1: %\documentstyle[12pt,aasms4]{article}
2: \documentstyle[emulateapj,epsf,apjfonts]{article}
3:
4: \received{}
5: \accepted{31 October 2000}
6: %\journalid{337}{15 January 1989}
7: %\articleid{11}{14}
8:
9: \slugcomment{To appear in The Astronomical Journal, February 2001}
10:
11: \lefthead{Hrivnak et al.}
12: \righthead{UV Study of OO Aql}
13:
14: \begin{document}
15:
16: \title{An Ultraviolet Study of the Short-Period Binary OO Aquilae}
17:
18: \author{Bruce J. Hrivnak\altaffilmark{1}}
19: \affil{Department of Physics and Astronomy, Valparaiso University,
20: Valparaiso, IN 46383; bruce.hrivnak@valpo.edu}
21:
22: \author{Edward F. Guinan\altaffilmark{1}, Laurence E. DeWarf, Ignasi Ribas}
23: \affil{Department of Astronomy and Astrophysics, Villanova University,
24: Villanova, PA 19085; edward.guinan@villanova.edu,
25: laurence.dewarf@villanova.edu, iribas@ast.villanova.edu}
26:
27: \altaffiltext{1}{Guest Observer, {\it International Ultraviolet Explorer}
28: satellite}
29:
30: \begin{abstract}
31:
32: OO Aql is a rare W UMa-type eclipsing binary in which the two solar-type stars
33: may have only recently evolved into contact. The binary has an unusually high
34: mass ratio (0.84), and a relatively long orbital period (0$\fd$506) for it
35: spectral type (mid-G). Twelve ultraviolet spectra of OO Aql were obtained in
36: 1988 with the {\it IUE} satellite, including a series of consecutive
37: observations that cover nearly a complete orbital cycle. Chromospheric
38: activity is studied by means of the Mg~{\sc ii} h$+$k emission at 2800 {\AA}.
39: The Mg~{\sc ii} emission is found to vary, even when the emission is normalized
40: to the adjacent continuum flux. This variation may be correlated with orbital
41: phase in the 1988 observations. It also appears that the normalized Mg~{\sc
42: ii} emission varies with time, as seen in spectra obtained at two different
43: epochs in 1988 and when compared with two spectra obtained several years
44: earlier. The level of chromospheric activity in OO Aql is less than that of
45: other W UMa-type binaries of similar colors, but this is attributed to its
46: early stage of contact binary evolution. Ultraviolet light curves were
47: composed from measurements of the ultraviolet continuum in the spectra. These
48: were analyzed along with visible light curves of OO Aql to determine the system
49: parameters. The large wavelength range in the light curves enabled a
50: well-constrained fit to a cool spot in the system.
51: \end{abstract}
52:
53: \keywords{binaries: eclipsing -- stars: activity --
54: stars: chromospheres -- stars: individual (OO Aquilae)}
55:
56:
57: \section{Introduction}
58:
59: The eclipsing binary star system OO Aquilae (HD 187183, $V_{\rm max}=9.2$,
60: $B-V=+0.76$, mid-G) is an unusual W UMa-type binary because it possesses a mass
61: ratio near unity (0.84; Hrivnak 1989). This suggests that the components have
62: only recently (on astronomical timescales) come into contact, since once in
63: contact the stars are expected to evolve toward low mass ratio (Webbink 1976;
64: Vilhu 1982).
65: The orbital period of 0$\fd$506 is long compared to other G spectral types
66: stars with W UMa-type light curves. This results in an unusually large angular
67: momentum for a cool W UMa-type system, again suggesting that it has only
68: recently evolved into contact (Mochnacki 1981). Thus OO Aql represents a rare,
69: transient phase in the evolution of contact binary star systems. On the basis
70: of his modern radial velocity study using the cross-correlation technique,
71: Hrivnak (1989) carried out a consistent analysis of both the light and velocity
72: curves of OO Aql. He determined the absolute parameters of the two stars with
73: high precision (M$_1$=1.04$\pm$0.02 M$_{\sun}$, R$_1$=1.39$\pm$0.02 R$_{\sun}$,
74: M$_2$=0.88$\pm$0.02 M$_{\sun}$, R$_2$=1.29$\pm$0.02 R$_{\sun}$). By comparing
75: its properties with stellar models, Hrivnak presented evidence that the stars
76: have an age of $\sim$8 Gyr. Thus OO Aql
77: consists of a pair of mid-G stars in contact, which appear to be somewhat more
78: evolved than the Sun. It falls into the subclass of A-type W UMa-type
79: binaries, in which the more massive component is eclipsed at primary minimum;
80: this is unusual among the cooler (G and K spectral type) W UMa-type binaries.
81:
82: W UMa-type binaries of G and K spectral type are known to display a high level
83: of chromospheric activity, as shown in the study of Rucinski (1985). Most
84: contact binaries of G$-$K spectral types have periods of 0$\fd$25 to 0$\fd$35
85: and mass ratios of 0.3 to 0.5. Thus OO Aql is clearly unusual. Another
86: contact binary with a large mass ratio (0.94) is VZ Psc (P=0$\fd$26, K2$-$5).
87: This system displays a high level of chromospheric activity, with strong and
88: variable Mg~{\sc ii} h+k and Ca~{\sc ii} H+K emission (Hrivnak, Guinan, \& Lu 1995). This
89: suggests that OO Aql might also have variable chromospheric activity. In OO
90: Aql, Ca~{\sc ii} emission is not evident in medium-resolution spectra, although there
91: may be infilling of the broad absorption profiles (Hrivnak 1989). However,
92: this apparent absence is most probably the result of a contrast effect, since
93: OO Aql is hotter and has a much higher continuum in this spectral region
94: (3900$-$4000 {\AA}) than does VZ Psc. Since the continuum is lower in the
95: ultraviolet, the Mg~{\sc ii} h+k emission at 2800 {\AA} provides a better
96: opportunity to measure the chromospheric activity in OO Aql. A single
97: observation of OO Aql made with the low-dispersion LWR camera of the {\it
98: International Ultraviolet Explorer} ({\it IUE}) satellite by Rucinski (1985) in
99: 1984 indicated an unusually low Mg~{\sc ii} emission level when compared with
100: other W UMa-type systems. No x-ray detections, which would measure the coronal
101: activity, have been reported for OO Aql, not even in the ROSAT Faint Source
102: Catalogue (Voges et al. 2000). However, it has been measured as a variable
103: radio source. It was one of only a few systems detected in a 3.6-cm radio
104: continuum survey of W UMa-type binaries, but was not detected at similar
105: sensitivity when observed two years later (Rucinski 1995). With this
106: background of well-determined but unusual properties (compared with other W
107: UMa-type binary systems), we were motivated to carry out an ultraviolet study
108: of OO Aql with the {\it IUE} satellite during a complete orbital cycle. Our
109: goal was to investigate the level of its Mg~{\sc ii} emission and to search for
110: phase-related and time-related variations. As a by-product of these spectra,
111: we also obtained ultraviolet light curves. In this paper we present the
112: results of this study.
113:
114: \section{{\it IUE} Observations and Reductions}
115:
116: Observations of OO Aql were made with the {\it IUE} satellite on 1988 August 8
117: and September 30 (UT). Nine spectra were obtained on the first date, covering
118: 80\% of the orbit, and three on the second date, which fill in part of the
119: phase gap and also include a spectrum during primary minimum. The
120: long-wavelength primary (LWP) camera, which covers 2000$-$3200 {\AA}, was used
121: with the large aperture in the low-dispersion mode. This resulted in a
122: spectral resolution of $\sim$6 {\AA}. Exposure times ranged from 18 to 35 min,
123: and were varied with the phase of the light curve to give approximately
124: constant signal-to-noise ratios (except for the last observation of the first
125: date, which was exposed a bit longer). The observing log is listed in Table 1.
126: The observing times were converted to heliocentric Julian date (HJD), including
127: the heliocentric time corrections. The rising ultraviolet continuum with
128: numerous absorption blends is clearly seen in these spectra from 2400 to 3000
129: {\AA}. In each case Mg~{\sc ii} h+k emission is seen rising in the middle of
130: the broad Mg~{\sc ii} (photospheric) absorption feature. An example of the LWP
131: spectrum of OO Aql in this region is shown in Figure 1.
132:
133: The primary goal of the observations was to obtain good spectra in the region
134: around the chromospheric Mg~{\sc ii} h+k emission lines at 2800 {\AA}. At this
135: low dispersion, the Mg~{\sc ii} doublet (2795.5, 2802.7 {\AA}) cannot be
136: resolved for these broad-lined stars. Examples of the Mg~{\sc ii} emission
137: profiles from spectra taken on the two observing dates are shown in Figure 2.
138: To measure the strength of the Mg~{\sc ii} emission, one needs to take into
139: account the underlying Mg~{\sc ii} absorption profile. One possible approach
140: to this would be to obtain a similar {\it IUE} spectrum of a mid-G dwarf and
141: broaden it as necessary to match the spectrum of the binary. A subtraction of
142: normalized spectra would then yield a measure of the Mg~{\sc ii} emission.
143: However, it would be difficult to obtain a spectrum which is an exact match for
144: OO Aql. Instead, since the absorption profiles are clearly seen except for the
145: central emission, one can extrapolate the wings to determine the central
146: absorption profile in each spectrum with consistency and reasonable accuracy.
147: This second method was used to determine the underlying absorption profile, and
148: by subtraction the emission strengths were determined. Reductions and
149: measurements were made using the standard NEWSIPS reduction program, but
150: including the new flux calibration of Massa \& Fitzpatrick (2000). The
151: measured values of the emission strengths are listed as f(Mg~{\sc ii}) in Table
152: 1.
153:
154: Several measurements of the continuum flux density were also made from each LWP
155: spectrum. These provide ultraviolet light curves, and also can be used to
156: investigate variations of the Mg~{\sc ii} emission relative to the ultraviolet
157: continuum. To investigate the ultraviolet light curves, the total ultraviolet
158: flux densities were measured in intervals of 50 {\AA} width, centered at 2575,
159: 2675, and 2975 {\AA}. These are spectral regions without dominant
160: absorption features or reseauxs. Also measured were two regions of 20 {\AA}
161: width, 2760$-$2780 {\AA} and 2820$-$2840 {\AA}, centered around the Mg~{\sc ii}
162: feature. These were later used to determine the Mg~{\sc ii} strength relative to the
163: adjacent continuum. All of these measurements are also listed in Table 1.
164:
165: In addition to obtaining ultraviolet light curves from the flux-calibrated
166: spectra, one can also use the fine error sensor (FES) on board the satellite to
167: obtain a visual light curve. To accomplish this, FES observations of OO Aql
168: were made immediately before and after most spectra. The observations each
169: consisted of two successive measurements with sample times of about 10 sec
170: each, which were averaged. To compensate for possible drifts in the
171: sensitivity of the FES with time, several observations were also made of a
172: nearby comparison star, HD 186656 ($V=8.70$, $B-V=+0.49$). Differential FES
173: magnitudes, transformed approximately to the standard V system using the
174: calibration of Imhoff \& Wasatonic (1986), are listed in Table 2. The
175: estimated uncertainty in the differential V(FES) magnitudes is $\pm$0.015 mag.
176: For a star of $T_{\rm eff}=6000$ K, the effective wavelength of the broad
177: bandpass of the FES is 5500 {\AA} (Imhoff 1989).
178:
179: \section{Analysis of the Ultraviolet and Visible Light Curves}
180:
181: In Figure 3, the visible (FES) and two ultraviolet light curves of OO Aql are
182: plotted. The V(FES) light curve shows the typical shape of a W UMa-type binary,
183: with continuously varying light and curvature in the maxima.
184: The orbital phases in Tables 1 and 2 and Figure 3 were calculated using
185: the results of the period study of this system by Demircan \& Gurol (1996).
186: Their linear ephemeris was used:
187: \begin{equation}
188: \mbox{Min I (HJD)} = 2438239.720 + 0.50678830\;E
189: \end{equation}
190: Their sinusoidal ephemeris gave similar results for the times of these
191: {\it IUE} observations, producing a phase difference of only $-0\fp002$.
192:
193: The ultraviolet light curves are all very similar to one another. The
194: measurements in each wavelength region were normalized to their maximum
195: value and converted to the magnitude scale. As examples, the light
196: curves at 2675 and 2975~\AA\ are shown in Figure 3. The magnitude scale of the
197: ultraviolet light curves has been arbitrarily shifted in this figure for
198: display purposes.
199:
200: The shape of a light curve in systems like OO Aql is determined by its orbital
201: and physical properties, which stay constant over rather long timescales, and
202: also by transient phenomena, that vary over timescales as short as a few
203: months, like starspots. As can be seen in Figure 3, the number of measurements
204: in both the FES and the ultraviolet light curves is unfortunately too small to
205: attempt a complete (orbital, physical, and radiative properties) solution of
206: the light curves. Several tests showed that the phase coverage is not dense
207: enough (notice the lack of coverage of the minima) for reaching
208: convergence. To circumvent the problem, we adopt the orbital and physical
209: properties of the system from the solution of light curves with better phase
210: coverage, and then check for consistency with our observations.
211:
212: Good-quality differential light curves in blue ($B$) and yellow ($V$) passbands
213: were published by Binnendijk (1968) from 1966 and by Lafta \& Grainger (1985)
214: from 1982. Binnendijk's measurements have excellent phase coverage, while Lafta
215: \& Grainger's observations show a small gap at orbital phases $0.85-0.95$.
216: They all show the secondary minimum to be nearly as deep as primary, and the
217: first maximum at phase 0.25 to be somewhat brighter than the second maximum.
218: We decided to consider both datasets for further analysis because they are both
219: of good quality. The analyses for the data sets, however, had to be done
220: independently, since the observations were made 16 years apart and the effects
221: of transient phenomena on the light curves, such as starspots, do not allow a
222: simultaneous and self-consistent analysis.
223:
224: The light curves were fitted using an improved version of the Wilson-Devinney
225: program (Wilson \& Devinney 1971; hereafter WD) that includes an atmosphere
226: model (Kurucz's ATLAS9) routine developed by Milone, Stagg, \& Kurucz (1992)
227: for the computation of the stellar radiative parameters. A contact
228: configuration (WD in mode 3) was chosen when running the models. The
229: bolometric albedo and the gravity brightening coefficients were set to values
230: of 0.5 and 0.32, respectively, as usually adopted for the convective envelopes
231: of these late-type stars. The mass ratio ($q\equiv M_2/M_1$) was
232: fixed to the spectroscopic value of $q=0.843$ and the temperature of the
233: primary star was set to 5700~K (Hrivnak 1989). The WD program was run in an
234: iterative mode, in which differential corrections were made to the fitted
235: parameters until convergence was achieved. Several sets of starting
236: parameters were used in order to explore the full-extent of the parameter
237: space and also to make a realistic estimation of the uncertainties.
238: Simultaneous solutions for the $B$ and $V$ light curves were run for both
239: datasets.
240:
241: The fitted orbital and physical parameters in the light curve analyses were the
242: orbital inclination ($i$), the temperature of the secondary ($T_2$), the
243: gravitational potential of the common surface ($\Omega_{1,2}$), the
244: luminosity of the primary ($L_1$), and a phase offset. In addition, a
245: dark spot on one of the components was included in the modeling to fit
246: the asymmetry of the light curves. Both in the case of Binnendijk's and Lafta
247: \& Grainger's light curves, similarly good fits were obtained with a spot
248: about 300~K cooler than the photosphere located on either the primary or
249: secondary components. This inability to distinguish between these two
250: options is not surprising since the difference in wavelength is small between
251: the {\it B} and {\it V} light curves and since the spot is visible at
252: an orbital phase when both components are in view.
253: Since the solution to the ultraviolet light curves that we discuss later
254: clearly indicates that a cool spot is present on the primary component,
255: we have also assigned the cool spot to the primary component in the
256: modeling of these visible light curves.
257: The latitude, the longitude, and the size of the spot were left free in the
258: differential corrections algorithm. Convergence for the full set of parameters
259: was reached very quickly for both data sets primarily because of the good phase
260: coverage.
261:
262: The orbital and physical parameters of OO Aql are not expected to change
263: between the two observing epochs. Therefore it is pleasantly reassuring to
264: find that the best-fitting orbital and physical properties obtained
265: from the analyses of the two independent datasets are found to agree within
266: their uncertainties. Only the spot parameters, as one may expect, differ between
267: the two light curve solutions. In both cases, the latitude\footnote{We are
268: listing the latitude rather than the co-latitude,
269: which is used internally in the WD program.} of the spot in
270: the secondary component is found to be around $+40^\circ$, but the spot
271: longitudes and sizes are different: a longitude of $60^\circ$ and spot
272: radius of $30^\circ$ for Binnendijk's light curves, and a longitude of
273: $90^\circ$ and a spot radius of
274: $65^\circ$ for Lafta \& Grainger's observations.
275: In both cases, these result in a second maximum that is fainter than the first.
276: The fits of the model light curves to the observed normal points are displayed
277: in Figure 4.
278: The fits are quite good; only slight systematic deviations among the
279: observations and the synthetic light curves are seen near phase 0.8 in the
280: light curves of Lafta \& Grainger (both $B$ and $V$). These may be
281: caused by the simplified spot model (circular cool spot with uniform
282: temperature) that the WD program uses. The final r.m.s. residuals of the fit
283: were about 0.011~mag and 0.012~mag for the light curves of Binnendijk and
284: Lafta \& Grainger, respectively, which corroborates their very similar quality.
285:
286: We decided to adopt the best-fitting WD solution to Binnendijk's light curves
287: as our final values, as listed in Table 3. The reason for this is twofold. On
288: the one hand, no systematic deviations are present in the $O-C$ residuals. On
289: the other hand, the spot size, and therefore its influence on the light curve,
290: is smaller in Binnendijk's light curves, which makes the determination of the
291: intrinsic properties of the stars more reliable. In any case, as we have
292: already noted, the parameters yielded by the analysis of Lafta \& Grainger's
293: light curves fall within the uncertainties quoted in Table 3.
294:
295: A detailed study of the light curves of Binnendijk has already been carried out
296: by Hrivnak (1989). We decided, however, to undertake a new analysis chiefly
297: because of the improved light curve fitting program presently available. The
298: new modeling has a more sophisticated treatment of the stellar atmospheres with
299: ATLAS9 Kurucz models (limb darkening, reflection, etc.) and it also includes a
300: starspot to account for the asymmetry of the light curves. Despite the more
301: accurate treatment, the values of the best-fitting parameters presented in
302: Table 3 are in agreement with those already published by Hrivnak (1989),
303: with the only significant difference being a higher inclination ($90^\circ$)
304: in his solution.
305: A summary of the physical properties of the two components derived with our
306: new solution is presented in Table 4.
307: These are in agreement with the results of the earlier analysis by
308: Hrivnak (1989)\footnote{A
309: computational error exists in Hrivnak's (1989) values for
310: log~$\it g$ (cgs); they should be 4.18 and 4.17 for the primary and
311: secondary components, respectively.}.
312:
313: Once the orbital and physical properties of the system were determined, the
314: sparsely-covered FES and ultraviolet light curves could be fit while fixing most
315: of the parameters. Only the zero-point of the magnitude scale in each of the
316: light curves and the spot parameters were left free in the differential
317: corrections algorithm. A total of five light curves (FES and four ultraviolet,
318: including one at $\overline{\lambda}$ = 2800 \AA\ formed from averaging the
319: continuum fluxes in the two narrower regions adjacent to the Mg~{\sc ii} feature)
320: were fitted simultaneously. Interestingly, despite the poor phase-coverage,
321: convergence of the spot parameters was achieved rapidly and with a high degree
322: of confidence. This can be explained in terms of the large wavelength range
323: (5500 to 2575 \AA) in which the effects of the spot are critical (the contrast
324: between the spot and the surrounding photosphere increases strongly from the
325: optical to the ultraviolet). The solution indicates a spot on the primary
326: component $\sim250$~K cooler than the photosphere located at a latitude of
327: $+45^\circ$, a longitude of $5^\circ$, and with a radius of $40^\circ$. Thus
328: the spot is facing the observer around primary minimum, although it is
329: partially eclipsed.
330:
331: Examples of the best-fitting synthetic light curves are shown together with
332: the observations in Figure 3. As can be seen, the fits are very good at all
333: wavelengths and do not show any noticeable systematics, indicating that our
334: physical model is essentially correct. The r.m.s. residuals are 0.032~mag,
335: 0.041~mag, 0.067~mag, 0.045~mag, and 0.049~mag, for the light curves at
336: 5500~\AA\ (FES), 2975~\AA, 2800~\AA, 2675~\AA, and 2575~\AA, respectively.
337: The shape of our best-fitting light curves is consistent with the observations
338: of Demircan, Derman, \& Ekmekci (1991) made around this time (1988--1990),
339: i.e. maxima of similar heights and slightly more unequal eclipse depths.
340: These contrast with the brighter maximum at phase 0.25 and the slightly more
341: similar eclipse depths seen in the light curves of Binnendijk and Lafta \&
342: Grainger. The results indicate that the starspot (or major spot group), once
343: centered around phase 0.75, had moved close to the phase of the primary minimum
344: when our observations were made.
345:
346: In Figure 4 a color curve formed from the $\lambda$2675 ultraviolet curve and
347: the V(FES) light curve is also plotted. The data points represent the
348: difference between the observed ultraviolet magnitudes and the magnitudes
349: interpolated from the best-fitting synthetic V(FES) light curve at the proper
350: phases. Note the color dependence in the phases around the primary eclipse
351: despite the similar effective temperature of both components. This color
352: dependence can be attributed to the increased effect of limb darkening when
353: the smaller star transits the larger one (limbs of both stars visible) and to
354: the presence of the dark spot partially in view near the phase of the primary
355: minimum, in addition to the gravity darkening which reddens both minima.
356:
357: \section{Analysis of the Mg \sc{ii} Emission}
358:
359: The values of the Mg~{\sc ii} flux density, as determined by measurements of
360: the emission above the underlying absorption profiles and listed in Table 1,
361: indicate that the flux varies. These are plotted in the top portion of Figure
362: 5 as a function of orbital phase. It can be seen that there is a clear phase
363: dependence in the Mg~{\sc ii} strength, with lower values corresponding to
364: phases of minimum light (eclipses) in the light curve. Another way to
365: investigate the presence of a variation in the emission is to simply measure
366: the total flux within the Mg~{\sc ii} spectral region, relative to the adjacent
367: continuum. This was also carried out, and shows a similar phase-dependent
368: variation in Mg~{\sc ii} emission.
369:
370: If the Mg~{\sc ii} h+k emission originates from many active regions distributed
371: around the atmosphere of the two components, then one would expect the net
372: emission to vary with the projected cross-sectional area of the two stars,
373: which varies with the orbital phase. A first-order correction for this effect
374: can be made by normalizing the Mg~{\sc ii} emission relative to the light
375: curve. The two adjacent continuum bands measured from 2760$-$2780 {\AA} and
376: 2820$-$2840 {\AA} were used for this purpose, and the Mg~{\sc ii} emission flux
377: relative to the average of these was calculated. This is listed in Table 1 as
378: R$_{uv}$, where R$_{uv}$ = f(Mg~{\sc ii})/f(uv cont). Rucinski (1985), in his
379: study of the Mg~{\sc ii} emission in W UMa-type binaries, normalized the Mg
380: emission relative to the bolometric flux, R$_{bol}$ = f(Mg~{\sc ii})/f(bol),
381: where
382: \begin{equation}
383: {\rm f(bol)} = 2.7\times10^{-5} \: 10^{-0.4\;{\rm m(bol)}} \;
384: \mbox{erg cm$^{-2}$ s$^{-1}$},
385: \end{equation}
386: and m(bol) = V + BC. We also calculated this ratio. The value of V is
387: interpolated from the synthetic V(FES) light curve,
388: and BC was estimated from the colors
389: of the system, $(B-V)=+0.76$ (Eggen 1967), and reddening, $E(B-V)=+0.03$
390: (Hrivnak 1989), to be $-$0.15 (Flower 1996).
391: Both of these normalized measures of the Mg~{\sc ii} emission show similar effects,
392: and in Figure 5 is plotted R$_{uv}$ versus orbital phase.
393:
394: A variation is seen in the normalized Mg~{\sc ii} emission R$_{uv}$. Initial
395: inspection of the bottom portion of Figure 5 suggested a phase dependence,
396: since the values based on
397: the five observations in the phase interval 0.80$-$0.05 are larger than those
398: of the seven observations in the phase interval 0.05$-$0.70 by 36$\%$ on
399: average. However, these may instead represent a variation with time, with the
400: average R$_{uv}$ value of the three observations from September 1988 larger
401: by 35$\%$ than the average R$_{uv}$ value of the nine observations from
402: August 1988. The highest Mg~{\sc ii} flux measurement, at phase$=$0.81, may be the
403: result of chromospheric flaring, which gradually decreased or rotated partially
404: out of view when the following two observations were made. Also included in
405: Table 1 and Figure 4 are the results from single observations of OO Aql made
406: four and six years earlier. These spectra were obtained from the {\it IUE}
407: archives and were reduced and measured in the same way as the others.
408: The one from 1982 has a normalized Mg~{\sc ii} flux similar to those of the 1988
409: observations, while the one from 1984 (Rucinski) is lower.
410: We conclude that the normalized Mg~{\sc ii} flux varies with time, and perhaps it
411: also varied with phase during our 1988 observations.
412: If the normalized Mg~{\sc ii} variation is indeed related to phase in the 1988
413: observations, one can investigate its relationship with the visibility of
414: the spot. If we assume that the highest emission value at phase$=0.81$
415: represents a flare and simply look at the other four higher emission
416: measurements in the bottom of Figure 5, the results seem to be ambiguous.
417: The normalized Mg~{\sc ii} emission is high as the spot comes into view
418: (phase $=0.85-0.95$), but the spot is partially eclipsed during the high Mg~{\sc ii}
419: observation at phase$=0.02$, and the observed emission is not high in the
420: phase interval $0.05-0.15$.
421:
422:
423: \section{Summary and Conclusions}
424:
425: Ultraviolet observations were obtained for the short-period contact binary
426: OO Aql on two dates in 1988, including nine consecutive observations that
427: covered most of one orbital cycle. Mg~{\sc ii} h+k emission is seen in each of
428: the spectra. These were measured for emission strength, and also
429: measurements were made of the continuum at several wavelength intervals
430: and used to compose ultraviolet light curves for the binary.
431: An analysis of the Mg~{\sc ii} emission indicates that real variations in the
432: strength of the feature exist for OO Aql, even when normalized
433: to the surrounding ultraviolet continuum. These may be correlated with the
434: orbital phase, or they may instead indicate overall changes in the
435: chromospheric activity level with time.
436:
437: Previous studies of phase-related activity in W UMa-type binaries have yielded
438: similar results. Eaton (1986) found only a slight phase dependence in the
439: normalize Mg~{\sc ii} emission for SW Lac (P=0$\fd$32, K0), another contact binary
440: with a relatively large mass ratio (0.73, Hrivnak 1992). The study of VZ Psc
441: (Hrivnak et al. 1995) revealed variations in the normalized Mg~{\sc ii} emission
442: that were not phase dependent, but that seemed to vary in time.
443: VW Cep (P=0$\fd$27, G5V+G8V) shows primarily
444: time-dependent variations in chromospheric Mg~{\sc ii} emission and also in the
445: transition region emission flux (Guinan \& Gim\'enez 1993). Thus variations in
446: the normalized Mg~{\sc ii} emission are common in W UMa-type binaries, but these are
447: typically not phase dependent, and instead are probably due to variations in
448: chromospheric activity arising from possible flaring.
449:
450: General trends have been found for W UMa-type and other short-period binaries
451: in studies of chromospheric activity versus color or orbital period. A summary
452: for W UMa-type binaries has been published by Rucinski (1985), which included
453: his observation of OO Aql. While these more recent observations show
454: OO Aql to have a somewhat higher level of chromospheric activity than at the
455: time of his 1984 observation, the basic interpretation is the same; OO Aql has
456: a much lower level of relative activity (by a factor of 2 relative to f(bol))
457: than do other contact binaries with similar colors. However, this is
458: presumably due to the recent evolution of its components into contact.
459: While in contact, OO Aql is expected to transfer mass from the less-massive
460: secondary to the more-massive primary component (Webbink 1976), reducing the
461: mass ratio and causing the system to evolve to evolve to an F spectral type.
462: The level of chromospheric activity in OO Aql is in fact comparable to that
463: in W UMa-type systems with the colors of F stars.
464:
465: Ultraviolet light curves are presented here for the first time for OO Aql.
466: They show a similar shape and amplitude as the visible light curves. Since the
467: two stars have similar temperatures, there is expected
468: to be little change in the light curves with wavelength.
469: Synthetic light curve solutions were carried out for $B$ and $V$ light
470: curves for two different epochs, and similar orbital and physical parameters
471: were obtained. These were used to model the new ultraviolet light curves,
472: and good fits were obtained.
473: The combination of ultraviolet and visible light curves led to a
474: well-determined fit for a cool spot on the primary component. It appears
475: that the light curves at the different epochs can each be fit by the
476: same orbital and physical parameters, with the presence of one major cool
477: spot on the primary component that varies in longitude and size with time.
478: Jeong et al. (1994) carried out a somewhat similar ultraviolet and
479: visible light curve analysis for SW Lac, also including a single cool spot.
480:
481: The high mass ratio and relatively long orbital period lead to a relatively
482: large orbital angular momentum for OO Aql, which is larger than that of most
483: contact binaries but less than that of the short-period detached binaries with
484: G-type components, such as ER Vul and UV Leo. It is likely that the detached
485: systems such as ER Vul (P=0$\fd$70, M$_1$=0.96 M$_{\sun}$, M$_2$=0.89
486: M$_{\sun}$, Budding \& Zeilik 1987) and UV Leo (P=0$\fd$60, M$_1$=0.99
487: M$_{\sun}$, M$_2$=0.92 M$_{\sun}$, Popper 1980) will evolve into contact
488: systems like OO Aql (P=0$\fd$51, M$_1$=1.05 M$_{\sun}$, M$_2$=0.88 M$_{\sun}$)
489: through angular momentum loss by magnetic breaking (Mochnacki 1981; Vilhu
490: 1982). However, ER Vul has a much higher level of variable starspot and
491: chromospheric activity than that seen in OO Aql, even though its rotational
492: period is longer and its spectral type earlier than those of OO Aql.
493: If OO Aql is typical of a system of solar mass stars that have recently
494: evolved into contact, then the previous high level of activity must be
495: significantly reduced in the contact phase. Subsequent evolution while in
496: contact is expected to decrease the mass ratio, and if this is so, then a
497: binary like OO Aql will gradually evolve from G to F spectral type as the
498: mass is transferred from the secondary to the primary component. As shown by
499: Rucinski (1985), the F spectral type systems show a lower activity, consistent
500: with what is now seen in OO Aql.
501:
502: Thus this study of OO Aql has presented an opportunity to learn more about the
503: properties of a close binary in a rather short-lived stage of its evolution,
504: shortly after the two stars have evolved into contact. If OO Aql is typical
505: of this stage, then it appears that the level of chromospheric activity must
506: be reduced rather quickly upon entering the contact configuration, with the
507: contact binary continuing to maintain this lower level with some smaller
508: variations over time.
509:
510:
511: \acknowledgments
512:
513: The assistance of the staff of the {\it IUE} in making these observations and
514: of the RDAF staff at the Goddard Space Flight Center in carrying out the
515: preliminary reductions is much appreciated. We thank John Pritchard for
516: providing the model atmosphere files for the ultraviolet passbands. This
517: research was supported in part by grants from NASA to B.J.H. (NAG 5-964, NAG
518: 5-2645) and to E.F.G. (NAG 5-2160); these are gratefully acknowledged. I.R.
519: acknowledges the Catalan Regional Government (CIRIT) for financial support
520: through a postdoctoral Fulbright fellowship.
521:
522: \begin{thebibliography}{}
523:
524: \bibitem[]{b1}
525: Binnendijk, L. 1968, AJ, 73, 32
526: \bibitem[]{b2}
527: Budding, E., \& Zeilik, M. 1987, ApJ, 319, 827
528: \bibitem[]{b3}
529: Demircan, O., Derman, E., \& Ekmekci F. 1991, IAU Inf. Bull. Var. Stars, 3551
530: \bibitem[]{b4}
531: Demircan, O., \& Gurol, B. 1996, A\&AS, 115, 333
532: \bibitem[]{b5}
533: Eaton, J. A. 1986, Acta Astr., 36, 79
534: \bibitem[]{b6}
535: Eggen, O. J. 1967, Mem. RAS, 70, 101
536: \bibitem[]{b7}
537: Flower, P. J. 1996, ApJ, 469, 355
538: \bibitem[]{b8}
539: Guinan, E. F., \& Gim\'enez, A. 1993, in The Realm of Interacting Binary
540: Stars, ed. J. Sahade, G. E. McCluskey, Jr., Y. Kondo (Dordrecht: Kluwer), 51
541: \bibitem[]{b9}
542: Hrivnak, B. J. 1989, ApJ, 340, 458
543: \bibitem[]{b10}
544: Hrivnak, B. J. 1992, Bull. AAS, 24, 686
545: \bibitem[]{b11}
546: Hrivnak, B. J., Guinan, E. F., \& Lu, W. 1995, ApJ, 455, 300
547: \bibitem[]{b12}
548: Imhoff, C. L. 1989, NASA IUE Newsletter, 40, 5
549: \bibitem[]{b13}
550: Imhoff, C. L., \& Wasatonic, R. 1986, NASA IUE Newsletter, 29, 45
551: \bibitem[]{b14}
552: Jeong, J. H., Kang, Y. W., Lee, W. B., \& Sung, E. C. 1994, ApJ, 421, 779
553: \bibitem[]{b15}
554: Lafta, S. L., \& Grainger, J. F. 1985, Ap\&SS, 114, 23
555: \bibitem[]{b16}
556: Massa, D. \& Fitzpatrick, E. L. 2000, ApJS, 126, 517
557: \bibitem[]{b17}
558: Milone, E. F., Stagg, C. R., \& Kurucz, R. L. 1992, ApJS, 79, 123
559: \bibitem[]{b18}
560: Mochnacki, S. W. 1981, ApJ, 245, 650
561: \bibitem[]{b19}
562: Popper, D. M. 1980, ARA\&A, 18, 115
563: \bibitem[]{b20}
564: Rucinski, S. M. 1985, MNRAS, 215, 615
565: \bibitem[]{b21}
566: Rucinski, S. M. 1995, AJ, 109, 2690
567: \bibitem[]{b22}
568: Vilhu, O. 1982, A\&A, 109, 17
569: \bibitem[]{b23}
570: Voges, W., Aschenbach, B., Boller, Th., et al. 2000, The ROSAT All-Sky
571: Survey Faint Source Catalogue, Max-Planck-Institut f\"ur Extraterrestrische
572: Physik, Garching (http://wave.xray.mpe.mpg.de/rosat/catalogues/rass-fsc/)
573: \bibitem[]{b24}
574: Webbink, R. F. 1976, ApJ, 209, 829
575: \bibitem[]{b25}
576: Wilson, R. E., \& Devinney, E. J. 1971, ApJ, 166, 605 (WD)
577: \end{thebibliography}
578:
579: \clearpage
580: \begin{deluxetable}{rrrcrrrrrrrr}
581: \tablewidth{0pt}
582: \tiny
583: \tablecaption{{\it IUE} Observing Log for OO Aquilae
584: \label{iue_log}}
585: \tablehead{\colhead{Image No.}
586: &\colhead{HJD}
587: &\colhead{Phase}
588: &\colhead{Exposure}
589: &\colhead{f(Mg~{\sc ii})\tablenotemark{a}}
590: &\colhead{f(2575)\tablenotemark{b}}
591: &\colhead{f(2675)\tablenotemark{b}}
592: &\colhead{f(2975)\tablenotemark{b}}
593: &\colhead{f(2770)\tablenotemark{b}}
594: &\colhead{f(2830)\tablenotemark{b}}
595: &\colhead{R$_{uv}$}
596: &\colhead{R$_{bol}$\tablenotemark{c}}\\
597: \colhead{}
598: &\colhead{(2,440,000+)}
599: &\colhead{}
600: &\colhead{Time(min)}
601: &\colhead{}
602: &\colhead{(50 {\AA})\tablenotemark{d}}
603: &\colhead{(50 {\AA})\tablenotemark{d}}
604: &\colhead{(50 {\AA})\tablenotemark{d}}
605: &\colhead{(20 {\AA})\tablenotemark{d}}
606: &\colhead{(20 {\AA})\tablenotemark{d}}
607: &\colhead{}
608: &\colhead{}}
609: \startdata
610: LWP 13801& 7381.601& 0.856& 25& 5.41& 1.60& 2.99& 6.32& 1.28& 1.58& 0.38&10.16\nl
611: LWP 13802& 7381.647& 0.946& 25& 3.53& 1.14& 1.79& 3.95& 0.72& 0.95& 0.42& 9.43\nl
612: LWP 13803& 7381.711& 0.073& 28& 3.64& 1.27& 2.32& 5.23& 0.99& 1.29& 0.32& 8.56\nl
613: LWP 13804& 7381.759& 0.168& 20& 5.01& 1.82& 3.30& 7.19& 1.33& 1.79& 0.32& 8.88\nl
614: LWP 13805& 7381.817& 0.283& 18& 5.82& 2.07& 3.65& 7.52& 1.50& 1.98& 0.33& 9.58\nl
615: LWP 13806& 7381.858& 0.362& 20& 4.17& 1.74& 3.33& 6.57& 1.29& 1.65& 0.28& 7.65\nl
616: LWP 13807& 7381.909& 0.463& 25& 2.82& 1.11& 1.91& 4.07& 0.79& 1.01& 0.31& 8.18\nl
617: LWP 13808& 7381.958& 0.561& 30& 3.98& 1.35& 2.52& 5.09& 1.09& 1.34& 0.33& 9.65\nl
618: LWP 13809& 7382.011& 0.664& 35& 5.45& 1.81& 3.54& 7.45& 1.50& 1.92& 0.32& 9.55\nl
619: LWP 14137& 7434.788& 0.805& 20& 9.05& 1.87& 3.53& 7.16& 1.49& 2.00& 0.52&15.50\nl
620: LWP 14138& 7434.851& 0.929& 20& 4.34& 1.20& 2.18& 4.88& 0.95& 1.30& 0.39&10.46\nl
621: LWP 14139& 7434.896& 0.018& 28& 2.94& 0.75& 1.54& 3.19& 0.58& 0.75& 0.44&10.24\nl
622: \tableline
623: LWR 13609& 5155.647& 0.580& 36& 4.09& 1.23& 2.30& 5.22& 0.96& 1.28& 0.37& 8.93\nl
624: LWP 03362& 5834.767& 0.626& 45& 3.91& 1.63& 3.34& 6.59& 1.34& 1.77& 0.25& 7.34\nl
625: \enddata
626: \tablenotetext{a}{Units of 10$^{-13}$ erg cm$^{-2}$ s$^{-1}$.}
627: \tablenotetext{b}{Units of 10$^{-12}$ erg cm$^{-2}$ s$^{-1}$.}
628: \tablenotetext{c}{Units of 10$^{-5}$.}
629: \tablenotetext{d}{Width of integrated flux density interval.}
630: \end{deluxetable}
631:
632: \begin{deluxetable}{rrr}
633: \tablewidth{0pt}
634: \tablecaption{{\it IUE} FES Light Curve of OO Aquilae (1988)
635: \label{fes}}
636: \tablehead{\colhead{HJD} & \colhead{Phase} & \colhead{$\Delta$V(FES)}\\
637: \colhead{(2,440,000+)}&\colhead{}&\colhead{}}
638: \startdata
639: 7381.5882& 0.830& +0.67\nl
640: 7381.6125& 0.878& +0.80\nl
641: 7381.6327& 0.918& +0.96\nl
642: 7381.6584& 0.969& +1.36\nl
643: 7381.6959& 0.043& +1.17\nl
644: 7381.7230& 0.096& +0.82\nl
645: 7381.7264& 0.103& +0.79\nl
646: 7381.7462& 0.142& +0.69\nl
647: 7381.7688& 0.187& +0.62\nl
648: 7381.7889& 0.226& +0.57\nl
649: 7381.8264& 0.300& +0.59\nl
650: 7381.8473& 0.342& +0.64\nl
651: 7381.8716& 0.390& +0.74\nl
652: 7381.8945& 0.435& +0.96\nl
653: 7381.9209& 0.487& +1.32\nl
654: 7381.9431& 0.531& +1.15\nl
655: 7381.9931& 0.629& +0.68\nl
656: 7382.0257& 0.694& +0.61\nl
657: 7434.8194& 0.867& +0.70\nl
658: 7434.8409& 0.909& +0.85\nl
659: 7434.8610& 0.949& +1.11\nl
660: 7434.8826& 0.991& +1.51\nl
661: \enddata
662: \end{deluxetable}
663:
664: \begin{deluxetable}{lrr}
665: \tablewidth{0pt}
666: \tablecaption{Light Curve Solution for OO~Aquilae \tablenotemark{a}
667: \label{lc_model}}
668: \tablehead{\colhead{Parameter}&\colhead{Value}}
669: \startdata
670: Inclination (deg) & $87.7\pm0.5$ \nl
671: $\Delta$T$\equiv T_1 - T_2$ (K) & $+20\pm25$ \nl
672: %$q\equiv\frac{M_2}{M_1}$ & $0.843$\rlap{\tablenotemark{b}}\nl
673: $q\equiv{M_2}$/${M_1}$ & $0.843$\rlap{\tablenotemark{b}}\nl
674: $({L_2}$/${L_1}$)$_{B}$ & $0.839\pm0.021$ \nl
675: $({L_2}$/${L_1}$)$_{V}$ & $0.844\pm0.018$ \nl
676: %$ \left. \frac{L_2}{L_1} \right|_{B}$ & $0.839\pm0.021$ \nl
677: %$ \left. \frac{L_2}{L_1} \right|_{V}$ & $0.844\pm0.018$ \nl
678: $\Omega_{1,2}$ & $3.388\pm0.024$ \nl
679: $f$ (fill-out factor) & $0.22\pm0.06$\rlap{\tablenotemark{c}}\nl
680: $r_1$ (volume) & $0.415\pm0.004$ \nl
681: $r_2$ (volume) & $0.385\pm0.004$ \nl
682: \enddata
683: \tablecomments{The uncertainties listed are standard deviations
684: derived from a sampling of the $\chi^2$ minimum.}
685: \tablenotetext{a}{Based on the solution obtained from the analysis of the
686: light curves of Binnendijk (1968).}
687: \tablenotetext{b}{Fixed at the spectroscopic value of Hrivnak (1989).}
688: \tablenotetext{c}{$f\equiv(\Omega_{1,2}-\Omega_{\rm inner})/
689: (\Omega_{\rm outer}-\Omega_{\rm inner})$.}
690: \end{deluxetable}
691:
692: \begin{deluxetable}{lrr}
693: \tablewidth{0pt}
694: \tablecaption{Physical Properties of the Components of OO~Aquilae
695: \label{star_properties}}
696: \tablehead{\colhead{}&\colhead{Primary}&\colhead{Secondary}}
697: \startdata
698: Mass (M$_{\odot}$) & $1.05\pm0.02$ & $0.88\pm0.02$ \nl
699: Radius (R$_{\odot}$) & $1.38\pm0.02$ & $1.28\pm0.02$ \nl
700: $\log g$ (cgs) & $4.18\pm0.01$ & $4.17\pm0.01$ \nl
701: $T_{\rm eff}$ (K) & $5700\pm300$ & $5680\pm300$ \nl
702: $\log (L/L_{\odot})$ & $0.26\pm0.09$ & $0.19\pm0.09$ \nl
703: $M_{\rm bol}$ (mag) & $4.11\pm0.22$ & $4.29\pm0.22$ \nl
704: $M_{\rm v}$ (mag) & $4.21\pm0.22$ & $4.39\pm0.22$ \nl
705: \enddata
706: \tablecomments{Adopted the radial velocity curve solution of Hrivnak (1989);
707: BC from Flower (1996).}
708: \end{deluxetable}
709:
710: \clearpage
711:
712: \begin{figure*}
713: \plotone{fig1.eps}
714: \figcaption[]
715: {A sample {\it IUE} spectrum of OO Aql, showing the continuum from
716: 2400 to 3000 {\AA} with the Mg~{\sc ii} emission marked.
717: \label{f1.eps}}
718: \end{figure*}
719:
720: \begin{figure*}
721: \plotone{fig2.eps}
722: \figcaption[]
723: {Spectra of OO Aql in the region of the Mg~{\sc ii} feature, obtained around
724: the same phase but on different dates
725: (LWP 13801, $0\fp85$; LWP 14137, $0\fp81$). Note the difference in the
726: emission strength between the two spectra.
727: \label{f2.eps}}
728: \end{figure*}
729:
730: \begin{figure*}
731: \plotone{fig3.eps}
732: \figcaption[]
733: {(upper panel) Visible (FES) and representative ultraviolet light curves of
734: OO Aql, and
735: (lower panel) an ultraviolet minus visible color curve formed from the
736: lower and upper light curves of the upper panel.
737: Different symbols are used to distinguish the different dates: 1988
738: Aug (filled circles), 1988 Sep (filled triangles). The magnitude scale of the
739: ultraviolet light curves has been arbitrarily shifted for display purposes.
740: \label{f3.eps}}
741: \end{figure*}
742:
743: \begin{figure*}
744: \plotone{fig4.eps}
745: \figcaption[]
746: {Observed normal-point, differential (variable $-$ comparison) B and V light
747: curves of OO Aql from the studies of Binnendijk (1968) for 1966 and Lafta \&
748: Grainger (1985) for 1982, fitted by our synthetic light curve solutions. (Note
749: that different comparison stars were used in the two studies.)
750: \label{f4.eps}}
751: \end{figure*}
752:
753: \begin{figure*}
754: \plotone{fig5.eps}
755: \figcaption[]
756: {Plots of the Mg~{\sc ii} emission for OO Aql, (upper panel) in flux units
757: (f(Mg~{\sc ii})), and (lower panel) normalized relative to the adjacent
758: ultraviolet continuum (R$_{uv}$).
759: Different symbols are used to distinguish the different dates:
760: 1988 Aug (filled circles), 1988 Sep (filled triangles), 1984 May (plus sign),
761: 1982 Jul (cross sign).
762: \label{f5.eps}}
763: \end{figure*}
764:
765: \end{document}
766: