1: %\documentclass[12pt,preprint]{aastex}
2: %\documentclass[10pt, letterpaper, onecolumn]{emulateapj}
3: \documentclass[10pt, letterpaper]{emulateapj}
4:
5: %\usepackage{graphicx}
6: %\usepackage{epstopdf}
7:
8:
9:
10:
11: \begin{document}
12:
13: \title{Interferometric Observations of V1663 Aquilae (Nova Aql 2005)}
14: \author{B. F. Lane \altaffilmark{1}, A. Retter \altaffilmark{2}, J. A. Eisner\altaffilmark{5},
15: M. W. Muterspaugh\altaffilmark{3}, R. R. Thompson \altaffilmark{4},
16: J. L. Sokoloski\altaffilmark{6}}
17:
18: \altaffiltext{1}{Kavli Institute, MIT Department of Physics, 70 Vassar Street, Cambridge, MA 02139; blane@mit.edu}
19: \altaffiltext{2}{Astronomy \& Astrophysics Dept., Penn State University, 525 Davey Lab, University Park, PA 16802-6305; retter@astro.psu.edu}
20: \altaffiltext{3}{Townes Fellow, Space Sciences Laboratory, University of California, Berkeley, CA 94720}
21: \altaffiltext{4}{ Palomar Testbed Interferometer, Palomar Mtn, CA 92060}
22: \altaffiltext{5}{Miller Fellow, University of California at Berkeley, 601 Campbell Hall, Berkeley, CA 94720}
23: \altaffiltext{6}{Columbia Astrophysics Lab, MC 5247, Columbia University, New York, NY 10027.}
24: %\altaffiltext{7}{Jolly Good Fellow}
25: \begin{abstract}
26: We have resolved the classical nova V1663 Aql using long-baseline
27: near-IR interferometry covering the period from $\sim$5--18 days after
28: peak brightness. We directly measure the shape and size
29: of the fireball, which we find to be asymmetric. In addition
30: we measure an apparent expansion rate of $0.21 \pm
31: 0.03$ ${\rm mas\,day^{-1}}$. Assuming a linear expansion
32: model we infer a time of initial outburst approximately 4 days prior to peak brightness.
33: When combined with published spectroscopic expansion velocities our
34: angular expansion rate implies a distance of
35: $8.9\pm3.6$ kpc. This distance measurement is independent of, but
36: consistent with, determinations made using widely available
37: photometric relations for novae.
38: \end{abstract}
39:
40: \keywords{techniques:interferometry--star:V1663 Aql--novae}
41:
42:
43: \section{Introduction}
44: Novae are violent stellar explosions exceeded in energy output only by
45: $\gamma$-ray bursts and supernovae. They are erratic outbursts that
46: occur in systems containing a white dwarf accreting mass from a
47: late-type stellar companion (e.g Prialnik \& Kovetz\nocite{pk05} 2005).
48: When the amount of accreted
49: material on the surface of the white dwarf reaches some critical value
50: a thermonuclear-runaway is ignited, giving rise to the observed nova
51: outburst in which material enriched in heavy elements is ejected into
52: the surrounding medium at high velocities. For certain elements, this
53: ejected material may influence observed abundances in the
54: ISM \citep{gehrz98,hernanz05}. Direct observations of the expansion of
55: the nova shell provide an opportunity to accurately determine the
56: distance to the nova. Such observations are usually only possible many
57: months or years after the outburst, when the expanding shell can be resolved
58: \citep{bode02}. However, several optical/IR interferometers are now capable of resolving
59: novae and other explosive variables, allowing detailed studies of the
60: initial "fireball", \citep{ches07,mon06} as well as later stages of
61: development \citep{lane05,lane07}.
62:
63:
64: Nova Aquilae 2005 (ASAS190512+0514.2, V1663 Aql) was discovered on 9
65: June 2005 by \citet{iauc8540}. At the time of discovery the magnitude
66: was $m_V$ = 11.05; the source reached $m_V \sim 10.8$ the following
67: day, and declined in brightness thereafter. A possible progenitor near
68: the source coordinates (sep. $\sim$ 4.5 arcseconds) is seen on Palomar
69: Optical Sky Survey plates (USNO-B1.0 0952-00410569, \citet{usno-b1}),
70: with magnitudes $m_R \sim 18.1$ and $m_I \sim 16.45$. Soon after
71: discovery \citet{iauc8544} obtained an optical spectrum with features
72: indicating a heavily reddened, peculiar nova. H-$\alpha$ emission
73: lines exhibited P Cygni line profiles and indicated an expansion
74: velocity of $700 \pm 150 {\rm km\,s^{-1}}$ \nocite{iauc8544} (Dennefeld et al. 2005,
75: confirmed via personal communication), somehwhat slow for a classical nova, but not outside
76: the range of observed values. Recently, \citet{pog06} published
77: spectra and analysis of published light-curves of this nova, deriving
78: a distance in the range 7.3--11.3 kpc, and an expansion velocity of $\sim2000$~${\rm km\,s^{-1}}$.
79:
80: We have used the Palomar Testbed Interferometer (PTI) to resolve the
81: $2.2 \mu$m emission from V1663 Aql and measure its apparent angular
82: diameter as a function of time. We are able to follow the expansion
83: starting $\sim 9$ days after the initial explosion; when combined with
84: radial velocities derived from spectroscopy we are able to infer a
85: distance to, and luminosity of, the object. We compare this result
86: with values inferred by a maximum magnitude-rate of decline (MMRD)
87: relation in the literature (see Poggiani 2006 for a summary).
88:
89: The Palomar Testbed Interferometer
90: (PTI) was built by NASA/JPL as a testbed for developing ground and
91: space-based interferometry and is located on Palomar Mountain near San
92: Diego, CA \citep{col99}. It combines starlight from two out of three
93: available 40-cm apertures and measures the resulting interference
94: fringes. The high angular resolution provided by this long-baseline
95: (85-110 m), near infrared ($2.2 \mu$m) interferometer is sufficient to
96: resolve emission on the milli-arcsecond scale.
97:
98: \section{Observations}
99:
100: We observed V1663 Aql on 10 nights between 15 June 2005 and 28 June
101: 2005; on six of those nights we obtained data on two or three
102: interferometric baselines. For a detailed description of the
103: instrument we refer the reader to \cite{col99}. Each nightly
104: observation consisted of one or more 130-second integrations during
105: which the normalized fringe visibility of the science target was
106: measured. The measured fringe visibilities of the science target were
107: calibrated by dividing them by the point-source response of the
108: instrument, determined by interleaving observations of calibration
109: sources (Table \ref{tab:calibs}); the calibration sources were chosen
110: to be single stars, close to the target on the sky and to have angular
111: diameters less than 2 milli-arcseconds, determined by fitting a
112: black-body to archival broadband photometry of the sources. For
113: further details of the data-reduction process, see \citet{colavita99b}
114: and \citet{boden00}. Calibrated fringe visibilities are listed in
115: Table \ref{tab:data}; though given as averages, the fits to the models
116: were done using the individual data points, without averaging on a nightly basis.
117: Note that HD~188310 has been claimed to be a binary
118: with a separation of $0.09\pm0.01$ arcseconds \citep{scardi00}. However,
119: this claim is based on a single observational epoch obtained under poor
120: observing conditions, and has to our knowledge not been confirmed
121: by any other group. We do not see any indications of binarity in
122: our data; the fringe visibilities of HD~188310 are stable to 2.5\%
123: when calibrated using HD~176303. In addition, we reduce the V1663
124: Aql data both with and without using HD~188310 as a calibrator and find the
125: results to be fully self-consistent. We include HD~188310 as a calibrator in
126: the results shown here as it was the only calibrator available on the last
127: night of observations.
128:
129: %\clearpage
130: \begin{table}
131: \begin{center}
132: \begin{tabular}{lccccc}
133: & & & & & \\
134: \tableline
135: \tableline
136: Calibrator & V & K & Spectral & Ang. Diameter & Separation \\
137: & & & Type & $\theta_{UD}$ (mas) & (deg) \\
138: \tableline
139: HD~176303 & 5.27 & 3.93 & F8V & $0.71 \pm 0.03$ & 8.5 \\
140: HD~188310 & 4.72 & 2.17 & G9III & $1.62 \pm 0.08$ & 13 \\
141: HD~186442 & 6.56 & 3.67 & K0III & $0.92 \pm 0.16$ & 11 \\
142: \tableline
143: \end{tabular}
144: \caption{\label{tab:calibs} Relevant parameters of the calibrators and check star. The
145: separation listed is the angular distance from the calibrator to V1663 Aql. Alternate catalog designations: HD~176303 = 11 Aql = HR 7172; HD~188310 = 59 Aql = $\xi$ Aql = HR 7595; HD 186442 = BD +09 4233 = SAO 125000.}
146: \end{center}
147: \end{table}
148:
149: %\clearpage
150: \begin{table}
151: \begin{center}
152: \begin{tabular}{lcccc}
153: & & & & \\
154: \tableline
155: \tableline
156: Epoch & $u$ & $v$ & No. Pts. &Cal. Visibility \\
157: (MJD) & (m) & (m) & & ($V^2$) \\
158: \tableline
159: 53536.424 & -75.3240 & -25.0938 & 1 &$ 0.698 \pm 0.117$ \\
160: 53537.393 & -79.8978 & -23.7842 & 8 &$ 0.749 \pm 0.030$ \\
161: 53538.383 & -44.1836 & 66.0275 & 10 &$ 0.718 \pm 0.033$ \\
162: 53539.346 & -81.4086 & -21.8080 & 14 &$ 0.672 \pm 0.050$ \\
163: 53540.392 & -79.3017 & -24.1591 & 4 &$ 0.610 \pm 0.035$ \\
164: 53540.333 & -51.6711 & -88.3761 & 5 &$ 0.567 \pm 0.026$ \\
165: 53541.395 & -78.3316 & -24.4178 & 3 &$ 0.568 \pm 0.038$ \\
166: 53541.339 & -48.6701 & -88.6301 & 5 &$ 0.484 \pm 0.024$ \\
167: 53542.396 & -77.4187 & -24.5759 & 3 &$ 0.538 \pm 0.037$ \\
168: 53542.320 & -53.4276 & -88.0947 & 6 &$ 0.534 \pm 0.048$ \\
169: 53542.458 & -59.4251 & 63.4296 & 5 &$ 0.519 \pm 0.074$ \\
170: 53543.400 & -76.2535 & -24.8719 & 4 &$ 0.499 \pm 0.029$ \\
171: 53543.452 & -59.1619 & 63.5118 & 5 &$ 0.454 \pm 0.046$ \\
172: 53548.363 & -80.2661 & -23.8083 & 5 &$ 0.537 \pm 0.078$ \\
173: 53548.300 & -54.5455 & -88.0372 & 5 &$ 0.539 \pm 0.096$ \\
174: 53549.377 & -77.8030 & -24.5517 & 2 &$ 0.614 \pm 0.194$ \\
175: 53549.305 & -52.5498 & -88.2695 & 4 &$ 0.538 \pm 0.173$ \\
176: \tableline
177: \end{tabular}
178: \caption{\label{tab:data} Calibrated fringe visibilities of
179: V1663 Aql, together with the projected baseline components.
180: The effective wavelength of observations was 2.2 $\mu$m.
181: $u,v$ are the components of the projected baseline in meters. All of the
182: data from a given baseline on a given night has been averaged. }
183: \end{center}
184: \end{table}
185: %\clearpage
186:
187:
188: PTI is equipped with a low-resolution spectrometer which provides
189: fringe visibility measurements and photon count rates in five spectral
190: channels across the K band. In addition to the fringe visibilities
191: measured in each channel (referred to as ``narrow-band'' visibilies),
192: we compute a wide-band average visibility as the photon-weighted
193: average of the five spectral channels. Using photon count rates from
194: PTI and K-band magnitudes for the calibrator sources provided by 2MASS
195: \citep{2mass}, we also derive a K-band apparent magnitude of V1663
196: Aql. It should be remembered that PTI was not designed
197: for high-precision photometry, and hence the K-magnitudes, while useful, should be
198: treated with some caution.
199:
200: \section{Models \& Results}
201:
202:
203: \subsection{Interferometric Data}
204:
205: The theoretical relation between source brightness distribution and
206: fringe visibility is given by the van Cittert-Zernike theorem\citep{tms01}.
207: For a uniform intensity disk model the normalized fringe visibility (squared) can be related
208: to the apparent angular diameter using
209: \begin{equation}
210: V^2 = \left( \frac{2 \; J_{1}(\pi B \theta_{UD} / \lambda)}{\pi B\theta_{UD} / \lambda} \right)^2
211: \label{eq:V2_single}
212: \end{equation}
213: where $J_{1}$ is the first-order Bessel function, $B$ is the projected
214: aperture separation and given by $B = \sqrt{u^2 + v^2}$ where $u,v$
215: are two orthogonal components of the projected baseline, $\theta_{UD}$
216: is the apparent angular diameter of the star in the uniform-disk
217: model, and $\lambda$ is the wavelength of the observation.
218:
219: We obtain an initial characterization of the data using a circularly symmetric
220: uniform-disk model fit to data from a single baseline on a nightly basis (Fig. \ref{fig:udfits}).
221: This indicates that the apparent source morphology was initially
222: expanding, but that the apparent expansion reversed sometime between
223: MJD 53545 and MJD 53548. The data also clearly indicate that while
224: the North-West and South-West data produce consistent angular diameters, the
225: North-South data indicates both a smaller apparent diameter and a
226: less smooth expansion. Clearly, a more sophisticated morphological model
227: is required.
228:
229: \begin{figure}[h]
230: \epsscale{1.0}
231: \plotone{f1.eps}
232: \caption[]{\label{fig:udfits} The apparent angular dimeter of the
233: V1663 Aql source, as a function of time, assuming a uniform (circular)
234: disk model. Each symbol represents a fit to data from a single baseline.
235: Although the data are all consistent with an expansion, the disagreement
236: between various baselines indicates the need for a more complex
237: morphological model.}
238: \end{figure}
239: %\clearpage
240:
241: The availability of data taken with multiple baselines with different
242: position angles allows us to distinguish between a circularly
243: symmetric source and elliptical disk models. For the
244: asymmetric cases we fit for three parameters: size ($\theta$),
245: inclination angle ($\phi$), and position angle ($\psi$). Inclination
246: is defined such that a face-on disk has $\phi=0$ and $\psi$ is
247: measured east of north. Following \citet{eis03}, we include $\phi$ and
248: $\psi$ in our models of the brightness distribution via a simple
249: coordinate transformation:
250: \begin{eqnarray}
251: u\prime & = & u \sin \psi + v \cos \psi \\
252: v\prime & = & \cos \phi \left ( v \sin \psi - u \cos \psi \right )
253: \end{eqnarray}
254: Substitution of ($u\prime, v\prime$) for ($u, v$) in the expressions
255: above yields models with inclination effects included. Note that the
256: ``inclination'' is not tied to a true disk-like morphology; the source
257: can well be (and in this particular case probably is) an ellipsoid,
258: in which case the major axis of the ellipsoid is $\theta$ and the
259: minor axis is given by $\theta \cos(\phi)$. We perform least-squares
260: fits of uniform and inclined disk (or equivalently ellipsoidal) models
261: to the measured fringe visibilities.
262:
263: As is apparent from Fig. \ref{fig:inclined} the data on nights with
264: more than one available baseline are not well matched by a simple
265: uniform disk model; in effect the visibility in the North-South
266: baseline is too high (and the corresponding uniform-disk diameter too
267: small) to be consistent with the North-West and South-West
268: baselines. Typical goodness-of-fit parameters are
269: $\chi_r^2\sim5$. However, it is possible to fit the data using an
270: inclined disk-model (Table \ref{tab:dfits}). While the angular size
271: scale changes from night to night, the consistency of the inclination
272: and position angles of the fits lend credence to the idea that we are
273: observing an expanding, asymmetric source.
274:
275: \begin{figure*}[htbp]
276: \epsscale{1.0}
277: \plottwo{f2a.eps}{f2b.eps}
278: \caption[]{\label{fig:inclined} (left) The measured fringe visibility of
279: V1663 Aql as a function of hour angle during the night of 22 June
280: 2005. Although PTI can only use one baseline at a time, during the
281: night two changeovers were performed, yielding data on all three
282: baselines within a relatively short period of time. An inclined-disk
283: model was fit to the data (shown as thick lines), yielding a good fit. The
284: fit parameters are given in Table \ref{tab:dfits}. A corresponding
285: uniform-disk fit (thin lines) is shown for comparison; while it can match
286: the N-W and S-W data it significantly under-predicts the N-S baseline
287: visibilities. (right) The $uv$-plane coverage (i.e. the projected baselines)
288: of the observations.}
289: \end{figure*}
290:
291: %\clearpage
292: \begin{table*}
293: \begin{center}
294: \begin{tabular}{lccccc}
295: & & & & \\
296: \tableline
297: \tableline
298: Date & $\chi_r $& Size & $\psi$ & $\phi$ & Baselines\\
299: & & (mas) & (deg) & (deg) & \\
300: \tableline
301: %19 June 2005 & 0.4 & $3.1\pm0.5$ & $105.9 \pm 0.1$ & $46.2\pm0.2$ & NS,NW\\
302: %20 June 2005 & 0.8 & $3.3\pm0.5$ & $105.9 \pm 0.2$ & $46.3\pm0.2$ & NS,NW \\
303: %21 June 2005 & 0.5 & $2.9\pm0.05$ & $105.5 \pm 0.05$ & $46.3\pm0.03$ & NS,NW,SW \\
304: %22 June 2005 & 0.6 & $3.1\pm0.05$ & $105.4 \pm 0.04$ & $46.3\pm0.03$ & NS,NW,SW \\
305: %27 June 2005 & 0.4 & $3.2\pm1.1$ & $105.9 \pm 0.4$ & $46.3\pm0.3$ & NS,NW \\
306: %28 June 2005 & 0.4 & $3.0\pm1.9$ & $105.1 \pm 2.6$ & $45.5\pm98$ & NS,NW \\
307: 19 June 2005 & 0.4 & $3.1\pm0.5$ & $129.5 \pm 6.6$ & $49.2\pm8.3$ & NS,NW\\
308: 20 June 2005 & 0.8 & $3.3\pm0.5$ & $127.7 \pm 7.9$ & $46.0\pm9.2$ & NS,NW \\
309: 21 June 2005 & 0.5 & $2.9\pm0.05$ & $ 103.8 \pm 2.9$ & $44.8\pm2.0$ & NS,NW,SW \\
310: 22 June 2005 & 0.6 & $3.1\pm0.05$ & $101.6 \pm 2.1$ & $45.11\pm2.0$ & NS,NW,SW \\
311: 27 June 2005 & 0.4 & $3.2\pm1.0$ & $ 127.1 \pm 19$ & $48.8\pm18$ & NS,NW \\
312: 28 June 2005 & 4.4 & $2.8\pm1.7$ & $ 80.7 \pm 161$ & $89.2\pm360$ & NS,NW \\
313:
314: \tableline
315: \end{tabular}
316: \caption{\label{tab:dfits} Inclined-disk model fits to data from a given
317: multi-baseline night. NS = North-South, NW = North-West and
318: SW South-West baselines, respectively}
319: \end{center}
320: \end{table*}
321: %\clearpage
322:
323: We examine the possibility of systematic errors mimicing the effect of
324: non disk-like emission using the following test: we use HD 186442 as a
325: ``check star'' by making observations of it interleaved with
326: observations of our calibrators and V1663 Aql. Such observations were
327: made on three nights, and on two of those nights we observed the check
328: star with all three available interferometric baselines. We calibrate
329: the check star data using the same procedure as applied to V1663 Aql, and
330: fit a uniform-disk model to the calibrated nightly data. We find that
331: the best-fit uniform disk diameter is $0.83 \pm 0.05$ milli-arcseconds
332: with a goodness-of-fit parameter ($\chi_r$) of 0.80 for 16 130-second
333: data points, and a mean scatter in the visibilities about the best-fit
334: model of 2.9\%, fully comparable with typical point source observations
335: using PTI.
336:
337: Given the reasonable stability of the position angle and inclination (or
338: equivalently, aspect ratio) parameters, we re-fit the data from all
339: the nights, holding these two parameters fixed at the average values
340: from Table \ref{tab:dfits} (excluding the last two points where
341: the fits are effectively indeterminate), while
342: letting the diameter vary on a nightly basis. The resulting best-fit
343: diameters are given in Table \ref{tab:fits} and shown in Figure
344: \ref{fig:expansion}. A simple linear expansion model fit to the disk
345: sizes indicates an expansion rate of $0.21 \pm 0.03$
346: milli-arcseconds/day, beginning on MJD $53527.4 \pm 1.9 $.
347:
348: %\clearpage
349: \begin{table}
350: \begin{center}
351: \begin{tabular}{lccccc}
352: & & & \\
353: \tableline
354: \tableline
355: Date & MJD & Size & $\chi_r $ \\
356: & & (mas) & \\
357: \tableline
358: %16 June 2005 &53537.393 & $2.12 \pm 0.03 $ & 0.8\\
359: %17 June 2005 &53538.383 & $2.22 \pm 0.03 $ & 0.8\\
360: %18 June 2005 &53539.346 & $2.43 \pm 0.03 $ & 0.9\\
361: %19 June 2005 &53540.359 & $2.90 \pm 0.04 $ & 1.1\\
362: %20 June 2005 &53541.36 & $3.29 \pm 0.06 $ & 1.5\\
363: %21 June 2005 &53542.385 & $3.01 \pm 0.05 $ & 1.0\\
364: %22 June 2005 &53543.366 & $3.20 \pm 0.06 $ & 1.6\\
365: %27 June 2005 &53548.325 & $3.00 \pm 0.05 $ & 0.6\\
366: %28 June 2005 &53549.329 & $2.90 \pm 0.35 $ & 1.8\\
367: 16 June 2005 &53537.393 & $2.13 \pm 0.03$& 0.7\\
368: 17 June 2005 &53538.383 & $2.23 \pm 0.03$& 0.6\\
369: 18 June 2005 &53539.346 & $2.44\pm 0.03$& 0.9\\
370: 19 June 2005 &53540.359 & $2.92 \pm 0.04$& 1.3\\
371: 20 June 2005 &53541.36 & $3.30\pm 0.06$& 2.3\\
372: 21 June 2005 &53542.385 & $3.03 \pm 0.05$& 1.2\\
373: 22 June 2005 &53543.366 & $3.21\pm 0.06$& 2.6\\
374: 27 June 2005 &53548.325 & $3.01\pm 0.05$& 0.4\\
375: 28 June 2005 &53549.329 & $2.57 \pm 0.35$& 3.3\\
376:
377: \tableline
378: \end{tabular}
379: \caption{\label{tab:fits} The best-fit major-axis angular sizes of
380: inclined-disk models fit to the data on a nightly basis. The inclination
381: of the disks was fixed at 46 degrees, and the position angle
382: was held at 115 degrees, corresponding to the average values
383: determined from the best fits in Table \ref{tab:dfits}. }
384: \end{center}
385: \end{table}
386: %\clearpage
387:
388: \begin{figure}[h]
389: \epsscale{1.0}
390: \plotone{f3.eps}
391: \caption[]{\label{fig:expansion} The best-fit inclined disk diameter as
392: a function of time, togther with a best-fit linear expansion model.
393: The expansion appears to stop sometime after MJD 53545; the subsequent
394: points have been left out of the fit. }
395: \end{figure}
396: %\clearpage
397:
398: \subsection{Narrow-band Interferometric Data}
399:
400: In addition to the wide-band fringe visibilities, we can compute the
401: best-fit disk diameters for data from each spectral channel separately;
402: results from three nights are show in Fig. \ref{fig:nb}. The lack of
403: significant variation in apparent angular diameter with wavelength
404: supports our assumption that the observed
405: visible emission lines are originating from the same region as the K-band
406: emission seen by PTI, at least during the first few weeks of the expansion.
407:
408: \begin{figure}[h]
409: \epsscale{1.0}
410: \plotone{f5.eps}
411: \caption[]{\label{fig:nb} The best-fit inclined disk diameter as
412: a function of wavelength, for data from three nights. The
413: inclination and posotion angles were held fixed at the
414: average best-fit values from Table \ref{tab:dfits}}
415: \end{figure}
416: %\clearpage
417:
418: \subsection{Photometry}
419: \label{sec:phot}
420: We derive the approximate K-band magnitude of V1663 Aql using photon
421: fluxes measured by PTI; the results are shown in
422: Fig. \ref{fig:phot}. We have also collected B and V-band photometry
423: published from VSNET\footnote{\tt http://www.kusastro.kyoto-u.ac.jp/vsnet/}
424: and IAU circulars. We fit a linear trend to the data in order to derive the rate
425: of declines in the K and V bands; we find
426: that the PTI K-band photometry implies $\dot{m}_K = 0.13 \pm 0.01 {\rm mag\, day^{-1}}$,
427: and the compiled V-band photometry gives
428: $\dot{m}_V = 0.126 \pm 0.004{\rm mag\, day^{-1}}$. The V-band data
429: indicates a time to decline 2
430: magnitudes from maximum ($t_2$) of $15.9\pm0.5$ days, and a 3-magnitude
431: decline $t_3 = 23.8 \pm 0.8$ days. We also fit a low-order polynomial
432: function to the $B$- and $V$-band photometry, allowing us to find an
433: approximate $B-V$ color for $t=t_2$ of 1.11. Based on the work by \citet{vdB87} the intrinsic B--V
434: color of novae two magnitudes below maximum light is $-0.02\pm0.04$,
435: and hence we infer $E(B-V)=1.1$. The corresponding V-band extinction
436: $A_V\sim 3.4$ mag.
437:
438: \begin{figure}[h]
439: \epsscale{1.0}
440: \plotone{f4.eps}
441: \caption[]{\label{fig:phot} Measured B-,V- and K-magnitudes
442: of V1663 aql following the nova outburst. K magnitudes were
443: measured by PTI, while B and V magnitudes were collected
444: from VSNET and IAU circulars. The best-fit linear magnitude trends
445: are also shown, together with the time of initial expansion
446: $T_0$ determined from fitting to the measured diameters.}
447: \end{figure}
448: %\clearpage
449:
450: We use the MMRD relation
451: from \citet{dV95} and our value of $t_2$ to derive $M_V = -8.31$
452: for V1663 Aql. Adopting a peak value of $m_{V,max}=10.8$ we
453: would predict a distance of $\sim 13.8$ kpc.
454:
455: \section{Discussion}
456:
457: Using our interferometrically measured angular diameters it is
458: possible to constrain three important parameters of this system:
459: the fireball shape, the time of initial expansion, and the distance to the system.
460: However, we note that we are using relatively simplistic models,
461: and our interpretations are therefore limited and subject to
462: a systematic uncertainty that is difficult to quantify. Nonetheless, the
463: measured fringe visibility and the way it changes with time and baseline
464: strongly imply that the source is expanding and likely asymmetric.
465:
466: First, we measure a non-zero asymmetry in the expanding
467: fireball as early as ten days after the initial outburst; this
468: would tend to indicate that the source of the asymmetry is
469: inherent to the explosion mechanism and not the result
470: of interactions with the interstellar or circumstellar medium.
471: As discussed in \citet{bode02} there is an apparent correlation between
472: nova speed class and apparent asymmetry; our measured
473: asymmetry (major/minor axial ratio = 1.44, for $t_3\sim24$d) is
474: larger than what would be expected from that relation (1.0--1.1).
475:
476: Second, we can extrapolate backward in time to find the time of the inital
477: explosion; we find that it occurred around MJD $53527.4 \pm 1.9
478: $. Comparing this with the time of maximum V-band luminosity (MJD
479: $53531.7 \pm 0.1$; Poggiani 2006) we find a delay of $\sim 4 \pm 2$ days
480: between the onset of expansion and the maximum observed luminosity.
481: In this context we note the somewhat odd V-band point from June 3 ($m_V=14$)
482: reported in \citet{iauc8540}; this is probably not the pre-outburst
483: magnitude of the nova. However, this point would then indicate the
484: the outburst began sometime before 3 June, i.e. even earlier than
485: our $T_0$. We have no other indications of a pre-maximum halt
486: in this nova and note that such phenomena are usually associated with
487: slow novae. On the other hand, \citet{k02} have observed
488: a long pre-maximum halt in the rapidly evolving nova V463 Sct, indicating
489: that such expectations cannot be absolute.
490:
491: Third, we can use our measured expansion rate of $0.21 \pm 0.03$
492: ${\rm mas\,day^{-1}}$ together with spectroscopically determined expansion speeds to
493: find a geometric distance to this nova, independent of any photometric
494: biases. Unfortunately there appears to be little consensus as to
495: the spectroscopically determined expansion velocity, with available
496: estimates differing by a factor of three.
497: Poggiani (2006) measured the H$\alpha$ line on 31 July 2005 and finds a
498: Half-Width at Zero Intensity (HWZI) $\sim 2000 {\rm km\,s^{-1}}$, indicating an
499: expansion velocity in that range, while the O I 8446 line from the same paper
500: appears more consistent with a HWZI of $\sim 1500 {\rm km\,s^{-1}}$.
501: On the other hand, \citet{iauc8544} find an expansion rate of 700${\rm km\,s^{-1}}$ from
502: spectroscopy obtained on 11 June 2005, while \citet{iauc8640} find a
503: Full-Width at Zero Intensity $\sim 2600 {\rm km\,s^{-1}}$
504: from spectroscopy obtained on 14 Nov 2005.
505: We adopt an intermediate value of 1375 $\pm 500{\rm km\,s^{-1}}$,
506: being the average and standard deviation of the four values listed above.
507: In this context we note that the expected expansion rate for a nova with
508: $t_3 = 23.8$ days is $\sim1000 {\rm km\,s^{-1}}$ (McLaughlin 1960).
509:
510: A second difficulty, explained by \citet{wh00}, is that for a non-spherical
511: nova shell the unknown orientation of the shell in the plane of the sky
512: biases any distance determination that does not account for this
513: inclination effect, and in fact the proper inclination can only be determined
514: with spatially resolved spectroscopy. In our case, we measure an ellipse with an apparent
515: minor/major axis ratio of $\cos(\theta)=0.69$, and do not have access to
516: any information that would unambiguously constrain the inclination
517: of the nova shell; we therefore use the estimator recommended by
518: \citet{wh00}, viz. the arithmetic mean of the apparent major and minor
519: axis expansion rates. This estimator will on average yield a result
520: within a few percent of the true distance as long the orientation
521: of the nova shell isn't close to 0 or 90 degrees. We find a
522: distance to this nova of $8.9\pm3.6$ kpc.
523: This is consistent with the range derived by Poggiani (2006), based on
524: a number of photometric MMRD relations
525: available in the literature (7.3--11.3 kpc), but marginally inconsistent
526: with the value we derive in Section \ref{sec:phot}.
527: We caution, however, that the expansion rate may in fact change
528: with time; this could explain the range of measured rates, as well as bias our result.
529:
530: Finally, an important feature in our data is that the apparent expansion
531: ceased sometime after MJD 53545 (day 18). We note that behavior is
532: similar to what was seen in RS Oph \citep{lane07}, where the apparent
533: expansion appeared to reverse around day 20. It is possible that the apparent reversal is
534: not due to the transition from optically thick to optically thin emission
535: from the expanding fireball \citep{gehrz88b}, but rather that the wind
536: mass-loss rate (and hence effective optical depth of the source material)
537: is changing. However, the details of this process
538: are not easily modeled and we will explore this further in subsequent
539: papers.
540:
541: \section{Conclusion}
542:
543: We have used long-baseline near-IR interferometry to resolve the
544: classical nova V1663 Aql, starting $\sim 9$ days after outburst. We
545: measure an apparent expansion rate of $0.21 \pm 0.03$ ${\rm mas day^{-1}}$, which
546: can be combined with previously determined expansion velocities to
547: produce a distance estimate to the nova of
548: $8.9\pm3.6$ kpc; the precision is
549: limited by the precision of the available spectroscopic radial velocities.
550: Such a large distance is consistent with the large reddening ($E(B-V) \sim
551: 1.1$) determined from photometry, as well as with distances found from
552: MMRD relations. This represents only the third time a nova
553: has been resolved using optical/IR interferometry, the previous cases being
554: Nova V1974 Cyg 1992 \citep{q93} and more recently RS Oph \citep{mon06,lane07}.
555: We anticipate that further instrumental improvements, in particular
556: high spectral resolution interferometry such as that recently deployed
557: on the Keck Interferometer \citep{eis07}, may break the inclination degeneracy
558: and thus yield very precise expansion distances to these interesting objects.
559:
560:
561: \acknowledgements
562: We wish to acknowledge the extraordinary
563: observational efforts of K. Rykoski. Observations with PTI are made
564: possible through the efforts of the PTI Collaboration, which we
565: gratefully acknowledge. This research has made use of services from
566: the Michelson Science Center, California Institute of Technology,
567: http://msc.caltech.edu. Part of the work described in this paper was
568: performed at the Jet Propulsion Laboratory under contract with the
569: National Aeronautics and Space Administration. This research has made
570: use of the Simbad database, operated at CDS, Strasbourg, France, and
571: of data products from the Two Micron All Sky Survey, which is a joint
572: project of the University of Massachusetts and the Infrared Processing
573: and Analysis Center/California Institute of Technology, funded by the
574: NASA and the NSF. MWM is grateful for the
575: support of a Townes Fellowship.
576:
577: \begin{thebibliography}{}
578:
579: %
580:
581: \bibitem[Beichman et al.(1998)]{2mass}
582: Beichman, C.~A., Chester, T.~J., Skrutskie, M., Low, F.~J., \& Gillett, F.\ 1998, \pasp,
583: 110, 480
584:
585: \bibitem[Bode(2002)]{bode02}
586: Bode, M.~F.\ 2002, Classical Nova Explosions, 637, 497
587:
588: %
589: \bibitem[Boden et al.(2000)]{boden00}
590: Boden, A., Creech-Eakman, M., Queloz, D., 2000, \apj, 536, 880-890.
591:
592: \bibitem[Bryan \& Royer(1992)]{br92}
593:
594: \bibitem[Chesneau et al.(2007)]{ches07}
595: Chesneau, O., et al.\ 2007, \aap, 464, 119
596:
597: %
598: \bibitem[Colavita et al.(1999)]{col99}
599: Colavita, M. M.,et al.~ 1999, \apj, 510 , 505.
600:
601: \bibitem[Colavita(1999b)]{colavita99b}
602: Colavita, M. M., 1999, \pasp, 111, 111.
603:
604: \bibitem[Dennefeld et al.(2005)]{iauc8544}
605: Dennefeld, M., Ricquebourg, F., \& Damerdji, Y.\ 2005, \iaucirc, 8544, 1
606:
607: \bibitem[della Valle \& Livio(1995)]{dV95}
608: della Valle, M., \& Livio, M.\ 1995, \apj, 452, 704
609:
610: \bibitem[Eisner et al.(2003)]{eis03}
611: Eisner, J.~A., Lane, B.~F., Akeson, R.~L.,
612: Hillenbrand, L.~A., \& Sargent, A.~I.\ 2003, \apj, 588, 360
613:
614: \bibitem[Eisner et al.(2007)]{eis07} Eisner, J.~A., et al.\
615: 2007, \apjl, 654, L77
616:
617: \bibitem[Gehrz et al.(1998)]{gehrz98}
618: Gehrz, R.~D., Truran, J.~W., Williams, R.~E., \& Starrfield, S.\ 1998, \pasp, 110, 3
619:
620: \bibitem[Gehrz(1988)]{gehrz88b}
621: Gehrz, R.~D.\ 1988, \araa, 26, 377
622:
623: \bibitem[Hernanz(2005)]{hernanz05}
624: Hernanz, M. 2005, Astronomical Society of the Pacific Conference Series, 330, 265
625:
626: \bibitem[Kato et al.(2002)]{k02}
627: Kato, T., Uemura, M., Haseda, K., Yamaoka, H., Takamizawa, K., Fujii, M., \& Kiyota, S.\ 2002,
628: \pasj, 54, 1009
629:
630: \bibitem[Lane et al.(2005)]{lane05} Lane, B.~F., Retter, A.,
631: Thompson, R.~R., \& Eisner, J.~A.\ 2005, \apjl, 622, L137
632:
633: \bibitem[Lane et al.(2007)]{lane07}
634: Lane, B.~F., et al.\ 2007, \apj, 658, 520
635:
636:
637: \bibitem[Monet et al.(2003)]{usno-b1}
638: Monet, D.~G., et al. 2003, \aj, 125, 984
639:
640:
641: \bibitem[Monnier et al.(2006)]{mon06}
642: Monnier, J.~D., et al. 2006, \apjl, 647, L127
643:
644: % nova shells and tau
645: \bibitem[Ney \& Hatfield(1980)]{ney80}
646: Ney, E. P., Hatfield, B. F., 1980, \aj, 85, 1292.
647:
648: % nova review
649: \bibitem[Prialnik \& Kovetz(2005)]{pk05}
650: Prialnik, D., \& Kovetz, A.\ 2005, AIP Conf.~Proc.~797: Interacting Binaries: Accretion,
651: Evolution, and Outcomes, 797, 319
652:
653:
654: % analysis
655: \bibitem[Poggiani(2006)]{pog06}
656: Poggiani, R.\ 2006, Astronomische Nachrichten, 327, 895
657:
658: %discovery of Nova Aql 1663
659: \bibitem[Pojmanski \& Oksanen(2005)]{iauc8540}
660: Pojmanski, G., \& Oksanen, A.\ 2005, \iaucirc, 8540, 1
661:
662: % spectroscopy
663: \bibitem[Puetter et al.(2005)]{iauc8640}
664: Puetter, R.~C., Rudy, R.~J., Lynch, D.~K., Mazuk, S., Venturini, C.~C., Perry, R.~B., \& Walp,
665: B.\ 2005, \iaucirc, 8640, 2
666:
667: %TMS
668: \bibitem[Thompson et al.(2001)]{tms01}
669: Thompson, A.~R., Moran, J.~M., Swenson, G.~W., 2001, Interferometry and
670: Synthesis in radio Astronomy, (2nd ed; New York, NY : Wiley)
671:
672: % MkIII of V1974Cyg
673: \bibitem[Quirrenbach et al.(1993)]{q93}
674: Quirrenbach, A., Elias, N.~M., Mozurkewich, D., Armstrong, J.~T., Buscher, D.~F., \& Hummel,
675: C.~A.\ 1993, \aj, 106, 1118
676:
677: % hd 188310 binary
678: \bibitem[Scardia et al.(2000)]{scardi00} Scardia, M., Prieur,
J.-L., Aristidi, E., \& Koechlin, L.\ 2000, \apjs, 131, 561
679:
680: \bibitem[van den Bergh \& Younger(1987)]{vdB87} van den
681: Bergh, S., \& Younger, P.~F.\ 1987, \aaps, 70, 125
682:
683: \bibitem[Wade et al.(2000)]{wh00}
684: Wade, R.~A., Harlow, J.~J.~B., \& Ciardullo, R.\ 2000, \pasp, 112, 614
685:
686:
687: \end{thebibliography}
688:
689:
690: \end{document}
691:
692:
693:
694:
695:
696:
697: