1:
2:
3: % RS Oph Nuller Paper revision history
4: % Original draft RK Barry, April 06, 2007
5: % Revised draft, WC Danchi, 25 May 2007
6: % revised draft, rkbarry, 30May07
7: % revised draft, wcd, 1 june 2007, includes new material on discussion of
8: % of unified model.
9: % revised draft, wcd, 6 june 2007, finished unified model discussion, corrections
10: % of typos, and other edits.
11: %
12: % revised draft, rkb, 6 June 2007, including new graphics, redid mathemetical description of observatory, incorporated comments by Wisniewski, edited
13: %spelling and presentation
14: %
15: % revised to version 10 18Jun2007 RKBarry
16: % incorporated Koresko's comments
17: % added a *lot* of references, rkb
18: % corrected the nova speed section, rkb
19: % corrected figure 3 - dates were wrong. rkb
20: % corrected nova spectral class, luminosity class section
21: % Added argument in discussion and abstract about seeing spectral lines in outer region prior to the
22: % blast wave getting there as an argument for spiral model. rkb
23: % Significantly changed most captions to make them stand-alones.
24: % revised to version 11 and handed off to W. Danchi
25: % Grabbed back from W. Danchi on July 20 and made a further adjustment %in the distance to nova section, etc.
26: % Version 13 -- modified from version 12 -- Danchi -- 23 July 2007
27: %Fixed tables 2 and 3 so they look better and conform to standard forms.
28: %Added Figure 5 of model of RS Oph with spiral shock pattern and wrote caption
29: %Added discussion of Figuue 5
30: %Added data to Table 3.
31: % Danchi -- 1 August 2007 -- try to finish ms. with additional info from
32: % numerical studies
33: % 2 August 2007 -- added new figures describing evolution of circumstellar
34: % material, and various types of luminosity evolution.
35: % 8 August 2007 -- finished cleaning up discussion and conclusion section.
36: % -- this version ready for a final read and cleaning and submission.
37: % 9 August 2007 -- Final clean up and prepare in form for ApJ submission.
38: %
39: % Barry - 4Sep07; made first pass through the referee's comments.
40: % modified and simplified the abstract.
41: % reworked table 3 and table 4.
42: % completely rewrote the abstract, introduction, and data and analysis sections
43: % provided a new figure - figure 4 - showing identification of Spitzer lines with KIN features
44: % r evisited references section to remove typos
45: % Danchi -- 10 October 2007 -- updated ms. to include appendix, and rewritten conclusions.
46: %
47: % Barry -- 30Nov07: completely rewrote several sections, updated figure 1 and figure 2, and title and
48: % shorttitle of paper. Added co-authors Colavita and Acheson. Updated references.
49: % added new definitions. updated tables 4 to include Halpha detect. Prettied table 5.
50: %Danchi -- 30Nov07: revised appendix on luminosity evolution for clarity, based on Traub
51: % comments, revised entire ms based on Traub comments.
52: % Danchi -- 2 Dec07: Added Figure 0 which is a conceptual diagram of the system response.
53: % Barry -- 10 Dec 07: passed along to Marc Kuchner to get input on KIN section.
54: % Barry -- 19 DEC 07: added Gene Serabyns edits, Wes's edits, Akesons edits, Akeson's author list,
55: % pulled AAVSO lightcurve and all references after a discussion with Jeno and Matt - it only made our
56: % arguments obscure and did not support our model.
57: % Danchi -- 22 Dec 2007 -- final minor corrections to text including reference to work of
58: % Lane et al. 2007 and on Eqn. A4.
59: % Danchi -- 04 January 2008 -- made minor revisions on the ms. based on comments
60: % from referee. Added acknowledgment of the referee. Slight mods from Jeno S. and
61: % qualification statement on line fluxes.
62: % Danchi -- 18 Jan 2008 -- fixed alphabetization problem with Vashist. Also added factor of
63: % 2 wording on 4th paragraph of section 4 as requested by Rachel.
64:
65: \documentclass[12pt,preprint]{aastex}
66: %\documentclass[12pt,manuscript]{aastex}
67: %\documentclass{emulateapj}
68: %\documentclass[10pt, preprint2]{aastex}
69: %\usepackage{epsf}
70: %\usepackage{psfig}
71: %\usepackage{graphicx}
72: %\usepackage{lscape}
73:
74: \newcommand{\vdag}{(v)^\dagger}
75: \def\boldsigma{\mbox{\boldmath $\sigma$}}
76: %\def\boldsigma{\bf $\sigma$}
77: \def\boldb{\bf B}
78: \def\hatsigma{\hat{\boldsigma}}
79: \def\ql{\textquoteleft\textquoteleft}
80:
81: \newcommand{\myemail}{richard.k.barry@nasa.gov}
82:
83: \shorttitle{RS Oph: First Science with the Keck Interferometer Nuller}
84: \shortauthors{R. K. Barry et al.}
85:
86:
87: \begin{document}
88:
89:
90: \title{Milliarcsecond N-Band Observations of the Nova RS Ophiuchi: \\
91: First Science with the Keck Interferometer Nuller}
92:
93: \author{ R. K. Barry\altaffilmark{1,2}, W. C. Danchi\altaffilmark{1}, W. A. Traub\altaffilmark{3}, J. L. Sokoloski\altaffilmark{4}, J. P. Wisniewski\altaffilmark{1}, E. Serabyn\altaffilmark{3}, M. J. Kuchner\altaffilmark{1},
94: R. Akeson\altaffilmark{5}, E. Appleby\altaffilmark{6}, J. Bell\altaffilmark{6}, A. Booth\altaffilmark{3}, H. Brandenburg\altaffilmark{5}, M. Colavita\altaffilmark{3}, S. Crawford\altaffilmark{3}, M. Creech-Eakman\altaffilmark{3}, W. Dahl\altaffilmark{6}, C. Felizardo\altaffilmark{5}, J. Garcia\altaffilmark{3}, J. Gathright\altaffilmark{6}, M. A. Greenhouse\altaffilmark{1}, J.Herstein\altaffilmark{5}, E. Hovland\altaffilmark{3}, M. Hrynevych\altaffilmark{6}, C. Koresko\altaffilmark{3}, R. Ligon\altaffilmark{3}, B. Mennesson\altaffilmark{3}, R. Millan-Gabet\altaffilmark{5}, D. Morrison\altaffilmark{6}, D. Palmer\altaffilmark{3}, T. Panteleeva\altaffilmark{6}, S. Ragland\altaffilmark{6}, M. Shao\altaffilmark{3}, R. Smythe\altaffilmark{3}, K. Summers\altaffilmark{6}, M. Swain\altaffilmark{3}, K. Tsubota\altaffilmark{6}, C. Tyau\altaffilmark{6}, G. Vasisht\altaffilmark{3}, E. Wetherell\altaffilmark{6}, P. Wizinowich\altaffilmark{6}, J. Woillez\altaffilmark{6} }
95:
96:
97: \altaffiltext{1}{NASA Goddard Space Flight Center, Exoplanets and Stellar Astrophysics Laboratory,
98: Greenbelt, MD 20771}\email{Richard.K.Barry@nasa.gov}
99: \altaffiltext{2}{Department of Physics and Astronomy, The Johns Hopkins University, Baltimore, MD 21218}
100: \altaffiltext{3}{Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109}
101: \altaffiltext{4}{Columbia University, Department of Physics, NY 10027 }
102: \altaffiltext{5}{Michelson Science Center, Caltech 100-22 Pasadena, CA 91125}
103: \altaffiltext{6}{W. M. Keck Observatory, California Association for Research in Astronomy, 65-1120 Mamalahoa Highway, Kamuela, HI 96743}
104: \begin{abstract}
105:
106: We report observations of the nova RS Ophiuchi (RS Oph) using the Keck Interferometer Nuller
107: (KIN), approximately 3.8 days following the most recent outburst that occurred on 2006 February
108: 12. These observations represent the first scientific results from the KIN, which operates in N-band from 8 to 12.5 $\mu$m in a nulling mode. The nulling technique is the sparse aperture equivalent of the conventional coronagraphic technique used in filled aperture telescopes. In this mode the stellar light itself is suppressed by a destructive fringe, effectively enhancing the contrast of the circumstellar material located near the star. By fitting the unique KIN data, we have obtained an angular size of the mid-infrared continuum of 6.2, 4.0, or 5.4 mas for a disk profile, gaussian profile (FWHM), and shell profile respectively. The data show evidence of enhanced neutral atomic hydrogen emission and atomic metals including silicon located in the inner spatial regime near the white dwarf (WD) relative to the outer regime. There are also nebular emission lines and evidence of hot silicate dust in the outer spatial region, centered at $\sim$ 17 AU from the WD, that are not found in the inner regime. Our evidence suggests that
109: these features have been excited by the nova flash in the outer spatial regime before the blast wave reached these regions. These identifications support a model in which the dust appears to be present between outbursts and is not created during the outburst event. We further discuss the present results in terms of a unifying model of the system that includes an increase in density in the plane of the orbit of the two stars created by a spiral shock wave caused by the motion of the stars through the cool wind of the red giant star. These data show the power and potential of the nulling technique which has been developed for the detection of Earth-like planets around nearby stars for the Terrestrial Planet Finder Mission and Darwin missions.
110:
111: \end{abstract}
112:
113: \keywords{novae: cataclysmic variables --- stars: dwarf novae ---
114: stars: individual(\objectname{RS Ophiuchi},
115: \object{HD162214})--- techniques: interferometric--- techniques: high angular resolution }
116:
117: \section{INTRODUCTION}
118:
119: Classical novae (CN) are categorized as cataclysmic variable stars that have had only one \textit{observed} outburst - an occurrence typified by a Johnson V-band brightening of between six and nineteen magnitudes \citep{war95}. These eruptions are well-modeled as thermonuclear runaways (TNR) of hydrogen-rich material on the surface of white dwarf (WD) primary stars that, importantly, remain intact after the event. Current theory tells us that CN can be modeled as binary systems in which a lower-mass companion -- the secondary -- orbits the WD primary such that the rate of mass transfer giving rise to the observed eruption is very low. Recurrent novae (RN) are a related class of CN which have been observed to have more than one eruption. Like CN, RN events are well-represented as surface TNR on WD primary stars in a binary system, but are thought to have much higher mass transfer rates commensurate with their greater eruption frequency. There are two types of systems that produce recurrent novae -- CVs, in which the WD accretes from a main sequence star that orbits the WD on a time scale of hours, and symbiotic stars, in which the WD accretes from a red-giant companion that orbits that WD on a time scale of years.
120: %Dwarf novae (DN) are distinct from either of these broad classes in that their eruptions, while high in
121: %frequency, are well-modeled as a sudden release of potential energy from a large clump of material as %it is transfered from secondary to the primary.
122:
123: CN and RN produce a few specific elemental isotopes by the entrainment of metal-enriched surface layers of the WD primary during unbound TNR outer-shell fusion reactions. In contrast, type Ia supernovae produce most of the elements heavier than helium in the Universe through fusion reactions leading to the complete destruction of their WD primary. Some theoretical models indicate that RN could be a type of progenitor system for supernovae. Importantly, these theories are predicated on two critical factors: 1.) the system primary must be a compact carbon-oxygen core supported solely by electron degeneracy pressure and 2.) there must be some mechanism to allow the WD mass to increase secularly towards the Chandrasekhar limit.
124: %It does appear that the RS Ophiuchi (RS Oph) system is a member of this subclass of RN and we are %very fortunate to have it as a nearby laboratory to advance our understanding of the astrophysics of %these great engines of nucleosynthesis.
125:
126: The nova RS Oph has undergone six recorded episodic outbursts of irregular interval in 1898 \citep{fle04}, 1933 \citep{ada33}, 1958 \citep{wal58}, 1967 \citep{bar69}, 1985 \citep{mor85} and now 2006. There are also two possible outbursts in 1907 \cite[]{sch04} and 1945 \cite[]{opp93}. All outbursts have shown very similar light curves. This system is a single-line binary, symbiotic with a red giant secondary characterized as K$5.7 \pm 0.4$ I-II \citep{ken87} to a K7 III \citep{mur99} in quiescence and a white dwarf primary in a $455.72\pm0.83$ day orbit about their common center of mass as measured using single-line radial velocity techniques \cite[]{fek00}. \citeauthor{opp93} examined all the outbursts and found that V band luminosity of RS Oph decreased averaged 0.09 magnitudes/day for the first 43 days after outburst. A 2-magnitude drop would then require on average 22 days, establishing RS Oph as a ÒfastÓ nova based on the classification system of \cite{pay57}.
127: %:speed of rsoph
128: % see page 130 in notebook 3 for speed calculation using AAVSO data.
129:
130: The most recent outburst of the nova RS Oph was discovered at an estimated V-band magnitude of 4.5 by H. Narumi of Ehime, Japan on 2006 February 12.829 UT \cite[]{nar06}. This is 0.4 mag brighter than its historical average AAVSO V-band {\it peak} magnitude so it is reasonable to take Feb 12.829 (JD 2453779.329) as day zero. The speed of an outburst is characterized by its $t_2$ and $t_3$ times
131: which are the intervals in days from the visible maximum until the system has dimmed by 2 and 3 magnitudes, respectively. For this outburst the $t_2$ and $t_3$ times are 4.8 and 10.2 days, respectively.
132: %As the WD approaches the Chandrasekhar limit the amount of mass initially ejected by the
133: %thermonuclear runaway (TNR) decreases. It would not, therefore be unreasonable to expect the
134: %initial V-band speed of the nova to gradually increase. Indeed it must be the case that the mass of the %WD has grown over time. If earlier in its secular evolution it was a 1.0 $M_{\sun}$ WD and still had a
135: %20 year recurrence time, its accretion rate would have had to be unacceptably high.
136:
137: %:distance to rsoph
138: The distance to the RS Oph system is of importance to the interferometry community as it effects interpretation of astrometric data (cf. \citet{mon06}). There has been a good deal of disagreement in the literature with a surprisingly broad range, from as near as 0.4 kpc \citep{hac01} to as far as 5.8 kpc \citep{pot67}.
139: %As late as 2001, \citeauthor{hac01} had estimated the nearer distance while early analysis of X-ray
140: %data from the Rossi X-ray Timing Explorer confirmed a distance of 1.6 kpc \citep{sok06}. Distance
141: %estimates using absorption line calculations in the nova's last epoch (\cite{hje86} \& \cite{sni85}) and, %using envelope expansion parallaxes from new radio observations from the current epoch \citep
142: %{rup08} establish a distance at the mid range.
143: \citet{bar08} have recently undertaken a thorough review of the various techniques that have been used to derive a distance to RS Oph and obtain a distance of $1.4^{+ 0.6}_{-0.2} $ kpc.
144: It is this value that we adopt for astrometric calculations in this paper.
145:
146: %The $t_3$ and $t_2$ times noted above do constrain the distance to RS Oph and this is worth noting here. If we restrict ourselves to the use of the $t_3$ time as a more settled figure, and apply the Maximum Magnitude Rate of Decline (MMRD) expression attributed to \citet{dow00} we obtain a range of distances of $D_{RS Oph} = 2.51 \pm 1.28$ kpc. To this broad range we apply an additional constraint. As first noted by \citet{cas85}, the distance to RS Ophiuchi cannot exceed approximately 2.0 kpc because it's spectrum does not contain evidence of absorption by material in the Carina arm of the Galaxy. Inclusion of the Galactic upper constraint and the MMRD lower constraint calculated above yields a range of distances to the nova RS Ophiuchi of $1.61 \pm 0.39$ kpc.
147:
148: %Note: see journal #3, pages 130 - 134
149:
150: The structure of this paper is as follows. We report high-resolution N-band observations of RS Oph using the nulling mode of the 85 meter baseline Keck Interferometer, beginning with a discussion of
151: the nulling mode itself in Section 2. We discuss the observations in Section 3, and the data and
152: analysis in Section 4. In Section 5 we introduce a new physical model of the system,
153: which unifies many of the observations into a coherent framework. The results of this paper and those of
154: other recent observations of RS Oph are discussed in the context of this model in Section 6.
155: Finally, Section 7 contains a summary of our major results and conclusions.
156:
157: % note to Bill - we don't specifically take Lane et al into account so I changed the words "all of the observations" to "many..."
158:
159: \section{THE KECK INTERFEROMETER NULLER}
160:
161:
162: The KIN is designed to detect faint emission due, e.g., to an
163: optically-thin dust envelope, at small angular distances from a bright
164: central star (Serabyn, Colavita and Beichman 2000). Its operation
165: differs from a more common fringe scanning optical interferometer in
166: that a nulling stage precedes the fringe scanning stage (Serabyn et
167: al. 2004, 2005, 2006; Colavita et al. 2006). The basic measurement thus
168: remains the fringe amplitude, but both the meaning of the fringe
169: signal in relation to the source and the processing of the fringe
170: information differ from the normal case of a standard visibility
171: measurement. Here we provide only a brief description of the
172: measurement process, because this has been and will be described in
173: depth elsewhere (Serabyn et al. 2004, 2005, 2006, Colavita et al. 2006;
174: Serabyn et al. 2008, in prep).
175:
176: To remove both the stellar signal and the thermal background in the
177: MIR, a two stage interferometer has been developed.
178: To implement this approach, each Keck telescope is first
179: split into two half-apertures, to generate a total of four collecting
180: sub-apertures. The starlight is then first nulled on the two long (85
181: m) parallel baselines between corresponding Keck subapertures. This
182: generates the familiar sinusoidal fringe pattern on the sky (Fig. \ref{KINconcept}),
183: except that the central dark fringe on the star is achromatic, and
184: fixed on the star, to achieve deep and stable rejection of the
185: starlight. After the nulling stage, the residual, non-nulled light
186: making it through the first stage fringe pattern is measured by a
187: fringe scan in a second stage combiner, the ``cross-combiner'', which
188: combines the light across the Keck apertures ($\sim$ 4 m
189: baseline). Thus, what is measured is the fringe amplitude of this
190: ``nulled source brightness distribution'' (Serabyn et al. 2008, in
191: prep.) This quantity is then normalized by the total signal. This is
192: measured by moving the nullers to the constructive phase, and again
193: scanning the cross-combiners. The basic measured quantity, the null
194: depth, $N$, is then the ratio of the signals with the star in the
195: destructive and constructive states. $N$ is related to the classical
196: interferometer visibility $V = (I_{max} -I_{min})/(I_{max} +I_{min})$,
197: the modulus of the complex visibility $\hat{V}$, by a simple formula:
198:
199: \begin{equation}
200: N ={ {1-V}\over{1+V}} ~~ .
201: \end{equation}
202:
203: The long nulling baselines produce fringes spaced at about 23.5 mas at
204: 10 $\mu$m, while the short baseline produces fringes spaced at $ \sim
205: 400$~mas,which is similar to the size of the primary beam and is
206: assumed to be large compared to the extent of the target object.
207: Modulating its phase therefore modulates the transmitted flux of the
208: entire astronomical source, as modified by the sinusoidal nuller
209: fringes, and so the amplitude of this modulated signal gives the flux
210: that passes through the fringe transmission pattern
211: produced by the long baseline nullers.
212:
213: \section{OBSERVATIONS}
214:
215: We observed the nova about an hour angle of about -2.0 on the Keck Interferometer in nulling mode on 2006 Feb 16 with a total of three observations between day 3.831 and 3.846 post-outburst bracketed with observations of two calibrators stars, $\rho$ Boo and $\chi$ UMa. Data were obtained at N-band
216: (8 to 12.5 $\mu$m) through both ports of the KALI spectrograph with the gating of data for long baseline phase delay and group delay turned on. The Infrared Astronomical Satellite (IRAS) Low Resolution Spectrometer (LRS) spectra for the calibrator stars were flux-scaled according to the broadband IRAS 12 $\mu$m fluxes. The nova data were flux calibrated and telluric features were removed using calibrator
217: data. These calibrators are well matched to the target flux and their size and point-symmetry are well known. A journal of observations is presented in Table \ref{observations}.
218:
219: Our data analysis involves removing biases and coherently demodulating the short-baseline fringe with the long-baseline fringe tuned to alternate between constructive and destructive phases, combining the results of many measurements to improve the sensitivity, and estimating the part of the null leakage signal that is associated with the finite angular size of the central star. Comparison of the results of null measurements on science target and calibrator stars permits the instrumental leakage - the ``system null leakage'' - to be removed and the off-axis light to be measured.
220:
221: %:Sources of Measurement Noise
222:
223: Sources of noise in the measurements made by this instrument have been well described elsewhere \citep{kor06}, however, we outline them here for reference. The null leakage and intensity spectra include contributions from the astrophysical size of the object, phase and amplitude imbalances, wavefront error, beamtrain vibration, pupil polarization rotation, and pupil overlap mismatch. There are also biases, mostly eliminated by use of sky frames, in the calculated fringe quadratures, caused mainly by thermal background modulation due to residual movement of the mirrors used to shutter the combiner inputs for the long baselines. Another source of error is the KALI spectrometer channel bandpass, which is large enough to produce a significant mismatch at some wavelengths between the center wavelength and the short-baseline stroke OPD. This effect, termed ÔwarpingÕ, distorts the quadratures and is corrected by a mathematical dewarping step accomplished during calibration. There is also the effect of the partial resolution of extended structure on the long baselines which will cause the flux to be undercounted by some amount depending on the spatial extent and distribution of the emitting region. Compared to the error bars this is a rather small effect for most normal stars, and is unlikely to have much influence on the actual spectrum, though, unless the emission lines are coming from a very extended shell Ð approaching 25 mas in angular size. We do not expect this to be the case for nova RS Oph at day 3.8.
224:
225: The most important contributor to measurement noise is sky and instrument drifts between target and calibrator.
226: Other potential sources of noise include the difference between the band center wavelength for the interferometer and that of the IRAS LRS calibrator, undercounting of stellar flux resulting from glitches in the short-baseline phase tracking, and the chromaticity of the first maximum of the long-baseline fringe. None of these are significant for the following reasons. First, the calibration is based not on broadband photometry but on the IRAS LRS spectrum. Second, short-baseline tracking glitches happen nearly as frequently on target observations as on calibrator observations so they should have minimal effect on overall calibrated flux. Third, the effect of chromaticity should be negligible because fringe detection is done on a per-spectral-channel basis. As a result, it is only affected by dispersion within the individual KALI spectral channels, which are about 0.3 $\mu$m in width.
227:
228: \section{DATA \& ANALYSIS}
229:
230: %: figure 1 description
231:
232: We developed a mathematical solution and software suite to model the observatory and source brightness distribution. We used this suite to conduct an exhaustive grid search and to
233: generate a monte carlo confidence interval analysis of solution spaces of these models. We explored three types of models for the source surface brightness distribution; Gaussian, disk, and shell. Limited (u,v) coverage permitted only rotationally-symmetric models with two parameters - size and flux. We used $\chi^2$ minimization to obtain the best fit models for both the inner and outer spatial regimes simultaneously. Table \ref{models} displays size measurements, flux values, and one-$\sigma$ confidence interval values. The error bars have been increased slightly beyond the $\chi^2$ values
234: to incorporate the effect of an adopted 0.005 one-sigma systematic error which would be correlated among measurements at different wavelengths.
235:
236: The measurements made with the two KALI ports are somewhat independent - the data they produce are combined for purposes of fringe tracking, but not for data reduction. The system null and the final calibrated leakage are computed separately for the two ports. The apparent inconsistency detected between the ports is the result of optical alignment drifts at the time of the measurement. In particular, the last calibrator measurement showed a sudden change in the system null for Port 1, while for all the other calibrator measurements the system nulls were stable. We therefore compared our source brightness distribution models against KALI port 2 data alone. For the best-fit models in Table \ref{models} we used {\it Spitzer} spectra to identify and remove emission features centered at 8.7, 9.4, 10.4, 11.4, and 12.5 $\micron$ in the KIN inner and 8.9, 9.8, and 11.4 $\micron$ in the KIN outer spatial regime. We removed the emission feature data because our intention was to model the continuum.
237:
238: Figure \ref{modelfit} shows two sets of $8 - 12.5$ $\mu$m spectra of RS Oph on day 3.8 post-outburst. The upper plot shows the outer spatial regime, which is the dimensionless null leakage spectrum, i.e., the intensity of light remaining after destructive interference divided by the intensity spectrum, plotted against wavelength in microns. The lower plot is the intensity spectrum, which is light principally from the inner 25 mas centered on the source brightness distribution orthogonal to the Keck Interferometer baseline direction - 38 degrees East of North. The null leakage spectrum may be broadly described as a distribution that drops monotonically with increasing wavelength overlaid with wide, emission-like features. The intensity spectrum, with an average flux density of about 22 Jy over the instrument spectral range, may be similarly described but with a continuum that has a saddle shape with a distinct rise at each end. The underlying shape curves upward for wavelengths shorter than approximately 9.7 $\mu$m and longer than 12.1 $\mu$m. Overlaid on each of these are traces representing simple models of source brightness distributions fit to the data, described above.
239:
240: %:Figure 2 description
241:
242: %Note I intend to go through the spectra again and carefully derive fluxes and equivalent widths to include in the line ID table.
243:
244: Figure \ref{kn_spectra} shows the inner and outer spatial regimes from the KIN data together with {\it Spitzer} data from day 63 \citep{eva07b} . The absolute fluxes of these data are unscaled and given with a broken ordinate axis for clarity.
245: The outer KIN spectrum flux has been multiplied by 2 to correct
246: for the transmission through the fringe pattern for extended
247: sources.
248: In efforts to identify the sources of these emission features we took the high resolution {\it Spitzer} spectra and, using boxcar averaging, re-binned the data until it and the KIN data had equivalent resolution. The binned {\it Spitzer} spectrum is quite similar in character to the KIN spectra. Importantly, the sum of the inner and outer spatial regime KIN spectra is nearly identical to the binned {\it Spitzer} spectrum with the exception that the absolute scale magnitude is, on day 3.8, about an order of magnitude greater than that of the {\it Spitzer} spectrum. This is as expected because
249: the measured flux drops with time after the peak due to cooling and the summed inner and outer spatial regime data should very nearly reproduce the transmission of a filled aperture telescope of equivalent diameter.
250:
251: The wavelength range sampled by the KALI spectrometer, 8 - 12.5 $\mu$m, covers many important discrete transitions including molecular rotation-vibration, atomic fine structure, and electronic transitions of atoms, molecules, and ions. This range also samples several important transitions in solids such as silicates found in dust and polycyclic aromatic hydrocarbons (PAHs). With the exception noted below, spectral features in the KIN data are clearly not resolved by the instrument and are often, in the associated {\it Spitzer} spectrum, Doppler broadened and blended. Also, because the {\it Spitzer} spectra were not taken contemporaneously with the KIN spectra and because of the transient nature of the RN, identification of KIN features with
252: {\it Spitzer} lines is necessarily tentative.
253:
254: %pg 34, vol 4 rkb journal
255: %: Figure 3 description : Spitzer spectra and continuum discussion
256: Figure \ref{continuum} shows {\it Spitzer} spectra on days 63, 73 and 209 after peak V-band brightness. The continuum decreases monotonically with time. The continuum values are approximately 1.4, 1.1, and 0.25 Jy for April 14, 26 and September 9, respectively. The first two spectra on April 14 and 16 show strong atomic lines with no obvious evidence of dust emission. However it is expected that thermal bremsstrahlung from the central source will overwhelm any other faint sources. Additionally, there are several narrow emission features in this latter spectrum which appear to be similar to those in the earlier spectra. In contrast, the spectrum taken on September 9 shows a distinct broad emisison feature from 8.9 to 14.3 $\mu$m peaking at 10.1 $\mu$m. We proceed on the assumption that this is emission from dust in the vicinity of the nova.
257:
258: Referring again to the KIN spectrum in Figure \ref{kn_spectra}, strong continuum radiation is apparent in both inner and outer spatial regimes. In contrast, while the continuum was still clearly visible in J, H, I, K bands on February 24 and detectable on April 9 \citep{eva07a}, it has subsided by day 63 in Figure \ref{kn_spectra} and Figure \ref{continuum}. Comparing the spectra from RS Oph to those from V1187 Sco \citep{lyn06} note that the continuum given for V1187 Sco is non-Planckian showing an excess longwards of 9 $\mu$m and is strongly red as compared to a F5V Kurucz spectrum. The V1187 Sco and RS Oph continua have slopes that agree to within 10\%. While both free-bound and free-free transition processes lead to emission of continuum radiation, in the MIR spectral range thermal bremsstrahlung free-free emission dominates. We attribute the drop in continuum radiation to the transition to line-emission cooling mechanisms. Also, by the time {\it Spitzer} data were taken, the object had become less dense and so the emission coefficient for thermal bremsstrahlung (proportional to number density of protons and electrons) had dropped considerably. The continuum emission is described in detail by \citet{bar08b}.
259: %: table 3 description: high resolution emission lines
260: % reference pages 20 - 37 in rkb journal volume 4.
261:
262: Table \ref{spitzlinelist} gives identification of all narrow features in Figure \ref{continuum} where it is possible to do so, as follows. We generated line lists of atomic species assuming that all ionization stages of all elements were possible for the three {\it Spitzer} spectra. We assumed standard cosmic abundances \citep{gre84}. After continuum normalization, each emission line was fitted with a Gaussian and, where necessary, de-blended using IRAF. Our fitted emission features were compared with
263: the {\it Spitzer} list.
264:
265: %We then obtained line wavelengths and fluxes and made a comparison to the line list we calculated
266: %previously.
267: %Where there was ambiguity due to the closeness of generated lines we identified the species with the %greatest cosmic abundance as the source. If atomic species in the calculated line list had significantly %similar abundances we compared transition probability and opted to report the species with the
268: %highest probability. In the unlikely event that the Einstein transition probabilities {\it and} the cosmic
269: %abundances were similar, we report whichever species had the lowest-lying starting energy state on %the assumption that those species would have the greatest occupation numbers although the
270: %environment was doubtlessly not in thermal equilibrium. For the species recorded for the {\it Spitzer} %data taken on September 9, 2006, an additional restriction was levied on the generated atomic line list %in the event that two or more calculated species shared very similar emission line wavelength. In this %case, based on the evidence of the solid-state feature in the spectrum, we assumed that the
271: %environment of RS Ophiuchi at that time had become substantially nebular in character. Here we
272: %assumed that metals would more likely have condensed onto dust grains in the vicinity of the nova
273: %and that allowed transitions of hydrogen and helium, magnetic dipole and electric quadrupole of all
274: %other elements, and ground-state forbidden transitions would be the likely source of line emission
275: %\citep{cow86}. Importantly, we still allowed the possibility of cosmic-abundance atomic line emission
276: %as we had no way of knowing the particular depletion state in the vicinity of the nova a priori. Finally, %we suppressed all hydrogen fine structure and auto-ionization states in this line identification list as
277: %the former would be irretrievably blended and the latter would provide no new information.
278:
279: %: table 4 description: KIN features
280: % reference pages 38 - 40 in rkb journal volume 4.
281: Table \ref{kinfeaturelist} gives our identification of particular emission sources with the continuum-normalized spectral features in the KIN inner and outer spectra. The center wavelength of each of the broad features is listed in the first column, while the second column displays a width for each feature defined as the cutoff wavelength at the intersection of the the feature and the unity continuum line - the full-width zero intensity (FWZI) level of the feature. The flux and the one-$\sigma$ uncertainty
282: contained in that feature through measurement of the total area under it and above the unity continuum level is in the third column.
283: The fourth column shows the particular atomic species found in the {\it Spitzer} spectrum with each KIN spectral feature. Because {\it Spitzer} lacks the spatial resolution to discriminate the inner from the outer regions of the nova and the KIN lacks the spectral resolution to discriminate among atomic species we assumed that features in the KIN spectra would most reasonably be identified with {\it Spitzer} features that were in later spectra. In particular, {\it Spitzer} emission lines features identified with corresponding ones in KIN inner spatial regime, dominated by light originating in the close vicinity the WD, are well-represented by a cosmic distribution of atomic elements. We assumed that any condensates within the blast radius of the nova would be sublimated away and dissociated into atoms, and that any nucleosynthesis that occurred in the outer layers of the WD during TNR would negligibly impact abundances. The KIN inner spatial regime spectral features were keyed to {\it Spitzer} spectral features in the 4/16 and 4/26 spectra. Similarly, the KIN outer spatial regime, light predominantly from a region $\sim$ 17 AU from the WD, is keyed to the {\it Spitzer} spectrum taken on 9/9 \citep{eva07c} as it is assumed that the abundances of atomic species in the latter, nebular {\it Spitzer} spectrum, would reasonably be representative of the environment observed by that KIN channel.
284:
285: Note that other symbiotic novae (e.g. V1016 Cyg, RR Tel) have broad emission features around 10 and 18 $\mu$m, evident in our {\it Spitzer} spectrum, that have been successfully fitted by crystalline silicate features (e.g. \citet{sac07, sch01, eyr98}). Motivated by this fact, we calculated the temperature and emission SED of various species of dust at the range from the pseudo-photosphere of the WD at which the KIN's outer spatial regime has maximum transmission. The center of the first constructive fringe, when the null fringe is located on top of the WD, at 9.8 $\mu$m and a projected baseline of 84 m, is $\sim$ 12 mas from the WD. At a distance to the object of 1.4 kpc \citep{bar08} this corresponds to about 17 AU. At day 3.8, the pseudo-photosphere of the WD has a luminosity of 1.6$\times 10^5$ $L_{\odot}$ (radius, 18.1 $R_{\odot}$, and
286: temperature, 27050 K), computed by
287: interpolating between values for the post-outburst evolution displayed in Table 5.
288: The temperature of a black dust grain in thermal equilibrium is 272 K, using Eqn. A3 from
289: the Appendices in this paper, and is well below the sublimation temperature of silicate dust, $\sim$ 1500 K.
290: The dust feature in the {\it Spitzer} 9/9 spectrum is wide enough at FWHM that it falls across over seven spectroscopic elements in the KIN outer spatial channel meaning that the dust is both spectrally resolved and spatially localized.
291: %A detailed analysis of the continuum emission in these spectra is reported in \citet{bar07b}.
292:
293: %cf. Barry journal, volume 4, pg 8, 9, 12 - 16, 37and private communications with Dwek.
294: %:figure 4 description
295: Figure \ref{keckspitzlines} displays the association of identified {\it Spitzer} atomic emission lines with the continuum-normalized KIN inner and outer regime spectral features. Clearly the two traces are markedly different. Note that when the inner and outer spatial regimes are summed, the result closely follows the re-binned {\it Spitzer} spectra with the exception of the KIN outer spatial regime feature (lower trace) centered at 9.8 $\mu$m. Based on our assumptions of a primarily nebular environment in the vicinity of the outer spectrum, the atomic metals evident in the upper trace (KIN Inner spatial regime) would be unlikely to contribute to this feature. In any case, the total power in these metal lines in the wavelength range of this feature (cf. Table \ref{spitzlinelist}), including Ca I, Si I, C I, \& C II, evident in the upper trace is 0.05 Jy, while that in the broad feature centered at 9.8$\mu$m in the lower trace exhibits is 0.24 Jy. Our models suggest that the source of this feature may be hot silicate dust in the temperature range 800 - 1000 K.
296:
297: Note that our KIN data detects the faint emission from well outside of the blast radius, assuming an initial shock front velocity of 3500 km/s \citep{obr06} and negligible deceleration. The radiation from
298: this spatial region originates primarily from material around the nova that has been illuminated and warmed by photons from the nova flash and, as a result, must have existed {\it before} the nova event. This establishes that silicate dust, created in the vicinity of the RS Ophiuchi system some time previous to the 2006 outburst, is detected by our measurements, and is consistent with the conclusions of \citet{eva07c}.
299:
300: %In efforts to identify the sources of these emission features we took the high resolution {\it Spitzer} spectra and, using boxcar averaging, down-binned the data until it and the KIN data had the same equivalent resolution. We then identified spectral lines in the full-resolution {\it Spitzer} spectrum. These are given in Table \ref{spitzlinelist}.
301: %This table also shows identification of these spectral lines where they are evident in the KIN spectrum. We then fit a continuum to the KIN spectrum and determined the remaining flux in the particular continuum-normalized spectral channel for which the line was identified. We give this number as the bounding flux for that particular line together with notes about the width of the spectral channel - a value that varies across the N-band. While it is possible for forbidden lines to be excited soon after detonation in ionized, low-density regions of the wind ahead of the shock, we assumed that such transitions found in the {\it Spitzer} spectrum on day 63 were not strongly represented in the KIN spectra on day 3.8. All spectra were continuum normalized prior to analysis and only minimum $1 - \sigma$ detections and better are noted. Line fluxes were determined by Gaussian line fitting and de-blending where needed.
302:
303: %We note that \citet{eva07a} found spectra in the I, J, H, and K bands on days 11.81 and 20.75 dominated by hydrogen recombination lines including higher members of the Pfund series. Also present in these early spectra were OII and un-ionized helium and oxygen. In the day-63 {\it Spitzer} data there are Ne VI and Ne II lines (blended with $Pf\alpha$ and $Hu\alpha$ respectively) just outside of the KIN spectral range from which it may be determined that the evolution of the event has advanced into the coronal phase and the medium has become very tenuous as the meta-stable states giving rise to the forbidden lines are no longer collisionally de-excited. The {\it Spitzer} spectrum given is replete with hydrogen recombination lines. The final low-resolution datum at 12.2 $\mu$m is strongly influenced by and identified with the {\it Spitzer} 12.3719 $\mu$m HI 7-6 $Hu\alpha$ line blended with HI 11-8 at 12.3833 $\mu$m. As indicated in Table \ref{spitzlinelist} at least one neutral hydrogen feature is seen in the inner spatial regime that is not evident in the outer. PAHs, such as Coronene, emit strongly at 8.6 and 11.9 $\mu$m. Both of these could be represented in the data, with the latter, blended with other emission lines, contributing to the rise at the longwards end. In agreement with {\it Spitzer} spectrometry conducted on day 63 \citep{eva07b} in which no thermal emission from dust was detected, we see no evidence of silicate dust in the KIN inner spatial regime spectrum as the continuum-normalized spectrum has no contribution from silicon . The outer spatial regime, however, does show evidence of neutral Si emission at 9.407 - an important component of dust.
304:
305: %: description of figure 5.
306:
307: % see rkb notebook #3, page 13 for calculations
308:
309: %Figure \ref{lightcurve} displays the AAVSO V-band photometry for the current epoch of this cataclysmic variable star. Centered at about September 4, 2006 (JD 2453983) there is a strong depression in stellar magnitude. The lightcurve is labeled with the dates that {\it Spitzer} spectra were taken \citep{eva07c}. In particular, note that the {\it Spitzer} data of September 9, 2006 shows the broad, rounded spectral features noted earlier as being associated with circumstellar dust. In classical novae, fluctuations in V-band light have been shown to be closely related to rapid changes in mass loss from the primary and have also been shown to be contemporaneous with re-emission of the light at infrared wavelengths. Taken together, these are evidence for obscuration of the light source by dust. As discussed previously, the KIN measurements indicate that circumstellar dust exists between nova events and it could be effects from this dust that is evident in Figure \ref{lightcurve}.
310:
311: %\section{Discussion}
312:
313: %\section{Model of the evolution of circumstellar material in the post-outburst phase applicable to the KIN data and other high angular resolution observations}
314:
315:
316: \section{A physical model of the recurrent nova}
317:
318: One aspect of the RS Oph binary system that has been neglected in the current literature
319: regarding the recent outburst
320: is the effect of the motion of the two stars through the wind created by mass loss from the red giant
321: star. Garcia (1986) suggested that there could be a possible ring of material of
322: diameter less than 40 AU around the RS Oph binary system or possibly surrounding the red giant
323: component, based on his measurements of an absorption feature in the core of the Fe II emission
324: line profile at 5197 \AA. The observations were performed in 1982 and 1983, several
325: years before the 1985 nova outburst.
326:
327: Subsequently, and motivated by somewhat different observations,
328: \citet{mas99} computed three-dimensional hydrodynamical
329: models of morphologies of the envelopes of binaries with detached WD and RG/AGB components in general. Their purpose was to see if these models could
330: reproduce some of the observed characteristics of axisymmetric or bipolar pre-planetary
331: nebulae. Their study focused on a parameter space that encompassed outflow velocities
332: from 10 to 26 km/s, circular orbits with binary separations from 3.6 to 50 AU, and binary
333: companions having a mass range of 0.25 to 2 M$_{\odot}$. For binary separations of
334: about 3.6 AU and mass ratios of 1.5, it was possible to generate a single spiral shock that
335: winds 2-3 times around the binary before it dissipates at $>$ 25 times the radius of the
336: RG star. There is a density enhancement of about a factor of 100 over the normal density
337: in the wind in the plane of the orbit of the two stars, and an under density or evacuated region
338: perpendicular to the plane of the orbit. Observational support for this model was found recently
339: by \citet{bod07} who detected a double-ring
340: structure in HST data, which they interpreted as due to an equatorial density
341: enhancement and \citet{mau06} who observed a spiral pattern around the AGB star AFGL 3068 both of which were consistent with the model of \citeauthor{mas99}. The underlying binary, a red giant and white dwarf, was discovered by \citet{mor06}, who also determined the binary separation and
342: hence approximate orbital period, which was consistent with expectations from the appearance
343: of the spiral nebula pattern seen by \citeauthor{mau06} and the model.
344:
345: Figure \ref{dust-sub_01} shows the proposed geometry of the nebula in the plane of the orbit of the RS Oph system based
346: on the parameters adopted from \citet{dob94} for the system, including an orbital period of 460 days,
347: and an inclination angle of about 33 degrees, for the epoch just prior to the nova outburst.
348: The spiral shock model produces an
349: archimedian spiral nebula, with the separation between adjacent windings of about 3.3 mas
350: based on the period noted above and a wind speed from the red giant star of 20 km/s. Based on
351: these parameters
352: we estimate that about 17 such rings could be created between outbursts, with the overall size
353: of the nebula of the order of 100 mas. However,
354: there could be as few as 10 rings to as many as about 20 rings, depending on the parameters, some
355: of which are not well known.
356:
357: In order to better understand the Keck Nuller observations and other high angular resolution
358: observations, we have modeled the evolution of the circumstellar nebula following the outburst,
359: which we display in Fig. \ref{CSMatterEvol}. For the purposes of this discussion, we adopt the model of
360: Hachisu and Kato (2001),
361: but with the simplification of a flat accretion disk geometry, i.e., not warped as in their paper.
362: The details of our calculations are presented in the Appendices to this paper.
363: %(We note that the result of \citet{lan07} does not specifically apply to these results nor does it
364: %contradict them as \citeauthor{lan07} used a simplified free-free continuum model and did not
365: %address N-band dust emission.)
366:
367: The results are of fundamental importance to the interpretation of our KIN data. We begin
368: with the most immediate effect after the blast, which is the sublimation of dust within a zone
369: where the temperature of a blackbody dust grain would be $>$ 1500 K in equilibrium. Figure \ref{dust-sub_02}
370: displays the evolution of the dust sublimation radius, which is roughly 4-5 mas until about
371: day 70, after which it steadily declines to $<<$ 1 mas about 250 days after the outburst.
372: This means that much or all of the dust within this zone has entered the gaseous phase, except
373: for a small ``sliver'' of material in the shadow of the red giant star. This provides additional
374: hot gas (rich with metals) that is subsequently affected by the blast wave passing through
375: within the next few days.
376:
377: As the two stars move relative to each other in their orbit about their common center of mass,
378: the location of the shadow of the nova moves, and consequently the material in the shadow that
379: has not been affected by the blast wave from the nova will be sublimated during this luminous
380: phase, creating hot gas in the vicinity within a few mas of the stars. This material may still have
381: the type of repeating density structure that was initially present, and may be observable
382: with high angular resolution instrumentation.
383:
384: Whether or not this particular material is observable, depends in part
385: on the density distribution as a function of latitude of the plane of the orbit of the two stars.
386: If the material is uniformly distributed over 4$\pi$ steradians, then the fraction
387: of the total solid angle subtended by the red giant star as seen by the white dwarf star is
388: approximately 0.13-0.29\% for a RG star between 27 and 40 R$_{\odot}$ and a binary separation of
389: 1.72 AU. However, the hydrodynamical studies of Mastrodemos and Morris (1999) show that
390: the density falls off steeply as a function of latitude, and is as low as 1\% of the mid-plane density by
391: latitudes of about $\pm$ 40 degrees. This gives a scale height of about 9 degrees. Thus, the
392: red giant star subtends a much bigger fraction of the solid angle up to this scale height,
393: i.e., to as much as 1.9\%.
394:
395: Furthermore, studies of the supernova blast waves around red giant stars indicate that the
396: blast wave diffracts around the RG star, with a hole in the debris of angular size $\sim$ 31-34 degrees,
397: which did not depend strongly on whether the companion star was indeed a RG star or
398: a main sequence or subgiant star (Marietta, Burrows, \& Fryxell 2000). In this case the fraction
399: of the sky subtended by this hole in the debris field is $\sim$ 10-13 \%. If the material is concentrated
400: in the mid-plane, the effect is substantially bigger as noted in the previous paragraph. Thus
401: there are several reasons to expect that material in the shadow of the RG star can have an observable
402: effect.
403:
404: Stripping of material from the red giant companion by the blast wave for supernova has been studied by Wheeler, Lecar, and McKee (1975) and by Lyne, Tuchman, and Wheeler (1992), however, Lane et al. (2007) showed that stripping of material for the RG star is negligible for the less energetic blasts from RS Oph.
405:
406: Another effect is the heating of the surface of the red giant star
407: that faces the nova (see Appendices), as seen in Figure \ref{RG-face-temp}, for times past the maximum in the visible light curve. Our calculations indicate that this side of the red giant star increases from about
408: 3400 K to as much as 4200 K within a few days past the outburst. The temperature steadily declines
409: from day 70 until it reaches equilibrium with the other side around day 250. The surface of the
410: RG star is initially heated by the shock from the blast wave (not included in our calculations), however, the continued heating during the
411: high luminosity phase is important as it affects the interpretation of data from modern complex
412: instrumentation such
413: as stellar interferometers and adaptive optics systems, where various subsystems are controlled at wavelengths other than the measurement wavelength.
414:
415: %: 1730
416:
417: Figure \ref{CSMatterEvol} displays a schematic view of the system geometry from days
418: 4 to 90 after the recent outburst. Assuming typical wind velocities of about 20 km/s for the red giant wind,
419: there are roughly 17 rings separated by approximately 3.3 mas that form between RS Oph outbursts.
420: The top left panel displays the system geometry at 4 days post-outburst. A gray ring is drawn in the
421: center of the figure to indicate the size of the region affected by the blast at this epoch.
422: In (a) the outer part of the spiral
423: is overlaid with light gray to indicate that it is not known if the material stays in a coherent spiral
424: past the first few turns. The diameter of the shocked region is about 8.8 mas assuming the
425: blast wave travels at a velocity of $\le 1800$ km/s {\it in the plane of the orbit}, using the velocity measured
426: by \citet{che07}, at approximately the same epoch as our measurements. (We acknowledge that most out-of-plane blast-wave velocities noted in the literature are much higher than this.)
427: In this figure we assume the blast wave moves at constant velocity as it traverses the
428: spiral shock material. We show the extreme case in which
429: the blast wave is 100\% efficient in sweeping up material, thus creating a ring-like structure that propagates outward from the
430: system. Note this figure is meant only to be illustrative and differs in detail from estimates of the
431: position of the blast wave from observations, such as \citet{obr06} who obtained a value
432: of the shock radius of 8.6 mas at day 13.8 by which time the blast wave had apparently slowed considerably. We obtain a value of 10.5 mas for the shock radius assuming a constant velocity of $\sim$ 1800 km/s from one day past the initiation of the TNR process,
433: which we take to be about 3 days before maximum light in V band \citep{sta85}. The value of 8.6 mas at day 13.8 is
434: consistent with a somewhat lower mean velocity, of the order of 1400 km/s.
435: It is beyond the scope of this paper to compute the evolution of the
436: blast wave, however, this figure makes a connection with the previous observational evidence
437: for a ``ring" of material concentrated in the plane of the orbit as discussed by Garcia (1986) and with
438: the recent observations of \citet{che07} who observe different velocities perpendicular
439: to the plane of the orbit than in the plane of the orbit.
440:
441: The evolution of the luminosity of the red giant and nova is also significant in aiding our
442: understanding of the observations. Figure \ref{LumEvol} (a) displays computed V band light curves
443: up to day 250 after the outburst including the effect of the irradiation of the
444: accretion disk and the irradiation of the red giant star by the nova. Note this calculation overestimates
445: the total luminosity of the nova during the
446: period from about 10 days to about 50 days, and this is likely due to the simplified disk
447: geometry that we have employed in our own calculations. Most importantly these
448: light curves show that the V band luminosity is dominated by the nova for about the first 50-70 days, and after that the red giant star dominates the V band luminosity. This is significant as most telescopes
449: track on V band light (including interferometers) and the tracking center then moves by a
450: mas or so during the post-outburst evolution of the system. H band luminosity evolution
451: is plotted in Fig. \ref{LumEvol} (b), and is different than that of the V band evolution. At H band the nova
452: dominates the luminosity only for the first 5 days or so, and after that the H band light curve is
453: completely dominated by the red giant star. This means the phase center for fringe detection
454: is offset from that of the tracking center from days 5 onward by a mas or so, as mentioned above.
455: The N band luminosity, displayed in Fig. \ref{LumEvol} (c),
456: evolves like that of the H band, and the nova dominates the mid-infrared
457: light only up to day 4. After that there is an offset between the tracking center and N band
458: fringe center like that noted for the H band.
459:
460: \section{Implications to the Keck Nuller and other high-angular resolution observations}
461:
462: We now reexamine data from the 2006 outburst
463: within the framework of the spiral shock model of the geometry of RS Oph presented in the last section.
464: In its simplest form, this geometry evolves into a bipolar morphology similar to that seen
465: in the radio emission from the outburst that was discussed by O'Brien et al. (2006), as seen
466: in Figure \ref{CSMatterEvol}(d).
467:
468: Such a geometry provides a natural explanation for some of the differences between the
469: interferometric and other measurements, as the blast wave would be restricted and slowed in the orbital plane due to
470: the high density regions, while its flow would be relatively unimpeded perpendicular to
471: that plane. This corresponds well to what was observed by Chesneau et al. (2006),
472: where the observed velocity components came from distinctly
473: different position angles. These authors also calculate that the red giant star should be at position
474: angle 150-170 degrees at the time of the outburst. Recently, \citet{bra08} have derived spectroscopic orbits for both components of the RS Oph system based on the radial velocities of the M giant absorption lines and the broad emission wings of H$\alpha$. They have also constrained the orbit inclination, $i \geq 49 \pm 3$, using the estimated hot component mass, $M_{\rm h}\, \sin^3 i = 0.59\, \rm M_{\odot}$, and assuming that the white dwarf mass cannot exceed the Chandrasekhar limit. Thus the plane
475: of the orbit is such that the high velocity flow would be expected to be mostly East-West and
476: the the continuum emission would have an elliptical shape with the position angle measured
477: by Chesneau et al. (2006). Furthermore, the spiral shock wave tends to have the largest
478: densities near the white dwarf star and on the opposite side of the red giant star, the separation
479: being of the order of two to four times the separation between the two stars, i.e., a few
480: mas. The larger separation occurs when
481: the red giant star has been spun up so that it is synchronous with the orbital
482: motion. Thus it is possible that some of the data can be interpreted in terms of emission
483: from hot clumps of material within the spiral shocks. Detailed radiative transfer calculations
484: will need to be performed to refine this model.
485:
486: There is a wealth of recent observational work on RS Oph across the spectrum from X-ray \citep{sok06, bod06},
487: near-infrared \citep{eva07a, che07, das06}, radio \citep{obr06}, and now mid-infrared \citep[this work]{eva07b, eva07c}.
488: Generally speaking, the observational picture is consistent with the shocked wind model
489: as described by \citet{bod85}. Extensive modeling of the light curves has been
490: performed by \citet{hac00, hac01}, which include effects such as the irradiation of the
491: red giant star by the white dwarf, and the
492: accretion disk, which become important
493: about 4 days post-outburst, and which are the basis for
494: our own calculations presented in Figs. \ref{dust-sub_01},\ref{dust-sub_02}, and \ref{LumEvol}. In this general picture the recurrent nova
495: differs from a classical nova in that the high-velocity ejecta from the WD is impeded by the
496: wind from the red giant star, which in turn generates a shock wave that propagates through the
497: red giant wind. Observations of the shock wave in the near-infrared
498: by \citeauthor{das06} confirm this picture and
499: trace the evolution of the widths of Pa $\beta$ and O I lines
500: as a function of time post-outburst, and clearly show
501: the expected free-expansion phase of the shock ended about 4 days post-outburst. This is also consistent with independent measurements conducted by \citet{bod06} in which this phase was proposed to have ended by approximately day 6.
502:
503: Most of the observational work has used spectroscopic methods,
504: but only the interferometric observations
505: are capable of spatially separating the various components of RS Oph that contribute to the
506: emission that is seen spectroscopically. Chesneau et al. (2007) observed RS Oph 5.5 days post-outburst in the
507: continuum at 2.13 $\mu$m, Br $\gamma$ at 2.17 $\mu$m, and He I at 2.06 $\mu$m with the
508: AMBER instrument on the VLTI. They fitted their data with
509: uniform ellipses, gaussian ellipses, and a uniform ring. The models had excellent consistency
510: in terms of the position angle of the ellipse ($\sim$140 degrees) and the ratio between minor and
511: major axes (0.6). For the continuum at 2.13 $\mu$m, a uniform ellipse had a major axis of
512: 4.9 $\pm$ 0.4 and minor axis 3.0 $\pm$ 0.3 mas, while the gaussian ellipse had a major
513: axis of 3.1 $\pm$ 0.2 and minor axis of 1.9 $\pm$ 0.3 (FWHM). These measurements
514: are consistent with the expected size of the shocked region at the epoch of their measurements.
515:
516: Monnier et al. (2006) presented IOTA results in the near-infrared bands at H and K,
517: and had somewhat different conclusions
518: than some of the other workers because they were able to fit their visibility and closure phase
519: data best with a binary model with two sources separated by 3.13 $\pm$ 0.12 mas, position
520: angle of 36 $\pm$ 10 degrees, and brightness ratio of 0.42 $\pm$ 0.06. Their gaussian
521: fits at 2.2 $\mu$m gave a FWHM of 2.56 $\pm$ 0.24 mas for the same period of time as the AMBER observations.
522: A striking feature of those results is that the size of the emitting region
523: decreased 10-20\% between about days 4 and 65 post-outburst. For example, the size
524: at 2.2 $\mu$m actually decreased from about 2.6 to 2.0 mas (FWHM), while at 1.65 $\mu$m, the
525: size decreased from 3.3 to 2.9 mas. However the 2 $\mu$m
526: continuum sizes are in approximate agreement. Monnier et al. (2006) rule out an expanding fireball
527: model, however, they would have over-resolved the fireball anyway since it would be about 8.8 mas in diameter, at a distance of 1600 pc, or a substantially larger angular size if the distance
528: were smaller. Lane et al. (2007) observed RS Oph also with IOTA, PTI, and the Keck Interferometer
529: at H band over a longer time period, up to 120 days from the V band maximum. They observed
530: an increase in diameter from approximately 3 to 4 mas from days 0 to 20 and a decrease to
531: less than 2 mas in diameter around day 120. They interpreted the near-infrared size data in terms
532: of a simplifed model of free-free emission in the postnova wind as the
533: mass ejected in the wind decreased during that time period.
534:
535: The emitting region at 10 $\mu$m is somewhat
536: larger than that seen at 2 $\mu$m, for example, our data are fitted by a gaussian 4.0 $\pm$ 0.4 mas
537: (FWHM), while the IOTA data had a size of 2.6 $\pm$ 0.2 mas. By 4 days post-outburst, a spherical
538: shock wave would be expected to have a radius of about 3.9 mas assuming a speed of approximately 1730 km/s (O'Brien et al. 2006)
539: and a distance of 1.4 kpc. Thus, according to this simple model, all the interferometric
540: IR continuum emission should be coming from within the post-shock region. The AMBER/VLTI
541: results indicate a more complex picture of the velocity field of the expanding material, with
542: two indicated - a ``slow'' field between -1800 and 1800 km/s, and a ``fast'' one between
543: -3000/-1800 km/s and 1800/3000 km/s. The position angle of the emitting material for the two
544: velocity groups differ, with the ``fast'' component being well defined in the East-West (position
545: angle 90 or 270 degrees $\pm$ 5 degrees) direction
546: and the ``slow'' component with position angles from 55 to 110 degrees modulo 180 degrees.
547:
548: Recent observations by Bode et al. (2007) using the HST confirm the elliptical shape seen with
549: AMBER, with an axial ratio of 0.5-0.6. The HST observations also confirm the velocity structure
550: seen by AMBER, i.e., the ``fast'' velocity field in the East-West direction, of the order of 3000 km/s,
551: and the ``slow'' velocity about half that of the ``fast'' one. The HST results showed that the expansion
552: rate in the plane of orbit of the stars decreased, from about 0.62 mas/day on day 13.8 to 0.48 mas/day
553: on day 155. This is a reduction in velocity from about 1700 km/s to about 1300 km/s.
554:
555: The spiral shock model is clearly relevant to the interpretation of the IOTA data presented in Monnier
556: et al. (2006). As noted above, the increased density of gas and dust in the arms of the spiral
557: would certainly provide an impediment to the free expansion of the fireball, as well as provide
558: a reservoir of hot material that would emit strongly at 2 $\mu$m. Their data indicate a closure
559: phase that is consistent with zero or 180 degrees for the first epoch (days 4-11) and convincingly
560: non-zero only for hour angles from -1.5 to -1 hr on the second epoch (days 14-29), and
561: at +1/2 hr on the third epoch (days 49-65). Hence, the closure phase signals and the binary
562: interpretation may be more consistent with hot clumpy material that is cooling, some of which may
563: (re-) condense into dust
564: as the white dwarf's outburst luminosity declines from its most luminous state, i.e., as shown in
565: Fig. \ref{dust-sub_01}. Note that the sublimation radius decreases from $\sim$ 5 mas to about
566: 2 mas from day 70 to day 120. Furthermore, as the luminosity
567: changes, the star tracker on IOTA may be providing a different optical center to the fringe
568: detection system, since in the first few days during the outburst the optical emission is dominated
569: by that centered on the white dwarf star itself, whereas after it has cooled, the optical emission is dominated by
570: the red giant star. Thus the effects of changes in the optical tracking and
571: interferometric phase center will need to be included in the analysis
572: to properly interpret the data. The offset between the optical tracking center and the
573: fringe phase center can cause a miscalibration of the visibility.
574: Since the 2 $\mu$m emitting region is actually
575: decreasing in size with time, as seen in their Fig. 1 and Table 2, it seems more likely they are observing
576: the cooling of this hot material near the two stars than actually resolving the binary, however,
577: the effects of the material in the shadow of the red giant star must be included in the interpretation
578: of the near-infrared and mid-infrared data.
579:
580: %emission lines
581:
582: \section{Summary and Conclusions}
583: We analyzed data from the recurrent nova RS Oph for the epoch
584: at $\sim$ 4 days post-outburst using the new KIN instrument.
585: These data allowed us to determine the size of the emitting region around the RS Oph
586: at wavelengths from 8-12 $\mu$m.
587: By fitting the unique KIN inner and outer spatial regime data, we obtained an angular size of the mid-infrared continuum of 6.2, 4.0, or 5.4 mas for a disk profile, gaussian profile (FWHM), and shell profile respectively.
588: The data show evidence of enhanced neutral atomic hydrogen emission located in the inner spatial regime relative to the outer regime. There is also evidence of a 9.7 $\mu$m
589: silicate feature seen outside of this region, which is consistent with dust that had condensed
590: prior to the outburst, and which has not yet been disturbed by the blast wave from the nova.
591: Our analysis of the observations, including the new ones presented in this paper, are most
592: consistent with a new physical model of RS Oph, in which spiral shock waves associated with
593: the motion of the two stars through the cool wind from the red giant create density enhancements
594: within the plane of their orbital motion.
595:
596: Further observations are needed to clarify this new picture of the RS Oph system. One issue that
597: has not been fully resolved is whether or not the red giant star really overflows its Roche lobe.
598: If so a hot spot would be expected where the material streaming from the red giant envelope
599: hits the accretion disk, and UV or X-ray observations could search for this effect.
600: Another approach would be to observe RS Oph over several orbital cycles using infrared photometry
601: to look for variations of the light curve due to the departure of the red giant star from spherical
602: symmetry. High resolution
603: spectra could also help. Confirmation of the rotational velocity of the red giant
604: star measured by Zamanov et al. (2007) would be worthwhile, and could provide another
605: estimate of the absolute size of the red giant star, assuming that it is co-rotating with the
606: orbit. Further KIN observations within the next few years would also be helpful as they could
607: show evidence of the re-establishment of the spiral shock wave, and perhaps some information
608: about the shape of the circumstellar material and dust formation. Another epoch of HST
609: observations would also determine the deceleration of the outflow in the two directions, i.e.,
610: within the plane of the orbit of the two stars and along the poles.
611:
612: Theoretical studies of the motion of the blast wave in an environment with a high density
613: region in the plane of the orbit of the two stars are also worthwhile. In particular, it would be
614: important to understand how the blast wave is diffracted around the red giant star and
615: how it propagates in an medium with the periodic density enhancements in the plane due to
616: the spiral pattern.
617:
618: The recurrent nova RS Ophiuchi is a rich system for the study of circumstellar matter under
619: extreme physical conditions. Continued study will provide important insights into Type Ia
620: supernovae, of which RS Oph may be a progenitor.
621:
622: \acknowledgments
623:
624: We are grateful to the National Aeronautics and Space Administration, Jet Propulsion Laboratory, the California Association for Research in Astronomy, the Harvard-Smithsonian Center for Astrophysics
625: (including SAO grant G06-7022A to JLS), and to the National Aeronautics and Space Administration, Goddard Space Flight Center for support of this research. The
626: data presented herein were obtained at the
627: W.M. Keck Observatory, which is operated
628: as a scientific partnership among the California
629: Institute of Technology, the University
630: of California and the National Aeronautics
631: and Space Administration. The Observatory
632: was made possible by the generous financial
633: support of the W.M. Keck Foundation.
634: JPW acknowledges support provided by an NPP Fellowship (NNH06CC03B) at NASA Goddard
635: Space Flight Center. This work has made use of services produced by the Michelson Science
636: Center at the California Institute of Technology. One of the authors (RKB) would also like to acknowledge Eugene E. Rudd formerly of the United States Naval Research Laboratory for his continued support and encouragement. The authors thank the referee, Dr. Michael Bode, for his careful and thorough review of the manuscript, which has helped us to significantly improve it.
637:
638: Facilities: \facility{Spitzer}, \facility{IOTA}, \facility{Keck:I}, \facility{Keck:II}.
639:
640: \appendix
641: \section{Luminosity Evolution}
642: In this appendix we discuss the mathematical formulation used to derive the evolution of
643: the luminosity of the white dwarf star starting at the maximum magnitude in V band. Our
644: calculations are based on those presented in the paper by Hachisu and Kato (2001), who
645: expanded on their discussion of their light-curve model presented in Hachisu and Kato (2000).
646:
647: Our purpose is to evaluate the light curve not only at V band, but also at H, K, and N bands,
648: which have been used by the AMBER, IOTA, Keck, and PTI interferometers to observe RS Oph
649: after the 2006 outburst. Our discussion is in fact relevant to any modern instrument that has
650: wavefront sensing and control, and also tracking, at wavelengths different than that of the
651: observations of interest. For example, the IOTA interferometer performs its precision pointing at
652: V band, but observations are made at H band. Similarly the Keck Interferometer does
653: precision pointing at V band, wavefront control of the two large apertures at H band, and
654: the data are taken at N band. This difference in wavelengths is important because as
655: the V band luminosity of the white dwarf star decreases, the observed light from the binary
656: is mainly from the red giant star, which gives a shift in the center of light. This shift can affect
657: the calibration of these instruments as there is an offset, which will change over time as compared
658: to a calibrator, which has all the same optical center at all times for all of these wavebands.
659:
660: The evolution of the stellar parameters for the white dwarf are given in Table \ref{evo} of Hachisu
661: \& Kato (2001).
662: We assume the parameters for the red giant star remain constant with $R_{rg} = 40 R_{\odot}$,
663: and $T_{rg}$ = 3400 K.
664:
665: The blackbody flux density (erg cm$^{-2}$ s$^{-1}$ Hz$^{-1}$) at a frequency, $\nu$, for a star of temperature, T$_{\ast}$, radius,
666: $R_{\ast}$, and distance, $D$ is:
667: \begin{equation}
668: F_{\ast}(\nu) = 2 \pi (R_{\ast} / D)^2 (h { \nu}^3 / c^2) / [exp(h \nu / k T_{\ast}) - 1] ,
669: \end{equation}
670: where $h$, $k$, and $c$, are Planck's constant, the Boltzmann constant, and the speed of
671: light, respectively.
672:
673: The total luminosity (erg s$^{-1}$) is:
674: \begin{equation}
675: L_{\ast} = 4 \pi \sigma {R_{\ast}}^2 {T_{\ast}}^4 ,
676: \end{equation}
677: where $\sigma$ is Stefan's constant.
678:
679: \subsection{Sublimation Radius}
680: The temperature of a black grain in radiative equilibrium at a distance, $r$,
681: from a star with luminosity, $L_{\ast}$ is:
682: %of temperature, $T_{\ast}$, and radius, $R_{\ast}$, is:
683: \begin{equation}
684: T_{grain} = \left(\frac{L_{\ast}}{16 \pi \sigma r^2}\right)^{1/4} =
685: \frac{T_{\ast}}{\sqrt{2}} ~ \left( \frac{R_{\ast}}{r}\right)^{1/2}
686: \end{equation}
687: A simple rearrangement of the above equation yields a formula for the sublimation radius, $R_{sub}$,
688: assuming a sublimation temperature, $T_{sub}$:
689: \begin{equation}
690: R_{sub} = \left(\frac{L_{\ast}}{16 \pi \sigma T_{sub}^4}\right)^{1/2}
691: = \frac{R_{\ast}}{2} ~ \left(\frac{T_{\ast}}{T_{sub}}\right)^2 .
692: \end{equation}
693: For this paper we assume a sublimation temperature of 1500 K.
694:
695: The evolution of the sublimation radius for the red giant star (dashed lines) and nova (solid lines) beginning at the maximum in V band is
696: displayed in Fig. \ref{dust-sub_02}. Note that the sublimation radius remains approximately constant at about 5 mas for the first 70 days, after which it gradually reduces to $<$ 0.2 mas after about day 120.
697:
698: \subsection{Illumination of the Red Giant Star}
699: During the high luminosity phase of the outburst, the surface of the red giant star facing the nova
700: is heated substantially. We calculate the effect of the nova luminosity on the red giant star using
701: Eqn. (10) of Hachisu \& Kato (2001), which is based on thermal equilibrium between the
702: faces of the two stars, which we display here:
703: \begin{equation}
704: \sigma {T_{rgi}}^4 = \eta_{rg} \cos ( \theta ) L_{wd} / (4 \pi r^2) + \sigma {T_{rg}}^4
705: \end{equation}
706: where $T_{rgi}$ is the new effective temperature of the red giant star for facing side and where
707: we include an extra term from the white dwarf luminosity, $L_{wd}$. There are two additional
708: constants included in this new term, the first is $\eta_{rg}$, which is an effective emissivity for
709: the stellar surface, while the second, $\cos(\theta)$ is the average inclination angle of the surface.
710: The term, r, is the distance between the two stars, and $T_{rg}$ is the original temperature of
711: the red giant star. Equation A5 can be rearranged into a simple form after substituting for
712: $L_{wd}$:
713: \begin{equation}
714: {T_{rgi}}^4 = \left[ \eta_{rg} \cos ( \theta ) \left( \frac{R_{wd}}{r} \right)^2 \right] {T_{wd}}^4 + {T_{rg}}^4
715: \end{equation}
716: %
717: Equation A5 was evaluated as a function of time, using the evolution of the
718: white dwarf luminosity derived from Eq. A2 and the values from Table 5.
719: The result of our computation of this effect is displayed in Fig. \ref{RG-face-temp}, in which we use $ r = 325 R_{\odot}$,
720: $\cos(\theta) = 0.5$, $\eta_{rg} = 0.5$, and $T_{rg} = 3400$ K.
721:
722: \subsection{Accretion Luminosity}
723:
724: The computation of the evolution of the accretion luminosity is based on the treatment of
725: Hachisu \& Kato (2001) with the simplification of a flat accretion disk, instead of adding
726: the complexity of a warped accretion disk that is used in their paper.
727: These calculations are based on the well known \citet{Lynden74} and \citet{Pringle81} equations for the luminosity of an accretion disk, where the accretion luminosity is based on a numerical integration
728: of the flux from the accretion disk, assuming a temperature distribution across the disk.
729: Normally the temperature of the accretion disk is determined solely by the accretion rate onto
730: the disk for normal viscous heating of the disk. To this Hachisu \& Kato (2001) added a
731: term based on additional heating of the disk from the white dwarf star as it is in a high
732: luminosity state and is contracting and heating during the constant high luminosity period
733: after the outburst. The flux density of the disk at frequency $\nu$ is given by:
734: %
735: %Fdisk2[?_, Ms_, Mdot_, Rs_, Lwd_, etaDK_, ?2_, inc_, Rin_, Rout_, D1_]:=
736: %
737: % 2 Pi Cos[inc] NIntegrate[B?[?,Tdd2[Ms,Mdot,R,Rs, etaDK, ?2, Lwd]]*R,{R,Rin,Rout}] * D1^-2
738: %
739: \begin{equation}
740: F_{disk}\left(\nu\right) = \frac{2 \pi \cos(\theta_{inc})}{D^2} \int_{R_{in}}^{R_{out}} \rho B_{\nu}[T_{disk}(\rho)] d\rho
741: \end{equation}
742: %
743: The conventional expression for the radial distribution of the temperature of the disk is:
744: \begin{equation}
745: T_{disk1}\left(\rho\right) = \left(\frac{3GM_*\dot{M}}{8\pi\sigma \rho^3} \left[1-\left(\frac{R_*}{\rho}\right)^{1/2}\right]\right)^{1/4}
746: \end{equation}
747: %
748: where $T_{disk1}$ is the temperature in the accretion disk \citep{Pringle81} and D is the distance to the Earth. The quantity B$_{\nu}$ is the Planck function, M$_*$ is the mass of the white dwarf, R$_*$ is its radius, $\dot{M}$ is the accretion rate onto the disk, and $\sigma$ and G are the Stefan-Boltzmann constant and the gravitational constant, respectively. The maximum temperature in the accretion disk occurs at a radius R$_{max}$ = 1.36 R$_*$, and is given by
749: %
750: \begin{equation}
751: T_{max} = 0.488 \left(\frac{3GM_*\dot{M}}{8\pi \sigma R^3_*}\right)^{1/4}
752: \end{equation}
753: as discussed in \citet{Lynden74}, \citet{Pringle81}, and \citet{Hartmann98}. In this treatment we neglect the emission from the boundary layer, which occurs over a very small angular region around the surface of the white dwarf star and emits at a very high temperature, which would have a negligible contribution to the luminosity in the infrared region near 10 $\mu$m.
754: %
755: The evolution of the luminosity of the accretion disk will be computed using the formulation of
756: Hachisu and Kato (2001), as set up below.
757: %
758: % Tdd[Ms_, Mdot_, R_, Rs_] := ((3 G Ms Mdot / (8 Pi R^3 ?p )) *(1 - (Rs/R)^(1/2)))^(1/4)
759:
760: % B?[?_, T_] := (2 * h ?^3 / c^2) / (Exp[h ? / (k T)] - 1)
761:
762: % Tmax[Ms_, Rs_, Mdot_] := 0.488 (3 G Ms Mdot / (8 Pi Rs^3 ?p ))^(1/4)
763:
764: Let us now formulate the evolution of the accretion disk luminosity by adjusting some of the
765: parameters in the equation above as a function of time. In particular the outer radius of the
766: accretion disk $R_{disk}$ is parametrized by a power law decrease from day 7 until day 79
767: using the following formulae:
768: %
769: \begin{equation}
770: R_{disk} = \alpha {R_1}^{\ast}
771: \end{equation}
772:
773: \begin{equation}
774: \alpha = \alpha_0 { \left( \frac{\alpha_1}{\alpha_0} \right) }^{(t-t_0)/72}
775: \end{equation}
776: %
777: In these equations ${R_1}^{\ast}$ is the inner critical radius of the Roche lobe for the white
778: dwarf (nova) component, which we assume is 138.6 $R_{\odot}$
779: %
780: % Rdisk = ? a (aka R1star)
781: % ? = aa)^(t-t0)/72 where t0 < t < t0 +72 , where t0=7 days
782: %
783: The parameter $\alpha$ helps define the size of the accretion disk which varies between $\alpha_0$ = 0.1 and $\alpha_1$ = 0.008, and so the power law form of Eqn. A10 is to make an interpolation function.
784: Haschisu and Kato include irradiation of the accretion disk as the WD luminosity changes between days 7 and 79, where the accretion disk temperature is given by:
785: %
786: \begin{equation}
787: \sigma {\left[ T_{disk}(\rho)\right]}^4 = \eta_{disk} \cos(\theta_{inc}) \left( \frac{L_{wd}}{4 \pi \rho^2} \right) +
788: %\frac{3 G M_{wd} \dot{M}}{8 \pi \rho^3}
789: \left(\frac{3GM_*\dot{M}}{8\pi \rho^3} \left[1-\left(\frac{R_*}{\rho}\right)^{1/2}\right]\right)
790: \end{equation}
791: %
792: % ? Tdisk^4 = etaDK * (Lwd / (4 Pi r^2)) Cos[?] + 3 G Mwd Mdotacc /(8 Pi r^3)
793: %
794: The second term of this equation is the traditional accretion luminosity term, while the first term includes the luminosity of the white dwarf. The parameter $\eta_{disk}$ is an efficiency factor, and is assumed to be 0.5, and $\cos(\theta_{inc})$ is 0.1, based on the average inclination of the surface. They assume the outer radius of the accretion disk is at a temperature of 2000 K, and is not affected by radiation from the WD photosphere.
795: For the remainder of the calculation we will compute $\alpha$ in discrete time intervals corresponding to what we have done before over the days that this effect matters.
796:
797: The integral is numerically evaluated at the center wavelengths for V (0.55 $\mu$m), H (1.65 $\mu$m), and N (10.5 $\mu$m) bands,
798: using an inner radius given by the radius of the white dwarf star and the outer radius given by the
799: equation for $R_{disk}$, for the time steps of Table 5 and the white dwarf parameters in that
800: table. The results are plotted in Fig. 10. In this figure the quantities plotted include the the
801: contributions of the red giant and white dwarf alone (stars and solid squares), the red giant
802: including effects of irradiation by the white dwarf (diamonds), the accretion disk alone (solid circles and
803: dash-dot-dot line), the irradiated accretion disk (triangles and dashed line), and the total for the
804: white dwarf including the irradiated accretion disk (diamonds and solid line).
805:
806:
807: %:references
808:
809:
810: \begin{thebibliography}{}
811:
812: \bibitem[Adams \& Joy(1933)]{ada33}Adams, W. S., Joy, A. H., 1933, \pasp, 45, 249a
813:
814: \bibitem[Barbon, Mammano \& Rosino(1969)]{bar69}Barbon, R., Mammano, A., Rosino, L., 1969, Comm. Konkoly Obs., 65, 257
815:
816: \bibitem[Barry et al.(2008a)]{bar08} Barry, R.K., Mukai, K., Sokoloski, J. L., Danchi, W. C., Hachisu, I., Evans, A., Gehrz, R., \& Mikolajewska, J., 2008, in RS Ophiuchi (2006) and the recurrent nova phenomenon, eds A. Evans, M. F. Bode, T. J. O'Brien, Astronomical Society of the Pacific Conference Series, in press
817:
818: \bibitem[Barry et al.(2008b)]{bar08b} Barry, R. K., Skopal, A., \& Danchi, W. C., 2008, \apj, manuscript in prep.
819:
820: \bibitem[Blair et al.,(1983)]{bla83}Blair, W. P., Stencel, R. E., Feibelman, W. A., \& Michalitsianos, A. G., 1983, \apjs, 53, 573B
821:
822: \bibitem[Bode \& Kahn(1985)]{bod85} Bode, M. F., \& Kahn, F. D., 1985, \mnras, 217, 205
823:
824: \bibitem[Bode et al.(2006)]{bod06}Bode, M. F., O'Brien, T. J., Osborne, J. P., Page, K. L., Senziani, F., Skinner, G. K., Starrfield, S., Ness, J. -U., Drake, J. J., Schwarz, G., Beardmore, A. P., Darnley, M. J., Eyres, S. P. S., Evans, A., Gehrels, N., Goad, M. R., Jean, P., Krautter, J., \& Novara, G., 2006, \apj, 652, 629
825:
826: \bibitem[Bode et al.(2007)]{bod07}Bode, M. F., Harman, D. J., O'Brien, T. J., Bond, H. E., Starrfield, S., Darnley M. J., Evans, A., \& Eyres, S. P. S., 2007, \apj, 665, L63
827:
828: \bibitem[Brandi et al.(2008)]{bra08}Brandi, E., Quiroga, C., Ferrer, O. E., Miko\l ajewska, J., \& Garcia, L. G., 2008, in RS Ophiuchi (2006) and the recurrent nova phenomenon, eds A. Evans, M. F. Bode, T. J. O'Brien, Astronomical Society of the Pacific Conference Series, in press
829:
830: \bibitem[Cardelli, Clayton \& Mathis(1989)]{car89}Cardelli, J. A., Clayton, G. C., \& Mathis, J. S., 1989, \apj, 345, 245
831:
832: \bibitem[Cassatella et al.(1985)]{cas85}Cassatella, A., Harris, A. Snijders, M. A. J., \& Hassall, B. J. M., 1985, Proc. ESA Workshop: Recent Results on Cataclysmic Variables, ESA SP-236
833:
834: \bibitem[Chesneau et al.(2007)]{che07}Chesneau, O., Nardetto, N., Millour, F., Hummel, C., Domiciano de Souza, A., Bonneau, D., Vannier, M., Rantakyro, F., Spang, A., Malbet, F., Mourard, D., Bode, M. F., O\'Brien, T. J., Skinner, G., Petrov, R. G., Stee, P., Tatulli, E., \& Vakili, F., 2007, \aap, 464, 119
835:
836: \bibitem[Colavita et al.(2006)]{col06}Colavita, M.M. Serabyn, G. Wizinowich, P.L. \& Akeson, R.L. 2006, in Advances in Stellar Interferometry, eds. J.D. Monnier, M. Scholler and W.C. Danchi, Proc. SPIE 6268, 626803
837:
838: \bibitem[Cowie \& Songaila(1986)]{cow86}Cowie, L. L., \& Songaila, A., 1986, \araa, 24, 499
839:
840: \bibitem[Creech-Eakman et al.(2003)]{cre03}Creech-Eakman, M.J., Moore, J.D., Palmer, D.L., \& Serabyn, E., 2003, \procspie, 4841, 330c
841:
842: \bibitem[Danchi \& Barry(2007)]{dan07} Danchi, W.C., \& Barry, R. K., 2007, \apj, manuscript in
843: preparation
844:
845: \bibitem[Das et al.(2006)]{das06} Das, R., Banerjee, D.P.K., \& Ashok, N.M. 2006, \apj, 653, 141
846:
847: \bibitem[Dobrzycka \& Kenyon(1994)]{dob94}Dobrzycka, D., \& Kenyon, S. J., 1994, \apj, 108, 2259
848:
849: \bibitem[Downes \& Duerbeck(2000)]{dow00}Downes, R. A., \& Duerbeck, H. W., 2000, \apj, 120, 2007
850:
851: \bibitem[Evans et al.(2007a)]{eva07a}Evans, A., Kerr, T., Yang, B., Matsuoka, Y., Tsuzuki, Y., Bode, M. F., Eyres, S. P. S., Geballe, T. R., Woodward, C. E., Gehrz, R. D., Lynch, D. K., Rudy, R. J., Russell, R. W., O'Brien, T. J., Starrfield, S. G., Davis, R. J., Ness, J.-U., Drake, J., Osborne, J. P., Page, K. L., Adamson, A., Schwarz, G., \& Krautter, J. 2007, \mnras, 374, 1
852:
853: \bibitem[Evans et al.(2007b)]{eva07b} Evans, A., Woodward, C.E., Helton, A., Gehrz, R. D., Lynch, D. K., Rudy, R. J., Russell, R. W., Bode, M. F., Kerr, T., Yang, B., Matsuoka, Y., Tsuzuki, Y., Eyres, S. P. S., Geballe, T. R., O'Brien, T. J., Davis, R. J., Starrfield, S. G., Ness, J.-U., Drake, J., Osborne, J. P., Page, K. L., Schwarz, G., \& Krautter, J., 2007, \apj, 663L, 29E
854:
855: \bibitem[Evans et al.(2007c)]{eva07c}Evans, A., Woodward, C., Helton, A., van Loon, J. Th., Barry, R. K., Bode, M. F., et al., 2007, \apj, 671, L157
856:
857: \bibitem[Eyres et al.(1998)]{eyr98}Eyres, S. P. S., Evans, A., Salama, A, Barr, P., Clavel, J., Jenkins, N., Leech, K., Kessler, M., Lim, T., Metcalfe, L, \& Schulz, B., 1998, \apss, 255, 361
858:
859: \bibitem[Fekel et al.(2000)]{fek00}Fekel, F. C., Joyce, R. R., Hinkle, K. H., \& Skrutskie, M. F., 2000, \apj, 119, 1375
860:
861: \bibitem[Garcia(1986)]{gar86}Garcia, M.R., 1986, \aj, 91,1400
862:
863: \bibitem[Grevesse(1984)]{gre84}Grevesse, N., 1984, Phys. Scr., T8, 49
864:
865: \bibitem[Fleming(1904)]{fle04}Fleming, W., 1904, Harvard College Observatory Circular, 76
866:
867: \bibitem[Hachisu \& Kato(2000)]{hac00}Hachisu, I., \& Kato, M., 2000, \apj, 536, 93
868:
869: \bibitem[Hachisu \& Kato(2001)]{hac01}Hachisu, I., \& Kato, M., 2001, \apj, 558, 323
870:
871: \bibitem[Hachisu et al.(2006)]{hac06}Hachisu, I., Kato, M., Kiyota, S., Kubotera, K., Maehara, H., et al., 2006, \apjl, 651, L141
872:
873: \bibitem[Hartmann (1998)]{Hartmann98} Hartmann, L., 1998, in "Accretion Processes in Star Formation", [Cambridge UK, Cambridge University Press]
874:
875: \bibitem[Hjellming et al.(1986)]{hje86}Hjellming, R. M., van Gorkom, J. H., Taylor, A. R., Seaquist, E. R., Padin, S., Davis, R. J., \& Bode, M. F., 1986, \apj, 305, 71
876:
877: \bibitem[Kenyon \& Fernandez-Castro(1987)]{ken87}Kenyon, S. J., \& Fernandez-Castro, T., 1987, \apj, 93, 938
878:
879: \bibitem[Kenyon \& Gallagher(1983)]{ken83}Kenyon, S. J., \& Gallagher, J. S., 1983, \aj, 88, 666K
880:
881: \bibitem[Koresko et al.(2006)]{kor06}Koresko, C., Colavita, M., Serabyn, E., Booth, A., \& Garcia, J., 2006, \procspie, 6268, 626816-1
882:
883: \bibitem[Lane et al.(2007)]{lan07}Lane, B. F., Sokoloski, J., Barry, R. K., Traub, W. A., Retter, A., et al., 2007, \apj, 658, 520
884:
885: \bibitem[Lynch et al.(2006)]{lyn06}Lynch, D. K., Woodward, C. E., Geballe, T. R., Russell, R. W., Rudy, R. J., Venturini, C. C., Schwarz, G. J., Gehrz, R. D., Smith, N., Lyke, J. E., Bus, S. J., Sitko, M. L., Harrison, T. E., Fisher, S., Eyres, S. P., Evans, A., Shore, S. N., Starrfield, S., Bode, M. F., Greenhouse, M. A., Hauschildt, P. H., Truran, J. W., Williams, R. E., Perry, R. B., Zamanov, R., \& O'Brien, T. J., 2006, \apj, 638, 987
886:
887: \bibitem[Lynden-Bell \& Pringle(1974)]{Lynden74} Lynden-Bell, D. \& Pringle, 1974, MNRAS, 168, 603
888:
889: \bibitem[Mastrodemos \& Morris(1999)]{mas99}Mastrodemos, N., \& Morris, M., 1999, \apj,
890: 523, 357
891:
892: \bibitem[Mauron \& Huggins(2006)]{mau06}Mauron, N., \& Huggins, P.J., 2006, \aap, 452, 257
893:
894: \bibitem[Monnier et al.(2006)]{mon06}Monnier, J. D., Barry, R. K., Traub, W. A., Lane, B. F., Akeson, R. L., Ragland, S., Schuller, P. A., Berger, J. P., Millan-Gabet, R., Pedretti, E., Schloerb, F. P., Koresko, C., Carleton, N. P., Lacasse, M. G., Kern, P., Malbet, F., Perraut, K., \& Muterspaugh, M. W., 2006, \apjl, 647, 127
895:
896: \bibitem[Morris et al.(2006)]{mor06}Morris, M., Sahai, R., Matthews, K., Cheng, J., Lu, J.,
897: Claussen, M., \& Sanchez-Contreras, C., 2006, in Planetary Nebulae in our Galaxy and Beyond,
898: Proceedings of IAU Symposium No. 234, M. J. Barlow \& R. H. Mendez, eds., 469.
899:
900: \bibitem[Morrison(1985)]{mor85}Morrison, W., 1985, \iaucirc, 4030
901:
902: \bibitem[M\"{u}rset \& Schmid(1999)]{mur99} M\"{u}rset, U., \& Schmid, H.M., 1999, \aaps, 137, 473
903:
904: \bibitem[Narumi(2006)]{nar06}Narumi, H., Hirosawa, K., Kanai, K., Renz, W., Pereira, A.,Nakano, S., Nakamura, Y., \& Pojmanski, G. 2006, \iaucirc, 8671, 2
905:
906: \bibitem[O'Brien et al.(2006)]{obr06}O'Brien, T. J., Bode, M. F., Porcas, R. W., Muxlow, W. B., Eyres, S. P. S., Beswick, R. J., Garrington, S. T., Davis, R. J., \& Evans, A., 2006, \nat, 442, 279
907:
908: \bibitem[Oppenheimer \& Mattei(1993)]{opp93}Oppenheimer, B.D., \& Mattei, J.A., 1993, JAVSO, 22, 1050
909:
910: \bibitem[Payne-Gaposchkin(1957)]{pay57}Payne-Gaposchkin, C., 1957, The Galactic Novae, North-Holland, Amsterdam
911:
912: %\bibitem[Payne-Gaposchkin(1977)]{pay77}Payne-Gaposchkin, C., 1977, \aj, 82, 665
913:
914: \bibitem[Pottasch(1967)]{pot67}Pottasch, S. R., 1967, \bain, 19, 227 %
915:
916: \bibitem[Pringle (1981)]{Pringle81} Pringle, J. E. 1981, ARA\&A, 19, 137
917:
918: \bibitem[Rupin et al.(2008)]{rup08}Rupin, M. P., Mioduszweski, A. J., Sokoloski, J. L., Kaiser, C. R., \& Brocksopp, C., 2006, Proc. AAS
919:
920: \bibitem[Sacuto et al.(2007)]{sac07}Sacuto, s., Chesneau, O., Vannier, M., \& Cruzalebes, P., 2007, \aaps, 465, 469
921:
922: \bibitem[Schaeffer(2004)]{sch04}Schaeffer B., 2004, IAUC 8396
923:
924: \bibitem[Schild et al.(2001)]{sch01}Schild, H., Eyres, S. P. S., Salama, A., \& Evans, A., 2001, \aaps, 378, 146
925:
926: \bibitem[Serabyn et al.(2000)]{ser00}Serabyn, E., Colavita, M.M. \& Beichman, C.A. 2000, in Thermal Emission
927: Spectroscopy and Analysis of Dust, Disks, and Regoliths, ASP Conf. Ser. Vol. 196, eds. M.L. Sitko, A.L. Sprague \& D.K. Lynch, p. 357.
928:
929: \bibitem[Serabyn et al.(2004)]{ser04}Serabyn, E. et al. 2004, in SPIE Vol. 5491, New Frontiers in Stellar
930: Interferometry, ed. W.A. Traub, 136.
931:
932: \bibitem[Serabyn et al.(2005)]{ser05}Serabyn, E. et al. 2005, in SPIE Vol. 5905, Techniques and Instrumentation for Detection of Exoplanets II, ed. D.R. Coulter, 5905OT-1.
933:
934: \bibitem[Serabyn et al.(2006)]{ser06}Serabyn, E. et al. 2006, Proc. SPIE 6268, Advances in Stellar Interferometry, eds. J.D. Monnier, M. Scholler and W.C. Danchi, 626815.
935:
936: %\bibitem[Serabyn et al.(2008)]{ser08}Serabyn, E., Kuchner, M., Menneson, B., \& Traub, W., et al., 2007, \apj, manuscript in preparation
937: \bibitem[Serabyn et al.(2008)]{ser08}Serabyn, E., et al., 2008, \apj, manuscript in preparation
938:
939: \bibitem[Snijders (1985)]{sni85}Snijders, M. A. J., 1985, \apss, 130, 244
940:
941: \bibitem[Sokoloski et al.(2006)]{sok06}Sokoloski, J. L., Luna, G. J. M., Mukai, K., \& Kenyon, S. J., 2006, \nat, 442, 276
942:
943: \bibitem[Starrfield, Sparks \& Truran(1985)]{sta85}Starrfield, S., Sparks, W. M., \& Truran, J. W., 1985, \apj, 291, 136
944:
945: \bibitem[Wallerstein(1958)]{wal58}Walerstein, G., 1958, \pasp, 70, 537w
946:
947: \bibitem[Warner (1995)]{war95} Warner, B., 1995, Cataclysmic Variable Stars, [Cambridge, UK: Cambridge University Press], p. 27.
948:
949: \end{thebibliography}
950:
951:
952: %:table 1 - observing log: label "observations"
953:
954: \clearpage
955:
956: \begin{deluxetable}{ccrrrr}
957: %\tabletypesize{\scriptsize}
958: %\rotate
959: \tablewidth{0pt}
960: \tablecaption{Observing Log for RS Ophiuchi.\label{observations}}
961: \tablehead{
962: \colhead{Object} & \colhead{Type}& \colhead{Time}& \colhead{U}& \colhead{V}& \colhead{Airmass} \\
963: \colhead{} & \colhead{}& \colhead{(UT)}& \colhead{(m)}& \colhead{(m)}& \colhead{}
964: }
965: \startdata
966: Chi UMa & cal & $15\colon 07\colon 07$ & 23.85 & 80.61 & 1.38\\
967: Chi UMa & cal & $15\colon 15\colon 39$ & 21.92 & 81.25 & 1.41\\
968: RS Oph & trg & $15\colon 50\colon 15$ & 54.57 & 64.75 & 1.46\\
969: RS Oph & trg & $16\colon 03\colon 46$ & 55.35 & 64.37 & 1.39\\
970: RS Oph & trg & $16\colon 12\colon 35$ & 55.75 & 64.12 & 1.35\\
971: Rho Boo & cal & $16\colon 34\colon 24$ & 39.63 & 75.14 & 1.08\\
972: \enddata
973: \end{deluxetable}
974: %\clearpage
975:
976:
977: %:table 2 - continuum fitting: label "models"
978:
979: \begin{deluxetable}{ccccc}
980: \tablewidth{0pt}
981: %\tabletypesize{\scriptsize}
982: %\rotate
983: \tablecaption{RS Ophiuchi model fitting results.\label{models}}
984: %\caption{These are the results of exhaustive $\chi^{2}$ fitting of both the inner and outer spatial continua to three models; uniform disk, Gaussian and shell. Monte Carlo analysis was conducted to derive the confidence intervals given. K-band sizes are given for comparison. Here we note that the size in N-band is significantly larger than that of K-band for all models explored. \label{models}}
985: %\tablewidth{0pt}
986: \tablehead{
987: \colhead{Source Model} & \colhead{Angular Size (mas)}& \colhead{Radiant Flux} & \colhead{Major Size (mas) }& \colhead{Minor size\tablenotemark{a} (mas)} \\
988: \colhead{} & \colhead{(N band)}& \colhead{(Jy)} & \colhead{(K band)}& \colhead{(K band)}
989: }
990: \startdata
991: Uniform Disk & $6.2\pm0.6$& $22.4\pm3.9$ & $4.9\pm0.4$ & $3.0\pm0.3$\\
992: Uniform Gaussian\tablenotemark{b} & $4.0\pm0.4$ & $22.4\pm3.8$ & $3.1\pm0.2$ &$1.9\pm0.3 $ \\
993: Uniform Shell\tablenotemark{c} & $5.4\pm0.6$ & $22.4\pm3.8$ & $3.7\pm0.3$ & $1.9\pm0.2$\\
994: % reference page 29 in rkb notebook 3.
995: \enddata
996: \tablenotetext{a}{Sizes for continuum values at 2.3 $\mu$m after \citet{che07}.}
997: \tablenotetext{b}{Full width at half maximum.}
998: \tablenotetext{c}{Spherical shell with thickness 1.0 mas - optically thin.}
999:
1000: \end{deluxetable}
1001:
1002:
1003: %reference page 28-31, rkb notebook 4 for line ids for Spitzer 4/16 and 4/26.
1004: %reference page 24-27 and 32, rkb notebook 4 for line ids for Spitzer 9/9.
1005: %: table 3 - spitzer line list: label "spitzlinelist"
1006: \begin{deluxetable}{cccccc}
1007: %\rotate
1008: \tablewidth{0pt}
1009: \tablecaption{Mid Infrared \textit{Spitzer} Line List: N-band.\label{spitzlinelist}}
1010: \tablehead{
1011: \colhead{Wavelength} & \colhead{ID} & \colhead{{\it Spitzer} 4/16} & \colhead{{\it Spitzer} 4/26} & \colhead{\it Spitzer 9/9} \\
1012: \colhead{($\mu$m)} & \colhead{} & \colhead{(Jy)} & \colhead{(Jy)} & \colhead{(Jy)}
1013: }
1014: \startdata
1015: %reporting three digits after decimal for both wavelength and flux
1016: 7.460 & $Pf\alpha - HI \colon 6-5$& detected \tablenotemark{a}& detected & detected\\
1017: 7.652 & $[NeVI]$ & detected& detected & detected\\
1018: 8.180 & $FeI$ & 0.009 & 0.007 &--\\
1019: 8.760 & $HI \colon 10-7$ & 0.042& 0.038 & 0.020 \tablenotemark{b} \\
1020: 8.985 & $[V II]$, Si I, \& Ca I]& -- & -- & 0.010\\
1021: 9.017 & $FeI$ & 0.015 &0.021&--\\
1022: 9.288 & $CaI$&0.004 &0.004&--\\
1023: 9.407 & $SiI$ & 0.014 & 0.013 & -- \\
1024: 9.529 & $CI$&0.009 &0.008&--\\
1025: 9.720 &$CII$& 0.018&0.017&--\\
1026: 9.852 &$CaI$ &0.007&0.008&--\\
1027: 10.285 &$MgI$ &0.007&0.006&--\\
1028: 10.492&$HI \colon 12-8$&--&--&0.013\\
1029: 10.517 &$ NeI$ &0.026&0.026&--\\
1030: 10.833 &$ CI]$ &0.018&0.016&--\\
1031: 11.284&$HI \colon 9-7$&--&--&0.007\\
1032: 11.318 &$ HeI$ &0.060&0.053&--\\
1033: 11.535 &$ HeII$ &0.016&0.013&--\\
1034: 12.168 &$HI \colon 36-11$&0.011&0.009&--\\
1035: 12.372 & $Hu\alpha - HI \colon 7-6$ & 0.158 & 0.150 & 0.037 \\
1036: 12.557 &$SiI$ &0.040&0.036&--\\
1037: 12.803 & $[NeII]$&--&--&0.159\\
1038: 12.824 &$HeI$ &0.009&0.015&--\\
1039: 13.128 &$HeII$& 0.009&0.011&--\\
1040: 13.188 &$HI \colon 18-10$ &0.009&0.010&0.005\\
1041:
1042: \enddata
1043:
1044: \tablenotetext{a}{Detected in \textit{Spitzer} spectrum but not fit due to intrusion of band edge.}
1045: \tablenotetext{b}{Blended with neutral hydrogen lines at 8.721 and 8.665 $\mu$m }
1046: \tablecomments{Some of these species were blended with others, and, due to the difficulty in deblending Spitzer low-resolution channel spectra, may in some cases be misidentified. This is not critical, however, to the conclusions of this paper.}
1047:
1048: \end{deluxetable}
1049: %: table 4 - KIN line list: label "kinfeaturelist"
1050: \clearpage
1051: \begin{deluxetable}{cccccc}
1052: %\rotate
1053: \tablewidth{0pt}
1054: \tablecolumns{4}
1055: \tablecaption{Continuum-normalized, Mid-infrared KIN Emission Source Identification.\label{kinfeaturelist}}
1056: \tablehead{
1057:
1058: %\colhead{} & \multicolumn{3}{c}{Non-shell Stars} & \colhead{} &\multicolumn{3}{c}{Shell Stars} \\
1059: %\colhead{} &\colhead{KIN Inner Spatial Regime}\\
1060: \colhead{Center Wavelength} & \colhead{Spectroscopic Width} & \colhead{Flux} & Attributed to\\
1061: \colhead{($\mu$m)} & \colhead{($\mu$m, FWZM)} & \colhead{(Jy)}
1062: }
1063: \startdata
1064: %reporting three digits after decimal for both wavelength and flux
1065: % see rkb logbook #4, page 21 and 23. See page 38 for attributed to
1066:
1067: \cutinhead{KIN Inner Spatial Regime}
1068: 8.7 & 8.3 - 9.1\tablenotemark{b}& 0.06\tablenotemark{b} & H I: 10-7, FeI, Ca I\tablenotemark{c} \\
1069: 9.4 & 8.9 - 11.1 & 0.02 & Ca I, Si I, C I, C II\\
1070: 10.4 & 9.9 - 11.1 & 0.04 & Mg I, Ne I, C I]\\
1071: 11.4 & 11.1 - 11.8 & 0.07 & He I, He II\\
1072: 12.5 & 12.1 - 12.6 & 0.15 & Hu$\alpha$\\
1073:
1074: \cutinhead{KIN Outer Spatial Regime}
1075:
1076: 8.9 & 8.3 - 9.5 & 0.14& H I: 10-7, [V II], Si I, Ca I]\tablenotemark{d}\\
1077: 9.8 & 9.0 - 10.7 & 0.24&Silicate Dust\tablenotemark{e} \\
1078: 11.4 & 11.0 - 11.8 & 0.19 & H I: 9 - 7, He I, He II\\
1079: \enddata
1080:
1081: \tablenotetext{a}{Approximate FWZI continuum crossing points.}
1082: \tablenotetext{b}{Because \textit{Keck} and \textit{Spitzer} data were not taken simultaneously and because of the extreme transient nature of the RN outburst it is not possible to assert particular flux numbers to the atomic lines attributed to the KIN feature. The flux listed should be considered the maximum, bounding flux for any one of the atomic species shown here. }
1083: \tablenotetext{c}{All atomic line emission in the KIN inner spatial regime is assumed to be emitted by a species at approximate cosmic abundance as for RN only a very moderate amount of nucleosynthesis is theorized to occur.}
1084: \tablenotetext{d}{All atomic line emission in the KIN Outer spatial regime are assumed to be predominantly of nebular abundance with some contribution from uncondensed metals.}
1085: \tablenotetext{e}{Spectral feature is spectrally and at least partially spatially resolved. Silicate dust feature has more than seven spectral elements across it and emits relatively strongly at a distance centered $\sim$ 17 AU from the WD. All flux is assumed to be attributable to emission by the dust.}
1086:
1087: \end{deluxetable}
1088:
1089: %:table 5 luminosity evolution
1090: \begin{deluxetable}{l@{\extracolsep{10ex}}ll}
1091: %\tabletypesize{\scriptsize}
1092: %\rotate
1093: \tablewidth{0pt}
1094: \tablecaption{Post-outburst Evolution of the White Dwarf Star in RS Oph \tablenotemark{a}\label{evo}}
1095: \tablehead{
1096: \colhead{Time\tablenotemark{b}} & \colhead{Radius}& {Temperature} \\
1097: \colhead{(days)} & \colhead{(R$_{\odot}$)}& {(K)} \\
1098: }
1099: \startdata
1100: 0 & 45 & 12200 \\
1101: 2 & 21 & 20000\\
1102: 14 & 1.6 & 67000 \\
1103: 29 & 0.56 & 114000 \\
1104: 51 & 0.23 & 181000 \\
1105: 72 & 0.083 & 302000\\
1106: 119 & 0.0037 & 1020000 \\
1107: 251 & 0.003 & 350000 \\
1108: \enddata
1109: \tablenotetext{a}{Stellar parameters for the white dwarf star after the 2006 outburst
1110: of RS Oph based on Hachisu \& Kato 2001}
1111: \tablenotetext{b}{The zero time for these parameters is the V band maximum light, which occurs
1112: about 3-4 days after the thermonuclear runaway process begins.}
1113: \end{deluxetable}
1114:
1115: \clearpage
1116: %: figure 1 - "KINconcept" conceptual diagram
1117: \begin{figure}
1118: \begin{center}
1119: \includegraphics[width=4.5in]{f1.eps}
1120: \caption{The upper part of plot displays a conceptual view of the operation of the KIN system. There are two short baselines, separated by ~ 5 m, and two long baselines separated by ~ 85 m. A $\pi$ phase shift is applied to the pupils on the long baseline. The autocorrelation of the four pupils is shown as well as the equivalent intensity pattern, also called the transmission pattern. Radiation passing through the white striped regions is detected by the system. Lower part of the plot displays the sequence used for the measurement process. There are two chopping sequences, (a) and (b) display the $\sim$ 5 Hz chopping between the null and bright fringe patterns on the source using the 85 m baseline for starlight subtraction, used to measure the null response, while (c) and (d) show the ~$\sim$1 Hz chop sequence between the two short baselines used to remove the telescope and sky backgrounds. Note that the long-baseline fringes are not to scale with the field of view in this depiction. \label{KINconcept}}
1121: \end{center}
1122: \end{figure}
1123:
1124: \clearpage
1125: %: figure 2 - continuum modeling: label "modelfit"
1126: \begin{figure}
1127: \begin{center}
1128: \includegraphics[width=6.5in]{f2.eps}
1129: \caption{Plot of three best-fit continuum models against {\it Keck} Nuller data. All three models; disk, shell, and Gaussian, that minimized $\chi^{2}$ simultaneously against the null leakage and intensity spectra, lie effectively on top of one another. The top trace gives the dimensionless null leakage or interferometric observable which is the null fringe output divided by the intensity spectrum. The lower trace is the constructive fringe output or intensity spectrum. As described in the text we have removed several data points associated with emission features from the original set for the purpose of fitting the continuum. \label{modelfit}}
1130: \end{center}
1131: \end{figure}
1132:
1133: \clearpage
1134: %: figure 3 - KIN vs. downbinned Spitzer: label "kn_spectra"
1135:
1136: \begin{figure}
1137: \begin{center}
1138: \includegraphics [width=5.5in]{f3.eps}
1139: \caption[KIN data compared to \textit{Spitzer} data]{Plot of KIN spectra together with a {\it Spitzer} spectrum. Upper trace is the intensity spectrum or inner KIN spatial regime - light dominated by the inner 25 mas about the center of the source brightness distribution at mid band. The middle trace is the nulled fringe output which is the interferometric observable times the intensity spectrum multiplied by 2 (to correct for transmission through the fringe pattern for extended sources) to give the source brightness in the outer spatial regime. This is predominantly emission from material greater than about 12.5 mas ($\sim$17 AU at 1.4 kpc) from the center of the source. The lower trace is {\it Spitzer} data \citep{eva07b} from day 63 boxcar averaged to yield approximately the same spectral resolution as the {\it Keck} Interferometer Nuller. None of the data were continuum normalized. Note the features between 9 and 11 $\micron$ and how the inner and outer spatial regime spectra are different from one another. Note that the Spitzer spectra have very low spatial resolution and combine the light from the entire region detected by both KIN inner and outer spatial regimes.}
1140: \label{kn_spectra}
1141: \end{center}
1142: \end{figure}
1143:
1144:
1145: \clearpage
1146: %: figure 3 - Spitzer high resolution spectra: label "continuum"
1147: \begin{figure}
1148: \begin{center}
1149: \includegraphics[width=6.5in]{f4.eps}
1150: \caption{{\it Spitzer} space telescope spectra from days 63, 73 and 209. Here we see that the continuum drops rapidly as time advances with the spectral emission features almost identical on days 4/16 and 4/26. Data obtained on 9/9 is starkly different with a strong solid-state feature evident between 9 and 13 microns (see \citet{eva07c} for more details on the {\it Spitzer} data). \label{continuum}}
1151: \end{center}
1152: \end{figure}
1153:
1154:
1155: %: figure 4 - KIN continuum normalized spectra with line id: label "keckspitzlines"
1156: \begin{figure}
1157: \begin{center}
1158: \includegraphics[width=4.5in]{f5.ps}
1159: \caption{Continuum normalized \textit{Keck} Interferometer Nuller spectra 3.8 days after peak V-band brightness. Spectra are shown matched with lines identified in \textit{Spitzer} spectra. The inner spatial regime line identification was matched to the cosmic abundances seen in the earlier \textit{Spitzer} spectra while the outer spatial regime was biased towards the later spectrum on September 9, 2006 that included a solid state feature and were assumed to have primarily nebular abundances. Please see the text for more discussion of the line identification process. \label{keckspitzlines}}
1160: \end{center}
1161: \end{figure}
1162:
1163: %: figure 5 - AAVSO lightcurve
1164: %\begin{figure}
1165: %\begin{center}
1166: %\includegraphics[width=4.5in]{f4.eps}%
1167: %\caption{AAVSO photometry with $T_2$ = 4.75 days and $T_3$ = 10.19 days indicated. Note the end of the plateau phase on 4/26/06 - evidence of the hydrogen burning turnoff theorized by \citet{hac06}. Note the evident strong fluctuation of V-band brightness about 9/9/06. This is strong evidence of a dust creation event. \label{lightcurve}}
1168: %\end{center}
1169: %\end{figure}
1170:
1171: %: figure 5 - this is the proposed model figure
1172:
1173: \begin{figure}
1174: \begin{center}
1175: \includegraphics[width=4.5in]{f6.eps}
1176: \caption{Proposed model of circumstellar material surrounding the binary RS Oph before the nova
1177: eruption. The interaction of the white dwarf and red giant star in the slow dense wind of the red giant
1178: star can create a spiral shock with an enhanced density in the plane of the orbit of the two stars. The overall size of the dense in-plane material is of the order of 100 mas if RS Oph is at a distance of 1400 pc and the wind speed is about 20 km/s. \label{dust-sub_01}}
1179: \end{center}
1180: \end{figure}
1181:
1182: %:fig6 dust-sub_02
1183:
1184:
1185: \begin{figure}
1186: \begin{center}
1187: \includegraphics[width=4.5in]{f7.eps}
1188: \caption{Sublimation radius of dust (mas) as a function of the number of days post-outburst for the Nova
1189: and for the red giant companion. \label{dust-sub_02}}
1190: \end{center}
1191: \end{figure}
1192:
1193: %fig7
1194:
1195: \begin{figure}
1196: \begin{center}
1197: \includegraphics[width=4.5in]{f8.eps}
1198: \caption{Temperature of red giant star for the surface that is facing the Nova as a function of
1199: the number of days post-outburst. Also plotted is the temperature of the non-irradiated side of
1200: the red giant. Modeled temperature rise is due to heating from irradiation by the nova and neglects any effect due to the passage of the forward shock. \label{RG-face-temp}}
1201: \end{center}
1202: \end{figure}
1203:
1204: %:fig8
1205:
1206: \clearpage
1207: \begin{figure}
1208: \begin{center}
1209: %\epsscale{0.6}
1210: %\plottwo{RSOph-day4-alt}{RSOph-day21}
1211: %\plottwo{RSOph-day57}{RSOph-day90}
1212: \includegraphics[width=2.2in]{f9a.eps}
1213: \includegraphics[width=2.2in]{f9b.eps}
1214: \includegraphics[width=2.2in]{f9c.eps}
1215: \includegraphics[width=2.5in]{f9d.eps}
1216: \caption{Spiral shock wave model of RS Oph.
1217: (a) Top left panel displays the system geometry at 4 days post-outburst. A gray ring is drawn in the
1218: center of the figure to indicate the size of the shocked region at this epoch. The outer part of the spiral
1219: is overlayed with light gray to indicate that it is not known if the material stays in a coherent spiral
1220: past the first several turns. The diameter of the shocked region is about 5 mas assuming the
1221: blast wave travels at a velocity of 1730 km/s in the plane of the orbit.
1222: (b) Top right panel. The blast wave is now about 13 mas in diameter on day 21.
1223: (c) Bottom left panel. The blast wave is about 36 mas in diameter on day 57.
1224: (d) Bottom right panel. By day 90 the blast wave has traversed the entire spiral pattern. \label{CSMatterEvol}}
1225: %The motion of the white dwarf and red giant star
1226: %in the dense wind of the red giant star creates a spiral shock with a density enhancement of about 100
1227: %in the plane of the orbit. Assuming typical wind velocities of about 20 km/s for the red giant wind,
1228: %there are about 17 rings separated by about 3.3 mas that form between RS Oph outbursts.
1229: \end{center}
1230: \end{figure}
1231:
1232: %:fig9
1233:
1234: \begin{figure}
1235: \begin{center}
1236: \includegraphics[width=4.5in]{f10a.eps}
1237: \includegraphics[width=4.5in]{f10b.eps}
1238: \includegraphics[width=4.5in]{f10c.eps}
1239: \caption{(a) V Band luminosity as a function of time for Nova and red giant components of binary
1240: system.(b) H band luminosity evolution. (c) N band luminosity evolution. \label{LumEvol}}
1241: \end{center}
1242: \end{figure}
1243:
1244:
1245: \end{document}
1246:
1247:
1248: