1: %%%%%%%%%%%%%%%%%%%%%%5
2: % THINGS TO DO
3: %%%%%%%%%%%%%%%%%%%%%%%%5
4: % AASTeX 5.2.
5: \def\today{\ifcase\month\or January\or
6: February\or March\or April\or May\or June\or
7: July\or August\or September\or October\or November\or December\fi
8: \space\number\day, \number\year}
9: % \documentclass[12pt,preprint]{aastex}
10: \documentclass{emulateapj}
11:
12: %%\topmargin 0.3in
13:
14: % These are the initial parameters which we can change as our fits improve
15: \newcommand{\TransAntiBelieve}{0.735}
16: \newcommand{\Flimit}{0.65}
17:
18: %%%%%%%%%%%%%%%%%%%%%
19: %CANDIDATE
20: %%%%%%%%%%%%%%%%%%%%%%%%%
21: \newcommand{\PlanetPeriodLong}{0.7517}
22: \newcommand{\PlanetPeriod}{0.752}
23: \newcommand{\PlanetRadius}{0.134} % //0.133904
24:
25: % 11.19 Rjup to Rearth conversion
26: \newcommand{\PlanetRadiusEarth}{1.50} % //1.62255
27: \newcommand{\PlanetInc}{86.0}
28: \newcommand{\PlanetChiPerc}{0.438}
29: \newcommand{\PlanetEphemeris}{2403.711} % 2403.711124
30: \newcommand{\PlanetMassEarth}{1.8}
31: \newcommand{\PlanetMassJupiter}{0.006}
32:
33: \newcommand{\limbone}{0.82523168}
34: \newcommand{\limbtwo}{-0.97755590}
35: \newcommand{\limbthree}{1.6867374}
36: \newcommand{\limbfour}{-0.65702013}
37:
38: \newcommand{\Pmin}{0.4}
39: \newcommand{\Pmax}{6.9}
40: \newcommand{\NumPassedSelection}{3}
41:
42: \newcommand{\EtaChoice}{0.000219}
43:
44: \newcommand{\HowMany}{100}
45: %%%%%%%%%%%
46: % TROJAN STUFF
47: %%%%%%%%%%%
48: \newcommand{\RTrojanJupiter}{0.24}
49: \newcommand{\RTrojanEarth}{2.7}
50: \newcommand{\MTrojanEarth}{11}
51:
52: %ABSTRACT STUFF
53: \newcommand{\Rjupsmall}{0.15} % HD 189733
54: \newcommand{\Rjuplarge}{0.31} % HD 189733
55: \newcommand{\Rearthsmall}{1.7} % HD 189733
56: \newcommand{\Rearthlarge}{3.5} % HD 189733
57: \newcommand{\Rratiosmall}{0.020} % HD 189733
58: \newcommand{\Rratiolarge}{0.042} % HD 189733
59: \newcommand{\Massearthsmall}{2.6} % HD 189733
60: \newcommand{\Massearthlarge}{22.8} % HD 189733
61: % jupiter mass is 317.894
62: \newcommand{\Massjupitersmall}{0.008} % HD 189733
63: \newcommand{\Massjupiterlarge}{0.07} % HD 189733
64: \newcommand{\MassNeptune}{12.2} % just times massearth large by 16/30
65:
66:
67: \newcommand{\MearthsmallIce}{1.3}
68: \newcommand{\MearthsmallRock}{4.8}
69: \newcommand{\MearthsmallIron}{27}
70: \newcommand{\MearthlargeIce}{23}
71: \newcommand{\MearthlargeRock}{153}
72:
73:
74:
75:
76:
77:
78: \begin{document}
79:
80: \slugcomment{Draft \today}
81:
82: \shorttitle{Looking for Super-Earths in the HD 189733 system}
83:
84: \shortauthors{Croll et~al.}
85:
86: \title{LOOKING FOR SUPER-EARTHS IN THE HD 189733 SYSTEM:
87: A SEARCH FOR TRANSITS IN {\it MOST}\altaffilmark{1} SPACE-BASED PHOTOMETRY}
88:
89: \author{Bryce Croll\altaffilmark{2}, Jaymie M. Matthews\altaffilmark{3},
90: Jason F. Rowe\altaffilmark{3}, Brett Gladman \altaffilmark{3},
91: Eliza Miller-Ricci\altaffilmark{4},
92: Dimitar Sasselov\altaffilmark{4},
93: Gordon A.H. Walker\altaffilmark{5}, Rainer Kuschnig\altaffilmark{3}, Douglas N.C. Lin \altaffilmark{6},
94: David B. Guenther\altaffilmark{7}, Anthony F.J. Moffat\altaffilmark{8}, Slavek M.
95: Rucinski\altaffilmark{9}, Werner W. Weiss\altaffilmark{10}}
96:
97: \altaffiltext{1}{Based on data from the MOST satellite, a Canadian Space
98: Agency mission, jointly operated by Dynacon Inc., the University of Toronto
99: Institute of Aerospace Studies and the University of British Columbia, with
100: the assistance of the University of Vienna.}
101:
102:
103: \altaffiltext{2}{Deptartment of Astronomy and Astrophysics, University of Toronto, 50 St. George Street, Toronto, ON
104: M5S 3H4, Canada;
105: croll@astro.utoronto.ca}
106:
107: \altaffiltext{3}{Deptartment of Physics \& Astronomy, University of British
108: Columbia, 6224 Agricultural Road, Vancouver, BC V6T 1Z1, Canada;
109: matthews@phas.ubc.ca, rowe@phas.ubc.ca, gladman@phas.ubc.ca,
110: kuschnig@phas.ubc.ca}
111:
112: \altaffiltext{4}{Harvard-Smithsonian Center for Astrophysics,
113: 60 Garden
114: Street, Cambridge, MA 02138, USA; emillerr@cfa.harvard.edu, sasselov@cfa.harvard.edu}
115:
116: \altaffiltext{5}{1234 Hewlett Place, Victoria, BC V8S 4P7, Canada;
117: gordonwa@uvic.ca}
118:
119: \altaffiltext{6}{University of California Observatories, Lick Observatory, University of California, Santa Cruz, CA 95064, USA;
120: lin@ucolick.org}
121:
122: \altaffiltext{7}{Department of Astronomy and Physics, St. Mary's
123: University, Halifax, NS B3H 3C3, Canada; guenther@ap.stmarys.ca}
124:
125: \altaffiltext{8}{D\'ept de physique, Univ de
126: Montr\'eal, C.P.\ 6128, Succ.\ Centre-Ville, Montr\'eal, QC H3C 3J7,
127: Obs du mont M\'egantic, Canada; moffat@astro.umontreal.ca}
128:
129: \altaffiltext{9}{Dept. Astronomy \& Astrophysics, David Dunlap Obs., Univ.
130: Toronto, P.O.~Box 360, Richmond Hill, ON L4C 4Y6, Canada;
131: rucinski@astro.utoronto.ca}
132:
133: \altaffiltext{10}{Institut f\"ur Astronomie, Universit\"at Wien,
134: T\"urkenschanzstrasse 17, A--1180 Wien, Austria; weiss@astro.univie.ac.at}
135:
136: \begin{abstract}
137:
138: We have made a comprehensive transit search for exoplanets
139: down to
140: $\simeq$1.5 - 2 Earth radii in the HD 189733 system, based on 21-days of nearly
141: uninterrupted broadband optical photometry obtained with the {\it MOST}
142: ({\it Microvariability \& Oscillations of STars}) satellite in 2006.
143: We have searched these data for
144: realistic limb-darkened transits from exoplanets other than the known hot Jupiter, HD 189733b,
145: with periods ranging from about 0.4 days to one week. Monte Carlo statistical tests of
146: the data with synthetic transits inserted into the data-set allow us to rule out additional
147: close-in exoplanets with sizes ranging from about \Rjupsmall \ - \Rjuplarge \ $R_{J}$
148: (Jupiter radii), or \Rearthsmall \ - \Rearthlarge \ $R_{\earth}$ (Earth radii) on orbits
149: whose planes are near that of HD 189733b.
150: These null results constrain
151: theories that invoke lower-mass hot Super-Earth and hot Neptune planets in orbits similar to HD 189733b
152: due to the inward migration of this hot Jupiter.
153: This work also illustrates the feasibility of
154: discovering smaller transiting planets around chromospherically active stars.
155: \end{abstract}
156:
157: \keywords{ planetary systems -- stars: individual: HD 189733 -- methods: data analysis -- techniques: photometric}
158:
159: \setcounter{footnote}{0} % otherwise corrupted numbers
160:
161: \section{INTRODUCTION}
162: \label{SecIntro}
163:
164: With the launch of the {\it MOST} ({\it Microvariability
165: \& Oscillations of STars}) (\citealt{Walker};
166: \citealt{Matthews}) and COROT (\citealt{Baglin};
167: \citealt{Barge}) satellites, and the upcoming launch
168: of the Kepler (\citealt{Borucki}; \citealt{Basri})
169: space mission, the age of transit searches with continuous space-based
170: photometry is now upon us.
171: With these new telescopes we should now be able to utilize the transit method to probe
172: the terrestrial and giant-terrestrial regime
173: of transiting exoplanets.
174: We can seriously anticipate the discovery of a transiting Super-Earth \citep{ValenciaA} sized
175: planet orbiting a Sun-like star in the near-term future.
176:
177: Numerical simulations of migrating hot Jupiters have indicated that one might
178: expect Earth, Super-Earth and Neptune sized planets
179: in similar orbits,
180: or in mean-motion resonances with the hot Jupiter
181: (\citealt{Zhou}; \citealt{Raymond}; \citealt{Mandell}; \citealt{Fogg}).
182: The interior regions of these hot Jupiter exoplanetary systems,
183: and specifically the mean-motion resonances of these hot Jupiters,
184: have been probed with the transit-timing method in a handful of transiting exoplanetary systems
185: to date - these are TrES-1 (K0 V) \citep{Steffen}, HD 209458 (G0 V) (\citealt{Agol}; \citealt{Miller}),
186: and
187: HD 189733 (K0 V) (\citealt{WinnPsi}; \citealt{MillerTwo}; \citealt{PontPrep}).
188: The transit method,
189: although currently less sensitive in the mean-motion resonances of the hot Jupiter, is able to
190: probe a continuous range of orbital separations in the interior regions of these hot-Jupiter systems.
191: The transit method has only been used to sensitively probe the interior regions of one hot Jupiter
192: exoplanetary system to date; \citet{Croll} reported a null-result in the HD 209458 system from the first
193: space-based transit-search using nearly continuous photometry returned by the
194: {\it MOST} satellite.
195: In that work, \citet{Croll} ruled out hot Super-Earth and hot Neptune-sized planets larger than
196: 2-4 Earth radii
197: in the HD 209458 system with periods ranging from 0.5 days to 2 weeks.
198: Here we report a search using {\it MOST} photometry and similar methods
199: ruling out transiting exoplanets with periods from 0.4 days to one week
200: in the HD 189733 system with orbital inclinations similar to that of
201: the known hot Jupiter exoplanet, HD 189733b.
202:
203: The existence of a hot Jupiter planet, HD 189733b, in the star system HD 189733 (V = 7.67)
204: was first reported by \citet{Bouchy}.
205: The planet was
206: detected from radial velocity measurements in the ELODIE metallicity-biased search for transiting hot Jupiters.
207: Photometry of HD 189733 taken by \citet{Bouchy}
208: indicated the planet HD 189733b transited the star along our line of sight.
209: Follow-up work has determined the stellar and planetary characteristics of the detected hot Jupiter
210: to a high
211: degree of accuracy (\citealt{Bakos}; \citealt{WinnLambda}; \citealt{WinnPsi}; \citealt{Baines}).
212: A number of researchers have determined characteristics of the planet by
213: measuring the drop in flux
214: during the secondary eclipse. These results include the thermal emission of the planet \citep{Deming},
215: one of the first spectra of an extrasolar planet \citep{Grillmair},
216: as well as the first measurement of the day-night contrast of an extrasolar planet \citep{Knutson}.
217: Recent evidence indicates that HD 189733 is likely a binary system with a mid-M dwarf star, HD 189733B (period $\sim$ 3200 $yr$),
218: as a companion to the main K0 star, HD 189733 \citep{BakosPal}. The putative transiting planets that are
219: searched for in our manuscript are assumed to transit the primary star, HD 189733,
220: as does the known planet HD 189733b.
221: Radial-velocity measurements have already ruled out other planets than the known hot Jupiter
222: in this system that are more massive than approximately $\sim$32 $M_{\earth}$
223: (\citealt{WinnLambda} as quoted in \citealt{MillerTwo}).
224:
225: Recently, HD 189733 was discovered to display quasiperiodic flux variations at the $\sim$3\%
226: level (\citealt{WinnPsi};
227: Matthews et al. in preparation).
228: % \citealt{MatthewsHD}). % PREPCHANGE
229: This result is in concordance with the results
230: of \citet{Wright}, which indicated that HD 189733 was a chromospherically active star.
231: Detecting planets via the transit method in chromospherically active stars is a relatively new
232: topic.
233: In addition to the original discovery of HD 189733b
234: around this active star, and the pioneering work of the TEP network (\citealt{Deeg}; \citealt{Doyle}),
235: \citet{Hebb} recently found that they could rule-out Jupiter and Neptune-sized planets
236: with short periods via the transit method
237: around the active star AU Mic.
238: We extend these efforts to the Neptune and Super-Earth-sized regime by investigating the continuous
239: and accurate photometry returned by {\it MOST} to search for
240: Super-Earth and Neptune sized planets with short periods around this active star.
241:
242: {\it MOST}'s 21-day observations of HD 189733 are presented in
243: Matthews et al. (in preparation).
244: % \citep{MatthewsHD}. % PREPCHANGE
245: These observations have already been used to place limits on other hot Super-Earth and hot Neptune
246: exoplanets near the mean-motion resonances of the hot Jupiter
247: in the system through the transit-timing method \citep{MillerTwo}. We extend these efforts
248: here by
249: ruling out hot Super-Earth and hot Neptune-sized planets using the transit method.
250: The rotational modulation displayed in {\it MOST}'s observations of this system has also been fit with a starspot model
251: by \citet{CrollSpot}. Their model argues in favour of moderate
252: spin-orbit misalignment between the stellar spin axis and the orbit-normal of the hot Jupiter in this system.
253:
254: In $\S$\ref{SecMOST} the {\it MOST} photometry of HD 189733 is
255: described. The transit search technique is briefly summarized in
256: $\S$\ref{SecTrans}. The Monte Carlo statistics used to estimate the
257: sensitivity of the transit search are specified in $\S$\ref{SecMonte}.
258: The transit search routine is applied to the {\it MOST} HD 189733 data
259: set in $\S$\ref{SecHDTrans}, where we are able to rule-out hot Super-Earth and hot-Neptune sized planets
260: in this system.
261: Our results are summarized in
262: $\S$\ref{SecDiscuss}, including a discussion of the approximate mass of exoplanets that have been ruled
263: out by our results for various compositions of extrasolar planets.
264:
265: \section{{\it MOST} Photometry of HD 189733}
266: \label{SecMOST}
267:
268: The {\it MOST} satellite was launched on 30 June 2003 and its initial
269: mission is described by \citet{Walker} and \citet{Matthews}. A
270: 15/17.3-cm Rumak-Maksutov telescope feeds two CCDs, one originally
271: dedicated to tracking and the other to science, through a single,
272: custom, broadband filter (350 -- 700 nm). {\it MOST} was placed in
273: an 820-km altitude circular Sun-synchronous polar orbit with a period of
274: 101.413 minutes. From this vantage point, {\it MOST} can monitor
275: stars in a continuous viewing zone (covering a declination range
276: $+36^{\circ} \leq \delta \leq -18^{\circ}$) within which stars can
277: be monitored without interruption for up to 8 weeks. Photometry of
278: very bright stars ($V \leq 6$) is obtained in Fabry Imaging mode, in
279: which a Fabry microlens projects an extended image of the telescope
280: pupil illuminated by the target starlight to achieve the highest
281: precision (see Matthews et al. 2004). Fainter stars (down to about
282: $V \sim 12$) can be observed in Direct Imaging mode, where
283: defocused images of stars are monitored in Science CCD subrasters
284: (see \citealt{Rowe}).
285:
286: {\it MOST} observed HD 189733 in Direct Imaging mode for 21 days during 31 July $-$
287: 21 August 2006. For the first 14 days of this run, HD 189733 was the exclusive target;
288: for the final 7 days, it was shared with another Primary Science Target field during each
289: 101.4-min orbit of the satellite. Consequently, during the last third of the run, HD 189733
290: was observed only during intervals of high stray Earthshine, resulting in somewhat
291: increased photometric scatter. The duty cycles of the light curve are 94\% during the
292: first 14 days and 46\% during the last 7 days.
293: Early in 2006 the tracking CCD system of {\it MOST} failed due to a
294: particle hit. Thereafter both science and tracking were carried out with the
295: Science CCD system. To avoid introducing significant tracking errors
296: exposures were limited to 1.5 sec. 14
297: consecutive exposures were then stacked on board to improve signal-to-noise,
298: and downloaded
299: from the satellite every 21 sec. Approximately 112,000 of these combined observations were taken.
300:
301: The photometric reduction was performed by one of us (JFR) using techniques similar
302: to those described by \citet{Rowe} and
303: % \citet{Rowe2}. % PREPCHANGE
304: Rowe et al. (in preparation).
305: The reduction procedure of aperture
306: photometry is similar to that applied to groundbased CCD photometry, but is non-differential.
307: Our reduction pipeline corrects for cosmic ray hits (more frequent during satellite
308: passages through the South Atlantic Anomaly [SAA]), the varying background due to
309: scattered Earthshine modulated at the satellite orbital period, and flatfielding effects.
310: Further details are given by
311: Matthews et al. (in preparation).
312: % \citet{MatthewsHD}. %PREPCHANGE
313: It should also be noted that HD 189733, and its M-dwarf companion are
314: sufficiently well separated
315: (216 AU or 11.2'' [\citealt{BakosPal}]), that they do not contaminate the photometric signal
316: obtained by {\it MOST}.
317:
318: \subsection{Additional selection and filtering of the data}
319: \label{SecMOSTReduce}
320:
321: For this transit search, additional selection and filtering of the photometry was
322: done to optimize the data for the application of Monte Carlo statistics.
323: The filtering is an automated part of the transit search routine, and is thus briefly summarized
324: in $\S$\ref{SecTrans}. This filtering step is described in detail here.
325:
326: Data obtained during passages through the SAA are conservatively
327: excised from the light curve, due to the increased photometric scatter
328: during those MOST orbital phases, without reducing seriously the phase
329: coverage of the exoplanetary periods searched. The transits of
330: the known giant planet HD 189733b are removed, at the orbital period $P$ $\sim$ 2.21857d determined by \citet{Bakos}.
331: The first $\sim$0.8$d$ of the data-set was excluded from the analysis due to an obvious
332: temperature swing-in effect common to {\it MOST}.
333: The obvious modulation of the light curve, due to starspot activity,
334: was removed with a cubic spline \citep{Press}.
335: The cubic spline was generated with data binned every $\sim$500 minutes;
336: this trend was then subtracted from the unbinned data.
337: This value of $\sim$500 minutes is believed to be optimum,
338: as it effectively balances the competing desires of having a sufficiently
339: short binning time to effectively remove long period modulation,
340: with the desire of having a long binning time to ensure that
341: one is able to recover
342: transits with periods in the
343: upper range of our interesting period range ($P$ $>$ 5$d$).
344: Limited statistical tests have confirmed that this binning value
345: optimizes our sensitivity to transits of HD 189733 with periods in the upper range of our period range.
346: The signal to noise that an individual transit is recovered with can be highly
347: dependent on this binning time for transits of relatively long periods ($P$ $>$ 5$d$).
348: At these longer periods the cubic-spline
349: can filter out much of the amplitude of the transit signal a low percentage of the time.
350: This effect can also be
351: seen with the reduced efficiency at longer periods for the ninety-degree inclination angle case of the
352: Monte Carlo statistics (Figure \ref{FigMonte}).
353: % Because of the modulation of stray Earthshine
354: % with {\it MOST}'s 101.4-min orbital period, the data were phased to that
355: % period, and segments showing the most noticeable effects from stray
356: % light were also removed. The coverage in phase
357: % per 101.4-min {\it MOST} orbit following this cut
358: % were 63\%. THIS WAS NOT DONE, WHY NOT?
359:
360: {\it MOST} suffers from stray Earthshine with its 101.4-min orbital period, and this stray
361: light background can also be modulated at a period of 1 day and its first harmonic (due to the
362: Sun-synchronous nature of the MOST satellite orbit). For this reason,
363: sinusoidal fits with periods within 1\% of 1 and 0.5 d were subtracted from the data.
364: After these cuts, any remaining outliers
365: greater than $6\sigma$ were excised.
366: In general, there were few points removed by this sigma-cut as few points were such extreme outliers.
367: It is also important to note that the magnitudes of the injected transits ($\S$\ref{SecMonte}) were always
368: at a level much less than this sigma-cut.
369: The data were also
370: median-subtracted, for reasons outlined in \citet{Croll}.
371: The automatic filtering step removed $\sim$8\% of the original data (mostly due to the
372: SAA and stray light corrections).
373:
374: The HD 189733 light curves before and after these reduction steps are plotted in Figure
375: \ref{FigData}. These filtered data (and the original reduced data) can
376: be downloaded from the {\it MOST} Public Data Archive at
377: www.astro.ubc.ca/MOST.
378:
379:
380: \begin{figure}
381: \includegraphics[width=1.5in, height = 3.5in, angle=270]{f1a.eps} % 2.1, 7.0
382: \includegraphics[width=1.5in, height = 3.5in, angle=270]{f1b.eps}
383: \includegraphics[width=1.5in, height = 3.5in, angle=270]{f1c.eps}
384: \caption{
385: {The 2006 (A) HD 189733 {\it MOST} data-set before the automated filtering step.
386: The data following (B) the automatic filtering step, in which the transits of the known planet were removed, and
387: a cubic spline removed the obvious long-term variability.
388: (C) The same data are shown binned in 30 minute intervals at a different vertical scale.
389: }
390: \label{FigData}
391: }
392: \end{figure}
393:
394:
395: \section{Transit Search Algorithm}
396: \label{SecTrans}
397:
398: The transit search routine used here is fully described in \citet{Croll}, and is briefly summarized here.
399: The minor modifications to the routine are highlighted below. The transit routine consists of five steps: (i) an automated filtering step (described in detail in $\S$\ref{SecMOSTReduce}),
400: (ii) a modified version of the EEBLS
401: (Edge Effect Box-fitting Least Squares) algorithm of \citet{Kovacs} that searches for box-shaped transits, (iii) selection criteria
402: to select the the best candidates that can be tested for astrophysical plausibility, (iv) realistic transit-fitting
403: and refinement of the transit parameters, and finally (v) detection
404: criteria to differentiate false positives from likely transit candidates.
405:
406: The detection criteria used in this study are summarized here.
407: We use the transit to anti-transit (the best-fit brightening event with the shape of a transit)
408: ratio statistic (RS) of \citet{Burke} to quantify a believable transit.
409: The RS value that a transit has to exceed to be deemed a believable transit is:
410: $\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ $\ge$ \TransAntiBelieve, where
411: $\Delta\chi^{2}$ and $\Delta\chi^{2}_{-}$ refer to the normal $\chi^{2}$ (the sum of sqaured residuals)
412: for the best-fit transit, and anti-transit, respectively. This value
413: was motivated by the results of the Monte Carlo tests ($\S$\ref{SecMonte}).
414: The RS statistic serves to objectively quantify the significance
415: of the measured signal against the correlated
416: scatter in the data, and to thus stringently rule-out false positives.
417: The correlated signal to noise calculation of \citet{Pont} serves a similar purpose.
418: The RS will be used in this present application.
419:
420: The restriction on the quantity
421: f, a measure of the ratio of the strength of any one specific transit compared to the others,
422: as described in \citet{Croll}, and \citet{Burke}, was also relaxed such that transits were accepted
423: as long as f $<$ 0.85. This motivation for this less stringent criterion was based on the Monte Carlo
424: statistics that indicated that a more conservative
425: criterion on f would rule out realistic transits, without an appreciable decrease in the number
426: of false positives in our data-set.
427:
428: The precise values used for various parameters in the EEBLS step (ii) in the transit search routine
429: are summarized in Table \ref{TableBls}.
430: We have decreased the minimum value probed by the EEBLS routine in the application of this
431: transit search routine to
432: HD 189733, compared to its application to HD 209458 in \citet{Croll}. This is due to the conclusions
433: of \citet{Sahu}, who observed
434: a number of ultra-short period ($P$ $<$ 1.0 $d$) planet candidates in the SWEEPs survey, and noted that these planets
435: may preferrentially
436: be found around low-mass stars. HD 189733 has a stellar mass of 0.82 $M_{\odot}$ \citep{Bakos},
437: below the limit of 0.88 $M_{\odot}$ suggested by \citet{Sahu} to be indicative of a preference for ulta-short period planets.
438: The minimum period investigated by the EEBLS routine
439: is thus now \Pmin$d$, a value corresponding to a semi-major axis, $a_{p}$, of just under three times the stellar radius ($a_{p}$ $\approx$ 2.5$R_{*}$).
440: For comparison, OGLE-TR-56b has $a_{p}$$\approx$$4.4R_{*}$,
441: the smallest orbital radius in terms of its stellar radius of a
442: confirmed planet to date.
443: The maximum period of 6.9$d$ is chosen similar to the maximum value for HD 209458
444: in \citet{Croll}, as a value that would allow one to observe three distinct transits in the 21 $d$ data-set.
445: This period range of 0.4 to 6.9 $d$
446: corresponds to 0.01 to 0.07 AU.
447: The minimum fractional transit length is 0.75 times that of the calculated average
448: fractional transit length at that period - a value chosen to ensure the routine remains adequately
449: sensitive to transiting planets with non edge-on inclination angles. We also increase the number of bins, $Nb$,
450: used by the EEBLS algorithm for short periods compared to the application of \citet{Croll} to HD 209458.
451: The number
452: of bins, $Nb$, used by the EEBLS algorithm for each period, $P$, is determined by the following formula:
453: $[0.7\times \exp(-$P$) + 1.0]$$\times$20.0/$Qmi$, where $Qmi$ is the minimum fractional transit length to
454: be tested. $Qmi$, and the maximum fractional transit length, $Qma$, are determined as given in Table
455: \ref{TableBls}, from our average estimate of the fractional transit length for a hypothetical planet, $Qm_{P}$,
456: at various periods \citep{Croll}.
457:
458: In our transit selection step we now select the top three EEBLS SR signals, instead
459: of the top two, as well as up to two transit signals with SR signals that are
460: at least 1.8 times the noise floor. A full explanation of these details
461: is provided in \citet{Croll}.
462: The stellar
463: characteristics used to determine the realistic limb-darkened transit model (iv) are summarized in Table \ref{TableFit}.
464: The precise transit parameters are refined using a Marquardt-Levenberg routine (described in detail in \citealt{Croll}).
465: The method used to refine our transit parameters is very similar to the Newton-Raphson method applied
466: by \citet{Collier} for the same purpose.
467:
468: \begin{deluxetable*}{ccc}
469: \tabletypesize{\footnotesize}
470: \tablecaption{EEBLS input parameters \label{TableBls}}
471: \tablewidth{0pt}
472: \tablehead{\colhead{var}& \colhead{definition} & \colhead{value} \\}
473: \startdata
474: $Np$ & Number of Period Points searched & 15 000 \\
475: $\eta$ & Logarithmic period step & \EtaChoice \\
476: $P_{min}$ & Minimum period submitted to EEBLS algorithm & \Pmin$d$ \\
477: $P_{max}$ & Maximum period submitted to EEBLS algorithm & \Pmax$d$ \\
478: $Qmi$ & Minimum fractional transit length to be tested & $0.75$$\times$$Qm_{P}$ \\
479: $Qma$ & Maximum fractional transit length to be tested & $1.1$$\times$$Qm_{P}$ \\
480: $Nb$ & Number of bins in the folded time series at each test period & $(0.7\times \exp(-$P$) + 1.0)$$\times$20.0/$Qmi$ \\
481: \enddata
482: \end{deluxetable*}
483:
484: \begin{deluxetable}{ccc}
485: \tabletypesize{\footnotesize}
486: \tablecaption{System parameters \label{TableFit}}
487: \tablewidth{0pt}
488: \tablehead{\colhead{var}& \colhead{definition} & \colhead{value} \\}
489: \startdata
490: $R_{*}$ & Stellar radius & 0.755 $R_{\odot}$ \tablenotemark{a} \\
491: $M_{*}$ & Stellar Mass & 0.82 $M_{\odot}$ \tablenotemark{b}\\
492: $c_{1}$ & Non-linear limb-darkening parameter 1 & \limbone \tablenotemark{c} \\
493: $c_{2}$ & Non-linear limb-darkening parameter 2 & \limbtwo \tablenotemark{c}\\
494: $c_{3}$ & Non-linear limb-darkening parameter 3 & \limbthree \tablenotemark{c}\\
495: $c_{4}$ & Non-linear limb-darkening parameter 4 & \limbfour \tablenotemark{c}\\
496: \enddata
497: \tablenotetext{a}{Parameter obtained from \citet{PontPrep}}
498: \tablenotetext{b}{Parameter obtained from \citet{Bakos}}
499: \tablenotetext{c}{These limb-darkening parameters are fully described in \citet{Mandel} and obtained from
500: \citet{MillerTwo}.}
501:
502: \end{deluxetable}
503:
504:
505: \begin{figure}
506: \includegraphics[height = 3.5in, width = 2.25in, angle=270]{f2a.eps} % 6, 4
507: \includegraphics[height = 3.5in, width = 2.25in, angle=270]{f2b.eps}
508: \caption{{The EEBLS spectrum of the {\it MOST} HD 189733 data-set, plotted in frequency (top),
509: and period (bottom).
510: The \NumPassedSelection \ candidates that passed the transit selection
511: criteria are shown by the solid vertical lines.}
512: \label{FigTransSelect}}
513: \end{figure}
514:
515:
516:
517:
518:
519:
520:
521:
522:
523:
524:
525:
526:
527:
528:
529:
530:
531:
532:
533:
534: \section{Monte Carlo statistics}
535: \label{SecMonte}
536:
537: \subsection{Monte Carlo setup}
538:
539: To assess the sensitivity of our routine and the {\it MOST} data-set
540: to smaller transiting planets in the HD 189733 system, simulated transits for planets of various radii
541: and orbital parameters were inserted
542: into the {\it MOST} photometry and Monte Carlo statistics of the
543: transit recovery rate were generated.
544: We thus proceed under the assumption that all signals in the data-set are noise,
545: and we place as sensitive of limits as possible under this paradigm.
546: Realistic limb-darkened transits due to planets with various radii,
547: $R_{p}$, orbital phases, $\phi$, periods, $P$, and inclinations, $i$,
548: were inserted into the 2006 {\it MOST} HD 189733 data. The full non-linear model of \citet{Mandel} was used
549: to ensure the inserted transits were as realistic as possible.
550: Circular
551: orbits were used. The M-dwarf companion star, HD 189733B, would have a negligible
552: effect on the orbit of a close-in putative planet in orbit around HD 189733, and thus the effect of this binary companion can be
553: safely ignored.
554:
555: These
556: modified data were then subjected to the transit search algorithm
557: as discussed in $\S$\ref{SecTrans}. It should be noted that this includes
558: the automated filtering step,
559: and thus the synthetic transits were injected
560: before the same filtering process our original data was subjected to,
561: therefore ensuring the validity of the following
562: Monte Carlo limits.
563: Transits were inserted with logarithmic period spacing (as discussed
564: in \citealt{Croll}), with $\eta_{inp} = 0.04$ in the period
565: range $0.45 d < P_{inp} < 6.7 d$. In total, 36 period steps were
566: used for the 90$^{o}$ inclination angle cases.
567: For the 86$^{o}$ and 82$^{o}$ inclination angle cases transits were inserted with this same logarithmic
568: period spacing until the period exceeded
569: the maximum period, or semi-major axis, that would produce a transit, a $<$ $(R_{*}+R_{P})$/cos$i$.
570: For each trial period, simulated transits corresponding to 20
571: different exoplanet radii were inserted, sampling the period-radius
572: space of interest. For each period and radius at least \HowMany \ phases were
573: inserted.
574: For each of these points, the phase $\phi$ was generated
575: randomly to be in the range 0 $\le$ $\phi$ $<$ 1.
576: Because a 2.4 Ghz Pentium processor with 1 Gbyte of memory can perform
577: the transit search algorithm on an individual MOST HD 189733 data set
578: in $\sim$1 minute, exploration of the entire grid just mentioned for
579: all inclinations involves $\simeq$3$\times$$10^5$ iterations and 0.6 CPU years.
580: The calculation was performed on the LeVerrier Beowulf cluster in
581: the Department of Physics and Astronomy at the University of British
582: Columbia using 46 dual-CPU compute nodes.
583:
584: An inserted transit
585: was judged to be detected if the parameters $\phi$ and $P$ returned
586: by the transit search algorithm were sufficiently close to the input
587: values: $\phi_{inp}$ and $P_{inp}$. The returned period had to satisfy
588: the following criteria: $|P - P_{inp}| < 0.05 d$ and $|P/P_{inp} - 1| <
589: 1\%$. The limits of the criterion for $\phi$ were dependent on the planetary orbital period,
590: because as the period of the putative transiting planet decreases,
591: the fractional transit length increases
592: accordingly. Thus the accuracy required in the determination of
593: $\phi$ was relaxed for shorter periods. The criterion on $\phi$
594: is: $|\phi-\phi_{inp}|$ $<$ $0.09 - 0.0106$ $\times$ ($P_{inp}$ - 0.4$d$)/$d$.
595: This corresponds to accepting an error in phase of up to 9\% at 0.4$d$, but only up to
596: 2\% at 7$d$.
597: Obvious multiples of the period of the inserted
598: planet, up to $4$ times the inserted period, $P_{inp}$,
599: as well as half-period ($P \approx \frac{1}{2}P_{inp}$), and one-third period ($P \approx \frac{1}{3}P_{inp}$)
600: solutions were also accepted. For the Monte Carlo tests when a transit was deemed recovered
601: the correct period was recovered 96\% of the time,
602: while the remainder returned a period that was double, triple, quadruple, one-half, or one-third
603: of the inserted period, $P_{inp}$.
604:
605: The level above which a
606: transit recovered from the data-set
607: could be considered believable was also determined due to these Monte Carlo tests, at a level
608: of $\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ $\ge$ $\TransAntiBelieve$ that
609: ruled out 95\% of the spurious candidates. Spurious candidates
610: are defined as those candidates where the parameters returned by the
611: transit search routine are different from the parameters inserted for the Monte Carlo tests.
612: The motivation for this criterion is to ensure that a systematic event is
613: misidentified as a believable transiting planet less than
614: 5\% of the time. Interestingly we can rigorously rule out these spurious candidates,
615: with our criterion set at a level where the most significant transit candidate
616: can be less significant than the most significant anti-transit (brightening) event.
617: Further discussion on this criterion is given in \citet{Croll}.
618: So as to not unduly skew this statistic we only generate this statistic using
619: putative planets with radii large enough that the routine has a realistic chance of detecting these
620: planets. We judge this to be planets greater than approximately the 25\% contour of our Monte Carlo statistics
621: without the RS criterion. Thus only planets with radii greater than the following limit are used:
622: $R_{Pinp} > (0.10 + \frac{0.05}{6.3} \times (P_{in}-0.4d)/$d$ )$ $R_{J}$.
623:
624: \subsection{Monte Carlo results}
625:
626: The fractions of times that the artificially inserted transits were
627: recovered from the data for various radii, periods, and inclination
628: angles are given in Figure \ref{FigMonte}. Also indicated in Figure \ref{FigMonte} is the 68\% contour limit
629: that would be placed without the $\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ $\ge$ \TransAntiBelieve \
630: criterion.
631: The close agreement between the 68\% contour with or without this criterion
632: indicates that this criterion does not seriously impact the sensitivity of the Monte Carlo statistics generated, while providing
633: a robust limit so as to avoid false positives and spurious detections.
634:
635: We also explore the distribution of spurious signals across period-radii space.
636: These values are displayed in Figure \ref{FigMonteFalse}, and indicate that spurious signals
637: are negligible (below 1\%) other than the intermediate areas where the routine
638: has a small to moderate chance to correctly determine
639: the inserted transit.
640: This result is expected, as it is similar to the result of \citet{Croll} for HD 209458.
641:
642: The performance of our routine is also expected to be slightly degraded
643: at harmonics and subharmonics of the periods that the sinusoids
644: were removed from the data at ($P$=1.0$d$, and $P$=0.5$d$).
645: Due to the fact that
646: the fractional length of a transit is comparitively large for short periods of 0.5 - 1.0 $d$,
647: it can be difficult to
648: completely differentiate and properly recover a transit signal near the
649: harmonics and subharmonics of these periods in all cases.
650: The period ranges demonstrating these drops in survey sensitivity
651: are very narrow ($\frac{ \Delta P }{ P }$ $\approx$ 2\%).
652: Limited statistical tests confirm that the loss in sensitivity near these periods is quite small,
653: similar to that of HD 209458 \citep{Croll}.
654:
655: The harmonics
656: and subharmonics of the orbital period of the known planet HD 189733b, $P$ $\approx$
657: 2.218573d \citep{Bakos},
658: are also expected to show decreased sensitivity to transits
659: due to the fact that the transits of the known planet were excised from the data.
660: This reduction of sensitivity
661: at these periods is unfortunate considering that it is expected that low-mass planets show a preference for
662: these mean-motion resonant orbits (\citealt{Thommes}; \citealt{Papaloizou} and \citealt{Zhou}).
663: The fractional length in phase that was removed at $P$ $\approx$ $2.218573$ to exclude the transit of the
664: known planet was 4.0\%
665: of the total.
666: Thus for the subharmonics of the known planet ($P$ $\approx$ 4.44$d$, 6.66$d$) in
667: approximately 4.0\% of the
668: cases the routine will be unable to recover a putative transiting planet, as the
669: transits will coincide with the transits of the known planet, and this information
670: will have been completely removed.
671: For
672: the harmonics of the orbital period of HD 189733b, the
673: situation is better, since not all of the transit events
674: would have been removed from the data.
675: For periods near the first harmonic ($P$ $\approx$ 1.11$d$), every second occurance of a transit will be lost
676: in only 8.0\% of the cases. For the second harmonic ($P$ $\approx$ 0.74$d$),
677: every third transit would be missed in 12.0\% of the cases.
678: For periods near these values the sensitivity limits can be expected to be
679: only slightly degraded from those shown in Figure \ref{FigMonte}. Even for exact harmonics
680: the routine's sensitivity should be only marginally worse than the limits quoted in Figure \ref{FigMonte},
681: as the remaining transits should allow the correct period (or a multiple of the period) to be recovered.
682: The mean-motion resonances from the known exoplanet HD 189733b have been examined
683: using the transit-timing technique
684: with ground-based photometry by \citet{WinnPsi}, HST photometry by \citet{PontPrep},
685: and with {\it MOST} photometry
686: by \citet{MillerTwo}.
687: It should also be noted that the detection limits presented
688: here are valid for circular orbits, but are largely applicable to orbits of other eccentricities,
689: as summarized for the analogous case of HD 209458 in \citet{Croll}.
690:
691:
692: We have made the simplifying assumption that the hypothetical planet
693: that transits the star to produce the synthetic transits of our Monte Carlo statistics does
694: not pass over prominent starspots along our line of sight.
695: The known planet has not been found to
696: occult a prominent starspot in the {\it MOST} photometry \citep{MillerTwo},
697: although a preliminary investigation of HST photometry
698: of HD 189733 indicates that the planet does occult small
699: to moderate sized starspots during their observations
700: \citep{PontPrep}.
701: Given that the rotational
702: period of the star is greater
703: than the period
704: range that is investigated here,
705: the limits presented here would likely only be slightly adversely affected
706: if this hypothetical planet does pass in front of a prominent starspot. This is because
707: the impact on the transit dip of the planet occulting
708: a large starspot would be relatively short, and should thus not severely affect our limits.
709:
710: Thus the search routine described above in $\S$\ref{SecTrans}
711: should be able to detect planets with radii greater than the limits
712: given in Figure \ref{FigMonte} with the {\it MOST} data. In the most optimistic case of an edge-on $90^{o}$ inclination angle transit, for planets
713: with periods from approximately half-a-day to one week, this limit is
714: approximately \Rjupsmall \ to \Rjuplarge \ $R_{J}$, or \Rearthsmall \
715: to \Rearthlarge \ $R_{\earth}$, respectively, with 95\% confidence.
716: These limits would scale up, or down linearly if the expected stellar radius,
717: $R_{*}$ = 0.755 $R_{\odot}$, is an over- or under-estimate, and are thus perhaps better
718: expressed as $R_p$/$R_{*}$ = \Rratiosmall \ and $R_p$/$R_{*}$ = \Rratiolarge \ for periods
719: from half-a-day to one week, respectively.
720: For periods at the upper range of those investigated here ($P$ $>$ 6.5$d$)
721: the routine is unable to pick out
722: the transit in a low percentage of cases, because the cubic spline
723: has removed significant power at that period; this effect can be observed
724: as a reduction of sensitivity for longer periods
725: in the ninety-degree inclination angle ($i$=90$^{o}$) panel of Fig. \ref{FigMonte}.
726: The performance of our automatic routine in this period-radius space probably underestimates
727: the true sensitivity of our survey, as it would be obvious that the cubic spline
728: has overcorrected in these cases through an individual inspection of the phased curve.
729: Also, limited statistical tests have shown that marginally improved limits could
730: probably be set for the shortest periods (P $\le$ 3$d$) by excluding the last seven days
731: of observations with increased noise, and only producing these limits with the first 14 days
732: of the highest quality data.
733:
734: % If one assumes a mean density of $\rho$ $\approx$ 3000 kg m$^{-3}$
735: % - a value averaging the various extrasolar Super-Earth models discussed below -
736: % their respective masses would be \Massjupitersmall \ and \Massjupiterlarge \ $M_{J}$,
737: % or \Massearthsmall \ and \Massearthlarge \ $M_{\earth}$, respectively.
738: % If the putative planet were gaseous, and thus a possible hot Neptune analogue ($\rho$ $\approx$ 1600 kg m$^{-3}$),
739: % our upper radius limit, $R_{P}$ = \Rearthlarge \ $R_{\earth}$, would result
740: % in a planetary mass of \MassNeptune \ $M_{\earth}$.
741: % Thus the mass-period parameter space
742: % we have ruled out in this study, assuming this hypothetical mean density,
743: % is the most sensitive limit set to date in this system for near edge-on inclination angles for our period range.
744:
745:
746:
747: \begin{figure}
748: \epsscale{0.6}
749: \includegraphics[angle=270, scale = 0.52]{f3a.eps}
750: \includegraphics[angle=270, scale = 0.52]{f3b.eps}
751: \includegraphics[angle=270, scale = 0.52]{f3c.eps}
752: \caption{
753: {Confidence limits of transit detection as a function of planet radius and orbital period, for different orbital
754: inclinations, based on Monte Carlo statistics. The crosses represent the radii and periods at which synthetic
755: transits were inserted in the data. The dotted, thick-solid, and thin-solid lines represent the 99, 95 and 68\%
756: confidence
757: contours, respectively.
758: The thick dot-dashed curve represents the 68\% confidence contour if the criterion $\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ $\ge$ \TransAntiBelieve \
759: was not used.
760: The near vertical dot-dash line in the later panels indicates
761: the maximum period that produces a transit for that given inclination angle.
762: Note the logarithmic
763: period scaling on the x-axis.
764: At least \HowMany \ phases were inserted for each of the radii-phase points.
765: }
766: \label{FigMonte}
767: }
768: \end{figure}
769:
770:
771:
772: \begin{figure}
773: \epsscale{0.6}
774: \includegraphics[angle=270, scale = 0.52]{f4a.eps}
775: \includegraphics[angle=270, scale = 0.52]{f4b.eps}
776: \includegraphics[angle=270, scale = 0.52]{f4c.eps}
777: \caption{
778: {The likelihood of spurious transit detections returned by the Monte Carlo statistical analysis. The
779: dotted, thick-solid and thin-solid lines represent the 10, 5 and 1\% spurious signal contours, respectively.
780: The format of the figure is otherwise identical to Figure \ref{FigMonte}. Note that the
781: spurious signals occur in the intermediate regions of the period-radii space of interest,
782: where it is not guaranteed that the correct transit will be recovered,
783: but the inserted transit still causes significant deviations to the light-curve.
784: }
785: \label{FigMonteFalse}
786: }
787: \end{figure}
788:
789:
790:
791:
792:
793: \section{HD 189733 transit search}
794: \label{SecHDTrans}
795:
796: The {\it MOST} HD 189733 2006 data-set was submitted to the analysis outlined in $\S$\ref{SecTrans}.
797: There were no transiting planet candidates that met the detection criteria.
798: As the Monte Carlo tests indicate we would be able to detect
799: exoplanets in this system with periods
800: from half-a-day to one week with radii larger than the radius limits given in Figure
801: \ref{FigMonte}, we can safely rule out these sized-exoplanets in this system.
802:
803: We also present the
804: details of the candidate with the greatest improvement in $\chi^{2}$, even though evidence for its
805: existence is marginal.
806: The phased diagram of this candidate (Figure \ref{FigCandidate}) is not particularly convincing,
807: and
808: this candidate did not meet the RS detection criterion,
809: ($\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ $\approx$ \PlanetChiPerc \ $\not\ge$ $\TransAntiBelieve$).
810: For these reasons we do not report it as a probable transiting candidate.
811: The specifics of this putative candidate are
812: given in Table \ref{TableCandidate}.
813: This event, with a period of $P \approx \PlanetPeriod$$d$,
814: has modest statistical significance, and thus
815: we report it is a possible, but unlikely, candidate. This candidate
816: would correspond to a putative hot Super-Earth, with
817: period P = \PlanetPeriod $d$, radius $R_{p}$ = \PlanetRadius $R_{J}$ (\PlanetRadiusEarth $R_{\oplus}$),
818: and inclination $i$ = \PlanetInc$^{o}$. Interestingly the period of this planet places it quite close to the
819: 1:3 interior mean-motion resonance to the known planet ($P$$\sim$0.74$d$).
820: For $\rho$ $\approx$ 3000 kg m$^{-3}$ the mass of this putative planet would be
821: approximately \PlanetMassEarth \ $M_{\earth}$ (\PlanetMassJupiter \ $M_{J}$).
822: A transit time combined with the supposed period is given in Table \ref{TableCandidate}.
823: Evidence for a transit at this period is marginal, but additional {\it MOST} photometry of the HD 189733
824: system should confirm or deny its existence.
825: % The transit timing investigation of HD 189733 of
826: % \citet{MillerTwo} should also be able to rule out a possible candidate near this mean-motion resonance.
827:
828: The most significant brightening event was one observed with a period of approximately $P \approx 6.28 d$.
829: It is likely statistical in nature, and thus is not expected to be related to any specific astrophysical process.
830:
831: Although our transit search covers a continuous set of periods, we do
832: not expect the existence of another planet with a semimajor axis `too close'
833: to that of the known planet HD189733b, since the gravitational perturbations
834: of the known Jupiter-scale planet will destabilize the orbit of another
835: object.
836: Taking the known planet's data - $a_b$=0.0313~AU, and mass
837: $M_b$=1.15~$M_{Jup}$ \citep{Bouchy} in orbit around a star of mass
838: $M_*$=0.82 solar masses \citep{Bakos} - orbits in an
839: annulus
840: $ a_{in} = (1-\Delta) a_b < a_b < (1+\Delta) a_b = a_{out} $
841: will be unstable. The fractional width $\Delta$ of the
842: annulus is given by \citep{Gladman}:
843: $ \Delta = 2.40 ( M_b / M_* )^{1/3} $
844: where the assumption is being made that the mass of the second
845: planet is much less than the known planet (if not, the width
846: of the zone scales as the cube root of the sum of the masses of the planets).
847: Here $\Delta$=0.26, and so we calculate that planets with periods in the
848: range 1.41 -- 3.14 days will be intrinsically unstable due to the
849: presence of the known planet.
850: Thus, some of the parameter space searched in Figure \ref{FigMonte} was very
851: unlikely to have been occupied by a second planet - Trojans lagging or leading \citep{Ford} HD 189733b being a
852: notable exception.
853:
854: We can also exclude transits from planets with longer orbital periods. There were no
855: transits from edge-on ($i$ = $90^{o}$) planets with radii greater than $\sim$0.35
856: $R_{J}$ (3.9 $R_{\earth}$) during {\it MOST}'s observations of HD 189733,
857: as a planet of this size would be readily visible with even a cursory
858: inspection of the data. A planet of this size would correspond to $\sim$ 20
859: $M_{\earth}$, as calculated from the
860: models of \citet{Fortney} assuming a composition similar to Neptune. We can
861: also place a limit on transiting planets with orbital periods from 7 $d$ $<$ P $<$
862: 10.5 $d$ - in this region we would expect two transits in our data-set. Limited
863: statistical tests indicate that we can rule out planets with radii greater than
864: approximately $\sim$0.32 $R_{J}$ (3.6 $R_{\earth}$) in this region - this
865: corresponds to $\sim$14 $M_{\earth}$ \citep{Fortney}, again assuming a
866: composition similar to Neptune. We do not provide a formal limit from Monte
867: Carlo estimates for this range of periods as we prefer to restrict our formal transit
868: search to periods that will result in three transits in our data-set.
869:
870: As an additional sanity check this transit search method was applied to the current data-set
871: without the removal of the transits
872: of the known planet, HD 189733b.
873: As the transits are apparent in the unbinned data (Figure \ref{FigData}), this causes
874: the cubic-spline to over-correct and remove a moderate portion
875: of the signal at this period. Even so, the routine correctly uncovers the transit of HD 189733b with reasonable accuracy.
876: The actual parameters of HD 189733b \citep{Bakos} were recovered within
877: 0.04\% and 0.3\% in period and phase, respectively, while the inclination angle
878: and radius were recovered within 1.0$^{o}$ and 0.20$R_{J}$ of the actual parameters.
879: The moderate discrepancy between
880: the inserted and recovered radius is due to the fact that the cubic spline overcorrects and
881: removes a portion of the transit-signal.
882:
883: %%%%%%%%%%%%
884: % Actual Period P=2.218573
885: % Eph 2453629.39420 -2451545 = 2084.3942 + 144*P = 2403.868712
886: % i = 86
887: % R = 1.154 (all of these from Bakos, Knutson 2006)
888: %%%%%%%%%%%%
889:
890:
891: \begin{deluxetable}{cc}
892: \tabletypesize{\footnotesize}
893: \tablecaption{Transit candidate of marginal significance \label{TableCandidate}}
894: \tablewidth{0pt}
895: \tablehead{\colhead{parameter}& \colhead{value} \\}
896: \startdata
897:
898:
899: $P$ & \PlanetPeriodLong d \\
900: $R_{p}$ & \PlanetRadiusEarth $R_{\oplus}$ (\PlanetRadius $R_{J}$) \\
901: $i$ & \PlanetInc$^{o}$ \\
902: Ephemeris Minimum (JD-2451545) & \PlanetEphemeris \\
903: $\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ & \PlanetChiPerc \\
904: Mass (assuming $\rho$ $\approx$ 3000 kg m$^{-3}$) & \PlanetMassEarth \ $M_{\earth}$ (\PlanetMassJupiter \ $M_{J}$)\\
905: \enddata
906: \end{deluxetable}
907:
908:
909:
910:
911:
912:
913: \begin{figure}
914: \includegraphics[width=1.5in, height = 3.5in, angle=270]{f5a.eps} % 2.3, 6.0
915: \includegraphics[width=1.5in, height = 3.5in, angle=270]{f5b.eps}
916: \caption{
917: {The best transit candidate as identified by our analysis of {\it MOST}'s HD 189733 2006
918: data-set, but with marginal significance. The thick red-line represents the transit model,
919: while the thin green-dashed line represents the constant
920: brightness model.
921: At top, the unbinned data are shown in blue, and the binned data are shown in black;
922: at bottom the binned data only.
923: As this candidate failed the improvement in transit over anti-transit ratio statistic (RS) criterion,
924: we report it as a possible, but unlikely candidate.
925: The period and radius of this putative planet would be
926: approximately \PlanetPeriod$d$ and \PlanetRadius $R_{J}$ (\PlanetRadiusEarth$R_{\oplus}$).}
927: \label{FigCandidate}
928: }
929:
930:
931: \end{figure}
932:
933:
934:
935:
936:
937:
938:
939:
940:
941:
942: \section{Discussion}
943: \label{SecDiscuss}
944:
945: {\it MOST}'s 2006 observations of HD 189733
946: have been searched for evidence of other exoplanets in the system.
947: The transit-search routine described in \citet{Croll} has been used to search for
948: realistic limb-darkened transits to take advantage of the precise, near-continuous photometry
949: returned by {\it MOST} and the fact that the
950: stellar characteristics of the star, HD 189733, are well established.
951: Monte Carlo statistics were generated using this routine and indicate that
952: this routine in combination with the aforementioned {\it MOST} data have placed robust
953: limits in the size of transiting bodies that have been ruled out in this system
954: for a range of periods and inclination angles. In the most optimistic case
955: of edge-on transits, planets with radii greater than
956: \Rjupsmall \ $R_{J}$ ($\sim$\Rearthsmall \ $R_{\oplus}$) to \Rjuplarge \ $R_{J}$ ($\sim$\Rearthlarge \ $R_{\oplus}$) with
957: periods from half-a-day to seven days, respectively,
958: have been ruled out with 95\% confidence through this analysis. For orbits co-planar with
959: that of the known planet ($i$ $\approx$ 86$^{o}$)
960: the limits are similar, except for longer periods ($P$$>$4$d$). For these longer periods the planets cease
961: to transit
962: the star along our line of sight and thus we cannot place firm limits for periods greater than 4-days at
963: $i$ $\sim$ 86$^{o}$, or for $i$ $<$ 86$^{o}$.
964: Specifically, we have been able to rule out
965: transiting planets in this system with radii greater
966: than those given in Figure \ref{FigMonte}.
967: This work has constrained theories that predict hot Super-Earths, and hot Neptunes
968: due to the inward migration of
969: HD 189733b, such as those proposed by \citet{Zhou}, \citet{Raymond}, \citet{Mandell},
970: and \citet{Fogg}.
971: % Further {\it MOST} photometry should be able to improve the
972: % radii-period limit that was set in this work, and further constrain or confirm these thories that predict
973: % hot Earths, and hot Neptunes in nearby orbits to the
974: % known planet.
975:
976: It should also be noted that we are able to place very sensitive limits on the size of planets that can be ruled
977: out in this system despite the prominent $\sim$3\% rotational modulation observed in this system. By removing
978: the observed rotational modulation using a cubic spine with a binning time longer than the expected duration
979: of the transit of the planet across the star, we are able to remove the effects of the prominent modulation
980: while remaining sensitive to the transits
981: of planets as small as those expected from hot Super-Earths and hot Neptunes. These methods should
982: be applicable to current or future space-based transit search missions, such as COROT and Kepler.
983:
984:
985: We are also able to place limits on the size of Trojan planets that have been ruled out in this analysis,
986: assuming the Trojan consistently transits the star with an inclination angle close to that
987: of the known planet ($i$$\approx$86$^{o}$).
988: We use the definition of Trojans given by \citet{Ford} as objects occupying the L4 or L5 stable Lagrange points.
989: Trojans lagging or leading
990: HD 189733b with a radius above $R$ = \RTrojanJupiter \ $R_{J}$ (\RTrojanEarth $R_{\oplus}$) should have
991: been detected with 95\%
992: confidence. Using a mean density of $\rho$$\approx$ 3000 kg m$^{-3}$ this corresponds
993: to \MTrojanEarth \ $M_{\oplus}$.
994: A detailed study of this photometry should be able to radically improve this limit on the size of Trojans that can be ruled out with
995: this photometry, assuming the Trojan consistently transits the star along our line of sight.
996:
997:
998: \subsection{Mass constraints on other Exoplanets in the system}
999:
1000: The above limits on the smallest planetary radii excluded by {\it MOST}
1001: for transiting planets allow some estimates on the type of planets
1002: that are being excluded in this system.
1003: For Gas Giant planets, including hot Neptunes, radii change with time as the planet
1004: cools, so we make use of the fact that the age of the HD 189733 system
1005: is greater than 1 Gyr \citep{Melo}. We follow the suggestion of \citet{WinnPsi}
1006: and adopt an age of $\sim$4.5 Gyr.
1007: We thus can refer to cooling
1008: models similar to the planets in the Solar System. One exception would
1009: be the consideration of possible planets in orbits smaller than that of
1010: HD189733b, where tidal heating will lead to larger radii per given mass
1011: and composition.
1012:
1013: Our radius limit of \Rearthsmall \ - \Rearthlarge \ $R_{\earth}$ already places us in the realm of Super-Earth
1014: and hot Neptune planets. \citet{Fortney} has recently provided theoretical predictions for exoplanetary
1015: radii for a diverse range of possible exoplanetary compositions, including hot Super-Earth, and hot Neptune planets.
1016: The Super-Earth models of \citet{Fortney} agree well with the more detailed Super-Earth
1017: models of \citeauthor{ValenciaA} (\citeyear{ValenciaA}, \citeyear{ValenciaB}).
1018: For our Super-Earth exoplanetary limits we use equation 7 and 8 of \citet{Fortney}.
1019: Our inner period limit of $R_{P}$ $<$ \Rearthsmall $R_{\earth}$ allows us to rule out
1020: planets more massive than \MearthsmallIce $M_{\earth}$ for a pure ice planet, \MearthsmallRock $M_{\earth}$ for a pure rock planet,
1021: and \MearthsmallIron $M_{\earth}$ for a pure iron planet.
1022: Our outer period limit of $R_{P}$ $<$ \Rearthlarge $R_{\earth}$ only allows to place a useful limit on pure ice Super-Earth planets.
1023: At this limit we are able to rule out planets more massive
1024: than \MearthlargeIce $M_{\earth}$ for pure ice planets. Pure rock and pure iron planets of this size are moderately less massive,
1025: and comparable
1026: in mass, respectively, to HD 189733b itself - planets of this mass, of course,
1027: have already been ruled out in this system with radial-velocity
1028: measurements (\citealt{WinnLambda} as quoted in \citealt{MillerTwo}).
1029: We note that our limit on ice planets may be more sensitive than the limits quoted
1030: here as one would expect that ice planets at such small orbital periods would have steam atmospheres, and would thus
1031: be significantly inflated at these small orbital seperations \citep{Kuchner,Fortney}.
1032:
1033: We are also able to place limits on the mass of hot Neptune-type planets that have been ruled out in our analysis.
1034: Hot Neptune planets should have cores similar in density to giant ice-planets, and an outer atmosphere of H/He \citep{Fortney} - they
1035: should thus be moderately less dense than giant-terrestrial ice planets, and thus be of a similar density to Neptune in our own solar
1036: system \citep{Fortney}.
1037: Our inner period radius limit ($R$ $<$ \Rearthsmall $R_{\earth}$) excludes hot Neptunes entirely, as
1038: hot Neptune exoplanets are expected to be of similar size to Neptune (3.9 $R_{\earth}$)
1039: (\citealt{Fortney}; \citealt{ValenciaA}; \citealt{ValenciaB}).
1040: At our outer period radius limit ($R$ $<$ \Rearthlarge $R_{\earth}$) we can rule out hot Neptune
1041: exoplanets more massive than
1042: \MassNeptune \ $M_{\earth}$.
1043:
1044: The radius limits presented here,
1045: combined with the mass-limits from transit-timing analyses (\citealt{WinnPsi}; \citealt{MillerTwo}; \citealt{PontPrep}),
1046: place very firm constraints
1047: on the sizes and masses of bodies that could still reside in close orbits to HD 189733.
1048: Re-observations of this system planned with {\it MOST}
1049: should be able to further reduce the size of planets that could remain undetected in this
1050: system, and further constrain theories that predict hot Super-Earth, and hot Neptune planets in similar orbits to
1051: that of the known hot Jupiter exoplanet HD 189733b. HD 189733 thus joins the growing list of hot Jupiter exoplanetary systems
1052: (HD 209458 [\citealt{Croll,Miller,Agol}], TrES-1 [\citealt{Steffen}])
1053: for which firm constraints have been placed
1054: on the size of other exoplanets in similar orbits to that of the known hot Jupiter by a variety of methods.
1055:
1056:
1057:
1058: \acknowledgements
1059: We would like to thank the anonymous referee for a constructive review.
1060: The Natual Sciences and Engineering Research Council of Canada supports the research of D.B.G., J.M.M., A.F.J.M., J.F.R., S.M.R., G.A.H.W., and B.G..
1061: Additional support for A.F.J.M. comes from FCAR
1062: (Qu\'ebec). R.K. and A.W., are supported by the Canadian Space Agency. W.W.W. is supported by the Austrian Space Agency and the Austrian Science Fund (P17580).
1063: The Canadian Foundation for Innovation, and the BC Knowledge Development Fund
1064: provided funding for the
1065: LeVerrier Beowulf cluster.
1066:
1067: \begin{thebibliography}{}
1068:
1069: \bibitem[Agol \& Steffen (2007)]{Agol}
1070: Agol, E. and Steffen, J.~H. 2007, \mnras, 374, 941
1071:
1072: \bibitem[Baglin (2003)]{Baglin}
1073: Baglin, A. 2003, Advances in Space Research, Volume 31, 345-349.
1074:
1075: \bibitem[Baines et al.(2007)]{Baines}
1076: Baines, E.~K. et al. 2007, astro-ph/0704.3722
1077:
1078: \bibitem[Bakos et al.(2006a)]{Bakos}
1079: Bakos, G.A., Knutson, H., Pont, F., Moutou, C., Charbonneau, D., Shporer, A., Bouchy, F., Everett, M., Hergenrother, C., Latham, D.W., Mayor, M., Mazeh, T., Noyes, R.W., Queloz, D., Pal, A., Udry, S. 2006, \apj, 650, 1160
1080:
1081: \bibitem[Bakos et al.(2006b)]{BakosPal}
1082: Bakos, G.A., Pal, A., Latham, D.W., Noyes, R.W., Stefanik, R.P. 2006. \apjl, 641, L57
1083:
1084: \bibitem[Barge et al.(2005)]{Barge}
1085: Barge, P., Baglin, A., Auvergne, M., Buey, J.-T., Catala, C., Michel, E., Weiss, W. W., Deleuil, M., Jorda, L., Moutou, C.,
1086: COROT Team 2005, SF2A-2005: Semaine de l'Astrophysique Francaise, eds. F. Casoli, T. Contini, J.M. Hameury \& L. Pagani, EdP-Sciences, Conference Series, p. 193.
1087:
1088: % Kepler Paper Review Article Good
1089: \bibitem[Basri et al.(2005)]{Basri}
1090: Basri, G., Borucki, W.~J., Koch, D. 2005, New Astronomy Review, 49, 478.
1091:
1092: \bibitem[Borucki et al.(2004)]{Borucki}
1093: Borucki, W.; Koch, D.; Boss, A.; Dunham, E.; Dupree, A.; Geary, J.;
1094: Gilliland, R.; Howell, S.; Jenkins, J.; Kondo, Y.; Latham, D.; Lissauer,
1095: J.; Reitsema, H. 2004, Second Eddington Workshop: Stellar structure and
1096: habitable planet finding, eds. F. Favata, S. Aigrain \& A. Wilson,
1097: ESA SP-538, Noordwijk: ESA Publications Division, 177 - 182.
1098:
1099: \bibitem[Bouchy et al.(2005)]{Bouchy}
1100: Bouchy, F., Udry, S., Mayor, M., Moutou, C., Pont, F., Iribarne, N., Da Silva, R., Ilovaisky, S., Queloz, D., Santos, N.C., Segransan, D., Zucker, S. 2005, \aap, 444, L15
1101:
1102: \bibitem[Burke et al.(2006)]{Burke}
1103: Burke, C.J., Gaudi, B.S., DePoy, D.L., Pogge, R.W. 2006, \aj, 132, 210
1104:
1105: \bibitem[Burke et al.(2007)]{BurkeSeven}
1106: Burke, C.J. et al. 2007, \apj, astro-ph/0705.0003
1107:
1108: \bibitem[Collier-Cameron et al.(2006)]{Collier}
1109: Collier-Cameron et al. 2006, \mnras, 373, 799
1110:
1111: \bibitem[Croll et al.(2007)]{Croll}
1112: Croll, B., Matthews, J.M., Rowe, J.F., Kuschnig, R., Walker, A., Gladman, B.,
1113: Sasselov, D., Cameron, C., Walker, G.A.H., Lin, D.N.C., Guenther, D.B., Moffat, A.F.J.,
1114: Rucinski, S.M., Weiss, W.W. 2007, \apj, 658, 1328
1115:
1116: \bibitem[Croll et al.(submitted)]{CrollSpot}
1117: Croll, B. et al. \apj, submitted 13 Apr 2007
1118:
1119: \bibitem[Deeg et al.(1998)]{Deeg}
1120: Deeg, H.~J. et al. 1998, \aap, 338, 479
1121:
1122:
1123: \bibitem[Deming et al.(2006)]{Deming}
1124: Deming, D., Harrington, J., Seager, S., Richardson, L.~J. 2006, \apj, 644, 560
1125:
1126: \bibitem[Doyle et al.(2000)]{Doyle}
1127: Doyle, L.~R. et al. 2000, \apj, 535, 338
1128:
1129: \bibitem[Fogg \& Nelson (2007)]{Fogg}
1130: Fogg, M.~J., Nelson, R.~P. 2007, \aap, 461, 1195
1131:
1132: \bibitem[Ford \& Gaudi (2006)]{Ford}
1133: Ford, E., Gaudi, B.S. 2006, \apjl, 652, L137
1134:
1135: \bibitem[Fortney (2007)]{Fortney}
1136: Fortney, J.~J., Marley, M.~S., Barnes, J.~W., 2007, \apj, 659, 1661
1137:
1138: \bibitem[Gaudi \& Winn(2007)]{Gaudi}
1139: Gaudi, B.~S., Winn, J.~N. 2007, \apj, 655, 550
1140:
1141: \bibitem[Gladman(1993)]{Gladman}
1142: Gladman, B. 1993, Icarus, 106, 247
1143:
1144: \bibitem[Grillmair et al.(2007)]{Grillmair}
1145: Grillmair, C.~J., Charbonneau, D., Burrows, A., Armus, L., Stauffer, J., Meadows, V., Van Cleve, J., Levine, D.,
1146: 2007, \apjl, 658, L115
1147:
1148: \bibitem[Hebb et al.(2007)]{Hebb}
1149: Hebb, L. et al. 2007, astro-ph/0704.3584
1150:
1151: \bibitem[Kov{\'a}cs et al.(2002)]{Kovacs}
1152: Kov{\'a}cs, G., Zucker, S., Mazeh, T. 2002, \aap, 391, 369
1153:
1154: \bibitem[Knutson et al.(2007)]{Knutson}
1155: Knutson, H.A. et al. 2007, Nature, 447, 183
1156:
1157: \bibitem[Kuchner (2003)]{Kuchner}
1158: Kuchner, M.J. 2003, \apjl, 596, L105
1159:
1160: \bibitem[Mandel \& Agol(2002)]{Mandel}
1161: Mandel, K., Agol, E. 2002, \apj, 580, L171
1162:
1163: \bibitem[Mandell et al.(2007)]{Mandell}
1164: Mandell, A.~M., Raymond, S.~N., Sigurdsson, S. 2007, astro-ph/0701048
1165:
1166: \bibitem[Matthews et al.(2004)]{Matthews}
1167: Matthews, J.M., Kusching, R., Guenther, D.B., Walker, G.A.H., Moffat, A.F.J., Rucinski, S.M., Sasselov, D., Weiss, W.W. 2004, Nature, 430, 51
1168:
1169: % \bibitem[Matthews et al. (in preparation)]{MatthewsHD}
1170: % Matthews, J.M. et al. 2007, \apj, in preparation
1171:
1172: \bibitem[Melo et al.(2006)]{Melo}
1173: Melo, C., Santos, N.C., Pont, F., Guillot, T., Israelian, G., Mayor, M., Queloz, D., Udry, S. 2006, \aap, 460, 251
1174:
1175: \bibitem[McLaughlin (1924)]{McLaughlin}
1176: McLaughlin, D.B. 1924, \apj, 60, 22
1177:
1178: \bibitem[Miller-Ricci et al.(2007a)]{Miller}
1179: Miller-Ricci, E., Rowe, J.F., Sasselov, D., Matthews, J.M., Guenther, D.B., Kuschnig, R., Moffat, A.F.J., Rucinski, S.M., Walker, G.A.H., Weiss, W.W.
1180: 2007, \apj, submitted 17 Oct 2006
1181:
1182: \bibitem[Miller-Ricci et al. (2007b)]{MillerTwo}
1183: Miller-Ricci, E., Rowe, J.F., Sasselov, D., Matthews, J.M., Guenther, D.B., Kuschnig, R., Moffat, A.F.J., Rucinski, S.M., Walker, G.A.H., Weiss, W.W.
1184: 2007, \apj, submitted 12 Apr 2007
1185:
1186: \bibitem[Papaloizou \& Szuszkiewicz (2005)]{Papaloizou}
1187: Papaloizou, J.C.B., Szuszkiewicz, E. 2005, \mnras, 363, 153
1188:
1189: \bibitem[Pont et al.(2006)]{Pont}
1190: Pont, F. et al. 2006, \mnras, 373, 231
1191:
1192: \bibitem[Pont et al.(submitted)]{PontPrep}
1193: Pont, F. et al. 2007, \aap, astro-ph/0707.1940
1194:
1195: \bibitem[Press et al.(1992)]{Press}
1196: Press, W.H., Flannery, B., Teukolsky, S.A., Vetterling, W.T. 1992. Numerical Recipes in Fortran, 2nd ed. (Cambridge University Press, Cambridge, England).
1197:
1198: \bibitem[Raymond et al.(2006)]{Raymond}
1199: Raymond, S., Mandell, A., \& Sigurdsson, S. 2006, Science, 313, 1413
1200:
1201: \bibitem[Rossiter (1924)]{Rossiter}
1202: Rossiter, R.A. 1924, \apj, 60, 15
1203:
1204: \bibitem[Rowe et al.(2006a)]{Rowe}
1205: Rowe, J.F., Matthews, J.M., Seager, S., Kuschnig, R., Guenther, D.B., Moffat, A.F.J., Rucinski, S.M., Sasselov, D., Walker, G.A.H., Weiss, W.W. 2006, \apj, 646, 1241.
1206:
1207: % \bibitem[Rowe et al. (in preparation)]{Rowe2}
1208: % Rowe, J.F., Matthews, J.M., Seager, S., Kuschnig, R., Guenther, D.B., Moffat, A.F.J., Rucinski, S.M., Sasselov, D., Walker, G.A.H., Weiss, W.W. 2006b, \apj, in preparation.
1209:
1210: \bibitem[Sahu et al.(2006)]{Sahu}
1211: Sahu, K.C., Casertano, S., Bond, H.E., Valenti, J., Smith, T.E., Minniti, D., Zoccali, M., Livio, M., Panagia, N., Piskunov, N., Brown, T.M., Brown, T., Renzini, A., Rich, R.M., Clarkson, W., Lubow, S. 2006, Nature, 443, 5
1212:
1213: \bibitem[Steffen \& Agol (2005)]{Steffen}
1214: Steffen, J.H., Agol, E. 2005, \mnras, 364, L96
1215:
1216: \bibitem[Thommes (2005)]{Thommes}
1217: Thommes, E. 2005, \apj, 626, 1033
1218:
1219: \bibitem[Valencia et al.(2006)]{ValenciaA}
1220: Valencia, D., O'Connell, R., \& Sasselov, D. 2006a, Icarus, 181, 545
1221:
1222: \bibitem[Valencia et al.(2007)]{ValenciaB}
1223: Valencia, D., Sasselov, D., \& O'Connell, R. 2007, \apj, 656, 545
1224:
1225: \bibitem[Walker et al.(2003)]{Walker}
1226: Walker, G. A. H., et al. 2003, \pasp, 115, 1023
1227:
1228: \bibitem[Winn et al.(2006)]{WinnLambda}
1229: Winn, J.~N., Johnson, J.~A., Marcy, G.~W., Butler, R.~P., Vogt, S.~S., Henry, G.~W., Roussanova, A.,
1230: Holman, M.~J., Enya, K., Narita, N., Suto, Y., Turner, E.~L. 2006, astro-ph/0609506
1231:
1232: \bibitem[Winn et al.(2007)]{WinnPsi}
1233: Winn, J.~N., Holman, M.J., Henry, G.W., Roussanova, A., Enya, K., Yoshii, Y., Shporer, A., Mazeh, T., Johnson, J.A., Narita, N., Suto, Y. 2007, \aj, astro-ph/0612224
1234:
1235: \bibitem[Wright et al.(2004)]{Wright}
1236: Wright, J.~T., Marcy, G.~W., Butler, R.~P., Vogt, S.~S. 2004, \apjs, 152, 261
1237:
1238: \bibitem[Zhou et al.(2005)]{Zhou}
1239: Zhou, J.L., Aarseth, S.J., Lin, D.N.C., Nagasawa, M. 2005, \apjl, 631, L85
1240:
1241: \end{thebibliography}
1242:
1243:
1244:
1245:
1246:
1247:
1248: \end{document}
1249:
1250:
1251:
1252: % Eccentric orbits will increase, decrease, or have little effect on the fractional length of the transit,
1253: % depending on whether the tranist occurs near apastron, periastron, or in between, repsectively.
1254: % Obviously the cases where the fractional transit length is increased will favourably affect the detection limits presented,
1255: % while the opposite cases will adversely affect the detection limits.
1256: % For low-eccentricity orbits this effect should be negligible, and thus the detection limits presented here should be largely
1257: % applicable to these orbits.
1258: % For non edge-on inclination angles, highly eccentric orbits will adversely affect the limits presented,
1259: % as a greater fraction of the time the planet will not transit the star along our line of sight.
1260: % For edge-on, high-eccentricity orbits the limits will likely be modestly, adversely affected due to the cases where the transit occurs near
1261: % periastron.
1262:
1263:
1264:
1265:
1266:
1267: % \begin{figure}
1268: % \includegraphics[height = 4.8in, width = 3.2in, angle=270]{HIST256_HISTO.eps}
1269: % \caption{
1270: % {A histogram showing the improvement of the transit over the anti-transit model ($\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$)
1271: % for all Monte Carlo candidates that are spurious (when the transit recovered by the transit routine is not the inserted transit).
1272: % The threshold level for $\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ above which
1273: % a transit can be considered significant is determined via these statistics as it is set
1274: % at a level, $\Delta\chi^{2}\%$/$\Delta\chi^{2}_{-}\%$ $\approx$ \TransAntiBelieve,
1275: % that would rule out 95\% of these spurious signals. The solid-vertical line thus
1276: % marks this threshold value.
1277: % }
1278: % \label{FigHisto}
1279: % }
1280: % \end{figure}
1281:
1282:
1283:
1284:
1285: % \begin{figure}
1286: % \includegraphics[angle=270, scale = 0.54]{contour_special_HD189733_0.5.eps}
1287: % \includegraphics[angle=270, scale = 0.54]{contour_special_HD189733_1.eps}
1288: % \includegraphics[angle=270, scale = 0.54]{contour_special_HD189733_1.5.eps}
1289: % \includegraphics[angle=270, scale = 0.54]{contour_special_HD189733_2.eps}
1290: %
1291: % \caption{
1292: % {The transit search sensitivity near the periods of 0.5 (top left) and 1.0$d$ (top right),
1293: % where sinusoidal terms were filtered
1294: % from the photometry. Subharmonics of these periods, 1.5 (bottom left) and 2.0$d$ (bottom right), are also shown.
1295: % The format is the same as Fig. \ref{FigMonte}. \HowMany \ phases were inserted for each period-radius point.
1296: % The automated routine probably slightly underestimates the true survey sensitivity near these periods, as
1297: % the non-sinusoidal nature of a transit would likely be obvious for relatively deep ($R_{P}$ $\approx$ 0.25 $R_{J}$)
1298: % transits.
1299: % }
1300: % \label{FigCutOut}
1301: % }
1302: % \end{figure}
1303: