1: % iaus2esa.tex -- sample pages for Proceedings IAU Symposium document class

2: % (based on v1.0 cca2esam.tex)

3: % v1.04 released 17 May 2004 by TechBooks

4: %% small changes and additions made by KAvdH/IAU 4 June 2004

5: % Copyright (2004) International Astronomical Union

6: 

7: \NeedsTeXFormat{LaTeX2e}

8: 

9: \documentclass{iaus}

10: \usepackage{graphicx}

11: 

12: \title[Emergent Exoplanet Flux] %% give here short title %%

13: {Emergent Exoplanet Flux: \\ Review of the Spitzer Results}

14: 

15: \author{Drake Deming}   %% give here short author list %%

16: 

17: \affiliation{Planetary Systems Laboratory \\ Code 693,

18: NASA's Goddard Space Flight Center \\

19: Greenbelt MD 20771 USA\\ email: {\tt ddeming@pop600.gsfc.nasa.gov}}

20: 

21: \pubyear{2008}

22: \volume{253}  %% insert here IAU Symposium No.

23: \pagerange{1--11}

24: % \date{?? and in revised form ??}

25: \setcounter{page}{1}

26: \jname{Transiting Planets}

27: \editors{F.~Pont et al., eds.}

28: \begin{document}

29: 

30: \maketitle

31: 

32: \begin{abstract}

33: Observations using the {\it Spitzer Space Telescope} provided the

34: first detections of photons from extrasolar planets. {\it Spitzer}

35: observations are allowing us to infer the temperature structure,

36: composition, and dynamics of exoplanet atmospheres.  The {Spitzer}

37: studies extend from many hot Jupiters, to the hot Neptune orbiting

38: GJ\,436. Here I review the current status of {\it Spitzer} secondary

39: eclipse observations, and summarize the results from the viewpoint of

40: what is robust, what needs more work, and what the observations are

41: telling us about the physical nature of exoplanet atmospheres.

42: \end{abstract}

43: 

44: \firstsection % if your document starts with a section,

45:               % remove some space above using this command.

46: \section{Introduction}

47: 

48: The powerful astrophysical leverage provided by transits enables us to

49: study extrasolar planets directly, i.e., by detection of their

50: emergent radiation.  The {\it Spitzer Space Telescope} has provided

51: the bulk of these detections.  The first {\it Spitzer} measurements of

52: exoplanet secondary eclipses were announced in 2005.  Two independent

53: groups (Charbonneau et al. 2005; Deming et al. 2005) measured eclipses

54: for two different planets, using two different {\it Spitzer}

55: instruments, and obtained very similar results (Figure~1).

56: 

57: \begin{figure}[ht]

58: % \vspace*{-2.0 cm}

59: \begin{center}

60: \includegraphics[width=2.8in]{f1.eps}

61: % \vspace*{-1.0 cm}

62:  \caption{First detections of exoplanet thermal emission using the

63:  {\it Spitzer Space Telescope}. Plotted are the secondary eclipses of

64:  TrES-1 at 8\,$\mu$m (top, Charbonneau et al. 2005), and HD\,209458b

65:  at 24\,$\mu$m (bottom, Deming et al. 2005)}.

66:    \label{fig1}

67: \end{center}

68: \end{figure}

69: 

70: Since each eclipse was independently measured to $\sim 6\sigma$

71: significance, exoplanet thermal emission was securely detected.  The

72: discovery of transits in HD\,189733b (Bouchy et al. 2005) provided an

73: opportunity to measure exoplanet thermal emission at higher 

74: signal-to-noise ratio.  Initial observations of HD\,189733b at

75: 16\,$\mu$m (Deming et al. 2006) showed an eclipse of the planet at

76: 32$\sigma$ significance (Figure~2), and subsequent work using the IRAC

77: instrument has detetcted the planet's flux to 60$\sigma$ precision at

78: 8\,$\mu$m (Knutson et al. 2007).  This extraordinary level of

79: precision in measuring exoplanet thermal emission allows many

80: intersting studies that could hardly have been imagined when the first

81: extrasolar planets were detected by radial velocity studies.

82: 

83: \begin{figure}[ht]

84: % \vspace*{-2.0 cm}

85: \begin{center}

86: \includegraphics[width=2.2in]{f2.eps}

87: % \vspace*{-1.0 cm}

88:  \caption{Eclipse of HD\,189733b at 16\,$\mu$m (Deming et al. 2006)}.

89:    \label{fig2}

90: \end{center}

91: \end{figure}

92: 

93: In this review I summarize highlights from {\it Spitzer} secondary

94: eclipse measurements, with some discussion of transmission

95: spectroscopy during transit. The quality of work in this field has

96: been uniformly high, but I will summarize the results from the

97: viewpoint of what is robust, and what I believe needs more work and

98: clarification.

99: 

100: \section{Spectral Energy Distributions from Photometry}

101: 

102: The depth of a planetary eclipse measures the brightness of the planet

103: at that wavelength, in units of the stellar brightness.  Combining

104: results from multiple eclipses at different wavelengths allows us to

105: reconstruct the spectral energy distribution of the planet at

106: photometric spectral resolution (typically, $\lambda/\delta \lambda

107: \sim 3$).  The results are usually shown in `contrast' units, i.e.,

108: planet divided by star, since that is what we actually measure. The

109: contrast amplitude of exoplanet eclipses is greatest at the longest

110: wavelengths.  In the Rayleigh-Jeans limit, the eclipse amplitude

111: ($A_{\lambda}$) is (Charbonneau 2003):

112: \begin{equation}

113: A_{\lambda} = (R_p^2/R_s^2)(T_p/T_s),

114: \end{equation}

115: 

116: \noindent where $(R_p^2/R_s^2)$ is the ratio of planet-to-star area, and

117: $(T_p/T_s)$ is the ratio of planet-to-star brightness temperature.

118: Assuming that the planet and star resemble blackbodies (a reasonable

119: approximation at the longest wavelengths), then:

120: \begin{equation}

121: T_p =\alpha T_s \theta^{1/2},

122: \end{equation}

123: 

124: \noindent where $\theta$ is the angular diameter of the star as seen from the

125: planet, and $\alpha$ is a constant that contains the planet's albedo,

126: circulation properties, etc.  An important corollary of Eqs. 2.1 and

127: 2.2 is the effect of smaller parent stars (e.g., M-dwarfs).  As we

128: proceed down the main sequence, $R_s$ and $\theta$ decrease, but the

129: exponents imply that the $(R_p^2/R_s^2)$ term dominates over

130: $\theta^{1/2}$.  Hence planets orbiting small stars will generally

131: exhibit deeper eclipses, and their emergent flux will be more

132: detectable for that reason. Moreover, the habitable zone moves closer

133: to lower main sequence stars, and the transit probability increases

134: inversely as the planet's orbit radius. This circumstance is a major impetus for

135: finding terrestrial planets in the habitable zones of M-dwarfs

136: (Charbonneau and Deming, 2007).

137: 

138: The time of the secondary eclipse is very sensitive to the

139: eccentricity of the orbit, specifically to $e\,{\cos}\,\omega$

140: (Charbonneau 2003).  For example, the eclipse of GJ\,436b occurs at

141: phase $0.585\pm0.005$ (Deming et al. 2007), more than five hours after

142: the mid-point between transits.  Since the eclipse duration is

143: $\sim$1-hour, the sensitivity of the eclipse time to moderately small

144: eccentricity ($e=0.15$ for GJ\,436b) is obvious.  

145: 

146: % Combined with

147: % transit timing measurements, secondary eclipse times can also provide

148: % powerful constraints on the presence of other planets (D. Fabricky,

149: % private communication).

150: 

151: \subsection{Molecular Absorption}

152: 

153: The actual flux from close-in planets will peak near 2 to 5\,$\mu$m,

154: not at the wavelengths of greatest contrast.  Moreover, the shorter

155: infrared (IR) wavelengths are key to inferring the composition and

156: temperature structure of the planet's atmosphere.  Figure~3 shows the

157: spectrum of HD\,189733b, in flux (not contrast) units, from Barman

158: (2008). The IR spectra of hot Jupiters are believed to be shaped

159: predominantly by water absorption (Burrows et al. 2005, Seager et

160: al. 2005), but other molecules such as methane also play a role (e.g.,

161: Swain et al. 2008a), and methane in particular could become more

162: important for cooler planets like GJ\,436b. For close-in planets

163: orbiting luminous stars, strong irradiation could flatten the

164: temperature gradient and weaken absorption features in the spectrum at

165: the time of eclipse (Fortney et al. 2006). {\it Spitzer} results from

166: spectroscopy initially suggested that water absorption might not be a

167: prominent feature in eclipse spectra (Grillmair et al. 2007,

168: Richardson et al 2007, see discussion below). However, the HD\,189733b

169: results from IRAC (Charbonneau et al. 2008) are in good accord with

170: Barman's standard model, and provide convincing evidence that water

171: absorption shapes the 2- to 5\,$\mu$m spectra of at least one hot

172: Jupiter.

173: 

174: Important observations of {\it transits} have also been made using

175: {\it Spitzer} (Richardson et al. 2006, Gillon et al. 2007, Nutzman et

176: al. 2008, Agol et al. 2008), including evidence for water absorption

177: during transit (Tinetti et al. 2007, Beaulieu et al. 2008).

178: Ehrenreich et al. (2007) have suggested that better correction for

179: instrument systematics is needed before we can conclude that water

180: absorption is detected via IRAC photometry during transit.  While I

181: believe there is good evidence for water absorption in transit,

182: additional work to better understand {\it Spitzer's} instrumental

183: systematics is certainly warranted.

184: 

185: \begin{figure}[ht]

186: % \vspace*{-2.0 cm}

187: \begin{center}

188: \includegraphics[width=2.5in]{f3.eps}

189: % \vspace*{-1.0 cm}

190:  \caption{Modeled spectrum of HD\,189733b, giving flux vs. wavelength

191:  in microns, from Barman (2008), with measurements from Charbonneau et

192:  al. (2008).  The standard model including water absorption provides

193:  the best fit to the measurements.}

194:    \label{fig3}

195: \end{center}

196: \end{figure}

197: 

198: \subsection{Circulation and Dynamics}

199: 

200: Hot Jupiters are believed to rotate synchronously with their orbital

201: period, keeping one side perpetually pointed toward the star.  An

202: important question is the degree to which they transport heat to the

203: anti-stellar hemisphere via strong atmospheric circulation.  Models of

204: this circulation (e.g., Showman \& Guillot 2002, Cho et al. 2003, 2008, 

205: Cooper and Showman 2006, Langton and Laughlin 2008) can be checked

206: using {\it Spitzer} around-the-orbit observations, either made

207: continuously (Knutson et al. 2007), or via periodic sampling

208: (Harrington et al. 2006, Cowan et al. 2007). Observations of two

209: non-transiting hot Jupiters, Ups And\,b at 24\,$\mu$m (Harrington et

210: al. 2006) and HD\,179949b at 8\,$\mu$m (Cowan et al. 2007) suggested

211: large day-to-night temperature contrasts.  Other planets give much

212: lower day-night contrast (Knutson et al. 2007, Cowan et al. 2007). One

213: possibility is that the difference for Ups And\,b is related to the

214: greater formation height of 24\,$\mu$m radiation. But note that

215: Knutson et al. (2008a) have observed HD\,189733b at 24\,$\mu$m, and

216: find a result commensurate with their 8\,$\mu$m results.  Another

217: possibility is that these type of phased observations could be

218: affected by a temporary `hot spot', and would not typically show such

219: a large day-night difference. However, in that case we would also

220: expect greater-than-predicted variability at secondary eclipse for

221: some transiting planets, and even low-level variability has not yet

222: been observed. {\it Spitzer} will re-observe Ups And\,b at 24\,$\mu$m

223: (B. Hansen, private communication) if the cryogen lasts long enough.

224: 

225: In addition to close-in planets on circular orbits, {\it Spitzer} has

226: great potential to observe the time-dependent heating (Iro \& Deming

227: 2008) for planets on very elliptic orbits.  Recently, Laughlin et

228: al. (2008) observed the flash-heating of HD\,80606b at periastron,

229: and {\it Spitzer} may be able to make more observations of this type

230: during the warm mission.

231: 

232: \subsection{Inverted Temperature Gradients}

233: 

234: Although HD\,189733b shows water absorption, and agrees well with

235: standard models, there is evidence that other exoplanets show

236: different atmospheric structures.  HD\,209458b exhibits an atmospheric

237: temperature inversion (i.e., temperature rises with increasing

238: height). The first hint of this inversion was found by Richardson et

239: al. (2007), who derived emission features in their eclipse spectrum at

240: relatively `high' spectral resolution $\lambda/\delta\lambda \sim

241: 100$, albeit at low signal-to-noise.  Definitive evidence of the

242: inversion comes from the IRAC eclipse measurements by Knutson et

243: al. (2008b).  Figure~4 shows the IRAC measurements for both

244: HD\,189733b (Charbonneau et al. 2008, Barman 2008) and HD\,209458b

245: (Knutson et al. 2008b, Burrows et al. 2007), compared with models from

246: Burrows et al. (2008).  In comparing the observations and models on

247: Figure~4, I've scaled the models by an arbitrary contrast factor to

248: produce the best fit by eye.  This serves to illustrate the nature of

249: these two different exoplanet spectra, but I caution that for the

250: quantitative fits, readers should consult the original papers.

251: 

252: \begin{figure}[ht]

253: % \vspace*{-2.0 cm}

254: \begin{center}

255: \includegraphics[width=3.0in]{f4.eps}

256: % \vspace*{-1.0 cm}

257:  \caption{Measurements of HD\,189733b from Charbonneau et al. (2008),

258:  and HD\,209458b from Knutson et al. (2008b), compared to a standard

259:  model (HD\,189733b) and a temperature-inverted model (HD\,209458b)

260:  from Burrows et al. (2008). The model and observations for

261:  HD\,189733b have been offset upward by 0.002 for clarity. Error bars are $\pm 2\sigma$.}

262:    \label{fig4}

263: \end{center}

264: \end{figure}

265: 

266: In addition to IRAC, {\it Spitzer} eclipse observations of both

267: planets have been obtained at 16\,$\mu$m and 24\,$\mu$m (not plotted).

268: These longer wavelengths are also sensitive to water absorption and

269: atmospheric temperature gradients, but less so than at IRAC

270: wavelengths. When comparing spectral energy distributions for

271: different planets, remember that all four IRAC wavelengths were observed

272: simultaneously only for HD\,189733b and HD\,209458b.  More commonly,

273: due to the nature of the instrument (Fazio et al. 2005), eclipses are

274: measured at 3.6 and 5.8\,$\mu$m simultaneously, and then in another

275: eclipse at 4.5 and 8\,$\mu$m simultaneously. In principle, variability

276: of the planet's thermal emission (Rauscher et al. 2007) could

277: contaminate the measured spectrum.  To date, there is no evidence for

278: variability large enough to produce significant spectrum errors (see

279: Agol et al. 2008).

280: 

281: The hallmark of an inverted spectrum can be seen by comparing 5.8 and

282: 8.0\,$\mu$m measurements, as well as comparing 3.6 and

283: 4.5\,$\mu$m. The inverted atmosphere has a higher 5.8\,$\mu$m flux

284: than 8.0\,$\mu$m, because 5.8\,$\mu$m has high opacity due to water

285: vapor, and the water bands are present in emission. In the

286: non-inverted atmosphere the 3.6\,$\mu$m flux is elevated because

287: the lesser opacity at this wavelength allows planet flux to well up

288: from the deeper atmosphere, where temperatures are higher.  The

289: non-inverted 4.5\,$\mu$m band shows lower flux because water and CO

290: opacity cause the radiation to be emitted from higher layers of the

291: atmosphere, where the temperature is lower.  The inverted atmosphere

292: may have a high altitude absorbing layer (Burrows et al. 2008), and

293: this raises the temperature at the 4.5\,$\mu$m height, and lowers it

294: at the 3.6\,$\mu$m height, reversing the relative magnitudes of the

295: emergent fluxes.

296: 

297: An open question is the physical cause of the inversion, and whether

298: this is a common phenomenon in hot Jupiter atmospheres.  Burrows et

299: al. (2008) attribute the inversion to the presence of a high altitude

300: optical absorbing layer, but the composition and origin of this layer

301: are unknown.  Doubtless the high level of stellar irradiation that hot

302: Jupiters experience plays a major role in inverting the temperature

303: gradient.  Fortney et al. (2008) define two classes of hot Jupiters

304: depending on whether the stellar irradiation drives the formation of a

305: hot stratosphere (i.e., region of higher temperature) via TiO/VO

306: absorption. TiO and VO have bands in the optical where stellar fluxes

307: are high, and have been implicated in perturbations to exoplanet

308: atmospheric temperature structure (Hubeny et al. 2003).  Figure~5

309: shows many of the known exoplanets in mass vs. irradiance

310: space, with the predicted boundary between the pM class

311: (stratospheres) and pL class (no stratospheres) indicated. I have

312: circled the three planets that are known or strongly suspected to have

313: inverted temperature gradients.  Besides HD\,209458b, HD\,149026b is

314: believed to have a hot stratosphere on the basis of the very high

315: brightness temperature at 8\,$\mu$m (Harrington et al. 2007).

316: Recently, XO-1b was found to have an inverted gradient (Machalek et

317: al. 2008).  Although XO-1b is predicted to be pL class, it is at the

318: upper range of that class, so there is no firm counterexample to the

319: Fortney et al. (2008) theory at this point.  During this conference,

320: new information is becoming available on TrES-2 (O'Donovan et

321: al. 2008) and TrES-4 (Knutson et al. 2008c), and {\it Spitzer} data

322: for several other planets are under analysis.  Hence we should soon

323: learn how well the curent pM/pL classification theory corresponds to reality.

324: 

325: \begin{figure}[ht]

326: % \vspace*{-2.0 cm}

327: \begin{center}

328: \includegraphics[width=2.5in]{f5.eps}

329: % \vspace*{-1.0 cm}

330:  \caption{Atmospheric structure classification for hot Jupiters, from

331:  Fortney et al. (2008).  The pM class planets are predicted to have

332:  hot stratospheres (i.e., temperature inversions), whereas the pL

333:  class planets should not. The planets that are currently known or

334:  strongly suspected to have inversions are circled with solid lines,

335:  and the non-inverted planets (TrES-1, and HD\,189733b) are circled with dashed

336:  lines.}

337:    \label{fig5}

338: \end{center}

339: \end{figure}

340: 

341: \subsection{A Hot Neptune}

342: 

343: {\it Spitzer} secondary eclipse observations extend down to the hot

344: Neptune orbiting GJ\,436 (Deming et al. 2007a, Demory et al. 2007).

345: Figure~6 shows this eclipse at 8\,$\mu$m.  The inferred brightness

346: temperature ($712\pm36$K) is modestly above the predicted brightness

347: temperature for thermal equilibrium with the star (Deming et

348: al. 2007a), but the uncertainties are relatively large.  Additional

349: {\it Spitzer} eclipse observations were recently made at at 8- and

350: 24\,$\mu$m (J. Harrington, private communication) and

351: `around-the-orbit' observations (by Knutson et al.) are pending.  The

352: totality of {\it Spitzer} observations may be sufficient to define the

353: total luminosity of the planet, and thereby determine whether it emits

354: significant energy due to tidal dissipation in its moderately

355: eccentric orbit.

356: 

357: 

358: \begin{figure}[ht]

359: % \vspace*{-2.0 cm}

360: \begin{center}

361: \includegraphics[width=3.0in]{f6.eps}

362: % \vspace*{-1.0 cm}

363:  \caption{Secondary eclipse of the hot Neptune planet GJ\,436b

364:  observed by {\it Spitzer} at 8\,$\mu$m (Deming et al. 2007a). The top

365:  panel shows the unbinned data prior to correction of the detector

366:  ramp, and the lower panel shows binned data and a fit of an eclipse

367:  curve centered at phase 0.587, indicative of an eccentric orbit.}

368:    \label{fig6}

369: \end{center}

370: \end{figure}

371: 

372: \section{Spectroscopy}

373: 

374: {\it Spitzer} secondary eclipse observations have been extended to

375: spectroscopy as well as photometry (Grillmair et al. 2007, Richardson

376: et al. 2007, Swain et al. 2008b).  The principle of these measurements

377: is simple.  Suppose that an absorption (emission) feature occurs in

378: the planets atmosphere at a given wavelength.  Then the depth of the

379: eclipse at that wavelength will be smaller (larger) than at other

380: wavelengths.  Hence the emergent spectrum of the planet can be

381: constructed from the wavelength dependence of the eclipse depth.  In

382: practice, this is a much more difficult observation than {\it Spitzer}

383: photometry, for two reasons.  First, there are many fewer photons per

384: wavelength channel because the light is dispersed, so the

385: signal-to-noise ratio is lower than for photometry.  Second,

386: spectroscopy is more affected by instrument systematic effects, as

387: discussed below.  In spite of these difficulties, the results are of

388: great interest.  Two exoplanets are sufficiently bright to make {\it

389: Spitzer} spectroscopy practical: HD\,189733b (Grillmair et al. 2007),

390: and HD\,209458b (Richardson et al. 2007, Swain et al. 2008b).  The

391: initial results indicated that these spectra were remarkably flat from

392: $\sim 7$ to $\sim 13\,\mu$m, not showing absorption due to water vapor

393: that was expected shortward of $\sim 8\,\mu$m.  However, additional

394: spectroscopy of HD\,189733b (Grillmair, private communication) agrees

395: very well with {\it Spitzer} photometry and with model atmosphere

396: predictions.  HD\,209468b having an inverted temperature gradient is

397: consistent with the flatness observed in the spectrum by Richardson et

398: al. (2007).

399: 

400: \begin{figure}[ht]

401: % \vspace*{-2.0 cm}

402: \begin{center}

403: \includegraphics[width=3.0in]{f7.eps}

404: % \vspace*{-1.0 cm}

405:  \caption{Spectrum of HD\,209458b, derived using an analysis very

406:  similar to Richardson et al. (2007), based on two eclipses (different

407:  symbols).  The two emission features evident in the spectrum are a

408:  broad bump near 9.8\,$\mu$m that Richardson et al. attribute to

409:  silicate clouds, and a sharp emission feature at 7.78\,$\mu$m

410:  possibly due to a C-C stretching resonance.}

411:    \label{fig7}

412: \end{center}

413: \end{figure}

414: 

415: Grillmair et al. (2007) did not find evidence for discrete spectral

416: absorption/emission features in the spectrum of HD\,189733b, but

417: Richardson et al. (2007) concluded that the spectrum of HD\,209458b

418: contained two discrete features, both present in emission.  Near

419: 9.8\,$\mu$m they found evidence for a relatively broad emission bump

420: that they attributed to the Si-O stretching resonance, and at

421: 7.78\,$\mu$m they find a sharp emission feature that may be due to a

422: C-C stretching resonance.  To illustrate these results, I've

423: re-analyzed the Richardson et al. data, using a similar method, and

424: this result is shown in Figure~7.  These data have also been analyzed

425: by Swain et al. (2008b), using an entirely different method.  Swain et

426: al. find emission near 7.8\,$\mu$m, but not at 9.8\,$\mu$m.  This

427: difference in results seems consistent with the nature of the

428: systematic effects in the IRS instrument: broad features (9.8\,$\mu$m)

429: are more sensitive to the way the instrument systematics are treated

430: in the analysis, whereas sharp features (7.8\,$\mu$m) are insensitive

431: to the analysis method.  It is important to optimize the spectroscopic

432: eclipse technique, so that we can use it with JWST to measure the

433: spectra of potentially habitable planets transiting M-dwarfs

434: (Charbonneau and Deming, 2007).

435: 

436: \section{{\it Spitzer} Instrument Systematics}

437: 

438: {\it Spitzer} observations have proven to be a remarkably stable and

439: sensitive way to measure exoplanet thermal emission. {\it Spitzer}

440: exoplanet aperture photometry (for $\lambda \leq 8\,\mu$m) achieves

441: noise levels closely approaching the photon noise limit, and the

442: errors average down as the inverse square root of exposure time.  For

443: the longer {\it Spitzer} wavelengths, where the zodiacal thermal

444: background is significant, the most precise photometry often requires

445: PSF-weighted optimal photometry, depending on the brightness of the

446: star. Unlike ground-based photometry, it is generally {\it not} necessary to

447: use `comparison stars' with {\it Spitzer}.  In fact, it can even be

448: detrimental to rely on comparison stars, because {\it Spitzer} does

449: have instrument systematic effects that could vary with position on

450: the detector. There are several effects that are currently recognized

451: and accounted for in {\it Spitzer} analyses, and most of them are now

452: described in the {\it Spitzer} instrument documentation.  The ones

453: most relevant to exoplanet eclipses are:

454: 

455: {\it The Ramp}.  Photometry at 8- and 16\,$\mu$m exhibits a gradually

456: increasing intensity, equivalent to an increasing gain in the

457: instrument response. This apparent gain increase is flux-dependent:

458: bright sources reach maxmium intensity more rapidly than faint

459: sources. This so-called `ramp' (Deming et al. 2006) is obvious in the

460: top panel of Figure~6.  Knutson et al. (2007) hypothesize that it is

461: due to charge-trapping, which is (so far) the most promising

462: hypothesis. Note that the ramp is {\it not} simply due to build-up of a

463: latent image, since none is present when the telescope is nodded

464: (Deming et al. 2006). In the charge-trapping hypothesis, the first

465: electrons generated by photons are captured by ionized impurites in

466: the detector material, and do not contribute to the signal on the

467: observed time scale.  As the detector is exposed to additional

468: radiation, the charge traps saturate, and the signal readout reaches

469: an asymptotic level.  This explanation is broadly consistent with the

470: known characteristics of the ramp, with some exceptions.  Observations

471: at 5.8\,$\mu$m can exhibit a `negative' ramp, i.e. a decreasing

472: intensity with time (Machalek et al. 2008), and no ramp is seen in

473: 24\,$\mu$m photometry (Knutson and Charbonneau, private

474: communication). The existence of decreasing, as well as increasing,

475: ramps suggests a complex phenomenon that may depend on the gate and

476: bias voltages applied to the detector.  The lack of a ramp at

477: 24\,$\mu$m may be due to the fact that the relatively large zodiacal

478: background at this wavelength keeps the ramp perpetually at its

479: maximum value, but the zodiacal background is also strong for

480: 16\,$\mu$m photometry, which does exhibit a prominent ramp (Deming et

481: al. 2006). Spectroscopy using IRS also exhibits a ramp, at least when

482: obtaining spectroscopy at 7-14\,$\mu$m (Richardson et al. 2007).

483: 

484: The ramp is a relatively benign effect for eclipse photometry of

485: bright and high-contrast systems. The time scale for the ramp to reach

486: its maximum value is significantly longer than the duration of an

487: eclipse, so it's essentially a baseline effect that is included

488: when fitting to the eclipse depth.  But the ramp is more problematic

489: for fainter and low-contrast systems because small uncertainties in

490: the ramp curvature become significant relative to the eclipse

491: depth. The ramp is also problematic for `around-the-orbit'

492: observations where the planet signal will vary on a longer time

493: scale. One promising approach for this type of observation is to

494: `pre-flash' the detector by exposing it to a bright source immediately

495: prior to the exoplanet observations.  This saturates the ramp before

496: the exoplanet is observed; preliminary examination of observations

497: using a pre-flash (H. Knutson, private communication) suggest that the

498: technique is largely successful.

499: 

500: {\it Pixel Phase}. The pixels in the IRAC detectors are more

501: responsive when stellar images are centered on the pixels than when

502: they lie near the edges, and this is called the pixel phase effect.

503: One exception to this is very bright stars that are near

504: saturation. Detector non-linearity can produce a lower signal when

505: very bright stars are centered on a given pixel, but this circumstance

506: is normally avoided by using shorter integration times.  The pixel

507: phase effect is ubiquitous in the 3.6 and 4.5\,$\mu$m channels of

508: IRAC, and may be present to a much lesser degree at 5.8- and

509: 8\,$\mu$m. Because there is pointing jitter in the telescope ($\sim$

510: tens of milli-arcsec), the pixel phase effect leads to a variable

511: intensity when performing aperture photometry. This is corrected in

512: eclipse data by decorrelating the intensity versus distance from pixel

513: center (e.g., Charbonneau et al. 2005, 2008).  Pixel phase is also

514: normally decorrelated from photometry performed for other {\it

515: Spitzer} research (Morales-Calderon et al. 2006), not just exoplanet

516: photometry.

517:  

518: {\it Spectroscopic Slit Losses}.  The IRS spectrometer (Houck et

519: al. 2005) has light losses at its entrance slit, like most

520: astronomical slit spectrometers.  Diffraction causes stellar images at

521: the slit to increase in size proportional to wavelength, so the slit

522: losses increase with wavelength also. Because there is telescope

523: pointing jitter on a time scale of $\sim1$\,hour, and possibly on

524: longer times scales, the intensity in a stellar spectrum will vary

525: slightly with both wavelength and time.  Even a slight variation in

526: the measured stellar spectrum distorts the spectra that are inferred

527: for exoplanets using the eclipse technique. There is no robust and

528: independent method to ascertain the exact telescope pointing for a

529: given spectrum.  Hence exoplanet observers correct for this distortion

530: in different ways (Richardson et al. 2007, Swain et

531: al. 2008b). Because the effect varies slowly with wavelength, it

532: primarily affects broad features in exoplanet spectra, not sharp

533: features like the 7.8\,$\mu$m emission inferred in HD\,209458b

534: (Richardson et al. 2007, and Figure 7). Hence the appearance of broad

535: features (like the 9.8\,$\mu$m peak on Figure 7), and indeed their

536: reality in exoplanet spectra, can depend on the method used to analyze

537: the data. Sharp features should be robust against instrument

538: systematics, but I note that the 7.8\,$\mu$m feature has relatively

539: low signal-to-noise.

540: 

541: \section{Warm Spitzer}

542: 

543: The depletion of cryogen (in $\sim$ April 2009) will terminate the

544: cold portion of the {\it Spitzer} mission.  However, the 3.6 and

545: 4.5\,$\mu$m channels of IRAC will still operate at full sensitivity,

546: because the observatory will remain radiatively cooled at $\sim

547: 35$K. The requirement to operate the `warm' mission at low cost

548: dictates that operations must be simple.  This favors relatively large

549: programs, and exoplanet transit and eclipse science is poised to take

550: full advantage of {\it Warm Spitzer}.  A `not-yet-obsolete' discussion

551: of some possible exoplanet applications is given in Deming et al. (2007b).

552: 

553: \section{Tabular Summary of the Spitzer Results}

554: 

555: The legacy of exoplanet science from cryogenic {\it Spitzer} has

556: already revolutionized exoplanet science, but has not yet reached full

557: fruition.  The results are of uniformly high quality, but the

558: observations are not easy.  The table in Figure~8 summarizes my

559: personal opinion as to which of the {\it Spitzer} results are already

560: robust, and which aspects need clarification by additional work.

561: 

562: \medskip

563: {\bf Acknowledgements.}  Adam Burrows, Travis Barman, Dave

564: Charbonneau, Jonathan Fortney, Heather Knutson, and Francis O'Donovan

565: all made valuable comments on a draft of this review.  I thank Travis

566: and Jonathan for providing Figures 3 and 5.

567: 

568: \begin{figure}[ht]

569: % \vspace*{-2.0 cm}

570: \begin{center}

571: \includegraphics[width=3.0in]{f8.eps}

572: % \vspace*{-1.0 cm}

573:  \caption{Tabular summary of the Spitzer results, categorized by

574:   results that are already robust, and those where more work is

575:   needed. Observations of eccentric planets fall into both categories,

576:   because the current work should be extended to more planets.}

577:    \label{fig8}

578: \end{center}

579: \end{figure}

580: 

581: 

582: \begin{thebibliography}{}

583: 

584: \bibitem[Agol, E. et al.(2008)]{}

585: {Agol,~E., Cowan,~N.~B., Bushong,~J. et al.} 2008, \textit{this volume}.

586: 

587: \bibitem[Barman, T. (2008)]{}

588: {Barman,~T.} 2008, \textit{ApJ}, 676, L61.

589: 

590: \bibitem[Beaulieu, J.~P., et al. (2008)]{}

591: {Beaulieu, J.~P., Carey,~S., Ribas,~I., et al.} 2008, \textit{ApJ} 677, 1343.

592: 

593: \bibitem[Bouchy, F. (2005)]{}

594: {Bouchy,~F., Udry,~S., Mayor,~M., et al.} 2005, \textit{A\&A} 444, L15.

595: 

596: \bibitem[Burrows, A. et al. (2005)]{}

597: {Burrows,~A., Hubeny,~I., \& Sudarsky,~D.} 2005, \textit{ApJ} 625, L135. 

598: 

599: \bibitem[Burrows, A. et al. (2007)]{}

600: {Burrows,~A., Hubeny,~I., Budaj,~J., Knutson,~H.~A., \& Charbonneau,~D.} 2007, 

601:   \textit{ApJ}, 668, L171.

602: 

603: \bibitem[Burrows, A. et al.(2008)]{}

604: {Burrows,~A., Hubeny,~I., Budaj,~J., et al.} 2008, \textit{ApJ}, 668, L171.

605: 

606: \bibitem[Charbonneau, D. (2003)]{}

607: {Charbonneau,~D.} 2003, \textit{A.~S.~P. Conf. Ser.}, 294, 449. 

608: 

609: \bibitem[Charbonneau, D., et al. (2005)]{}

610: {Charbonneau,~D., Allen,~L.~E., Megeath,~S.~T., et al.} 2005, \textit{ApJ} 626, 523.

611: 

612: \bibitem[Charbonneau, D. \& Deming, D. (2007)]{}

613: {Charbonneau,~D., \& Deming,~D.} 2007, astro-ph/0706.1047.

614: 

615: \bibitem[Charbonneau, D., et al. (2008)]{}

616: {Charbonneau,~D., Knutson,~H.~A., Barman,~T., et al.} 2008, \textit{ApJ}, in press, 

617:     astro-ph/0802.0845.

618: 

619: \bibitem[Cho,~J.~Y.-T., et al. (2003)]{}

620: {Cho,~J.~Y.-T., Menou,~K., Hansen,~B., et al.} 2003, \textit{ApJ}, 587, L117.

621: 

622: \bibitem[Cho,~J.~Y.-T., et al. (2008)]{}

623: {Cho,~J.~Y.-T., Menou,~K., Hansen,~B., et al.} 2008, \textit{ApJ}, 675, 817.

624: 

625: \bibitem[Cooper,~C.~S. \& Showman,~A.~P.]{}

626: {Cooper,~C.~S., \& Showman,~A.~P.} 2005, \textit{ApJ}, 629, L45.

627: 

628: \bibitem[Cowan,~N.~B. et al. (2007)]{}

629: {Cowan,~N.~B., Agol,~E., \& Charbonneau,~D.} 2007,  \textit{Mon. Not. Roy. Astr. Soc.},

630:   379, 641.

631: 

632: \bibitem[Deming, D. et al. (2005)]{}

633: {Deming,~D., Seager,~S., Richardson,~L.~J., et al.} 2005, \textit{Nature} 434, 740.

634: 

635: \bibitem[Deming, D. et al. (2006)]{}

636: {Deming,~D., Harrington,~J., Seager,~S., et al.} 2006, \textit{ApJ} 644, 560.

637: 

638: \bibitem[Deming, D. et al. (2007a)]{}

639: {Deming,~D., Harrington,~J., Laughlin,~G., et al.} 2007a, \textit{ApJ} 667, L199. 

640: 

641: \bibitem[Deming, D. et al. (2007b)]{}

642: {Deming,~D., Agol,~E., Charbonneau,~D., et al.} 2007b, \textit{AIP Conf. Proc.} 943, 89, 

643:   astro-ph/0710.4145.

644: 

645: \bibitem[Demory,~B.-O., et al. (2007)]{}

646: {Demory,~B.-O., Gillon,~M., Barman,~T., et al.} 2007, \textit{A\&A}, 475, 1125.

647: 

648: \bibitem[Ehrenreich,~D. et al. (2007)]{}

649: {Ehrenreich, D., Hebrard,~G., Lecvelier des Estangs,~A., et al.} 2007, \textit{ApJ}, 668, L179.

650: 

651: \bibitem[Fazio,~G.~G. et al. (2004)]{}

652: {Fazio,~G.~G., Hora,~J.~L., Allen,~L.~E., et al} 2004, \textit{ApJ Suppl.} 154, 10.

653: 

654: \bibitem[Fortney, J.~J. et al. (2006)]{}

655: {Fortney,~J.~J., Cooper,~C.~S., Showman,~A.~P., et al.} 2006, \textit{ApJ}, 652, 746.

656: 

657: \bibitem[Fortney,~J.~J. et al. (2008)]{}

658: {Fortney,~J.~J., Lodders,~K., Marley,~M.~S., et al.} 2008, \textit{ApJ}, 678, 1419.

659: 

660: \bibitem[Gillon, M., et al. (2007)]{}

661: {Gillon,~M., Demory,~B.-O., Barman,~T., et al.} 2007, \textit{A\&A} 471, L51.

662: 

663: \bibitem[Grillmair,~C.~J. et al. (2007)]{}

664: {Grillmair,~C.~J., Charbonneau,~D., Burrows,~A., et al.} 2007, \textit{ApJ} 658, L115.

665: 

666: \bibitem[Harrington,~J., et al. (2006)]{}

667: {Harrington,~J., Hansen,~B.~M., Lusczc,~S.~H., et al.} 2006, \textit{Science}, 314, 623.

668: 

669: \bibitem[Harrington,~J., et al. (2007)]{}

670: {Harrington,~J., Luszcz,~S., Seager,~S., et al.} 2007, \textit{Nature}, 447, 691.

671: 

672: \bibitem[Houck,~J.~R., et al. (2004)]{}

673: {Houck,~J.~R., Roellig,~T.~L., van~Cleve,~J., et al.} 2004, \textit{ApJ Suppl.} 154, 18.

674: 

675: \bibitem[Hubeny,~I., et al. (2003)]{}

676: {Hubeny,~I., Burrows,~A., \& Sudarsky,~D.} 2003, \textit{ApJ}, 594, 1011.

677: 

678: \bibitem[Iro,~N. \& Deming,~D. (2008)]{}

679: {Iro,~N. \& Deming,~D.} 2008, \textit{this volume}, astro-ph/0807.0266.

680: 

681: \bibitem[Knutson,~H.~A. et al. (2007)]{}

682: {Knutson,~H.~A., Charbonneau,~D., Allen,~L.~E., et al.} 2007, \textit{Nature}, 447, 183.

683: 

684: \bibitem[Knutson, H.~A., et al. (2008a)]{} 

685: {Knutson,~H.~A., Charbonneau,~D., Cowan,~N.~B., et al.} 2008a, \textit{ApJ, submitted},

686:   astro-ph/0802.1705.

687: 

688: \bibitem[Knutson,~H.~A., et al. (2008b)]{}

689: {Knutson,~H.~A., Charbonneau,~D., Allen,~L.~E., et al.} 2008b, \textit{ApJ}, 673, 526.

690: 

691: \bibitem[Knutson,~H.~A., et al. (2008c)]{}

692: {Knutson,~H.~A., et al.} 2008c, \textit{this volume}.

693: 

694: \bibitem[Langton,~J., \& Laughlin,~G. (2008)]{}

695: {Langton,~J., \& Laughlin,~G.} 2008, \textit{ApJ}, 674, 1106.

696: 

697: \bibitem[Laughlin,~G., et al. (2008)]{}

698: {Laughlin,~G., Deming,~D., Langton,~G.} 2008, \textit{Nature}, submitted.

699: 

700: \bibitem[Machalek,~P., et al. (2008)]{}

701: {Machalek,~P., McCullough,~P., Burke,~C.~J., et al.} 2008, \textit{ApJ}, in press, 

702:   astro-ph/0805.2418.

703: 

704: \bibitem[Morales-Calderon,~M., et al. (2006)]{}

705: {Morales-Calderon,~M., Stauffer,~J.~R., Kirkpatrick,~J.~D., et al.} 2006, \textit{ApJ}, 653, 1454.

706: 

707: \bibitem[Nutzman,~P., et al. (2008)]{}

708: {Nutzman,~P., Charbonneau,~D., Winn,~J.~N., et al.} 2008, \textit{ApJ}, submitted, 

709:    astro-ph/0807.1318.

710: 

711: \bibitem[O'Donovan,~F.~T., et al. (2008)]{}

712: {O'Donovan,~F.~T., Charbonneau,~D., Harrington,~J., et al.} 2008, \textit{this volume}.

713: 

714: \bibitem[Rauscher,~E., et al. (2007)]{}

715: {Rauscher,~E., Menou,~K., Cho,~J.~Y.-K., et al.} 2007, \textit{ApJ}, 662, L115.

716: 

717: \bibitem[Richardson,~L.~J. et al. (2006)]{}

718: {Richardson,~L.~J., Harrington,~J., Seager,~S., et al.} 2006, \textit{ApJ}, 649, 1043.

719: 

720: \bibitem[Richardson,~L.~J. et al. (2007)]{}

721: {Richardson,~L.~J., Deming,~D., Horning,~K., et al.} 2007, \textit{Nature}, 445, 892.

722: 

723: \bibitem[Seager,~S. et al. (2005)]{}

724: {Seager,~S., Richardson,~L.~J., Hansen,~B.~M.~S., et al.} 2005, \textit{ApJ}, 632, 1122.

725: 

726: \bibitem[Showman,~A.~P. \& Guillot,~T. (2002)]{}

727: {Showman,~A.~P., \& Guillot,~T.} 2002, \textit{A\&A}, 385, 166. 

728: 

729: \bibitem[Swain,~M.~R. et al. (2008a)]{}

730: {Swain,~M.~R., Gautam,~V., \& Tinetti,~G.} 2008a, \textit{Nature}, 452, 329.

731: 

732: \bibitem[Swain,~M.~R. et al. (2008b)]{}

733: {Swain,~M.~R., Bouwman,~J., Akeson,~R.~L., et al.} 2008b, \textit{ApJ}, 674, 482.

734: 

735: \bibitem[Tinetti,~G., et al. (2007)]{}

736: {Tinetti,~G., Vidal-Madjar,~A., Liang,~M.-C., et al.} 2007, \textit{Nature}, 448, 169.

737: 

738: \end{thebibliography}

739: 

740: \end{document}

741: