1: %\documentstyle[aasms4,natbib]{article}

2: \documentclass[12pt,preprint]{aastex}

3: \def\kms{{\rm km\,s^{-1}}}

4: \def\pc{{\rm pc}}

5: \def\erf{{\rm erf}}

6: \def\masyr{{\rm mas}\,{\rm yr}^{-1}}

7: \def\usno{{\rm USNO}}

8: \def\nltt{{\rm NLTT}}

9: \def\rpm{{\rm RPM}}

10: \def\lim{{\rm lim}}

11: \def\pyr{{{\rm yr}^{-1}}}

12: \def\bol{{\rm bol}}

13: \begin{document}

14: 

15: \title{Sensitivity of Transit Searches to Habitable Planets}

16: 

17: 

18: \author{Andrew Gould, Joshua Pepper, D.\ L.\ DePoy}

19: \affil{Department of Astronomy, The Ohio State University,

20: 140 W.\ 18th Ave., Columbus, OH 43210}

21: \email{gould,pepper,depoy@astronomy.ohio-state.edu}

22: 

23: \singlespace

24: 

25: \begin{abstract}

26: Photon-limited transit surveys in $V$ band are in principle about 20 times 

27: more sensitive to planets of fixed size in the habitable zone around M stars

28: than G stars. In $I$ band the ratio is about 400.  The advantages that

29: the habitable zone lies closer and that the stars are smaller (together 

30: with the numerical superiority of M stars)  more

31: than compensate for the reduced signal due to the lower luminosity of

32: the later-type stars.  That is, M stars can yield reliable transit 

33: detections at much fainter apparent magnitudes than G stars.

34: However, to achieve this greater sensitivity, the later-type

35: stars must be monitored to these correspondingly fainter magnitudes, which

36: can engender several practical problems.  We show that with modest

37: modifications, the {\it Kepler} mission could extend its effective

38: sensitivity from its current $M_V=6$ to $M_V=9$.  This would not capture

39: the whole M dwarf peak, but would roughly

40: triple its sensitivity to Earth-like planets in the habitable zone.  

41: However, to take

42: advantage of the huge bump in the sensitivity function at $M_V=12$

43: would require major changes.

44: 

45: \end{abstract}

46: \keywords{planets}

47: \clearpage

48:  

49: \section{Introduction

50: \label{sec:intro}}

51: 

52: 	The one confirmed transiting planet discovered to date, 

53: HD209458b \citep{char00}, lies only $10\,R_\odot$ from its host

54: G star, and therefore well inside the so-called ``habitable zone'',

55: where water could exist in its liquid state.  All ongoing transit 

56: surveys are primarily sensitive to such ``hot Jupiters'' because

57: their large diameters give rise to relatively strong $(\sim 1\%)$

58: signals, while their proximity to their host increases the probability

59: of transits occurring \citep{how00,brown99,mal02,udal02,

60: burke02,str00,str02}.

61: 

62: 	While gas giants are not expected to themselves be habitable

63: regardless of their location, they could have water-laden moons like

64: Europa, Ganymede, and Callisto, which could support life if the

65: planet lay in the habitable zone.  Future ground-based surveys could

66: be sensitive to such habitable-zone giant planets.  

67: Moreover, the planned {\it Kepler} 

68: mission\footnote{http://www.kepler.arc.nasa.gov/} has as its primary

69: goal the detection of planets down to Earth size in or near the habitable

70: zone.  The {\it Kepler} targets would be even more direct analogs of the Earth.

71: 

72: 	Here we investigate the sensitivity of transit surveys to

73: planets in the habitable zone as a function of stellar type.  We

74: derive two remarkable conclusions.  First, transit surveys are most

75: sensitive to habitable planets orbiting middle M stars

76: ($M_V\sim 12$), despite the fact that they are most sensitive overall 

77: to planets orbiting early G stars ($M_V\sim 4$).  Second, to achieve

78: this sensitivity to dim stars, the survey's magnitude limit must be

79: an extremely strong function of spectral type.  Specifically, it must

80: be about 7 magnitudes fainter for middle M stars than G stars.

81: Since such variable requirements can significantly affect survey design,

82: they deserve careful analysis and consideration.  For example, we show

83: that unless the {\it Kepler} mission extends its magnitude limit for M stars

84: well beyond its current (type independent) limit of

85: $V= 14$, it will lose most of its potential sensitivity to habitable planets.

86: 

87: \section{Sensitivity to Habitable Planets

88: \label{sec:sensitivity}}

89: 

90: \citet{pgd} showed that if every star of luminosity $L$ and radius $R$

91: has a planet of radius $r$ in a circular orbit of semi-major axis $a$, 

92: then a photon-noise limited transit survey will detect

93: \begin{equation}

94: N(L,R,r,a) = {\Omega d_0^3\over 3}n(L,R)\eta

95: {R_0\over a_0}

96: \biggl({L \over L_0}\biggr)^{3/2}

97: \biggl({R \over R_0}\biggr)^{-7/2}

98: \biggl({a \over a_0}\biggr)^{-5/2}

99: \biggl({r \over r_0}\biggr)^{6}

100: \label{eqn:ndetect}

101: \end{equation}

102: planets.  Here $n(L,R)$ is the number density of stars of the specified

103: type, $\Omega$ is the angular area of the survey, and $d_0$ is the

104: distance out to which an equatorial transit can just be 

105: reliably detected for the fiducial parameters $L_0$, $R_0$, $r_0$, and

106: $a_0$.  The numerical factor $\eta\simeq 0.719$ arises because the

107: volume probed by the survey is smaller by a factor $y^{3/2}$

108: for non-equatorial transits, where $y$ is the ratio of the transit

109: chord to the stellar diameter.  For the most part we will be comparing the 

110: detectability

111: of planets of the same size around different stars.  Hence, except when

112: we are evaluating the normalization of equation (\ref{eqn:ndetect})

113: for a specific example, we will ignore the last factor.

114: 

115: The equilibrium 

116: temperature of a planet (and thus whether or not water can exist

117: in liquid phase on the planet surface) is the same if

118:  $a\propto L_\bol^{1/2}$, 

119: where the

120: bolometric luminosity $L_\bol$ is to be distinguished from $L$, the

121: luminosity in the band of observation.  Hence, the relative sensitivity

122: of a transit survey to habitable planets is

123: \begin{equation}

124: N \propto n L^{3/2} R^{-7/2} L_\bol^{-5/4}.

125: \label{eqn:ndetprop}

126: \end{equation}

127: Here we assume that all planets

128: have the same albedo and neglect atmospheric effects.

129: If we compare, for example, G stars ($M_V=5$) with middle

130: M stars ($M_V=12$), the ratios of the various factors are

131: %6 x 0.000063 x 128 x 316 = 15

132: $N_{12}/N_{5} \sim 6\times 630^{-3/2}\times 4^{7/2}\times 80^{5/4}\sim 10$.

133: Note that if we were comparing detectability at the {\it same

134: semi-major axis} rather than the {\it same habitability}, the last

135: factor would not have entered, and the ratio would have been 0.05.

136: Hence, while the sensitivity to planets in general is completely

137: dominated by G stars, the sensitivity to habitable planets is completely

138: dominated by M stars.  That is, the lower luminosities (and so smaller

139: semi-major axes) combined with the smaller radii and greater numerical

140: density of M stars more than compensate for the reduced photon counts.

141: 

142: 	In Figure \ref{fig:one}, we show the sensitivity to habitable

143: ``Earths'' and to Earths all at the same semi-major axis (taken to be 1 AU).

144: That is, the histograms show the total number of planets $N$ that will be

145: detected as a function of $M_V$ assuming that each star in the

146: field has one Earth-size planet in the habitable zone or, respectively,

147: one such planet at 1 AU.  (The two histograms cross at $M_V=5$ because

148: the habitable zone is then at 1 AU).

149: The absolute normalization of this plot is set according to the 

150: characteristics of the {\it Kepler} mission

151: ($7.8\times 10^8\,e^-\,\rm hr^{-1}$ at $V=12$, $\Omega=105\,\rm deg^2$,

152: $A_V=0.3$, mission-total S/N=8 required for detection), but the

153: form of the histogram would be the same for any photon-limited survey.

154: The normalization for any other planet size should be multiplied by

155: a factor $(r/r_\oplus)^6$, and the normalization for any other

156: fixed semi-major axis should be multiplied by $(a/{\rm AU})^{-5/2}$ 

157: (see eq.\ [\ref{eqn:ndetect}]).  The figure

158: is constructed assuming that detection is in $V$ band.  

159: The effect of substituting other bands is approximately to change the

160: slope of the histogram.  For example, since the slope of the main

161: sequence is $d M_V/d(V-I) =3.37$ \citep{reid91}, substitution of $I$

162: band would lead to an increase of slope 

163: $d\Delta \log N/d M_V = (3/2)\times 0.4/3.37 = 0.178$.  That is, middle M

164: stars would gain relative to G stars by an additional factor of

165: $10^{7\times 0.178} = 18$.

166: 

167: 	To compute these histograms, we follow the procedure of \citet{pgd}.

168: An important feature of the color-magnitude diagram is that while

169: the main sequence is fairly narrow for $M_V>6$, it broadens for brighter stars

170: (due to faster stellar evolution),

171: so that a star of a given $M_V$ can have a large range of colors.

172: Thus, for the well-defined lower main sequence, $M_V>6$, we

173: consider the luminosity function \citep{bessell93,zheng01}

174: in 1 mag bins, and evaluate the stellar radius at the center of each bin

175: using the linear color-magnitude relation of \citet{reid91}, the

176: color/surface-brightness relation of \citet{vanb}, and the $VIK$

177: color-color relations from \citet{bessell88}.  

178: 

179: 	On the other hand, for the upper main sequence $M_V<6$ we evaluate 

180: the histograms directly using the

181: Hipparcos catalog \citep{hip}.  For example, the luminosity function for

182: $M_V=4$ is computed by summing $\sum_i [(4/3)\pi D_i^3]^{-1}$ over all stars

183: within the Hipparcos completeness limit, $V<7.3$, having $3.5<M_V<4.5$,

184: and lying within 50 pc.  The distance $D_i$ is the minimum of 50 pc

185: and the distance at which the star would have $V=7.3$.  Then, the

186: constant-semi-major-axis histogram is computed by summing (and

187: appropriately normalizing) 

188: $\sum_i L_i^{3/2}R_i^{-7/2}[(4/3)\pi D_i^3]^{-1}$,

189: while the habitable-zone histogram is found by summing

190: $\sum_i L_i^{3/2}R_i^{-7/2}L_{\bol,i}^{-5/4}[(4/3)\pi D_i^3]^{-1}$.

191: The stellar radii are determined from Hipparcos/Tycho $(B_T,V_T)$ photometry

192: and the color/surface-brightness relation of \citet{gm}, ultimately

193: derived from \citet{vanb}.

194: We evaluate $L_\bol$ using bolometric corrections as a

195: function of $V-I$ color derived from \citet{bm} and \citet{bessell88} at the

196: bright end and \citet{rg} at the faint end.

197: 

198: \section{Magnitude Limits for Habitable Planets

199: \label{sec:maglims}}

200: 

201: 	To achieve the sensitivities calculated in \S~\ref{sec:sensitivity}, 

202: one must analyze the light curves of all the stars being probed.  

203: This statement would appear so obvious as not to be worth mentioning.  However,

204: as we now show, dim stars are being probed at substantially fainter

205: apparent magnitudes than are their more luminous cousins.  Hence, to

206: avoid losing most of the sensitivity of a transit experiment

207: one must set different magnitude limits for stars of different $M_V$.

208: 

209: Let $m_\lim$ be the apparent magnitude at which the survey achieves the 

210: minimum acceptable signal-to-noise ratio (S/N) for a planet of radius $r$

211: and semi-major axis $a$ circling an $M_V=5$ star.  Consider an identical

212: planet circling an $M_V=12$ star in the same orbit and at the same apparent

213: magnitude.  Since the star's radius is

214: a factor 4 smaller, the S/N will be a factor $4^{2}\times 4^{-1/2}=4^{3/2}$

215: times larger.  The first factor comes from the fact that the planet occults

216: a larger fraction of the stellar surface, the second from the reduced 

217: total time spent in transit (``duty cycle'') 

218: due to the smaller radius.  Now, the bolometric luminosity of

219: the dimmer star is down by a factor 80, so to keep in the habitable zone,

220: the planet must be moved closer by a factor $80^{1/2}$.  This increases the

221: transit duty cycle by the same factor and so increases the S/N by $80^{1/4}$.

222: Combining these two effects with the usual dependence of S/N on flux implies

223: S/N $\propto {\rm flux}^{1/2}L_\bol^{-1/4}R^{-3/2}$.  Hence, to maintain the 

224: same S/N, one must increase the magnitude limit by,

225: \begin{equation}

226: \Delta m_\lim = 0.5\Delta M_\bol - 7.5\Delta \log R.

227: \label{eqn:deltavlim}

228: \end{equation}

229: For the example just given,

230: $\Delta m_\lim = 6.9$, i.e., a factor 570 in flux.  Note that

231: $\Delta m_\lim$ does not depend on the passband of observation.  The dashed

232: curve in Figure \ref{fig:one} shows the difference 

233: in the maximum distance modulus relative to the Sun 

234: at which a habitable planet (of given radius) can be detected.  Specifically, 

235: $\Delta(m-M)_0 = (V_\lim - M_V) - (V_{\lim,\odot} - M_{V,\odot})$.

236: For $V$ band observations, this offset is roughly

237: constant over 10 magnitudes.  That is, the limiting magnitude required

238: to achieve the sensitivities shown by the bold histogram increases 

239: approximately

240: in lock step with $M_V$.  This behavior can have major consequences for the

241: design of transit experiments.

242: 

243: \section{Discussion

244: \label{sec:discuss}}

245: 

246: Photon-limited observations are not routinely achievable over very large

247: dynamic ranges.  A number of effects can intervene at fainter magnitudes.

248: First, once the stars fall below the sky, the number of detected

249: systems no longer falls as $L^{3/2}$ (see eq.\ [\ref{eqn:ndetect}])

250: but as $L^3$.  That is, once the sky is reached, detections would

251: fall off by a factor $\sim 4$ per magnitude relative to the bold histogram

252: in Figure \ref{fig:one}.  Second, if the exposure times are set so as

253: not to saturate the brightest target stars (say 2 mag brighter than the

254: mag limit appropriate for $M_V=5$), then the flux levels will be

255: a factor $\sim 4000$ times lower at the mag limit appropriate for $M_V=12$.

256: For many observing setups, this would render the flux levels 

257: comparable to the read noise.

258: Third, a field that is very uncrowded at one magnitude may well be extremely

259: crowded 7 magnitudes fainter, and this may interfere with doing

260: photon-limited photometry.  Hence, it is unlikely that all of the

261: potential shown in Figure \ref{fig:one} can be achieved in any given

262: practical experiment.  Nevertheless, this potential is so great that it

263: is worth thinking about how to achieve as much as possible.

264: 

265: We illustrate the type of trade-offs involved by considering the

266: {\it Kepler} mission, the only transit experiment proposed to date

267: whose primary target is planets in the habitable zone.  

268: The {\it Kepler} design calls for recording only stars $V<14$, presumably

269: in order to minimize data transmission.  The mag limit required to detect 

270: Earth-like planets around Sun-like stars at S/N=8

271: is $V_{\lim,\odot}=13.6$.  Hence, from the dashed curve in 

272: Figure \ref{fig:one},

273: the required mag limit at $M_V=6$ is 

274: $V_\lim=V_{\lim,\odot} - M_{V,\odot} + M_V + \Delta(m-M)_0 = 14.4$, 

275: i.e., already somewhat fainter than

276: the $V=14$ limit adopted by {\it Kepler}.  The volume probed, and hence the

277: detections,

278: fall relative to the bold histogram by a factor $(10^{0.2})^3\sim 4$ 

279: for each magnitude beyond $M_V=6$.  See Figure \ref{fig:two}.

280: Hence, assuming that every star has one planet in

281: the habitable zone, only a total of 20 will be 

282: detected\footnote{Actually, we have not taken account of stellar

283: variability in this analysis, nor of binaries.  Each of these might

284: plausibly reduce detections by 25\%, the former by making transits

285: unrecognizable, the later by making habitable planets dynamically

286: unstable and by increasing the total light in the aperture.  Hence,

287: the true detection rate for this optimistic scenario should be reduced

288: by roughly 50\%.}.

289: 

290: Keeping the photometry information for stars to $V=17$, for example, 

291: should increase the number of habitable-planet detections

292: by almost a factor 3, 

293: while keeping stars with $V<20$ would increase this

294: number by a factor 9. See Figure~\ref{fig:two}.  Of course, 

295: there are many obstacles to 

296: keeping information on stars this faint.

297: 

298: For example, there would be an enormous number of giant stars contaminating 

299: such a deep sample.

300: In fact, however, 

301: distinguishing the later-type dwarfs from the much more numerous

302: giants of similar colors is straightforward.  As shown by \citet{gm}, 

303: they can easily be identified on a reduced proper motion (RPM) diagram

304: constructed using data from USNO-A \citep{usnoa1,usnoa2} and

305: 2MASS \citep{2mass}.  With USNO-B \citep{usnob} it should be possible

306: to extend coverage to fainter magnitudes and to achieve higher precision

307: as well.  The total number of such stars $M_V\leq 12$ is only 

308: $\sim 3\times 10^4$, far fewer than the $\sim 10^5$ giants in the field

309: that could be eliminated using the same RPM diagram \citep{gm}.

310: 

311: A more difficult problem is crowding.  The {\it Kepler} point spread function

312: (PSF) is deliberately defocused to $10''$, meaning that the crowding limit

313: is about $10\,\rm arcmin^{-2}$.  At the adopted {\it Kepler} sight line,

314: $(l,b)=(69.6,5.7)$, this limit is reached at $R_{\rm USNO}\sim 18.2$

315: (as determined by a query of the USNO-A catalog).

316: If {\it Kepler} were redirected to $(l,b)=(70,15)$, the density of

317: stars down to $R_{\rm USNO}=18.2$ would be $3.5\,\rm arcmin^{-2}$,

318: approximately 3 times lower.  Hence, recovery of faint stars would be

319: much easier.  Presumably, {\it Kepler} has chosen to look right in

320: the Galactic plane because of the higher overall star density.  However,

321: since the stars useful for transits mostly lie within 500 pc, a field

322: at $b=15^\circ$ would lie only 130 pc above the plane, where the

323: density of such target stars is only slightly lower than at the plane.

324: Hence, the crowding problem could be substantially ameliorated at small

325: cost.

326: 

327: The last problem is sky.  At high ecliptic latitude, the sky in space

328: is $V\sim 23.3\,\rm arcsec^{-2}$, or $V\sim 17$ in a $10''$ PSF.  Hence,

329: {\it Kepler} completeness appears to be fundamentally limited to stars $M_V<9$.

330: As shown in Figure~\ref{fig:two}, detections from stars at $M_V\geq 10$ 

331: would be highly suppressed.  Only contracting the PSF (which has been

332: deliberately defocused to improve the photometry) could overcome this

333: difficulty.  (Note that at $V\sim 17$, crowding would not be a serious

334: problem even for the current Galactic-plane line of sight.)

335: 

336: While the above remarks apply specifically to {\it Kepler}, any transit

337: experiment attempting to detect habitable planets would face similar

338: constraints and trade-offs.

339: 

340: Finally, we note that the results reported here are a strong function of

341: planet size.  As remarked in \S~\ref{sec:sensitivity}, the absolute

342: normalizations in Figure \ref{fig:one} scale as $r^6$.  By an argument

343: similar to the one given in \S~\ref{sec:maglims}, 

344: $\Delta m_\lim = 10\Delta \log r$.  For example, for $r=0.63\,r_\oplus$,

345: the normalizations in Figure \ref{fig:one} would be reduced by a factor 16,

346: while the required $V_\lim$ at each $M_V$ would be brighter by 2 mag.

347: Hence, in {\it Kepler}'s current configuration, it would retain full 

348: sensitivity to $M_V\sim 8$ stars, and so would be able to detect $\sim 3$ 

349: planets (see Fig.~\ref{fig:two}), 

350: while if its coverage were extended 3 mag to $V_\lim\sim 17$, it could

351: probe the M-dwarf peak, and so detect $\sim 8$ planets.

352: 

353: 

354: %\begin{equation}

355: %\label{eqn:}

356: %\end{equation}

357: 

358: 

359: 

360: 

361: 

362: \acknowledgments

363: We thank Scott Gaudi for his detailed comments on the manuscript. 

364: This work was supported by grant AST 02-01266 from the NSF.

365: 

366: %\clearpage

367: 

368: \begin{thebibliography}{}

369: 

370: \bibitem[Bessell \& Brett(1988)]{bessell88}  Bessell M.S., \& Brett J.M. 1988, 

371: \pasp, 100, 1134

372: 

373: \bibitem[Bessell \& Stringfellow(1993)]{bessell93}  Bessell M.S., 

374: \& Stringfellow G.S. 1993, \araa, 31, 433

375: 

376: \bibitem[Binney \& Merrifield(1998)]{bm} Binney, J., \& Merrifield, M. 1998,

377: Galactic Dynamics, (Princeton: Princeton Univ Press)

378: 

379: \bibitem[Brown \& Charbonneau(1999)]{brown99}  Brown, T. M., \&

380: Charbonneau, D. 1999, \baas, 31, 1534

381: 

382: \bibitem[Burke et al.(2002)]{burke02}  Burke, C.J., DePoy, D.L., Gaudi, B.S.,

383: Marshall, J.L., \& Pogge, R.W.\ 2002, in ASP Conf.

384: Ser., Scientific Frontiers in Research on Extrasolar Planets, eds.

385: D. Deming and S. Seager, in press (astro-ph/0208233)

386: 

387: \bibitem[Charbonneau et al.(2000)]{char00}  Charbonneau, D., Brown, T. M.,

388: Latham, D. W., \& Mayor, M. 2000, \apj, 529, L45

389: 

390: %\bibitem[Charbonneau et al.(2002)]{char02}  Charbonneau, D., Brown, T. M.,

391: %Noyes, R. W., \& Gilliland R. L. 2002, \apj, 568, 377

392: 

393: %\bibitem[Cody \& Sasselov(2002)]{cody02}  Cody, A. M., Sasselov, D. D.

394: %2002, \apj, 569, 451

395: 

396: \bibitem[Gould \& Morgan(2003)]{gm} Gould, A., \& Morgan, C.W. 2003,

397: \apj, 585, 000 (astro-ph/0209561)

398: 

399: \bibitem[ESA(1997)]{hip} European Space Agency (ESA). 1997, The Hipparcos and

400: Tycho Catalogues (SP-1200; Noordwijk: ESA)

401: 

402: %\bibitem[H{\o}g et al.(2000)]{t2} H{\o}g, E.~et al.\ 2000, \aap, 355, L27.

403: 

404: \bibitem[Howell et al.(2000)]{how00}  Howell, S. B., Everett, M., Davis,

405: D. R., Weidenschilling, S. J., McGruder, C. H., III, Gelderman, R. 2000,

406: BAAS, 32, 3203

407: 

408: %\bibitem[Hui \& Seager(2002)]{hui02}  Hui, L., Seager, S. 2002, 

409: %\apj, 572, 540

410: 

411: \bibitem[Mallen-Ornelas(2002)]{mal02}  Mallen-Ornelas, G., Seager, S.,

412: Yee, H.K.C., Minniti, D., Gladders, M.D., 

413: Mallen-Fullerton, G., \&  Brown, T. 2002, \apj, in press (astro-ph/0203218)

414: 

415: 

416: \bibitem[Monet(1996)]{usnoa1} Monet, D.~G.\ 1996, American Astronomical

417: Society Meeting, 188, 5404.

418: 

419: \bibitem[Monet(1998)]{usnoa2} Monet, D.~G.\ 1998, American Astronomical

420: Society Meeting, 193, 112003

421: 

422: \bibitem[Monet et al.(2002)]{usnob} Monet, D.~G., et al.\ 2002, \aj, in press

423: (astro-ph/0210694)

424: 

425: \bibitem[Pepper, Gould, \& DePoy(2002)]{pgd} Pepper, J., Gould, A., \&

426: DePoy, D.L.\ 2002, \apj, submitted (astro-ph/0208042)

427: 

428: \bibitem[Reid(1991)]{reid91} Reid, N. 1991, \aj, 102, 1428

429: 

430: \bibitem[Reid \& Gilmore(1984)]{rg} Reid, I.N., \& Gilmore, G.F. 1984,

431: \mnras, 206, 19

432: 

433: %\bibitem[Salim \& Gould(2003)]{faint} Salim, S.~\& Gould, A.\ 2003, \apj, 

434: %in press

435: 

436: %\bibitem[Schlegel, Finkbeiner, \& Davis(1998)]{schlegel}

437: %Schlegel, D.J., Finkbeiner, D.P., \& Davis, M.\ 1998, \apj, 500, 525

438: 

439: \bibitem[Skrutskie et al.(1997)]{2mass} Skrutskie, M.F.~et al.\ 1997, in The

440: Impact of Large-Scale Near-IR Sky Survey, ed.\ F.\ Garzon et al.\ (Kluwer:

441: Dordrecht), p.\ 187

442: 

443: \bibitem[Street et al.(2000)]{str00}  Street, R. A. et al. 2000, in ASP Conf.

444: Ser., Vol. 219, Euroconference on Disks, Planetesimals and Planets, 

445: eds. F. Garzon, C. Eiroa, D. de Winter, \& T. J. Mahoney (San Francisco: ASP), 572

446: 

447: \bibitem[Street et al.(2002)]{str02}  Street, R. A. et al. 2002, in ASP Conf.

448: Ser., Scientific Frontiers in Research on Extrasolar Planets, eds.

449: D. Deming and S. Seager, in press (astro-ph/0208233)

450: 

451: \bibitem[Udalski et al.(2002)]{udal02}  Udalski, A. et al. 2002, AcA, 52, 1

452: 

453: 

454: \bibitem[van Belle(1999)]{vanb} van Belle, G.T.\ 1999, \pasp, 111, 1515

455: 

456: %\bibitem[Zacharias et al.(2000)]{ucac} Zacharias, N.~et al.\ 2000, \aj, 120,

457: %2131

458: 

459: \bibitem[Zheng et al.(2001)]{zheng01}  Zheng, Z., Flynn, C., Gould, A., 

460: Bahcall, J.N., Salim, S. 2001, \apj, 555, 393

461: 

462: 

463: \end{thebibliography}

464: \clearpage

465: 

466: \begin{figure}

467: %\epsscale{0.7}

468: \plotone{f1.ps}

469: \caption{\label{fig:one}

470: Total number of planets detected in the habitable zone ({\it bold histogram}) 

471: and in 1 AU orbits ({\it solid histogram}) as functions of

472: absolute magnitude $M_V$ (assuming every star has one such planet).

473: The absolute normalizations have been set to Earth-size planets and to 

474: the characteristics of the {\it Kepler} mission, but the form

475: is completely general assuming observations in $V$ band.  In $I$ band,

476: the slope would be tilted upward toward faint stars.  Extinction has

477: been taken into account, but stellar variability and binarity have not.

478: The calculation uses different methods for early ($M_V\leq 6$) and late

479: ($M_V\geq 6$) stars.  The solid and dashed histograms compare the two

480: methods in the one bin of overlap.  The {\it dashed curve} shows the

481: difference (relative to Sun-type stars) in the limiting distance modulus 

482: to which stars of different $M_V$ must be monitored to achieve the sensitivity

483: shown by the ``habitable zone'' histogram.  Since this curve is basically

484: flat, dimmer stars must be observed to the same distance and hence to much

485: fainter apparent magnitudes.

486: }\end{figure}

487: 

488: \begin{figure}

489: %\epsscale{0.7}

490: \plotone{f2.ps}

491: \caption{\label{fig:two}

492: Total number of planets detected normalized to characteristics of {\it Kepler}

493: and under various assumptions.  

494: {\it Bold solid histogram}: each star has one Earth-size planet in the 

495: habitable zone and all stars are monitored regardless of magntiude (same

496: as Fig.\ \ref{fig:one}).

497: {\it Bold dashed histogram}: same planet distribution, but now only stars

498: $V<14$ are monitored.

499: {\it Thin dashed histogram}: same planet distribution, 

500: $V<17$ stars are monitored.

501: {\it Dotted histogram}: same planet distribution, 

502: $V<20$ stars are monitored.

503: {\it Thin solid histogram}: planets have radius $r=0.63\,r_\oplus$ and only

504: $V<14$ stars are monitored.

505: }\end{figure}

506: 

507: \end{document}

508: 

509: 

510: 

511: 

512: 

513: 

514: