1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass{aastex}
3: %\usepackage{emulateapj5,times,mathptm}
4: \slugcomment{Accepted for publication in the {\it Astrophysical Journal}}
5: \pagenumbering{arabic}
6:
7: \shortauthors{Mori}
8: \shorttitle{An X-ray measurement of Titan's atmospheric extent}
9:
10: \begin{document}
11: \title{An X-ray measurement of Titan's atmospheric extent from
12: its transit of the Crab Nebula}
13:
14: \author{Koji Mori}
15: \affil{Department of Astronomy and Astrophysics, Pennsylvania State
16: University, 525 Davey Laboratory, University Park, PA. 16802, USA}
17: \email{mori@astro.psu.edu}
18:
19: \author{Hiroshi Tsunemi and Haruyoshi Katayama\altaffilmark{1}}
20: \affil{Department of Earth and Space Science, Graduate School of
21: Science, Osaka University, 1-1 Machikaneyama, Toyonaka, Osaka 560-0043
22: Japan}
23:
24: \author{David N.\ Burrows and Gordon P.\ Garmire}
25: \affil{Department of Astronomy and Astrophysics, Pennsylvania State
26: University, 525 Davey Laboratory, University Park, PA. 16802, USA}
27:
28: \and
29:
30: \author{Albert E.\ Metzger}
31: \affil{Jet Propulsion Laboratory, 4800 Oak Grove Drive, Pasadena, CA
32: 91109, USA}
33:
34: \altaffiltext{1}{present address: National Space Development Agency of
35: Japan, Tsukuba Space Center, 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505,
36: Japan}
37:
38: \begin{abstract}
39:
40: Saturn's largest satellite, Titan, transited the Crab Nebula on 5
41: January 2003. We observed this astronomical event with the {\it Chandra}
42: X-ray Observatory. An ``occultation shadow'' has clearly been detected and
43: is found to be larger than the diameter of Titan's solid surface. The
44: difference gives a thickness for Titan's atmosphere of 880 $\pm$ 60 km.
45: This is the first measurement of Titan's atmospheric extent at
46: X-ray wavelengths. The value measured is consistent with or slightly
47: larger than those estimated from earlier Voyager observations at other
48: wavelengths. We discuss the possibility of temporal variations in the thickness
49: of Titan's atmosphere.
50:
51: \end{abstract}
52:
53: \keywords {planets and satellites: individual (Titan) --- X-rays:
54: general}
55:
56: \section {\label {sec:intro} INTRODUCTION}
57:
58: Titan is the only satellite in the solar system with a thick atmosphere.
59: Its atmosphere has a pressure near the surface is about 1.5 times greater than
60: that of the Earth at sea level and extends much further than that of the
61: Earth (Coustenis \& Lorenz 1999). Titan's atmosphere is known to
62: resemble the primitive environment of the Earth's atmosphere in terms of
63: chemistry, providing us with a laboratory to study the origin of life
64: (Owen et al.\ 1997). The atmospheric structure of Titan has been
65: investigated at radio (Lindal et al.\ 1983), infrared (IR) (Lellouch et
66: al.\ 1989), optical (Sicardy et al.\ 1990; Hubbard et al.\ 1990), and
67: ultraviolet (UV) wavelengths (Strobel, Summers, \& Zhu 1992; Smith et
68: al.\ 1982). These results came from the Voyager~1 spacecraft during its
69: encounter with Titan and from a stellar occultation by Titan. Due to the
70: wavelength dependence of transmissivity, the radio, IR, and optical
71: observations measure the thermal structure of Titan's atmosphere below
72: an altitude of 500 km while the UV observation measures it above an
73: altitude of 1000 km. No direct information has hitherto been obtained at
74: intermediate altitudes where the X-ray observation is effective.
75:
76: On 5 January 2003, the Saturnian system passed across the $2'$ wide
77: X-ray bright region of the Crab Nebula. The Saturnian system has a
78: conjunction with the Crab Nebula every 30 years. However, because of
79: an average offset of a few degrees, it rarely transits the Crab
80: Nebula. Although a similar conjunction occurred once in January of
81: 1296, the Crab Nebula, which is a remnant of SN1054, must have been
82: too small to be occulted. Therefore, this may be the first transit
83: since the birth of the Crab Nebula. The next similar conjunction will
84: take place in August of 2267, making its occurrence in 2003 a
85: ``once-in-a-lifetime'' event. The Saturnian system is a million times
86: brighter in visible light than the Crab Nebula whereas the Crab Nebula
87: is a million times brighter in X-rays than the Saturnian system. This
88: prevented us from observing the transit in the optical, but provided a
89: unique opportunity for an X-ray observation which had never been
90: performed. The Crab Nebula is one of the brightest synchrotron sources
91: in the sky and, thus, makes an ideal diffuse background light source
92: to study X-ray shadows of interesting objects.
93:
94: Here, we report results from an observation of this historical event
95: with the Advanced CCD Imaging Spectrometer (ACIS) aboard the {\it
96: Chandra} X-ray Observatory. {\it Chandra} has an angular resolution of
97: $0.^{\!\!\prime\prime}5$ and can resolve Saturn, its rings, and the
98: satellite Titan, whose angular size is about $1''$. Unfortunately,
99: it was not possible to observe the transit of Saturn
100: due to {\it Chandra}'s concurrent passage through the Earth's radiation
101: zone. Only Titan was observable because {\it Chandra} had passed through
102: the radiation zone by the time Titan transited 12 hours later. We
103: describe the observation in \S \ref{sec:obs}. The data analysis and
104: results are presented in \S \ref{sec:ana}. Comparison with previous
105: observations and implications of this observation will be discussed in \S
106: \ref{sec:discussion}.
107:
108:
109: \section {\label{sec:obs} OBSERVATION}
110:
111: The observation was performed from 09:04 to 18:46 (UT) on 5 January
112: 2003. Figure~\ref{fig:sky} shows the Titan transit path on the Crab
113: Nebula. In order to avoid event pile-up and telemetry saturation,
114: which made difficulties for analysis of previous {\it Chandra}
115: observations of the Crab Nebula (Weisskopf et al. 2000; Hester et
116: al. 2002), we inserted a transmission grating, shortened the CCD frame
117: time from the nominal value of 3.2 seconds to 0.3 seconds, and adopted a
118: small subarray window ($50'' \times 150''$). These observational modes
119: did work efficiently to result in no telemetry saturation and no
120: apparent event pile-up. We will carefully investigate the possibility of
121: the event pile-up effect on our result in \S \ref{sec:ana3}. Since the
122: restricted window could not cover the whole Titan transit path, we
123: changed the pointing direction twice during the observation. In spite of the
124: time loss due to the two maneuvers, we obtained an effective exposure
125: time of 32279 seconds, which corresponds to 92\% of the total observing time
126: duration.
127:
128:
129: \section {\label{sec:ana} ANALYSIS AND RESULTS}
130:
131: \subsection {Reprojection To the Titan Fixed Frame}
132:
133: No hint of the Titan transit can be seen in the standard sky image in
134: Figure~\ref{fig:sky}. Titan moved too fast to cast an observable shadow against
135: the Crab Nebula in the sky image. In order to search for a shadow, we reprojected each photon's
136: position to a frame fixed with respect to Titan. We used
137: CIAO\footnote{{\it Chandra} Interactive Analysis of Observations. See
138: http://cxc.harvard.edu/ciao/} tool {\tt sso\_freeze} for the
139: reprojection. Before applying {\tt sso\_freeze}, we removed the
140: pixel randomization, which was applied by default in the standard data
141: processing, and applied our subpixel resolution method (Tsunemi et al.\ 2001;
142: Mori et al.\ 2001)\footnote{See
143: http://asc.harvard.edu/cont-soft/software/subpixel\_resolution.1.4.html}
144: to obtain the best available spatial resolution.
145: Figure~\ref{fig:shadow} shows the {\it Titan fixed frame} image. A clear
146: shadow can be seen, which was made by Titan's occultation of the Crab
147: Nebula. Measurement of the effective radius of the occultation shadow
148: provides us with information on Titan's atmospheric extent.
149:
150: \subsection {Effective Radius of the Occultation Shadow}
151:
152: Figure~\ref{fig:data_profile} shows a radial profile of the photon
153: number density from the center of the occultation shadow (black
154: points). From this radial profile, we computed the effective radius of
155: the occultation shadow, $R_{shadow}$.
156:
157: \subsubsection{\label{sec:ana1} Method}
158:
159: We assumed that the distribution of photons in
160: Figure~\ref{fig:shadow} is symmetric with respect to the center of the
161: occultation shadow. The {\it observed} photon number density as a
162: function of the distance from the center of the occultation shadow,
163: $D_{obs}(r)$, can be given by a convolution of the {\it intrinsic}
164: photon number density, $D_{int}(r)$, and the point spread function (PSF),
165: $PSF(r)$:
166:
167: \begin{equation}
168: D_{obs}(r) = \int D_{int}(r') PSF(|{\bf r-r'}|) d{\bf r'}.
169: \label{eqn:density}
170: \end{equation}
171:
172: \noindent
173: The {\it Chandra} PSF has a very sharply peaked core with extended wings
174: and is strongly energy dependent due to larger scattering of higher
175: energy photon by the mirror surface (Chandra Proposers' Observatory
176: Guide 2002). In general, it is quite difficult to obtain a precise
177: analytical form of the PSF, which prevents us from solving
178: equation~\ref{eqn:density} to obtain $D_{int}$ directly from $D_{obs}$.
179: Instead, its numerical form can be generated through a raytrace code
180: which has been determined by the {\it Chandra} team on the basis of both
181: ground-based and on-orbit calibrations. Therefore, we performed Monte
182: Carlo simulations incorporating the numerical form of the PSF. We
183: approximated $D_{int}$ by a step function with a threshold radius,
184: $R_{disk}$. We fitted the data points in Figure~\ref{fig:data_profile}
185: with $D_{obs}(r)$, varying $R_{disk}$ to obtain the best fit.
186: $R_{shadow}$ is defined by $R_{disk}$ which gives the minimum
187: $\chi^{2}$. Justification for the approximation using the step function
188: will be discussed in \S \ref{sec:discussion}.
189:
190: The PSF was derived from raytrace tools ChaRT\footnote{{\it Chandra} Ray
191: Tracer. See http://asc.harvard.edu/chart/} and MARX\footnote{See
192: http://space.mit.edu/CXC/MARX/}. ChaRT provides the best available
193: mirror response at any off-axis angle and for any spectrum. MARX reads
194: the output of ChaRT and creates the PSF taking into account the detector
195: responses and aspect reconstruction uncertainties. We obtained the PSF
196: appropriate for the observation conditions: at off-axis angle of about
197: $45''$ and for the observed spectrum. The output PSF from MARX includes
198: the pixel randomization. Since the pixel randomization broadens PSF, we
199: sharpened it by the appropriate amount. To account for our application
200: of the subpixel method, we further sharpened the PSF by 5\% in terms of
201: the Half Power Diameter (HPD), which is the expected improvement for
202: this off-axis angle on the ACIS-S CCD (Tsunemi et al.\ 2002; Mori et
203: al.\ 2002). The HPD of the resultant PSF was $0.^{\!\!\prime\prime}844$.
204: We will discuss the validity of this PSF in \S \ref{sec:ana3}.
205:
206: In order to compare the results of the Monte Carlo simulation with the
207: data, we must also account for gradients in the sky background level and
208: for instrumental effects like trailing events. A big advantage of our
209: observation was the ability to determine the background level across the
210: shadow based on knowledge of the surface brightness of the Crab Nebula
211: along the Titan transit path. We derived the background profile across
212: the shadow by averaging two profiles which were taken centered at $3''$
213: ahead of and $3''$ behind of the occultation shadow center. The
214: background profiles are presented as red points in
215: Figure~\ref{fig:data_profile}. Then, the gradient of the background was
216: determined by fitting the red points within the shadow ($r<
217: 1.^{\!\!\prime\prime}5$) and the black points outside of it with a
218: quadratic function, which is shown with a dotted line in
219: Figure~\ref{fig:data_profile}.
220:
221: The trailing events are defined as events detected during the
222: CCD readout. Since the positions of the trailing events are recorded
223: improperly along the readout direction, they cause an offset in the
224: image. The contribution of the trailing events to the total photon number
225: density can be evaluated because its level is proportional to the
226: fraction of the total effective exposure time spent in transferring the charge
227: across the X-ray bright region, $F_{trailing}$. In the case of ACIS, it takes
228: 40 $\mu$sec to transfer the charge from one pixel to another. The width
229: of the X-ray bright region along the readout direction is about 210 pixels
230: ($\approx 103''$; see Fig.~\ref{fig:sky}). Therefore, it takes
231: 0.0084 seconds to transfer the charge
232: across the X-ray bright region. Since CCD frame time is 0.3 seconds,
233: $F_{trailing}$ becomes $0.0084/(0.0084+0.3)\approx 2.7$\%. The resultant
234: estimate for the background level due to the trailing events is shown as a dashed line in
235: Figure~\ref{fig:data_profile}.
236:
237: \subsubsection{Result}
238:
239: The first 30 data points from the center ($r< 1.^{\!\!\prime\prime}5$)
240: were used for the $\chi^{2}$ test because data points further from the
241: center hardly affected the result. Figure~\ref{fig:chi2} shows
242: $\chi^{2}$ as a function of $R_{disk}$. The $\chi^{2}$ values are well
243: described with a quadratic function. The best fit value was obtained
244: as $R_{shadow} = 0.588 \pm 0.011$ arcseconds with $\chi^{2}$\,/\,d.o.f
245: of 30.36\,/\,29. The uncertainty is 68.3\% (1$\sigma$) confidence
246: level. The best fit curve is shown in Figure~\ref{fig:data_profile}
247: (red line) along with a curve simulating $R_{shadow} =
248: 0.^{\!\!\prime\prime}437$ (blue line) which corresponds to the radius
249: of the Titan solid surface (2575 km). The difference between those two
250: curves is attributed to X-ray absorption by Titan's atmosphere.
251:
252: \subsubsection{ \label{sec:ana3} Systematic checks}
253:
254: The validity of the PSF, the effect of event pile-up, and the
255: contribution from the intrinsic background of ACIS were examined. The
256: accuracy of the adopted PSF was checked using a point
257: source in the Orion Nebula Cluster, CXOONC~J053514.0-052338, located
258: at a J2000 position of $\alpha=$ 05$^{\mathrm h}$35$^{\mathrm
259: m}$14\fs06, $\delta=-$05\arcdeg23\arcmin38\farcs4 (Getman et al.\ in
260: preparation). The point source has an off-axis angle of $45''$,
261: comparable with that of Titan in our observation. We processed the
262: data of the point source in the same manner described in \S
263: \ref{sec:ana1}; removing the pixel randomization and applying the
264: subpixel resolution method. The PSF was obtained again through
265: ChaRT/MARX, but appropriate for the spectrum of the point source, and
266: was sharpened again in the same manner described in \S \ref{sec:ana1}.
267: We confirmed that the PSF originally obtained from ChaRT/MARX had a
268: HPD 15\% larger than that of the point source, but that the PSF
269: sharpened for our data processing effects was consistent with the
270: point source data to within a few percent. Next, we studied the
271: dependency of $R_{shadow}$ on the PSF by performing simulations using
272: PSFs with 10\% larger and smaller HPDs than that of adopted one. In
273: an ideal case, even if a wrong PSF was used, $R_{shadow}$ would be
274: always the same although the $\chi^{2}$ value would increase and the
275: fit would not be statistically acceptable. In reality, $R_{shadow}$
276: depended on the HPD; $R_{shadow}$ became larger and smaller by about
277: 0.018 arcseconds which is comparable to about a 1.6 $\sigma$
278: statistical error. However, considering that the adopted PSF is
279: accurate to a few percent, we can safely assume that this dependency
280: is negligible compared with the statistical error. We note that those
281: simulations using broader and narrower PSFs resulted in higher
282: $\chi^{2}$ value by about 3 than the simulation described in \S
283: \ref{sec:ana1}, also supporting the validity of the adopted
284: PSF. Therefore, we conclude that the adopted PSF is reliable and its
285: uncertainty hardly affected the result.
286:
287: In order to examine the pile-up effect, the data were divided into two
288: time intervals according to flux level, ``high flux time interval''
289: and ``low flux time interval''. The division is indicated in
290: Figure~\ref{fig:sky}. The average background count rates in the high
291: and low flux time intervals are 8.1 and 4.0 $\times 10^{-2}$ counts
292: sec$^{-1}$ arcsec$^{-2}$, respectively. Separate calculations of
293: $R_{shadow}$ for the two data sets were performed in the same way
294: described above. Plots of $\chi^{2}$ as a function of $R_{disk}$ are
295: shown in Figure~\ref{fig:chi2}. The resultant values were $R_{shadow}=
296: 0.599 \pm 0.016$ and $0.581 \pm 0.016$ arcseconds for the high and low
297: flux time intervals, respectively. Since the values are statistically
298: consistent with each other and with the result obtained using all the
299: data, we conclude that event pile-up is unlikely to have affected the
300: result.
301:
302: The contribution of the intrinsic ACIS background cannot be a problem
303: because it is about 1.5 $\times$ 10$^{-6}$ counts sec$^{-1}$
304: arcsec$^{-2}$ (Chandra Proposers' Observatory Guide 2002). It is three
305: order of magnitude lower than that of the trailing events.
306:
307: \section {\label{sec:discussion} DISCUSSION}
308:
309: Taken with the distance of 1.214 $\times$ 10$^{9}$ km to Titan at the
310: time of this observation, $R_{shadow} = 0.588 \pm 0.011$ arcseconds
311: gives a thickness for Titan's atmosphere of $880 \pm 60$ km. This
312: value can be compared with estimates from the models for Titan's
313: atmosphere which have been constructed based on Voyager~1 observations
314: at radio, IR, and UV wavelengths. Figure~\ref{fig:comparison}a shows
315: profiles of the tangential column density (cm$^{-2}$) along the
316: line-of-sight as a function of altitude (distance above Titan's
317: surface), which are compiled from density (cm$^{-3}$) profiles
318: provided by Yelle et al.\ (1997) (red) and Vervack, Sandel, \& Strobel
319: (manuscript in preparation) (blue). Nitrogen dominates, making up more
320: than 95\% of the atmospheric constituents, and methane is the second
321: major component, although proportions slightly differ in the two
322: models. In those models, the profiles in the altitude range of
323: 500--1000 km, where no data were available, were interpolated between
324: the data of higher and lower altitudes assuming hydrostatic
325: equilibrium. The step function we used to determine the atmospheric
326: thickness is a simplification of the transmission curve of X-rays
327: through the atmosphere. In reality, X-rays suffer photoelectric
328: absorption as they pass through the atmosphere, with a transmissivity
329: determined by the atmospheric composition, the tangential column
330: density for a given altitude, and photon energy. We have calculated
331: transmission curves by using the atmospheric compositions of the above
332: two models, the tangential column densities in
333: Figure~\ref{fig:comparison}a, and integrating over the observed Crab
334: spectrum. The curves are shown in Figure~\ref{fig:comparison}b. The
335: step function with a threshold of $R_{shadow}$ is also shown.
336: Figure~\ref{fig:comparison}c shows a $\chi^{2}$ curve which is
337: identical with the black curve shown in Figure~\ref{fig:chi2}, but the
338: definitions of the x-axis are different. We compared the simulation
339: treating the atmosphere as a solid disk (disk model) with the
340: simulations based on the transmission curves shown in
341: Figure~\ref{fig:comparison}b (atmosphere model) as follows. We
342: calculated $\chi^{2}$ for the atmosphere model simulations and defined
343: the ``characteristic altitude'' $h$ for those models by
344:
345: \begin{equation}
346: h = \int_{0}^{\infty}\, \left( 1-T(h') \right) \,dh',
347: \end{equation}
348:
349: \noindent
350: where $T(h')$ is transmissivity as a function of altitude, $h'$. We
351: find that $h$ = 750 and 780 km for the Yelle et al.\ and Vervack et
352: al.\ models, respectively. The points ($h$, $\chi^{2}$) for those two
353: atmosphere models fall very close to the $\chi^{2}$ curve for the disk
354: model, as shown in Figure~\ref{fig:comparison}c. In other words, our
355: step-function transmissivity is indistinguishable from a more
356: realistic transmissivity with $h$ at the threshold of the step in the
357: simulation. This fact strongly justifies the approximation of
358: transmission curve with a step function in our simulation and suggests
359: that $R_{shadow}$ and $h$ can be directly compared with each
360: other. Then, our result for the thickness of Titan's atmosphere is
361: consistent with or slightly larger than those estimated from Voyager~1
362: observations (1.9 and 1.4 $\sigma$ separation).
363:
364: Although the statistical significance is not so large, it is still worth
365: discussing a temporal variation of Titan's upper atmosphere. Distances
366: between the Sun and Saturn were 1.42 and 1.35 $\times$ 10$^{9}$ km at
367: the time of Voyager's encounter (November 1980) and our observation
368: (January 2003), respectively. The solar luminosity is known to be fairly
369: stable with an uncertainty of 0.3\% regardless of its 11 years active
370: cycle (Fr\"{o}ehlich \& Lean 1998; Quinn \& Fr\"{o}ehlich
371: 1999). Accordingly, the closer distance to the Sun in our observation
372: might have resulted in the higher temperature and the larger extent
373: of the atmosphere. Although the size of this effect cannot be
374: ascertained without detailed modeling, the increase in solar flux
375: incident on Titan may account for some part of the slightly larger value
376: of Titan's atmosphere measured here. More detailed information will be
377: obtained by the Cassini/Huygens mission in 2005.
378:
379: Finally, we note that the spectrum taken from the shadow has no
380: absorption nor emission line feature and is statistically
381: indistinguishable from the spectrum taken from the surrounding
382: region. Those facts indicate that photons penetrating the atmosphere
383: hardly contributed to the spectrum taken from the shadow; if they did,
384: excess absorption due to Titan's atmosphere would be seen in the
385: shadow spectrum. The absence of excess absorption is reasonable
386: considering that the PSF is much broader (the HPD corresponds to about
387: 5000 km) than the corresponding atmospheric thickness interval of
388: about 500 km within which the tangential optical depth changes from
389: unity to zero (see Fig.~\ref{fig:comparison}b).
390:
391:
392: \acknowledgments
393:
394: We thank H. Marshall and S. Wolk for supporting observation planning,
395: R. Yelle and R. Vervack for providing their results, and K. Getman and
396: E. Feigelson for providing their data. K.\ M.\ and D.\ N.\ B.\ thank
397: J. Kasting for useful discussions. K.\ M.\ acknowledges the support of
398: JSPS through the fellowship for research abroad. This work was
399: supported in part by the NASA through Chandra Award GO3-4002A and was
400: carried out as a part of ``Ground-based Research Announcement for
401: Space Utilization'' promoted by the Japan Space Forum.
402:
403: \begin{thebibliography}{}
404:
405: \bibitem[Coustenis \& Lorenz 1999]{} Coustenis A., \& Lorenz, R.~D.\
406: 1999, in Encyclopedia of the Solar System, P.\ Weissman, L.\ McFadden,
407: T.\ Johnson, Eds.\ (London, Academic Press) pp.\ 377-404
408:
409: \bibitem[Chandra Proposers' Observatory Guide 2002]{} Chandra Proposers'
410: Observatory Guide, Version 5.0, 2002,
411: http://asc.harvard.edu/proposer/POG/index.html
412:
413: \bibitem[Fr\"{o}ehlich \& Lean 1998]{} Fr\"{o}ehlich, C., \& Lean, J.\ 1998,
414: Geophys. Res. Let. 25, 4377
415:
416: \bibitem[Hester et al.\ 2002]{} Hester, J.~J., Mori, K., Burrows,
417: D.~N.\ et al.\ 2002, \apjl, 577, L49
418:
419: \bibitem[Hubbard et al.\ 1990]{} Hubbard, W.~B., Hunten, D.~M., Reitsema, H.~J.,
420: Brosch, N., Nevo, Y., Carreira, E., Rossi, F., \& Wasserman,
421: L.~H.\ 1990, Nature, 343, 353
422:
423: \bibitem[Lellouch et al.\ 1989]{} Lellouch, E., Coustenis, A., Gautier,
424: D., Raulin, F., Dubouloz, N. \& Frere, C.\ 1989, Icarus, 79, 328
425:
426: \bibitem[Lindal et al.\ 1989]{} Lindal, G.~F., Wood, G.~E., Hotz, H.~B.,
427: Sweetnam, D.~N., Eshleman, V.~R., \& Tyler, G.~L.\ 1983, Icarus, 53, 348
428:
429: \bibitem[Mori et al.\ 2001]{} Mori, K., Tsunemi, H., Miyata, E., Baluta,
430: C., Burrows, D.~N., Garmire, G.~P., \& Chartas, G.\ 2001, in ASP
431: Conf.\ Ser.\ 251, New Century of X-Ray Astronomy, ed.\ H. Inoue \&
432: H. Kunieda (San Francisco: ASP), 576
433:
434: \bibitem[Owen et al.\ 1997]{} Owen, T., Raulin, F., McKay, C.~P.,
435: Lunine, J.~I., Lebreton, J.~-P., \& Matson, D.~L.\ 1997, in Huygens,
436: Payload and Mission, ESA SP-1177
437:
438: \bibitem[Quinn \& Fr\"{o}ehlich]{} Quinn, T.~J., \& Fr\"{o}ehlich,
439: C.\ 1999, Nature 401, 841
440:
441: \bibitem[Sicardy et al.\ 1990]{} Sicardy, B. et al.\ 1990, Nature, 343, 350
442:
443: \bibitem[Smith et al.\ 1982]{} Smith, G.~R., Strobel, D.~F., Broadfoot,
444: A.~L., Sandel, B.~R., Shemansky, D.~E., \& Holberg, J.~B.\ 1982,
445: J. Goephys. Res., 87, 1351
446:
447: \bibitem[Strobel et al.\ 1992]{} Strobel, D.~F., Summers, M.~E., \& Zhu,
448: X.\ 1992, Icarus, 100, 512
449:
450: \bibitem[tsunemi et al.\ 2001]{} Tsunemi, H., Mori, K., Miyata, E.,
451: Baluta, C., Burrows, D.~N., Garmire, G.~P., \& Chartas, G.\ 2001, \apj,
452: 554, 496
453:
454: \bibitem[Weisskopf et al.\ 2000]{} Weisskopf, M.~C., Hester, J.~J.,
455: Tennant, A.~F., et al.\ 2000, \apjl, 536, L81
456:
457: \bibitem[Yelle et al.\ 1997]{} Yelle, R.~V., Strobel, D.~F., Lellouch,
458: E., \& Gautier, D.\ 1997, Eur.\ Space Agency Sci.\ Tech. Rep., ESA
459: SP-1177, 243
460:
461:
462: \end{thebibliography}
463: \begin{figure}
464: \epsscale{1.0}
465: \plotone{f1.eps}
466: \caption{
467: {\it Chandra} image in equatorial coordinates with intensity
468: displayed on a logarithmic scale. The solid curve running from east to
469: west represents the Titan transit path seen from {\it Chandra}. Red
470: and white colors show the assigned ``high flux time interval'' and ``low
471: flux time interval'', respectively (see text). The observation started at
472: 09:04 and ended at 18:46 (UT) on 5 January 2003. The right bottom
473: arrow defines one arc-minute of scale. A hole at the pulsar and a
474: narrow line through the pulsar are instrumental effects.
475: \label{fig:sky}}
476: \end{figure}
477:
478: \begin{figure}
479: \epsscale{1.0}
480: \plotone{f2.eps}
481: \caption{
482: {\it Chandra} image in the frame fixed with respect to Titan.
483: Bottom and right histograms show the density profiles (counts
484: bin$^{-1}$) along a horizontal and vertical line through the shadow,
485: respectively. The image was made with $0.^{\!\!\prime\prime}246$ pixels
486: to enhance the detection of the shadow. The image size is $25'' \times
487: 25''$.
488: \label{fig:shadow}}
489: \end{figure}
490:
491: \begin{figure}
492: \epsscale{1.0}
493: \plotone{f3.eps}
494: \caption{
495: The radial profile of the photon number density from the center of the
496: occultation shadow (black points). The red curve shows the best fit of
497: the disk model simulation to the radial profile data, while the blue
498: curve shows the simulated shadow profile of Titan's solid surface. The
499: red points represent the background across the shadow. The dotted and
500: dashed lines indicate the slope of the background and the contribution
501: of the trailing events, respectively. The dotted line was determined
502: based on the red points within the shadow and black points outside of
503: it.
504: \label{fig:data_profile}}
505: \end{figure}
506:
507: \begin{figure}
508: \epsscale{1.0}
509: \plotone{f4.eps}
510: \caption{
511: $\chi^{2}$s as a function of $R_{disk}$ (points). Solid curves show
512: the fits of these points with a quadratic function. Black, red, and
513: blue colors represent the results for data sets of the entire time
514: interval, the high flux time interval, and the low flux time interval,
515: respectively.
516: \label{fig:chi2}}
517: \end{figure}
518:
519: \begin{figure}[]
520: \epsscale{1.0}
521: \plotone{f5.eps}
522: \caption{
523: (a) Tangential column density along the line-of-sight as a function of
524: altitude. Red and blue curves were compiled from density profiles
525: provided by Yelle et al.\ and Vervack et al. (b) X-ray
526: transmissivities for the observed spectrum as a function of
527: altitude. The curves are calculated from the tangential column
528: densities shown in Figure~\ref{fig:comparison}a with the same color
529: coding. The black line represents the transmissivity assumed in our
530: best fit disk model. (c) $\chi^{2}$ curve as a function of altitude
531: obtained for disk model simulations. Red and blue stars are plotted at
532: ($h$, $\chi^{2}$) for simulations with the red and blue transmissivity
533: curves shown in Figure~\ref{fig:comparison}b, respectively. $h$
534: represents the characteristic altitude (see text). The horizontal
535: dashed lines indicate 1 and 2 $\sigma$ confidence levels for the disk
536: model.
537: \label{fig:comparison}}
538: \end{figure}
539:
540: \end{document}
541:
542: