1:
2: %% The first piece of markup in an AASTeX v5.0 document
3: %% is the \documentclass command. LaTeX will ignore
4: %% any data that comes before this command.
5:
6: %% The command below calls the default manuscript style,
7: %% which will produce a double-spaced document on one column.
8: %% Examples of commands for other substyles follow. Use
9: %% whichever is most appropriate for your purposes.
10:
11: %\documentclass{aastex}
12:
13: %% preprint produces a one-column, single-spaced document:
14:
15: \documentclass[twocolumn]{emulateapj}
16: \usepackage{lscape}
17: %\documentclass{emulateapj}
18:
19: %% preprint2 produces a double-column, single-spaced document:
20:
21: %\documentclass[preprint2]{aastex}
22:
23: %% You can insert a short comment on the title page using the command below.
24:
25: \slugcomment{Accepted for publication in PASP}
26:
27: %% If you wish, you may supply running head information, although
28: %% this information may be modified by the editorial offices.
29: %% The left head contains a list of authors,
30: %% usually a maximum of three (otherwise use et al.). The right
31: %% head is a modified title of up to roughly 44 characters. Running heads
32: %% will not print in the manuscript style.
33:
34: \shorttitle{Closure Phase of Planet Transit Events}
35:
36: \shortauthors{van Belle}
37:
38: %% This is the end of the preamble. Indicate the beginning of the
39: %% paper itself with \begin{document}.
40:
41: \begin{document}
42:
43: %% LaTeX will automatically break titles if they run longer than
44: %% one line. However, you may use \\ to force a line break if
45: %% you desire.
46:
47: \title{Closure Phase Signatures of Planet Transit Events}
48:
49: %% Use \author, \affil, and the \and command to format
50: %% author and affiliation information.
51: %% Note that \email has replaced the old \authoremail command
52: %% from AASTeX v4.0. You can use \email to mark an email address
53: %% anywhere in the paper, not just in the front matter.
54: %% As in the title, you can use \\ to force line breaks.
55:
56:
57: \author{Gerard T. van Belle\altaffilmark{1}}
58: \affil{European Southern Observatory, Garching, Germany 85748\\
59: gerard.van.belle@eso.org}
60:
61:
62: %% Notice that each of these authors has alternate affiliations, which
63: %% are identified by the \altaffilmark after each name. Specify alternate
64: %% affiliation information with \altaffiltext, with one command per each
65: %% affiliation.
66:
67: \altaffiltext{1}{For preprints, please email gerard.van.belle@eso.org.}
68:
69: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
70: %% style, abstract will output a Received/Accepted line after the
71: %% title and affiliation information. No date will appear since the author
72: %% does not have this information. The dates will be filled in by the
73: %% editorial office after submission.
74:
75: \begin{abstract}
76:
77: Planet transit events present as attractive targets for the ultra-high-resolution capabilities afforded by optical interferometers. Herein is presented an evaluation of the possibility of detection of such events through measurement of high-precision closure phases with the MIRC instrument on the CHARA Array. Recovery of the transit position angle upon the sky appears readily achievable with the existing capabilities of the instrument, along with characterization of other system parameters, such as stellar radius, planet radius, and other parameters of the transit event. This technique is the only one presently available that can provide a transiting planet's orbital plane position angle, and can directly determine the planet's radius independent of any outside observations, appearing able to improve substantially upon other determinations of that radius. Additional directly observed parameters - also not dependent upon transit photometry or spectroscopy - include impact parameter, transit ingress time, transit velocity and stellar radius.
78:
79: \end{abstract}
80:
81: %% Keywords should appear after the \end{abstract} command. The uncommented
82: %% example has been keyed in ApJ style. See the instructions to authors
83: %% for the journal to which you are submitting your paper to determine
84: %% what keyword punctuation is appropriate.
85:
86: \keywords{techniques: interferometric, techniques: high angular resolution,
87: stars: planetary systems, stars: individual: HD189733}
88:
89: %% From the front matter, we move on to the body of the paper.
90: %% In the first two sections, notice the use of the natbib \citep
91: %% and \citet commands to identify citations. The citations are
92: %% tied to the reference list via symbolic KEYs. The KEY corresponds
93: %% to the KEY in the \bibitem in the reference list below. We have
94: %% chosen the first three characters of the first author's name plus
95: %% the last two numeral of the year of publication as our KEY for
96: %% each reference.
97:
98: \section{Introduction}
99:
100: Recent discoveries of stars exhibiting the telltale signs of
101: planet transits has begun to add a new layer of understanding to
102: the rapidly developing field of exoplanetology. While the
103: technique of radial velocity detection has produced the greatest
104: yield to date of planet detections \citep{2006ApJ...646..505B}, the detected transit
105: events have served to define the specific nature of those planets,
106: including parameters such as density, atmospheric composition, and
107: aspects of system dynamics \citep{2006ApJ...650.1140B}.
108:
109: Advances in the state of the art in astronomical optical interferometry
110: can be directed at these recent transit discoveries and also contribute
111: to filling in the pieces of the exoplanet puzzle. Specifically, measurements
112: of interferometric closure phase during a planet transit event can determine the inclination and
113: orientation of the planetary
114: orbit upon the sky, in addition to refining the angular diameter measurements
115: of both the planet and the star.
116: Just as the Rossiter-McLaughlin effect \citep{1924ApJ....60...15R,1924ApJ....60...22M}
117: in radial velocity measurements can contribute to our knowledge of
118: transiting planet system parameters
119: \citep{2006ApJ...653L..69W},
120: transit event closure phases can further the physical description of
121: these systems through direct detection of the transit. Interferometric phase in general
122: is a powerful tool that is beginning to be exploited to its fullest potential
123: in astronomy \citep{2007NewAR..51..604M}.
124:
125: Planet transit closure phase observations described herein are the only presently available technique that provide a measurement of the transiting planet's orbital plane orientation upon the sky. These closure phase observations also uniquely determine the other observables of the system - impact parameter, transit velocity, stellar radius, planet radius, transit ingress time - without the need for supporting observations such as transit photometry. For example, the previous direct determination of HD189733b's diameter \citep{baines2007} measured that parameter through a combination of interferometric measurements and transit photometry; this technique is independent of such outside measurements.
126:
127: We will begin with a review of instrument capabilities in \S \ref{sec_instruments}, an examination of
128: known extrasolar planet candidates in \S \ref{sec_potential_targets}, a `quick-and-dirty' partial analytic solution of the
129: problem in \S \ref{sec_partial_analytic}, a more thorough discussion of closure phase
130: leading to a numeric model and analysis of planet transits in \S \ref{sec_deviations}, and finally a full
131: Monte Carlo simulation to recover synthetic transits in \S \ref{sec_modeling}.
132:
133: \section{Instrument Capabilities}\label{sec_instruments}
134:
135: The Georgia State University Center for High Angular Resolution Astronomy (CHARA) Array
136: is a six-element optical interferometer located atop Mount Wilson in southern California.
137: The CHARA Array consists of six 1-m telescopes laid out on a `Y' array, two telescopes
138: per arm, with a baselines of $>300$ meters on the 3 longest baselines. Initial
139: science operations and a facility description can be found in
140: \citet{2005ApJ...628..439M} and \citet{2005ApJ...628..453T}.
141:
142: Commissioning on-sky tests have recently begun at the CHARA Array with the
143: Michigan Infrared Combiner (MIRC), a high-precision multi-telescope beam combiner
144: \citep{2006SPIE.6268E..55M}. MIRC's capability to combine 4 or 6 telescopes
145: simultaneously and provide 3 to 10 closure phase measurements on sources represents a major
146: step forward in capability for the facility. Moreover, initial MIRC tests
147: are indicating a remarkable ability to measure closure phases with
148: precision unprecedented in the field of optical interferometry,
149: at a level of $\sigma_\Phi \sim 0.03^o$
150: \citet{2006SPIE.6268E..55M}; the first science demonstrated by this capability includes
151: direct imaging the surface of the rapidly rotating star Altair \citep{2007Sci...317..342M}.
152: Other instruments, such as the VLTI AMBER instrument \citep{2006SPIE.6268E..53R},
153: also provide the capability to
154: make closure phase measurements, although initial indications are that
155: the closure phase precision of AMBER is not quite as capable as MIRC,
156: with $\sigma_\Phi \sim $few degrees \citep{2007NewAR..51..724W}.
157:
158:
159: \section{Potential Planet Transit Targets}\label{sec_potential_targets}
160:
161: The obvious candidate for observations of a planet transit event is
162: HD189733 \citep{2006ApJ...650.1160B,2005A&A...444L..15B}.
163: With an angular size of $376 \pm 31$ $\mu$as \citep{baines2007}, it is the
164: planet-transit hosting star with the largest angular size discovered to date.
165: The discovery paper of \citet{2005A&A...444L..15B} cites the following
166: system parameters: (a) a planet-star radius ratio of $0.172 \pm 0.003$,
167: (b) an orbital inclination of $i=85.79 \pm 0.24$, (c) an
168: orbital radius of $0.0313 \pm 0.0004$ AU, and (d) a transit duration of roughly 1.7 hours.
169: The geometry of the transit is depicted in Figure \ref{fig_HD189733geometry}. The best
170: known values for the host star and planet are found in Tables 1 and 2, respectively.
171:
172:
173: \begin{figure*}
174: \begin{center}
175: \includegraphics[scale=0.66,angle=0]{fig1.eps}
176: \end{center}
177: \caption{\label{fig_HD189733geometry} On-sky geometry for a model representative of HD189733, based on
178: data found in \citet{2005A&A...444L..15B} and \citet{baines2007},
179: as discussed in \S \ref{sec_potential_targets}.
180: In particular, the planet radius of $R_{\rm planet} = 64 \mu$as and impact parameter $b = 120 \mu$as
181: are derived from the stellar radius, the planet-star radius ratio, and orbital inclination.
182: The values of $c_1 = 181$ and $c_2 = 95 \mu$as mark the distance of the beginning and end of the ingress event
183: from the transit meridian.
184: The on-sky orientation of the transit chord is, at present, unknown.}
185: \end{figure*}
186:
187: \input{tab1.tex}
188: % Stellar angular diameter is 0.37 mas (source?)
189: % Stellar radius of $0.758 \pm 0.016$ $R_\odot$.
190:
191: \input{tab2.tex}
192:
193: The next best candidate known at the time of this draft is GJ 436 \citep{2004ApJ...617..580B,2007A&A...472L..13G}, with roughly
194: the same anticipated angular size as HD189733.
195: HD149026 \citep{2005ApJ...633..465S}, HD17156 \citep{2007ApJ...669.1336F,2007A&A...476L..13B}, and HD209458 \citep{1999IAUC.7307....1H,2000ApJ...529L..45C} are also worth considering, although their
196: stellar angular diameters of 170-250 $\mu$as are significantly less favorable for detection
197: when considering the currently available capabilities of CHARA-MIRC.
198:
199: \section{Partial Analytic Solution for Transit Interferometric Visibility}\label{sec_partial_analytic}
200:
201: The complex interferometric visibility of a binary star can be written as:
202: \begin{equation}\label{eqn_Vbinary}
203: V_{\rm binary} = e^{-2 \pi i (u \alpha_1 + v \beta_1)}
204: {V_A + r V_B e^{-2 \pi i (u \Delta \alpha + v \Delta \beta)}
205: \over
206: 1 + r}
207: \end{equation}
208: where $r$ is the brightness ratio \citep{1971MNRAS.151..161H}.
209: $V_A$ and $V_B$ are the visibility functions associated with a uniform disk, $V = 2 J_1 (x) / (x)$,
210: where $x=\theta_{UD} \pi B / \lambda$, $\theta_{UD}$ is the uniform disk angular diameter, $B$ is the
211: projected baseline, and $\lambda$ is the wavelength of operation. Dropping the absolute phase term
212: and rewriting this in terms of relative separation vector $\Delta {\bf s}$ and baseline vector ${\bf B}$:
213: \begin{equation}\label{eqn_Vsingle}
214: V_{binary} =
215: {V_A + r V_B e^{{-2 \pi i{{\bf B} \cdot \Delta {\bf s} / \lambda} }}
216: \over
217: 1 + r}
218: \end{equation}
219:
220: Determining visibilities for the specific case of a planet transit can
221: be adopted from this formalization
222: with the following caveats:
223: \begin{itemize}
224: \item The `brightness ratio' $r$ is the negative value of the squared planet-to-star diameter ratio.
225: \item Equation \ref{eqn_Vsingle} assumes that both the planet and stellar disks are indeed uniform
226: disks, since $r$ is constant. Specifically, limb darkening of the star is ignored.
227: \item Equation \ref{eqn_Vsingle} is valid only for the portion of the transit event
228: when the planet is fully in front of the star, when $r$ is again constant. The ingress and egress portions
229: of the transit event are not properly represented by this equation.
230: \end{itemize}
231:
232: This solution is useful in characterizing the order-of-magnitude
233: effects for observation planning; however, given the above caveats,
234: it is insufficient for proper evaluation of actual data.
235:
236: \section{Transit Closure Phase and Visibility Deviations}\label{sec_deviations}
237:
238: In its simplest realization, an optical interferometer
239: measures the Fourier components of an image upon the sky. The
240: location of the image's components in the Fourier transformed $\{u,v\}$ plane are dictated
241: by the baseline between the telescopes in the interferometer projected towards the source of interest, and the
242: wavelength of operation. As seen in \S \ref{sec_partial_analytic},
243: at least a partial analytic solution can be predicted from
244: the parameters of the experiment. In practice, only the visibility amplitude
245: (typically just referred to as the {\it visibility}) can be measured,
246: while atmospheric turbulence corrupts the direct measurement
247: of visibility phase.
248:
249: However, interferometers using three or more telescopes
250: can produce a measurement of the {\it closure phase}, a phase quantity
251: that remains uncorrupted by telescope-specific phase errors
252: \citep{1958MNRAS.118..276J,2000plbs.conf..203M}. The closure phase $\Phi$ is
253: the sum of visibility phases around a closed loop of baselines. For
254: three telescopes $i,j,k$, this is easily deduced from the observed
255: visibilities:
256: \begin{equation}
257: \Phi_{ijk} = arg(V_{ij}) + arg(V_{jk}) + arg(V_{ki})
258: \end{equation}
259:
260:
261: Each pair of telescopes
262: produces a source visibility and phase, where the phase is associated with the source's intrinsic
263: phase $\phi$, phase errors $\theta_1-\theta_2$ associated with the telescope pair, and noise. Typically
264: due to atmospheric corruption, for a pair of telescopes, phase information is useless. However,
265: for a three telescope array $\{l,m,n\}$, combination of the three measured phase pairs
266: $(\psi_{lm}=\phi_{lm}+(\theta_l-\theta_m),\ldots)$ results in cancelation of the phase error terms $\theta$
267: leaving only the sum of the three source phases intrinsic to the object - the closure phase, $\Phi_{lmn} = \phi_{lm}+\phi_{mn}+\phi_{nl}$.
268:
269: Use of the closure phase effectively cancels many of the corrupting effects of
270: the atmosphere and the instrument, and is a highly sensitive probe for interferometric image construction
271: on the smallest spatial scales. Significantly more complete discussions of the topic of
272: closure phase may be found in \citet{1984ARA&A..22...97P} and \citet{2007NewAR..51..604M}.
273: Closure phases have been used to explore disk asymmetries
274: in YSOs \citep{2006ApJ...647..444M} and are very sensitive to asymmetries in
275: images, which will prove quite useful in the application discussed here.
276:
277: For a star with a planet blocking out part of its disk during a transit event, the degree
278: of asymmetry is extreme - significantly much more so than a star with a starspot on its
279: surface: the spot temperature is merely some slight fraction of the rest of the photosphere and is
280: still emitting radiation at a fairly significant level. At the near-infared wavelengths being
281: considered here, a transiting planet
282: emits extremely little radiation with regards to the area of the stellar photosphere it is blocking
283: off from our line of sight.
284:
285: For the full envelope of expected visibility amplitudes and closure phases
286: for a {\it gedanken} experiment covering the interferometer response
287: during a planet transit event, the analytic solution of \S \ref{sec_partial_analytic}
288: is insufficient, in that it breaks down during the transit ingress and egress.
289: As a means to explore that full envelope, a numeric analysis can be performed
290: to compute directly the expected visibility components.
291:
292: For the specific case of HD189733, such an experiment may easily be
293: executed, leading to an expectation of the changes in observed visibility amplitudes and
294: closure phases during the planet transit event.
295: For the {\it gedanken} experiment, three baselines were postulated, with
296: coordinates (in meters) of $\{125.35, 305.94\}, \{-300.42, -89.62\}, \{175.07, -216.32\}$ (with zero
297: vertical separation), which
298: correspond roughly to the CHARA E1S1, W1E1, and S1W1 baselines, respectively.
299:
300: From the system parameters for HD189733 as cited by \citet{2005A&A...444L..15B},
301: a model of the transit event was constructed with a stellar radius of 185$\mu$as,
302: a planet radius of 32$\mu$as, and a transit chord that was offset from the center of
303: the stellar disk by 121$\mu$as. The rate of the transit was not considered in this section,
304: in this simple inspection of the effects upon the interferometer visibility signals,
305: although in \S \ref{sec_modeling} variables for that aspect of the event will be introduced.
306: The top panels of Figure \ref{fig_HD189733CP} shows the this transit event for three different orientations upon the sky,
307: $\alpha = 0^o,45^o,90^o$.
308:
309: For each of the three orientations, the visibility for each of the three baselines, along
310: with the closure phase, was computed. This computation was performed in the following manner:
311: An appropriately limb darkened
312: model star was created numerically on a $1024 \times 1024$ grid with a diameter corresponding
313: to the HD189733 parent star.
314: As introduced by \citet{1921MNRAS..81..361M} and discussed in the
315: context of stellar interferometry by
316: \citet{1974MNRAS.167..475B}, the conventional linear representation
317: of limb darkening across the disk of a star can be written as:
318: \begin{equation}
319: I_\lambda(\mu) = I_\lambda(1)[1-u_\lambda(1-\mu)]
320: \end{equation}
321: where $\mu=\cos \gamma$, $\gamma$ is the angle between the line of
322: sight and the stellar surface normal, and $I_\lambda(1)$ is the
323: specific intensity at the center of the disk. For HD189733 at $5000\pm100$K
324: \citep{baines2007}, the limb darkening parameter $u$ at 1.6 $\mu$m is roughly
325: equal to 0.35 \citep{1995A&AS..114..247C}; as we shall see in \S \ref{sec_modeling},
326: this technique is relative insensitive
327: to limb darkening of the planet host star. This latter fact is unsurprising
328: given the small angular size of the star relative to our notional
329: array; for an array with larger baselines ($B\sim 1$km), there would a
330: greater sensitivity to this parameter.
331:
332: A fully darkened spot corresponding to the radius of the transiting planet was then created on the
333: image for a given location along that transit, and the Fourier components were computed.
334: Rather than bear the full computational load of Fourier transforming the entire image
335: upon the sky, the approach of \citet{2006ApJ...645..664A} is followed, and
336: only the specific components corresponding to the 3 baselines in question were computed,
337: resulting in a much lighter computational load without a sacrifice in precision.
338:
339: This process was then repeated for various points along the planet transit. The computed values
340: for visibility and closure phase were
341: compared to the nominal values for the uneclipsed parent star, and
342: those deviations are plotted in the middle and lower panel of Figure \ref{fig_HD189733CP}.
343: This approach to computing the visibility amplitudes and closure phases
344: can be seen as being superior to the analytic solution in \S \ref{sec_partial_analytic},
345: since it takes into account stellar limb darkening, and is valid through the transit
346: event, including ingress and egress.
347:
348: As seen in Figure \ref{fig_HD189733CP}, during the transit event,
349: the visibility deviates from from the nominal unocculted star case, but
350: by only a marginal amount - on the order of $\pm0.01\%$. Such a measurement is
351: beyond the capabilities of any existing interferometer by two orders of magnitude. However, the
352: closure phase excursion is $\pm0.2^o$. As detailed in \citet{2006SPIE.6268E..55M}, closure phase
353: measurements at this level of precision appear possible: initial tests of the CHARA-MIRC
354: system showed closure phase formal error at the $\sigma_\Phi \sim 0.03^o$ level over the course of 3 hours.
355: Shorter integration times indicated a correspondingly higher level of scatter, but the magnitude
356: of this error gives a starting point from which to evaluate the possibility for observation
357: of a planet transit event using closure phases.
358:
359: \begin{figure*}
360: %\includegraphics[scale=0.66,angle=270]{fig2.eps.old}
361: \plotone{fig2.eps}
362: \caption{\label{fig_HD189733CP} Excursions in visibility amplitude
363: and closure phase data for HD 189733 as observed by CHARA for 3 different
364: orientations,
365: as discussed in \S \ref{sec_modeling}.
366: The top row is the image on the sky, the middle row is the visibility amplitude excursions
367: for each of the 3 CHARA baselines discussed herein, and the bottom row is the
368: closure phase difference during the transit event.}
369: \end{figure*}
370:
371:
372:
373: \section{Recovery of Transit Parameters from Closure Phase Measurements}\label{sec_modeling}
374:
375: Having demonstrated in \S \ref{sec_deviations} in at least a qualitative way that
376: closure phase excursions result from a planet transit event, it is useful to
377: further demonstrate that an event can be reconstructed from an ensemble of
378: closure phase measurements taken during a transit.
379: Free parameters of the fit are the primary descriptors of the image
380: upon the sky:
381: \begin{itemize}
382: \item The stellar radius $r_{star}$.
383: \item The planet to star radius ratio $R$.
384: \item The orbit orientation upon the sky $\alpha$, defined as the angular orientation of the chord of the transit event
385: across the disk of the star in right ascension and declination, as measured from north to east on the sky.
386: \item The `impact parameter' $b$, defined as the distance between the chord of the transit event across the
387: disk of the star, and the center of the stellar disk.
388: \item The `zero time' $JD_0$ of the transit event, defined here as the time of closest approach of
389: the planet disk to the center of the stellar disk.
390: \item The velocity $v$ of the planet disk across the disk of the star.
391: \end{itemize}
392:
393: Fixed input parameters of the fit are:
394: \begin{itemize}
395: \item The $i = 1 \dots N$ closure phases $\Phi_i$ and their associated errors $\sigma_{\Phi,i}$.
396: For this exercise, it will be assumed that these are the single closure phases associated with the
397: non-diurnally evolved CHARA baselines associated with the S1, S2, E1, and W1 stations. In practice, CHARA-MIRC is operated with
398: either four or six simultaneous baselines, which in the latter case can potentially provide even more closure phases than
399: being modelled here; this, in conjunction with baseline diurnal evolution, would serve to further
400: constrain the transit event parameters.
401: \item The closure phase errors $\sigma_{\Phi,i}$. These errors were assumed to have a normal distribution,
402: and for the various synthetic data sets created below, assumed to be of a magnitude ranging from $\overline{\sigma_{\Phi,i}}=0.005^o, 0.010^o, 0.020^o, 0.050^o, 0.100^o$. The
403: FWHM of the $\sigma_{\Phi,i}$ distribution was set at one-quarter of the average error value. Individual
404: closure phases $\Phi_i$ were randomized by an error of the scope of $\sigma_{\Phi,i}$ (but independent thereof).
405: \item The operational wavelength $\lambda_i$ of each observation of the closure phases. A single
406: wavelength of 1.6 $\mu$m will be assumed here for the synthetic data sets. In practice, the MIRC instrument spectrally
407: disperses the starlight and multiple closure phases per observation are available as a result.
408: \item The time $t_i$ of observation of each closure phase. For our synthetic data sets, we will postulate that sets of
409: closure phases are taken sequentially
410: at intervals of 0.001 of a Julian Day (roughly 85 seconds set-to-set), for a total duration of 0.124 days (about
411: 3 hours), which spans the duration of the transit event, plus about 40 minutes before and after the transit
412: event. This resulted in $N=124$ data points per observation set.
413: \item The $\{u,v\}$ coordinates of observing baselines for each closure phase,
414: For an actual observation, the baselines would need to be properly projected onto
415: the sky incorporating diurnal motion.
416: \item The limb darkening of the star.
417: \end{itemize}
418: The stellar limb darkening is also potentially
419: fittable free parameter; however, for the marginal resolution case of
420: CHARA observing HD189733, a model value is sufficient for the
421: fitting.
422:
423: For simplicity of this investigation, diurnal motion will be ignored and the $\{u,v\}$ coordinates
424: of the observing baselines will be constant. In practice, this motion will need to be accounted
425: for but will actually provide addition constraints upon the image reconstruction, much in the
426: same way that baseline evolution can serve to assist in constraining the parametrization of
427: binary star orbits \citep{1999ApJ...515..356B}.
428:
429: To create synthetic `observation sets' for testing our fitting and parameter recovery routines,
430: the free parameters $\{r_{star},R,b,{JD}_0,v\}$ discussed at the beginning of this section were set to the values seen in our
431: HD189733-like system in Figure \ref{fig_HD189733geometry} (with $JD_0=-6000$ sec, $R=0.17$, $b=121.0 \mu$as, $r_{\rm star}=185 \mu$as, and $v=0.07000 \mu$as/sec), and data sets were generated within the context of the fixed input parameters discussed above. For those data sets, the position
432: angle $\alpha$ of the planet transit across the stellar disk was also set to
433: the values $28^o$, $115^o$, and $170^o$, to test
434: the sensitivity of the parameter recovery on that particular parameter as well.
435:
436: Thus, to test 'goodness of fit' for a given set of six randomized free parameters
437: $\{r_{star},R,\alpha,b,JD_0,v\}$, a model transit event sequence was generated, projected upon the sky,
438: and resultant image sequence Fourier transformed for
439: comparison to each of the observed $\Phi_i$ data points,
440: and a $\chi^2/$DOF calculated. A
441: multi-dimensional optimization code was then utilized to loop about this goodness of fit routine and derive
442: the best $\{r_{star},R,\alpha,b,JD_0,v\}$ solution from any given
443: starting point, a process that took typically 500 iterations
444: \citep{pre92}.
445: An exhaustive search of the transit event parameter space was used
446: to explore the $\chi^2/$DOF space, using a two-fold approach.
447: First, a grid of reasonable $\{r_{star},R,\alpha,b,JD_0,v\}$ starting values was explored to see if the original $\{r_{star},R,\alpha,b,JD_0,v\}$
448: parameters could be recovered, with $\alpha$ varying over its full range of $\{0^o,360^o\}$, and the other parameters being explored over a range of $\pm 50\%$ of their `true' values. These latter ranges were expected to encompass the reasonable starting points for an actual investigation, based upon constraints that may be available from discovery photometry or spectroscopy. Each range was gridded with a density of 5 to 10 points per variable. Second, a large number of iterations $(N\sim1000)$ were also run for each
449: synthetic data set starting from fully randomized $\{r_{star},R,\alpha,b,JD_0,v\}$ starting values, also with the purpose of
450: recovering the original $\{r_{star},R,\alpha,b,JD_0,v\}$
451: parameters that described the generating values of the synthetic data set. The resulting $\chi^2$ manifold appeared
452: to be smoothly varying, as the recovery of the original parameters appeared to occur without local minima obstructing
453: the recovery of the global minima.
454: Once the best solution was established for a given data set, $1-\sigma$ errors
455: were established about the $\chi^2$ minimum through exploration of appropriate $\Delta \chi^2$ intervals.
456: The results for each of the three transit position angles,
457: with the five different levels of closure phase error, are seen in Table \ref{table3}.
458:
459:
460:
461: \section{Discussion and Conclusion}
462:
463: Examination of Table \ref{table3} illustrates that, even with a crude level of closure phase error ($\overline{\sigma_\Phi} \sim 0.1^o$), the position angle of the transit event is readily recovered to within a few degrees, and markedly better with modest improvements in closure phase error. Since closure phase is, in essence, an observable that quantifies the degree of asymmetry in an image upon the sky, we expect that this technique should work well even for grazing transit events. These events would be limited presumably by the shorter duration of the occultation event (and thus fewer closure phase data points to fit), and also by a lesser stellar surface area occulted by the planet's disk, resulting in a smaller closure phase signal, but the basic promise of the approach holds true. The sensitivity of this approach to limb darkening - mentioned, but largely dismissed in \S \ref{sec_modeling} - is only slight, due to the largely axisymmetric nature of limb darkening.
464:
465:
466: Of particular interest is the fact that, in the best conceivable case for each apparent transit position angle, the planet radius appears to be recoverable to the level of roughly one part in 30-40, which appears to best the previous interferometric measure of the planetary diameter \citep{baines2007} by a factor of 2.5. Additionally, the technique establishes that diameter in a way that is independent of the transit photometry - the \citet{baines2007} investigation relied upon transit photometry for a value of the planet-star radius, $R$. There is further the possibility that conducting such an observation in a wavelength-dependent sense could probe molecular opacity effects of the the planetary atmosphere though sensing apparent radius dependencies. For example, the recent detection of methane as an constituent of HD189733b's atmosphere \citep{2008arXiv0802.1030S} provides a tantalizing goal for this technique; in principle, a sufficiently precise measurement of this nature, using narrow channels inside of the $H$-band (say, comparing 0.05 $\mu$m-wide channels centered at 1.5, 1.66, and 1.73 $\mu$, respectively) could confirm this detection by detecting the wavelength dependence of the planetary radius.
467: However, this appears to require levels of closure phase precision beyond even the best cases considered here - an examination of the absorption depth data of \citet{2008arXiv0802.1030S} indicates such a detection to require radius measurement precision at the $0.2 - 1.1 \%$ level.
468:
469: These simulations use {\it only} the time-tagged closure phase data from a single transit event; supporting photometric and radial velocity signatures of the transit (and interferometer visibility measures), or multiple transits, have the potential to significantly improve the quality of the $\{r_{star},R,\alpha,b,JD_0\}$ fit parameters. Investigators wishing to mesh such data sets will of course have to pay particular attention to uniform time-tagging. It does seem possible, however, that data sets of such richness will be able to probe other system parameters: possible moons of the transiting planet, and the presence of other stellar planetary companions due to variations in the transit timing and impact parameter.
470:
471: This approach is the {\bf only} currently available technique that provide any value for the transit event orientation angle, $\alpha$; it is also an independent check on parameters such as stellar radius or planet-star radius ratio that is derived from other techniques, such as spectroscopy or photometric timing. Use of results from such interferometric observations could be highly useful for planning observations of TPF-I, Darwin, and other instruments that have a position-angle dependent response. For example, the extreme adaptive optics systems that are envisioned carrying out planet searches and/or characterizations through `dark hole' techniques \citep{2007ApJ...658.1386S,2008arXiv0803.3629O} could have their search times reduced through {\it a priori} knowledge of the planet's orbital plane position angle.
472:
473: \acknowledgements
474:
475: Special thanks to Theo ten Brummelaar, John Monnier, Mark Swain, and Ming Zhao for particularly helpful feedback during the development of this manuscript, which also benefited greatly from the input of an anonymous referee. The CHARA Array is funded by the National Science Foundation through grant AST 94-14449, the W. M. Keck Foundation, the David and Lucile Packard Foundation, and by Georgia State University. This research has made use of the SIMBAD literature database, operated at CDS, Strasbourg, France, the FUTDI database at AMNH, and of NASA's Astrophysics Data System.
476:
477:
478: \begin{thebibliography}{}
479:
480: \bibitem[Aufdenberg et al.(2006)]{2006ApJ...645..664A} Aufdenberg, J.~P., et al.\ 2006, \apj, 645, 664
481: \bibitem[Baines et al.(2007)]{baines2007} Baines, E.~K., van Belle, G.~T., ten Brummelaar, T.~A., McAlister, H.~A., Swain, M., Turner, N.~H., Sturmann, L., \& Sturmann, J.\ 2007, \apjl, 661, L195
482: \bibitem[Bakos et al.(2006)]{2006ApJ...650.1160B} Bakos, G.~{\'A}., et al.\ 2006, \apj, 650, 1160
483: \bibitem[Barbieri et al.(2007)]{2007A&A...476L..13B} Barbieri, M., et al.\ 2007, \aap, 476, L13
484: \bibitem[Boden et al.(1999)]{1999ApJ...515..356B} Boden, A.~F., et al.\ 1999, \apj, 515, 356
485: \bibitem[Bouchy et al.(2005)]{2005A&A...444L..15B} Bouchy, F., et al. 2005, \aap, 444, L15
486: \bibitem[Butler et al.(2004)]{2004ApJ...617..580B} Butler, R.~P., Vogt, S.~S., Marcy, G.~W., Fischer, D.~A., Wright, J.~T., Henry, G.~W., Laughlin, G., \& Lissauer, J.~J.\ 2004, \apj, 617, 580
487: \bibitem[Butler et al.(2006)]{2006ApJ...646..505B} Butler, R.~P., et al.\ 2006, \apj, 646, 505
488: \bibitem[Burrows et al.(2006)]{2006ApJ...650.1140B} Burrows, A., Sudarsky, D., \& Hubeny, I.\ 2006, \apj, 650, 1140
489: \bibitem[Charbonneau et al.(2000)]{2000ApJ...529L..45C} Charbonneau, D., Brown, T.~M., Latham, D.~W., \& Mayor, M.\ 2000, \apjl, 529, L45
490: \bibitem[Claret et al.(1995)]{1995A&AS..114..247C} Claret, A., Diaz-Cordoves, J., \& Gimenez, A.\ 1995, \aaps, 114, 247
491: \bibitem[Cody \& Sasselov(2002)]{2002ApJ...569..451C} Cody, A.~M., \& Sasselov, D.~D.\ 2002, \apj, 569, 451
492: \bibitem[Fischer et al.(2007)]{2007ApJ...669.1336F} Fischer, D.~A., et al.\ 2007, \apj, 669, 1336
493: \bibitem[Gillon et al.(2007)]{2007A&A...472L..13G} Gillon, M., et al.\ 2007, \aap, 472, L13
494: \bibitem[Hanbury Brown et al.(1974)]{1974MNRAS.167..475B} Hanbury Brown, R., Davis, J., Lake, R.~J.~W., \& Thompson, R.~J.\ 1974, \mnras, 167, 475
495: \bibitem[Henry et al.(1999)]{1999IAUC.7307....1H} Henry, G.~W., Marcy, G., Butler, R.~P., \& Vogt, S.~S.\ 1999, \iaucirc, 7307, 1
496: \bibitem[Herbison-Evans et al.(1971)]{1971MNRAS.151..161H} Herbison-Evans, D., Hanbury Brown, R., Davis, J., \& Allen, L.~R.\ 1971, \mnras, 151, 161
497: \bibitem[Jennison(1958)]{1958MNRAS.118..276J} Jennison, R.~C.\ 1958, \mnras, 118, 276
498: \bibitem[McLaughlin(1924)]{1924ApJ....60...22M} McLaughlin, D.~B.\ 1924, \apj, 60, 22
499: \bibitem[Milne(1921)]{1921MNRAS..81..361M} Milne, E.~A.\ 1921, \mnras, 81, 361
500: \bibitem[McAlister et al.(2005)]{2005ApJ...628..439M} McAlister, H.~A., et al.\ 2005, \apj, 628, 439
501: \bibitem[Monnier(2000)]{2000plbs.conf..203M} Monnier, J.~D.\ 2000, Principles of Long Baseline Stellar Interferometry, 203
502: \bibitem[Monnier et al.(2006a)]{2006SPIE.6268E..55M} Monnier, J.~D., et al.\ 2006a, \procspie, 6268
503: \bibitem[Monnier et al.(2006b)]{2006ApJ...647..444M} Monnier, J.~D., et al.\ 2006b, \apj, 647, 444
504: \bibitem[Monnier et al.(2007)]{2007Sci...317..342M} Monnier, J.~D., et al.\ 2007, Science, 317, 342
505: \bibitem[Monnier(2007)]{2007NewAR..51..604M} Monnier, J.~D.\ 2007, New Astronomy Review, 51, 604
506: \bibitem[Oppenheimer et al.(2008)]{2008arXiv0803.3629O} Oppenheimer, B.~R., et al.\ 2008, ArXiv e-prints, 803, arXiv:0803.3629
507: \bibitem[Pearson \& Readhead(1984)]{1984ARA&A..22...97P} Pearson, T.~J., \& Readhead, A.~C.~S.\ 1984, \araa, 22, 97
508: \bibitem[Press et al.(1992)]{pre92}Press, W.H., Teukolsky, S.A., Vetterling, W.T., Flannery, B.P., 1992, Numerical Receipes in C, Port Chester, Cambridge University Press
509: \bibitem[Rantakyr{\"o} et al.(2006)]{2006SPIE.6268E..53R} Rantakyr{\"o}, F.~T., et al.\ 2006, \procspie, 6268
510: \bibitem[Rossiter(1924)]{1924ApJ....60...15R} Rossiter, R.~A.\ 1924, \apj, 60, 15
511: \bibitem[Sato et al.(2005)]{2005ApJ...633..465S} Sato, B., et al.\ 2005, \apj, 633, 465
512: \bibitem[Serabyn et al.(2007)]{2007ApJ...658.1386S} Serabyn, E., Wallace, K., Troy, M., Mennesson, B., Haguenauer, P., Gappinger, R., \& Burruss, R.\ 2007, \apj, 658, 1386
513: \bibitem[Swain et al.(2008)]{2008arXiv0802.1030S} Swain, M.~R., Vasisht, G., \& Tinetti, G.\ 2008, ArXiv e-prints, 802, arXiv:0802.1030
514: \bibitem[Tango \& Davis(2002)]{2002MNRAS.333..642T} Tango, W.~J., \& Davis, J.\ 2002, \mnras, 333, 642
515: \bibitem[ten Brummelaar et al.(2005)]{2005ApJ...628..453T} ten Brummelaar, T.~A., et al.\ 2005, \apj, 628, 453
516: \bibitem[Weigelt et al.(2007)]{2007NewAR..51..724W} Weigelt, G., Driebe, T., Hofmann, K.-H., Kraus, S., Petrov, R., \& Schertl, D.\ 2007, New Astronomy Review, 51, 724
517: \bibitem[Winn et al.(2006)]{2006ApJ...653L..69W} Winn, J.~N., et al.\ 2006, \apjl, 653, L69
518:
519:
520: \end{thebibliography}
521:
522: \clearpage
523: \begin{landscape}
524: \input{tab3.tex}
525: \clearpage
526: \end{landscape}
527:
528:
529: \end{document}
530: