2001, \aj, 122, 3396
714:
715: \bibitem[Grossman(1972)]{Gros1972}
716: Grossman, L.\ 1972, \gca, 36, 597
717:
718: \bibitem[Habing et al.(1999)]{Habi99} Habing, H.J.,
719: et al. 1999, Nature, 401, 456
720:
721: \bibitem[Hanner et al.(1990)]{Hanner90} Hanner, M. S., Newburn, R. L.,
722: Gehrz, R. D., Harrison, T. E., Ney, E. P., \& Hayward, T. L. 1990,
723: \apj, 348, 312
724:
725: \bibitem[Harker \& Desch(2002)]{Hark2002}
726: Harker, D.~E.~\& Desch, S.~J.\ 2002, \apjl, 565, L109
727:
728: \bibitem[IRAS(1988)]{iras88} IRAS, The
729: Explanatory Supplement, 1988, US Government Publication Office
730:
731: \bibitem[Jager et al.(1998)]{Jag98}
732: Jaeger, C., Molster, F.~J., Dorschner, J., Henning, T., Mutschke, H., \&
733: Waters, L.~B.~F.~M.\ 1998, \aap, 339, 904
734:
735: \bibitem[Koerner(2001)]{Koerner01} Koerner, D.~W.\ 2001, ASP
736: Conf.~Ser.~231: Tetons 4: Galactic Structure, Stars and the Interstellar
737: Medium, eds. C.E. Woodward and M.D. Bicay [ASP: San Fransico], 563
738:
739: \bibitem[Li \& Greenberg(1997)]{LiGre97} Li, A.~\&
740: Greenberg, J.~M.\ 1997, \aap, 323, 566
741:
742: \bibitem[Lord(1993)]{Lord93} Lord, S. D. 1993, NASA
743: Technical Report TM--103957, (Moffett Field: NASA/Ames Research Center)
744:
745: \bibitem[Malfait et al.(1998)]{Malf98} Malfait, K., Waelkens, C.,
746: Waters, L. B. F. M., Vandenbussche, B., Huygen, E.,
747: de Graauw, M. S. 1998, \aap, 332, L25
748:
749: \bibitem[Mannings \& Sargent(1997)]{ms1997} Mannings, V.\ \&
750: Sargent, A.I.\ 1997, \apj, 490, 792
751:
752: \bibitem[Mannings \& Sargent(2000)]{ms2000} Mannings, V.\ \&
753: Sargent, A.I.\ 2000, \apj, 529, 391
754:
755: \bibitem[Mason et al.(2001)]{Mason01} Mason, C. G., Gehrz, R. D.,
756: Jones, T. J., Woodward, C. E., Hanner, M. S., \& Williams, D. M. 2001,
757: \apj, 549, 635
758:
759: \bibitem[Mathis(2000)]{Mathis00}
760: Mathis, J.~S.\ 2000, \jgr, 105, 10269
761:
762: \bibitem[Meeus et al.(2001)]{Meeus01}
763: Meeus, G., Waters, L. B. F. M., Bouwman, J., van den Ancker, M. E.,
764: Waelkens, C., Malfait, K. 2001, \aap, 365, 476
765:
766: \bibitem[Oudmaijer et al.(1992)]{Oudm92} Oudmaijer, R.~D., van
767: der Veen, W.~E.~C.~J., Waters, L.~B.~F.~M., Trams, N.~R., Waelkens, C., \&
768: Engelsman, E.\ 1992, \aaps, 96, 625
769:
770: \bibitem[Pantin, Lagage, \& Artymowicz(1997)]{Pant97} Pantin, E., Lagage, P.O., \& Atrymowicz, P. 1997, \aap, 372, 1123
771:
772: \bibitem[Pollack et al.(1994)]{Poll94}
773: Pollack, J.~B., Hollenbach, D., Beckwith, S., Simonelli, D.~P., Roush, T., \&
774: Fong, W.\ 1994, \apj, 421, 615
775:
776: \bibitem[Press et al.(1992)]{Press92}
777: Press, W.H., Teukolsky, S.A., Vetterling, W.T., \& Flannery, B.P.\ 1992,
778: Numerical Recipes (Cambridge Univ. Press)
779:
780: \bibitem[Rietmeijer et al.(2002)]{Riet2002} Rietmeijer,
781: F.~J.~M., Hallenbeck, S.~L., Nuth, J.~A., \& Karner, J.~M.\ 2002, Icarus, 156,
782: 269
783:
784: \bibitem[Tokunaga et al.(1986)]{Tokunaga86} Tokunaga, A. T., Golisch, W.
785: P., Griep, D. M., Kaminski, C. D., \& Hanner, M. S. 1986, \aj , 92, 1183
786:
787: \bibitem[van Boekel et al.(2004)]{vanb2004}
788: van Boekel, R., et al.\ 2004, \nat, 432, 479
789:
790: \bibitem[van Boekel et al.(2003)]{vanB2003} van Boekel, R.,
791: Waters, L.~B.~F.~M., Dominik, C., Bouwman, J., de Koter, A., Dullemond,
792: C.~P., \& Paresce, F.\ 2003, \aap, 400, L21
793:
794: \bibitem[van den Ancker et al.(1998)]{vda98}
795: van den Ancker, M.E., de~Winter, D., Tjin A Djie, H.R.E. 1998, \aap, 330,
796: 145
797:
798: \bibitem[Waters \& Waelkens(1998)]{WW98}
799: Waters, L.B.F.M.\ \& Waelkens, C.\ 1998, \araa, 36, 233
800:
801: \bibitem[Witteborn et al.(1991)]{Witteborn91}
802: Witteborn, F.C., Bregman, J.D., Rank, D.M., \& Cohen, M. 1991, in {\it
803: Proc. of the 1991 North American Workshop on Infrared Spectroscopy},
804: ed. R.E. Stencel (Univ. Colorado Press: Boulder], 29
805:
806: \bibitem[Wooden, Woodward, \& Harker(2004)]{Wood2004}
807: Wooden, D.~H., Woodward, C.~E., \& Harker, D.~E.\ 2004, \apjl, 612, L77
808:
809: \bibitem[Wooden et al.(2000a)]{Wooden2000a}
810: Wooden, D.H., Butner, H.M., Harker, D.E., \& Woodward, C.E.\ 2000a,
811: Icarus, 143, 126
812:
813: \bibitem[Wooden, Harker, \& Woodward(2000b)]{Wooden2000b} Wooden,
814: D.~H., Harker, D.~E., \& Woodward, C.~E.\ 2000b, ASP Conf.~Ser.~196: Thermal
815: Emission Spectroscopy and Analysis of Dust, Disks, and Regoliths,
816: Eds. A. Sprague and M. Sitko, [ASP: San Fransico], 99
817:
818: \bibitem[Yanamandra-Fisher \& Hanner(1999)]{yfh99}
819: Yanamandra-Fisher, P., \& Hanner, M. S. 1999, Icarus, 138, 107
820:
821: \end{thebibliography}
822:
823: \clearpage
824:
825: %Table 1
826: \input{tab1}
827:
828: %Table 2
829: \input{tab2}
830:
831: %Table 3
832: \input{tab3}
833:
834: % Table 4
835: \input{tab4}
836:
837: % Table 5
838: \input{tab5}
839:
840: %Table 6
841: \input{tab6}
842:
843: \clearpage
844:
845:
846: %Figure 1
847: \begin{figure}
848: \epsscale{0.5}
849: \plotone{f1.eps}
850: \caption{HIFOGS spectra of HD~150193, HD~100546, and HD~179218.
851: \label{fig:hifogs} }
852: \end{figure}
853:
854: %Figure 2
855: \begin{figure}
856: \epsscale{0.8}
857: \plotone{f2.eps}
858: \caption{Assembled SEDs for $a)$~HD~150193, $b)$~HD~100546, and
859: $c)$~HD~179218. Scaled to the HIFOGS spectrum
860: ({\it open circles}) for each SED is the {\it ISO} SWS spectrum
861: ({\it black line}), and the IRAS photometric points
862: ({\it gray squares}) corresponding to each object. The 100~\micron\
863: point for HD~150193 (panel $[a]$) is an upper limit and not used in the model
864: fitting.
865: \label{fig:sed} }
866: \end{figure}
867:
868: % Figure 3
869: \begin{figure}
870: \epsscale{0.8}
871: \plotone{f3.eps}
872: \caption{Cross-sectional schematic of our disk model, adapted from Fig.\ 1
873: in CO1. Identified in the sketch is the relative location of each mineral
874: species. The {\it vertical lines} indicate the radial zones in the disk in
875: which a particular mineral dominates the emission, and the {\it dashed lines}
876: indicate the separation between the disk interior and the disk surface
877: layer. The disk interior and disk surface layer ({\it hash lines}) are
878: indicated.
879: \label{fig:sketch} }
880: \end{figure}
881:
882: % Figure 4
883: \begin{figure}
884: \epsscale{0.75}
885: \plotone{f4.eps}
886: \caption{C01 model computed for three Herbig Ae stars:
887: HD 150193, HD 100546 and HD179218. The disk interior (blue line), disk
888: surface (red line), and stellar blackbody (orange line) are co-added to
889: produce the model SED (black line). The model SED is compared to the
890: assembled data sets including: ISO SWS spectra (indigo line), HIFOGS spectra
891: (green circles) and IRAS photometry points (turquoise squares). HD 179218
892: also has MLOF photometry points (brown circles).
893: \label{fig:model} }
894: \end{figure}
895:
896: % Figure 5
897: \begin{figure}
898: \epsscale{0.50}
899: \plotone{f5.eps}
900: \caption{Zoom in on the C01 model fits of the crystalline olivine features for
901: the three Herbig stars: HD 150193, HD 100546 and HD179218. The colors are the
902: same as in Fig.~\ref{fig:model}.
903: \label{fig:10model} }
904: \end{figure}
905:
906: \end{document}
907:
908: