0807.3542/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: %%\usepackage{graphicx}
3: \bibliographystyle{apj}
4: 
5: \newcommand{\msun}{M\ensuremath{_\odot}}	% Solar Mass
6: \newcommand{\mdot}{\ensuremath{\dot M}}		% mass accretion rate
7: \newcommand{\about}{\ensuremath{\sim}}		% about
8: \newcommand{\ctwo}{\ensuremath{\chi^2/\nu}}	% chi-square/nu
9: \newcommand{\cmd}{color-magnitude diagram}	% CMD
10: \newcommand{\bb}{blackbody}			% blackbody
11: \newcommand{\ns}{neutron star}			% neutron star
12: \newcommand{\bh}{black hole}			% black hole
13: \newcommand{\xrb}{X-ray binary}                 % X-ray binary
14: \newcommand{\xrt}{X-ray transient}      	% X-ray transient
15: \newcommand{\sxt}{soft X-ray transient}      	% Soft X-ray Transient
16: \newcommand{\aql}{Aquila~X-1}
17: \newcommand{\fu}{4U~1543-47}
18: 
19: \shorttitle{OIR Observations of \aql}
20: \shortauthors{Maitra \& Bailyn}
21: 
22: \begin{document}
23: \title{Outburst Morphology in the Soft X-ray Transient \aql}
24: 
25: \author{Dipankar Maitra}
26: \affil{Sterrenkundig Instituut ``Anton Pannekoek'', University of Amsterdam,
27: 	Kruislaan 403, 1098 SJ Amsterdam, The Netherlands}
28: \email{D.Maitra@uva.nl}
29: \and
30: 
31: \author{Charles D. Bailyn}
32: \affil{Yale University, Department of Astronomy, P.O.Box 208101, 
33: 	New Haven, CT 06520-8101, USA}
34: \email{charles.bailyn@yale.edu}
35: 
36: \begin{abstract}
37: We present optical and near-IR (OIR) observations of the major outbursts 
38: of the \ns\ soft \xrt\ binary system \aql, from summer 1998 -- fall 2007. 
39: The major outbursts of the source over the observed timespan seem to 
40: exhibit two main types of light curve morphologies, (a) the classical {\em 
41: Fast-Rise and Exponential-Decay} (FRED) type outburst seen in many soft 
42: X-ray transients and (b) the {\em Low-Intensity State} (LIS) where the 
43: optical-to-soft-X-ray flux ratio is much higher than that seen during a 
44: FRED. Thus there is no single correlation between the optical (R-band) and 
45: soft X-ray (1.5-12 keV, as seen by the ASM onboard RXTE) fluxes even 
46: within the hard state for \aql, suggesting that LISs and FREDs have 
47: fundamentally different accretion flow properties. Time evolution of the 
48: OIR fluxes during the major LIS and FRED outbursts is compatible with 
49: thermal heating of the irradiated outer accretion disk. No signature of 
50: X-ray spectral state changes or any compact jet are seen in the OIR, 
51: showing that the OIR \cmd\ (CMD) can be used as a diagnostic tool to 
52: separate thermal and non-thermal radiation from X-ray binaries where 
53: orbital and physical parameters of the system are reasonably well known. 
54: We suggest that the LIS may be caused by truncation of the inner disk in a 
55: relatively high \mdot\ state, possibly due to matter being diverted into a 
56: weak outflow. 
57: \end{abstract}
58: 
59: \keywords{accretion, accretion disks --- stars: neutron --- stars: --- 
60: X-rays: binaries --- individual (Aquila~X-1)}
61: 
62: \section{Introduction\label{sec:introduction}}
63: 
64: \aql\ is one of the most frequently recurring soft X-ray transients in the 
65: sky. It goes into outburst approximately once every year although the 
66: periodicity is not fixed \citep{simon2002}. The frequent outburst 
67: recurrence makes it an ideal source for studying accretion flow during an 
68: outburst as well as changes in accretion flow from outburst to outburst. 
69: The physical and orbital parameters of this system are relatively well 
70: known. The orbital period of the binary system was determined to be 
71: $18.71\pm0.06$ hours by \citet{wry2000} from optical photometry during 
72: quiescence. \citet{c1999} estimated the quiescent V-band magnitude to be 
73: $21.6\pm0.10$ magnitude. However, there is a contaminating star about 
74: 0.48\arcsec\ East of \aql\ which has a V magnitude of $19.42\pm0.06$ 
75: \citep{c1999}. The distance to the system is between 4-6.5 kpc as inferred 
76: both from (a) limits on optical brightness of the Roche lobe filling donor 
77: star of spectral type between K6-M0 \citep{c1999,r2001} and (b) peak X-ray 
78: fluxes during type-1 bursts \citep{ccg1987,r2001}.
79: 
80: In systems like \aql\ and other Roche lobe overflowing \sxt\ (SXT) binary 
81: systems, the thermal viscous disk instability model \citep[DIM;][]{l2001} 
82: is usually invoked to explain the typical {\em fast-rise and 
83: exponential-decay} (FRED) type outburst light curve morphology 
84: \citep{csl1997}. The DIM alone however cannot produce realistic FRED 
85: outbursts. Another mechanism that plays a key role during outbursts of 
86: SXTs is X-ray irradiation. The irradiation is caused by the intense X-rays 
87: from the inner regions of the accretion disk and \ns\ surface (if the 
88: system harbors a \ns) impinging on the outer disk/donor, and by 
89: re-emission of this impinged energy in lower energies 
90: \citep{c1976,v1990,vPM1994,m1995,d1999,obrien2002,h2002,r2006}. Another 
91: catagory of modifications to the DIM is generated by the possible 
92: truncation of the inner region of the accretion flow, where the disk 
93: inflow can be replaced by a radiatively inefficient accretion flow (RIAF).  
94: There is an increasing body of work on this topic 
95: \citep{i1977,r1982,ny1994,bb1999,m2001}, but the overall impact of these 
96: efforts on the outburst morphology is currently still unclear.
97: 
98: Collimated relativistic outflows, loosely termed as ``jets'', have been 
99: observed in several \sxt s \citep[see][for a review]{f2006} and their 
100: broad band spectral energy distribution (SED) from radio to X-rays have 
101: been fitted with jet+disk models comprising of a freely 
102: expanding ``Blandford - K\"onigl'' jet \citep{bk1979,fb1999,mnw2005} and a 
103: thermal-viscous ``Shakura-Sunyaev'' accretion disk \citep{ss1973}. 
104: Observations suggest that compared to black hole (BH) systems, the neutron 
105: star (NS) SXTs tend to have intrinsically weaker jets and also the 
106: transition from optically-thick to optically-thin jet spectrum in NS 
107: systems probably occurs in the mid-IR \citep[][this 
108: work]{mf2006,migliari2006}, compared to that in near-IR for BH systems 
109: \citep{cf2002,bb2004,h2005,r2006}.
110: 
111: During outburst, light curve morphology of many sources are not FREDs. 
112: The X-ray spectral and temporal features also show an enormous diversity
113: during outbursts of SXT systems, and are categorized into various {\em
114: states}. 
115: See e.g. \citet{mr2006} and \citet{hb2005} for extensive discussions on 
116: definition of states and \citet{vdk2006} for their timing properties.
117: Despite the multitude of states observed in the different systems and 
118: different outbursts, two ``canonical'' states can be identified during the 
119: outburst of most SXTs, characterised by presence or absence of a thermal and 
120: nonthermal component in the energy spectrum. In the temporal domain, presence 
121: or absence of a thermal component usually shows up respectively as low or 
122: high rms variability in the light curve. A host of other intermediate and 
123: extreme states are also seen, which could be a superposition of the thermal 
124: and nonthermal components with varying contributions. Precise 
125: determination of X-ray state requires modeling of spectral energy distribution
126: and Fourier power density spectra. Since this work primarily focusses mainly 
127: on OIR observations, we have used dates of X-ray state transitions of \aql\ 
128: from published studies of X-ray spectral and temporal properties 
129: of the source \citep{mc2003a,mc2003b,rks2004,mb2004,rodriguez2006,y2007} to
130: study correlation between X-ray state and OIR flux-density.
131: 
132: We 
133: have observed \aql\ for $\sim9$ years, through many outbursts, to explore 
134: the nature of outburst morphologies in a single object. Even though the 
135: OIR (as well as X-ray) light curve morphologies from outburst to outburst 
136: differ, we find systematic and repeated tracks traced by the source in OIR 
137: \cmd s. Since we have long-term data in two bandpasses (R and J) only, we 
138: have used a simple model to show that the color evolution and the 
139: resulting track are consistent with heating of a constant area blackbody 
140: and can be caused by an irradiated outer disc.
141: 
142: In \S\ref{sec:data} we describe the general reduction procedures adopted 
143: in this work. The light curve morphologies in J, R and ASM bands (of the 
144: RXTE satellite) and optical--soft X-ray correlations during the major 
145: outbursts are described in \S\ref{sec:ltc}. Evolution of \aql\ on the OIR 
146: \cmd\ is presented in \S\ref{sec:colors}. Finally we present our 
147: discussions in \S\ref{sec:discuss}.
148: 
149: \section{Observations and Data Analysis\label{sec:data}}
150: \subsection{Optical and Near-IR data\label{sec:oir-data}} 
151: 
152: Our long term monitoring campaign used the Yale 1.0m telescope at Cerro 
153: Tololo Inter-American Observatory (CTIO) in Chile to observe \aql\ from 
154: 1998-2002 on a daily basis (weather and Sun-angle constraints permitting).  
155: Since February 2003, the CTIO 1.3m telescope has been used for the 
156: monitoring. In both cases, images were obtained using the 
157: ANDICAM\footnote{http://www.astro.yale.edu/smarts/ANDICAM} instrument, 
158: which is a dual-channel imager capable of simultaneously recording images 
159: at an optical and an infrared wavelength. Prior to MJD 52430 (2002 June 
160: 5), the optical R-band data were recorded by a Lick/Loral-3 
161: $2048\times2048$ CCD.  Thereafter, the R-band has been recorded by a 
162: Fairchild 447 $2048\times2048$ CCD. The near-IR J-band data are recorded 
163: by a Rockwell $1024\times1024$ HgCdTe Astronomical Wide Area Infrared 
164: Imager (HAWAII) imager. Between MJD 51130--51367 (1998 November 13 -- 1999 
165: July 8th), the source was observed using a ``wide-R'' filter. The wide-R 
166: filter has a broader bandpass than the Johnson R filter used during the 
167: rest of the campaign. Even though a determination of the Johnson R-band 
168: flux cannot be made from the wide-R-band measurements, the light curve 
169: morphology changes little due to the change of filter. We calculated the 
170: quiescent wide-R-band differential magnitude (between MJD 51232 and MJD 
171: 51306) and applied an offset to the differential magnitudes obtained using 
172: wide-R filter so that the quiescent level was set to 18.8 magnitude, which 
173: is the quiescent R-band magnitude of \aql.
174: 
175: Under normal viewing conditions at CTIO, it is extremely difficult to 
176: separate the true optical counterpart of \aql\ from the bright contaminating star. 
177: Therefore the magnitudes reported in this work are the combined magnitudes 
178: of the \aql\ system and the contaminating star, and when we say magnitude of \aql\ 
179: we actually mean the combined magnitude of \aql\ and the contaminating star. In 
180: quiescence, light from the contaminating star dominates, but once an outburst 
181: begins, photons from \aql\ dominate.
182: 
183: The optical data reduction, including bias subtraction, overscan 
184: correction, and flat-fielding, was done through the standard IRAF data 
185: reduction pipeline developed for these instruments\footnote{For details 
186: see http://www.astro.yale.edu/smarts/smarts13m/optprocessing.txt}. For the 
187: J-band near-IR data, multiple dithered frames were taken and then 
188: flat-fielded, sky subtracted, aligned and average-combined using an 
189: in-house IRAF script. During a large portion of the year 2004, one of the 
190: four quadrants of the IR CCD was not working. In most of the observations 
191: during this period, one or more of the reference stars were not present in 
192: the frames. We used whichever comparison stars were present and adjusted 
193: the resulting differential magnitudes appropriately.
194: 
195: We used the field stars 2MASS 19111739+0035039 and 2MASS 19111633+0034333 
196: with J-band magnitudes of 12.404 and 12.490 respectively\footnote{Taken 
197: from the 2MASS point source catalogue at 
198: http://www.ipac.caltech.edu/2mass}, to obtain differential magnitude of 
199: the source. Flux calibration of these reference stars in the R-band was 
200: performed by comparing with Landolt standard stars on the photometric 
201: night of MJD 53904 (2006 June 18). In order to estimate time variability 
202: in flux from any of the reference stars, we calculated the difference 
203: between the observed magnitudes of the reference stars on every night. 
204: Smaller than 0.02 mag in R-band, this difference is the primary source of 
205: error in flux determination because during outbursts the Poisson error due 
206: to photon count-rate is much smaller. The $S/N$ is even better in J-band 
207: where interstellar extinction is lesser.
208: 
209: %For magnitude to flux conversion, following \citet{ac2000} and 
210: %\citet{c1985}, we assumed that R=0 converts to 2941 Jy and J=0 converts to 
211: %1603 Jy. \citet{wry2000} estimated that $83\pm2 \%$ of the quiescent 
212: For magnitude to flux conversion, following \citet{bessell1998} and
213: \citet{c1985}, we 
214: assumed that R=0 magnitude converts to 3064 Jy and J=0 magnitude converts to 
215: 1603 Jy. \citet{wry2000} estimated that $83\pm2 \%$ of the quiescent 
216: R-band flux is from the contaminating star. Assuming that the brightness of the 
217: contaminating star is constant in time, we subtracted this flux from the total 
218: observed flux to obtain the flux of \aql\ only, during quiescence as well 
219: as outbursts. Interstellar reddening and extinction was calculated 
220: assuming a hydrogen column density ($N_H$) of $3.4\times 10^{21}$ 
221: atoms/cm$^2$ in the direction of \aql\ \citep{dl1990}. From deep
222: optical photometry and spectroscopy of the source during quiescence 
223: \citet{c1999} estimated the color excess $E(B-V)=0.5\pm0.1$ which is 
224: consistent with the \citet{dl1990} value of $N_H$ in the direction of
225: \aql. For dereddening the OIR flux we used the standard $N_H$ to 
226: visual extinction ($A_V$) correlation given by $N_H/A_V=1.79\times10^{21}$ 
227: atoms/cm$^2$/mag \citep{ps1995} and the extinction law of \citet{ccm1989} 
228: which gives $A_R/A_V=0.748$ magnitude and $A_J/A_V=0.282$ magnitude 
229: for \aql. Taking reddening and extinction in the direction of \aql\ into 
230: account, the observed $R-J$ color is linearly related to the corresponding
231: intrinsic spectral index ($\alpha_{RJ;int}$) by the relation
232: \begin{equation}
233:  \alpha_{RJ;int} = \frac{log (f_{R}) - log (f_{J})}
234:                         {log (\nu_R) - log (\nu_J)}
235:  = 2.47 - 1.63\times (R-J)_{observed}.
236:  \label{eq:color-slope}
237: \end{equation}
238: 
239: \subsection{X-ray data\label{sec:asm-data}} 
240: 
241: The X-ray data was taken from the All Sky Monitor \citep[ASM;][]{l1996} 
242: aboard the Rossi X-ray Timing Explorer \citep[RXTE;][]{j1996} satellite. 
243: The ASM has been monitoring the sky in the 1.5-3, 3-5 and 5-12 keV energy 
244: bands continuously since 1996. Usually a given source is observed a few 
245: times every day, subject to constraints like angular proximity to the Sun 
246: etc.  The results are made publicly available by the MIT/ASM team through 
247: their website\footnote{http://xte.mit.edu}. We have used the one day 
248: average light curve for \aql.  During quiescence, the flux from \aql\ is 
249: below the detection limit of the ASM, however since we are interested only 
250: in the major outbursts, which are detectable by the ASM, we did not take 
251: into account data during quiescence, for our optical-X-ray correlation 
252: study. Also, during the early as well as the late stages of the outbursts, 
253: the count-rates in the individual energy channels are small; therefore we 
254: have only looked at the summed count-rates in this work. 
255: The ASM count-rates were converted to flux-density at 5.2 keV, assuming a 
256: Crab-like spectrum.  The Crab spectrum in the ASM energy range is a simple 
257: power law with a photon-index of -2.1 and the total flux within that range 
258: is $2.8\times10^{-8}$ ergs/cm$^2$/s \citep{s2002}, so the flux-density at 
259: 5.2 keV is 1.06 mJy. The flux conversion (from ASM count-rate to $\mu$Jy) is 
260: therefore 
261: a numerical factor by which the total ASM count rate is multiplied.
262: The choice of 5.2 keV was motivated by the fact that it is the geometric mean
263: of the half-transmission points of the ASM response shape\footnote{Taken from 
264: appendix F of the RXTE technical appendix 
265: (http://heasarc.nasa.gov/docs/xte/appendix\_f.html) and Ron Remillard, 
266: priv. comm.} near 2.3 and 12 keV. 
267: We used geometric mean (GM) because the GM compensates for the fact that 
268: almost all X-ray spectra slope downward in count rate toward higher energies.
269: %Therofore for the present work we decided to use total
270: %ASM counts rate and convert it to Crab-scaled flux-density as described above. 
271: 
272: Evolution of spectral shape during an outburst introduces uncertainities in
273: flux determination. 
274: Pivoting uncertainities (due to spectral evolution) in ASM's X-ray range 
275: are small compared to the optical to X-ray separation in frequency 
276: space. 
277: We used the online 
278: {\em WebPIMMS}\footnote{http://heasarc.gsfc.nasa.gov/Tools/w3pimms.html} tool
279: which shows that for a given total ASM count-rate, the difference in flux
280: in the ASM band-pass is less than 10\% between a power law of photon index 
281: 2.1 (i.e. Crab spectrum), a power law of photon index 1.7 or a blackbody at 
282: 2 keV \citep[typical photon index/blackbody temperature of \aql\ in hard/soft 
283: state; see e.g.][]{mc2003b,mb2004,rodriguez2006}.
284: Therefore we added
285: a 10\% systematic error, in quadrature to the Poisson error, during the ASM 
286: count-rate to flux conversion to account for the uncertainity in the spectral
287: shape.
288: The precise value for pinning and rescaling the flux conversion
289: is not very sensitive to the choice of keV reference, as long as this is 
290: done self-consistently (i.e. scale to Crab flux density at the reference keV, 
291: and implemented in this work).
292: 
293: In this process, any information about 
294: changes in the energy spectrum is lost, but we are primarily concerned 
295: with overall X-ray flux here, and using all the counts results in better 
296: signal-to-noise ratio. In order to assess the X-ray state of the source, we
297: have used the dates of state transitions from published literature
298: \citet{mc2003a,mc2003b,rks2004,mb2004,rodriguez2006,y2007}. For the most recent 
299: outbursts (e.g. outburst \#9, \#10) with no published dates of state 
300: transition(s), 
301: we used the (9.7-16 keV/6-9.7 keV) hardness ratio from archival RXTE/PCA 
302: pointed observations (Manuel Linares, priv. comm.) whose value correlates 
303: strongly with the X-ray state of the source. For outbursts with sparse 
304: coverage of pointed observations (e.g. outburst \#5) we used the 
305: (5-12 keV/3-5 keV) hardness ratio from the RXTE/ASM data to infer the spectral
306: state.
307: 
308: 
309: \section{Major Outbursts of \aql\ since 1998 June 20\label{sec:ltc}} 
310: 
311: In Fig.~\ref{longterm} we show the results of long-term multi-wavelength 
312: monitoring of \aql. The RXTE/ASM monitoring of \aql\ dates back to the 
313: beginning of 1996, but our nightly R-band monitoring started on MJD 50984 
314: (1998 June 20th). The J-band observations started MJD 51677 (2000 May 
315: 13th) onwards, but the coverage was somewhat sporadic till the end of 
316: 2002, as can be seen from the top panel of Fig.~\ref{longterm}. 
317: During quiescent periods, the system is detected in both of the OIR bands but 
318: not in ASM. Therefore in Fig.~\ref{longterm} we have only plotted those ASM
319: points for which the signal-to-noise ratio is better than four.
320: 
321: Since observing constraints sometimes make it difficult to date the 
322: precise starting and ending moments of an outburst, in 
323: Table~\ref{tab:outburst_log} we note the MJD when the ASM, R-band and 
324: J-band fluxes attained a maximum during an outburst. We define a major 
325: outburst as a period of activity of the source, spanning at least 20 days 
326: or longer, when the source flux was at least $3\sigma$ above quiescence in 
327: at least one of the observed (J, R and ASM) bandpasses. Eleven major 
328: periods of activity were observed by the ASM during this period. All these 
329: periods of major activity were also seen in OIR, except the one during 
330: late 2005--early 2006 when the source's angular proximity to the Sun was 
331: too close to observe with ground based OIR telescopes. Even the RXTE/ASM 
332: data, while showing unmistakeable signs of activity during this particular 
333: outburst, does not have sufficient coverage and $S/N$ to classify the light 
334: curve morphology.
335: 
336: Besides the major outbursts, several smaller and shorter periods of 
337: activity are seen in the OIR light curve (Fig.~\ref{longterm}). Because of 
338: the low signal-to-noise ratio, particularly in ASM data, and scarcity of 
339: PCA pointings during these periods of minor activity, they have not been 
340: discussed in the present work. However, the time of peak OIR flux from the 
341: source during these ``mini'' outbursts have been presented in 
342: Table~\ref{tab:outburst_log}.
343: 
344: \subsection{Lightcurve morphology: FREDs and LISs}
345: 
346: From the long-term OIR monitoring of \aql\ from MJD 50984 -- MJD 54411 
347: (1998 June 20 -- 2007 Nov 7) only three out of the major OIR outbursts are 
348: purely FREDs. The remaining are a combination of FREDs with another active 
349: state where the flux is above quiescence, but does not rise or decay on 
350: the timescale of a FRED. Instead, a plateau type light curve with 
351: significant variability on a timescale of days is seen in all bands. 
352: Sometimes this state persists significantly longer than the FREDs. For 
353: \aql\ the typical observed FRED rise times are \about\ few days to a week 
354: and the decay times are \about\ 3--4 weeks. On the other hand, the longer 
355: lasting, somewhat variable state can last for longer than a month. 
356: Following \citet{w2002}, who reported observation of such state in the 
357: neutron star system X1608-52, we call this a ``Low-Intensity State'' 
358: (LIS). In Fig.~\ref{zoom_lc_1-3} -- Fig.~\ref{zoom_lc_10} we show blowups of
359: the R-band and ASM light curves for each of the major outbursts in 
360: Fig.~\ref{longterm} (also tabulated in Table~\ref{tab:outburst_log}). FREDs and
361: LISs are shown by different symbols in Fig.~\ref{zoom_lc_1-3} -- 
362: Fig.~\ref{zoom_lc_10}. During a LIS, the OIR flux is near the 
363: maximum, but always slightly 
364: lower than the corresponding maximum flux observed during a pure FRED 
365: outburst. For LISs, the variability in the light curve is more complicated than
366: the typical `up to a maximum, then down to quiescence' morphology seen for 
367: FREDs. For a comparable optical brightness, the soft X-ray flux is a 
368: factor of seven or more lower in a LIS than in a FRED. This high 
369: optical-to-X-ray flux-ratio, discussed in greater detail in 
370: \S\ref{sec:r-asm}, is one of the defining characteristics of a LIS. 
371: 
372: \subsection{Optical--soft X-ray correlations\label{sec:r-asm}}
373: 
374: The strongest evidence for the existence of an independent LIS comes from 
375: optical--X-ray correlations. From the long-term OIR monitoring data, we 
376: estimated R-band flux-density from \aql\ as described in 
377: \S\ref{sec:oir-data}.  The RXTE/ASM one day averaged count-rates were 
378: converted to flux-density as well (\S\ref{sec:asm-data}). Both light 
379: curves were then searched for quasi-simultaneous optical--X-ray data where 
380: the observation times in each bandpass were not separated by more than one 
381: day. In Fig.~\ref{allfluxes} we show the optical--soft-X-ray flux 
382: correlation during the major outbursts for which we had quasi-simultaneous 
383: ASM and R-band data.
384: 
385: The optical and soft-X-rays increase approximately linearly with 
386: $f_{opt}/f_{X-ray}\sim 1.3-1.5$ for FRED outbursts and $\sim 9-12$ for the 
387: LISs. Given the considerable scatter in the data, we have not attempted to 
388: fit the data with any function and the flux-ratios quoted above, as shown in 
389: Fig.~\ref{allfluxes}, are a guide to the eye. For the same X-ray 
390: brightness, there is no strong hint of hysteresis, i.e. the optical 
391: brightness during outburst rise is not significantly different from than 
392: that during the outburst decline. 
393: %The existence of such a hysteresis loop 
394: %has been reported by \citet{r2006, r2007b}, in context of black hole X-ray 
395: %binaries where the enhanced near-IR flux is ascribed to an additional jet 
396: %component. 
397: The absence of optical--X-ray hysteresis in \aql, goes to 
398: support the hypothesis that either jets in \ns\ systems are generally 
399: weaker or the jet break occurs at a lower frequency 
400: \citep{mf2006,migliari2006,r2007a}.
401: 
402: Both FREDs and LISs start/end the outburst at quiescence and hence have a 
403: common start/end-point near the lower-left region of $f_{opt}-f_{X-ray}$ 
404: plane. For some of the weakest outbursts, the ASM fluxes are within a few 
405: sigma of detection threshold. At such low luminosities, both FREDs and 
406: LISs occupy similar regions on the $f_{opt}-f_{X-ray}$ plane. It 
407: has been observed that for some low luminosity outbursts, the source may never 
408: come out of the X-ray hard state during the entire period of activity
409: \citep[e.g. see][]{rodriguez2006}. Even 
410: though the peak optical brightnesses of the brightest LISs are comparable 
411: to the peak optical brightness during FREDs, LISs are typically fainter 
412: than FREDs. The distinction between LIS and FRED outbursts stands out 
413: clearly when optical and X-ray data from {\em all} the major outbursts 
414: since 1999 June 20 are plotted on $f_{opt}-f_{X-ray}$ plane 
415: (Fig.~\ref{allfluxes}). LISs and FREDs occupy separate regions of the 
416: $f_{opt}-f_{X-ray}$ plane and evolve on different timescales, suggesting 
417: different physical mechanisms for their origin.
418: 
419: \section{OIR color evolution\label{sec:colors}} 
420: 
421: In Fig.~\ref{r-j_1} and Fig.~\ref{r-j_2} the time evolution of the R and 
422: J-band magnitudes and 
423: that of the R$-$J color is shown for the major outbursts that were observed 
424: in both bandpasses. There seems to be no significant inter-band time lag 
425: during rise or decay of an outburst.  
426: Irrespective of the light curve morphology during the outburst, the R$-$J 
427: color is always bluer during an outburst compared to the quiescent value, 
428: and therefore $(R-J)_{outburst} < (R-J)_{quiescence}$ as seen in 
429: Fig.~\ref{r-j_1} and Fig.~\ref{r-j_2}.
430: This decrease in R$-$J color 
431: (becoming bluer) is characteristic of increased irradiation heating of the 
432: outer disk. 
433: 
434: During quiescence, the disk flux is very low compared to the 
435: combined flux from the contaminating star and the donor star. It is however 
436: interesting to note that the R$-$J color during an LIS is significantly 
437: redder than that during the peak of a FRED outburst.
438: 
439: Fig.~\ref{aql-cmd1} and Fig.~\ref{aql-cmd2} shows the OIR \cmd\ for the 
440: different outbursts. The 
441: R-band magnitude is plotted along the ordinate. As is typical for 
442: conventional \cmd s, the ordinate axis has been reversed so that higher 
443: luminosities (i.e. lower magnitude) are plotted near the top and 
444: vice-versa.  The R$-$J color is plotted along the abscissa.
445: 
446: During periods of quiescence, the source flux is low and essentially 
447: dominated by photons from the contaminating star and the donor, and it lingers 
448: near the bottom-right corner of the \cmd. As the source enters an 
449: outburst, the brightness increases and the color becomes bluer, due to an 
450: increasing contribution of the accretion disk flux. 
451: %The dates of state 
452: %transition between hard and soft X-ray states were estimated either from 
453: %previously reported observations 
454: %\citep{mc2003a,mc2003b,rks2004,mb2004,rodriguez2006,y2007} 
455: %or from 
456: %(9.7-16 keV/6-9.7 keV) X-ray colors from archival PCA observations 
457: %(Manuel Linares, priv. comm.) or the (5-12 keV/3-5 keV) hard color from the 
458: %RXTE/ASM data archive. 
459: In Fig.~\ref{aql-cmd1} and Fig.~\ref{aql-cmd2}, the observations when the 
460: source was in the hard 
461: state during the rise of an outburst, when it was in the hard state during 
462: the decay and when it was in the soft state are shown by different 
463: symbols. 
464: As noted in \S~\ref{sec:r-asm}, there are also outbursts
465: when the source spent the entire period of activity in an X-ray hard 
466: state \citep{rodriguez2006}. The OIR data during such outbursts are shown by
467: open squares. 
468: There seems to be no significant change in OIR color across the 
469: X-ray spectral state transitions.
470: 
471: We attempted to test the simple ansatz that the observed evolution on the 
472: \cmd\ is consistent with the thermal heating of a \bb. Our goal is to see 
473: whether the flux and color variations are correlated as would be expected 
474: if the only change is in the disk temperature.  For simplicity, and 
475: because we had only two band-passes, we considered the entire annulus to 
476: be a single temperature blackbody, even though real accretion disks are 
477: likely to have a significant temperature gradient. In this simple model, 
478: the R$-$J color is determined by the temperature of the annulus and the 
479: R-band brightness is proportional to the projected area of the annulus and 
480: inversely proportional to the square of the distance to the source. Since 
481: the observed brightness depends on the projected area, we multiplied the 
482: area of a disk of radius $R_{disk}$, by a multiplicative factor $f<1$, and 
483: used this area as the area of the annulus. Thus in this simple model the 
484: brightness is proportional to $f(R_{disk}/D_{disk})^2cos(i)$, where 
485: $D_{disk}$ and $i$ are the source distance and $i$ is the orbital 
486: inclination respectively. The only adjustable parameter here is $f$, which 
487: changes the overall projected area of the annulus. The parameter $f$ is 
488: changed from outburst to outburst, keeping other parameters constant. 
489: Fitting the ellipsoidal variations observed in the quiescent light curve, 
490: \citet{rwy2001} estimated that $36\degr\leq i\leq55\degr$. We assumed an 
491: inclination of $45\degr$. An upper limit on $R_{disk}$ can be presumed to 
492: be equal to the radius of the Roche lobe of the accretor and therefore can 
493: be estimated from the mass ratio and orbital period \citep{e1983}. 
494: Evolution of the source on the \cmd\ was generated by calculating the 
495: color and brightness of the annulus at varying temperature. We note that 
496: the track generated by the heating/cooling of this constant area blackbody 
497: broadly describes the evolution of the source during its outbursts.
498: For a purely viscous disk (no irradiation) the spectral index between R and
499: J bands (as defined in Eq.~\ref{eq:color-slope}) should lie between $1/3$ 
500: and $2$. The observed range in $R-J$ color during outbursts is approximately
501: between $1.35$ and $2.1$, which using Eq.~\ref{eq:color-slope} translates 
502: to a spectral slope between $0.27$ and $-0.95$ respectively. The observed 
503: range in spectral indices is quite different from what could be expected from 
504: a purely viscous disk and favors irradiation heating.
505: 
506: We found that even for the largest possible disk radius of \about 5 light 
507: seconds ($M_{donor}=0.5\msun$, $M_{accretor}=1.5\msun$, $P_{orb}=18.95$ 
508: hour), the model-predicted R-band flux is too small for the lowest 
509: distance estimate of 4 kpc by \citet{r2001}. Using $D_{disk}=3$ kpc seems 
510: to lie close to the observed datapoints (see Fig.~\ref{aql-cmd1} and 
511: Fig.~\ref{aql-cmd2}). Since we 
512: are not doing detailed spectral fits, we do not claim the distance to the 
513: source to be 3 kpc, but show that this very simple model can roughly 
514: predict the outburst behavior of the accretion disk. The model curves 
515: overplotted in Fig.~\ref{aql-cmd1} and Fig.~\ref{aql-cmd2} are not fits to the 
516: datapoints but 
517: created using a disk of radius 5 light-seconds, inclined at an angle of 
518: 45$\degr$ to the observer and at a distance of 3 kpc from the observer, 
519: viz. a set of parameters that is generally consistent with the observed 
520: values, and with a small range (0.6-0.8) in the value of the 
521: multiplicative factor $f$. The general consistency of the data over many 
522: outbursts supports the idea that the OIR flux is largely due to thermal 
523: emission from an irradiated disk. This is in contrast to recent 
524: observations of accreting black hole systems in which a strongly 
525: non-thermal component is often observed in the near IR 
526: \citep[][ Russell, Maitra \& Fender, in prep.]{bb2004,h2005,r2006}.
527: 
528: \section{Discussion\label{sec:discuss}}
529: 
530: Multi-wavelength monitoring of the neutron star soft X-ray transient \aql\ 
531: in near-IR, optical and soft X-ray data from RXTE/ASM over the past nine 
532: years has allowed us to study the evolution of the source with 
533: unprecedented coverage. The OIR and X-ray light curves show a wide range 
534: of outburst light curve patterns. However the major outbursts show two 
535: distinct light curve morphologies. The {\em fast-rise and 
536: exponential-decay} \citep[FRED;][]{csl1997} type outbursts characteristic 
537: of many soft X-ray transients is seen in \aql\ too. However, only three of 
538: the major outbursts seen since 1998 June show just a FRED type light curve 
539: profile. More often a prolonged, somewhat variable ``Low-Intensity state'' 
540: \citep[LIS;][]{w2002}, uniquely characterized by an extremely high 
541: optical-to-soft X-ray flux-ratio, is seen either before or after a FRED.
542: It has been observed that for some of these ``soft X-ray faint'' LIS 
543: outbursts, \aql\ never comes out of the X-ray hard state during the entire 
544: outburst \citep{rodriguez2006}.
545: Other than these major outbursts, small-scale flaring activities were also 
546: observed. While the origin of such small-scale short activity remains 
547: unknown, such activity has been observed in other sources like XTE 
548: J1550-564 by \citet{ss2005} who suggested that these small outbursts could 
549: be due to discrete accretion events. We suggest the following framework 
550: for understanding these results:
551: 
552:   {\em Irradiation dominates OIR flux for both FRED and LIS
553: 	  outbursts of \aql.}
554:   During the course of FRED and LIS outbursts of \aql\, the evolution of
555:   the source on the OIR \cmd\ is consistent with heating/cooling of an 
556:   irradiated outer accretion disk as a constant area \bb.
557:   A classical viscous accretion disk (e.g. a Shakura-Sunyaev disk) spectrum 
558:   has a frequency dependence of $F_{\nu}\sim \nu^{1/3}$ between 
559:   $kT(R_{out})/h \ll \nu \ll kT(R_{in})/h$ where  $T(R_{in})$ and $T(R_{out})$
560:   are the disc temperatures at the inner and outer edges of the disk. At 
561:   frequencies much smaller than $kT(R_{out})/h$ the disk spectrum is simply the
562:   Rayleigh-Jeans region of a \bb\ with $F_{\nu}\sim \nu^2$. For soft X-ray
563:   transients $kT(R_{in})/h$ is
564:   typically in far-UV or soft X-rays. Therefore in our 
565:   case the flux ratios (a.k.a. colors) of two bands, from a purely viscous 
566:   disc are expected to be constant as long as {\em both bands} are either on 
567:   the viscous ($\nu^{1/3}$) or Rayleigh-Jeans ($\nu^2$) regime. In the special 
568:   case when the break frequency between the viscous dissipation and the 
569:   Rayleigh-Jeans regime given by $\nu_{break}=kT(R_{out})/h$ lies between R
570:   and J bands, the spectral index should lie between $1/3$ and $2$.
571:   The observed systematic and continuous change in color (and the corresponding 
572:   range in the spectral index from $-0.95$ to $0.27$; \S\ref{sec:colors}) seen 
573:   in Fig.~\ref{r-j_1} and Fig.~\ref{r-j_2} is strongly 
574:   suggestive of an irradiative origin for the observed OIR emission and not 
575:   of a purely viscous origin. 
576: 
577:   Neither the OIR color nor its brightness change sharply during an
578:   X-ray spectral state transition. This confirms that for \aql\, the outer
579:   accretion disk, which emits most of the thermal OIR radiation, is not
580:   affected by X-ray spectral state transitions which change the physical
581:   environment of the inner regions of the accretion disk. This is in 
582:   sharp contrast with \bh\ binaries like 4U 1543-47, GX 339-4 
583:   \citep{bb2004,h2005} where near IR colors are strongly correlated with
584:   X-ray state changes and the correlation interpreted as evidence for 
585:   non-thermal emission.
586:   In FREDs, a slow increase in brightness above that 
587:   predicted by our constant area \bb\ model, as the source becomes bluer 
588:   and brighter, is most likely due to increase in the surface area
589:   of the disk (e.g most prominent for outburst \#2 and \#5 in 
590:   Fig.~\ref{aql-cmd1}). It is possible that at highest mass accretion rates 
591:   the high X-ray flux from the central 
592:   inner disk causes the flaring of the outer disk and hence increase the 
593:   effective surface area. In contrast, during the LISs, when the source 
594:   probably spends the whole period of activity in hard X-ray state, there 
595:   does not seem to be any significant change in the disk area (e.g. outburst
596:   \#7 in Fig.~\ref{aql-cmd2} ). 
597:   Therefore it is likely that the effect of X-ray state transition on the outer 
598:   accretion disk is both an increase of the outer disk temperature due to 
599:   enhanced irradiation from the inner disk and 
600:   an increase in the emitting area of the outer disk due to flaring. 
601:   Total irradiation (hence OIR flux) is a bit smaller in
602:   LIS than in peak FRED. This could be due either to lower \mdot, or 
603:   to radiative inefficiency lowering $L/\mdot$ ratio.
604: 
605:   {\em Broad band quasi-simultaneous observations in LIS suggest 
606: 	  outflow.} 
607:   Soft X-ray flux is very low in LIS, but hard X-rays are high 
608:   \citep[][Linares et al. in preparation,]{atel259}.
609:   While absence of soft, thermal X-ray photons suggests that the inner edge 
610:   of the accretion disk is cold and probably receded during the LIS,
611:   the high OIR brightness (see Fig.~\ref{allfluxes} for the dramatic 
612:   difference in $f_{opt}/f_{X-ray}$ between LIS and FRED outbursts )
613:   as well as type I X-ray bursts suggests that
614:   a significant amount of mass is being accreted by the outer accretion disk
615:   and fed to the inner region.
616: 
617:   Radio detection of the source during LIS \citep{atel286} with 
618:   flat spectral slope hints toward a jet outflow. 
619:   However, since the evolution of the source on OIR \cmd\ is consistent 
620:   with irradiation, it can be concluded that the jet, if present, is either 
621:   quite weak in 
622:   terms of the input power, or the transition from optically thick 
623:   to optically thin jet emission occurs at a wavelength much longer than 
624:   J-band. 
625:   Current models for jets from compact systems suggest that this might imply
626:   a large jet launching area or the jet 
627:   base, or much weaker particle acceleration \citep{mff2001}. Thus our 
628:   observations support the current hypothesis that jet-luminosity is not 
629:   significant in most neutron star X-ray binary systems \citep{mf2006,f2006}.
630: 
631:   {\em A possible scenario for the LIS might be that an \mdot\ 
632:   that would ordinarily be high enough to move the disk inwards, and thus 
633:   trigger a soft state, somehow doesn't do that, and launches an outflow 
634:   instead.}
635:   Thus the standard
636:   processes associated with the DIM are disrupted by the absence
637:   of an inner disk in this high \mdot\ state. If the accretion flow in 
638:   the inner region in a LIS is radiatively inefficient, it can drive 
639:   the observed outflow, as suggested by recent theoretical works 
640:   \citep[see e.g.][]{bb1999,m2001,n2005}.
641: 
642: 
643: \acknowledgments
644: 
645: It is a pleasure to thank the SMARTS observers J. Espinoza and D. Gonzalez 
646: for taking the data over the years, and R. Winnick, J.Nelan and M. Buxton 
647: who accommodated our many requests to revise the observing schedule. 
648: We would like to thank Ron Remillard of the ASM team for his help in 
649: converting ASM count-rates to flux-density.
650: DM would 
651: like to thank Manu Linares for providing PCA colors of all the RXTE 
652: observations of \aql\ ahead of publication, and Sera Markoff, Dave Russell and 
653: Rob Fender for discussion and useful comments. We would like to thank the 
654: anonymous referee for constructive criticisms which improved the paper.
655: DM is supported by Netherlands Organisation for Scientific 
656: Research (NWO) grant number 614000530. CDB acknowledges support 
657: from the National Science Fopundation grant NSF-AST 0707627. This work has 
658: made use of the Astrophysics Data System Abstract, the MIT/ASM data 
659: extraction website. This work has also used data products from the Two 
660: Micron All Sky Survey, which is a joint project of the University of 
661: Massachusetts and the Infrared Processing and Analysis Center/California 
662: Institute of Technology, funded by the National Aeronautics and Space 
663: Administration and the National Science Foundation.
664: 
665: {\it Facilities:} \facility{RXTE}, \facility{SMARTS}
666: 
667: 
668: \begin{thebibliography}{}
669: 
670: \bibitem[Bessell et al.(1998)]{bessell1998} Bessell, M.~S., Castelli, F., \& Plez, B.\ 1998, \aap, 333, 231
671: \bibitem[Blandford \& Begelman(1999)]{bb1999} Blandford, R.~D., \& Begelman, M.~C.\ 1999, \mnras, 303, L1
672: \bibitem[Blandford \& K\"onigl(1979)]{bk1979} Blandford, R.~D., 
673: \& K\"onigl, A.\ 1979, \apj, 232, 34 
674: \bibitem[Buxton \& Bailyn(2004)]{bb2004} Buxton, M.~M., \& Bailyn, C.~D.\ 2004, \apj, 615, 880
675: \bibitem[Campins et al.(1985)]{c1985} Campins, H., Rieke, G.~H., \& Lebofsky, M.~J.\ 1985, \aj, 90, 896
676: \bibitem[Cardelli et al.(1989)]{ccm1989} Cardelli, J.~A., Clayton, G.~C., \& Mathis, J.~S.\ 1989, \apj, 345, 245
677: \bibitem[Chen et al.(1997)]{csl1997} Chen, W., Shrader, C.~R., \& Livio, M.\ 1997, \apj, 491, 312
678: \bibitem[Chevalier et al.(1999)]{c1999} Chevalier, C., Ilovaisky, S.~A., Leisy, P., \& Patat, F.\ 1999, \aap, 347, L51
679: \bibitem[Corbel \& Fender(2002)]{cf2002} Corbel, S., \& Fender, R.~P.\ 2002, \apjl, 573, L35
680: \bibitem[Cunningham(1976)]{c1976} Cunningham, C.\ 1976, \apj, 208, 534
681: \bibitem[Czerny, Czerny, \& Grindlay(1987)]{ccg1987} Czerny, M., Czerny, B., \& Grindlay, J.~E.\ 1987, \apj, 312, 122
682: \bibitem[Dickey \& Lockman(1990)]{dl1990} Dickey, J.~M., \& Lockman, F.~J.\ 1990, \araa, 28, 215
683: \bibitem[Dubus et al.(1999)]{d1999} Dubus, G., Lasota, J.-P., Hameury, J.-M., \& Charles, P.\ 1999, \mnras, 303, 139
684: \bibitem[Eggleton(1983)]{e1983} Eggleton, P.~P.\ 1983, \apj, 268, 368
685: \bibitem[Falcke \& Biermann(1999)]{fb1999} Falcke, H., \& Biermann, P.~L.\ 1999, \aap, 342, 49
686: \bibitem[Fender(2006)]{f2006} Fender, R.\ 2006, In: Compact stellar X-ray 
687: sources. Edited by Walter Lewin \& Michiel van der Klis. Cambridge Astrophysics 
688: Series, No. 39. Cambridge, UK: Cambridge University Press, 381
689: \bibitem[Homan et al.(2005)]{h2005} Homan, J., Buxton, M., Markoff, S., Bailyn, C.~D., Nespoli, E., \& Belloni, T.\ 2005, \apj, 624, 295
690: \bibitem[Homan \& Belloni(2005)]{hb2005} Homan, J., \& Belloni, T.\ 2005, \apss, 300, 107
691: \bibitem[Hynes et al.(2002)]{h2002} Hynes, R.~I., Haswell, C.~A., Chaty, S., Shrader, C.~R., \& Cui, W.\ 2002, \mnras, 331, 169 
692: \bibitem[Ichimaru(1977)]{i1977} Ichimaru, S.\ 1977, \apj, 214, 840
693: \bibitem[Jahoda et al.(1996)]{j1996} Jahoda, K., Swank, J.~H., Giles, A.~B., Stark, M.~J., Strohmayer, T., Zhang, W., \& Morgan, E.~H.\ 1996, \procspie, 2808, 59
694: \bibitem[Lasota(2001)]{l2001} Lasota, J.-P.\ 2001, New Astronomy Review, 45, 449
695: \bibitem[Levine et al.(1996)]{l1996} Levine, A.~M., Bradt, H., Cui, W., Jernigan, J.~G., Morgan, E.~H., Remillard, R., Shirey, R.~E., 
696: \& Smith, D.~A.\ 1996, \apjl, 469, L33
697: \bibitem[McClintock et al.(1995)]{m1995} McClintock, J.~E., Horne, K., \& Remillard, R.~A.\ 1995, \apj, 442, 358
698: \bibitem[McClintock \& Remillard(2006)]{mr2006} McClintock, J.~E., \& Remillard, R.~A.\ 2006, In: Compact stellar X-ray sources. Edited by Walter Lewin \& Michiel van der Klis. Cambridge Astrophysics Series, No. 39. Cambridge, UK: Cambridge University Press, 157
699: \bibitem[Maccarone \& Coppi(2003a)]{mc2003a} Maccarone, T.~J., \& Coppi, P.~S.\ 2003, \mnras, 338, 189
700: \bibitem[Maccarone \& Coppi(2003b)]{mc2003b} Maccarone, T.~J., \& Coppi, P.~S.\ 2003, \aap, 399, 1151
701: \bibitem[Maitra \& Bailyn(2004)]{mb2004} Maitra, D., \& Bailyn, C.~D.\ 2004, \apj, 608, 444
702: \bibitem[Markoff et al.(2001)]{mff2001} Markoff, S., Falcke, H., \& Fender, R.\ 2001, \aap, 372, L25
703: \bibitem[Markoff et al.(2005)]{mnw2005} Markoff, S., Nowak, M.~A., \& Wilms, J.\ 2005, \apj, 635, 1203
704: \bibitem[Meier(2001)]{m2001} Meier, D.~L.\ 2001, \apjl, 548, L9
705: \bibitem[Migliari \& Fender(2006)]{mf2006} Migliari, S., \& Fender, R.~P.\ 2006, \mnras, 366, 79
706: \bibitem[Migliari et al.(2006)]{migliari2006} Migliari, S., Tomsick, J.~A., Maccarone, T.~J., Gallo, E., Fender, R.~P., Nelemans, G., \& Russell, D.~M.\ 2006, \apjl, 643, L41
707: \bibitem[Molkov et al.(2004)]{atel259} Molkov, S.~V., Lutovinov, A.~A., Cherepashchuk, A.~M., \& Sunyaev, R.~A.\ 2004, The Astronomer's Telegram, 259, 1
708: \bibitem[Narayan(2005)]{n2005} Narayan, R.\ 2005, \apss, 300, 177
709: \bibitem[Narayan \& Yi(1994)]{ny1994} Narayan, R., \& Yi, I.\ 1994, \apjl, 428, L13
710: \bibitem[O'Brien et al.(2002)]{obrien2002} O'Brien, K., Horne, K., Hynes, R.~I., Chen, W., Haswell, C.~A., \& Still, M.~D.\ 2002, \mnras, 334, 426
711: \bibitem[Predehl \& Schmitt(1995)]{ps1995} Predehl, P., \& Schmitt, J.~H.~M.~M.\ 1995, \aap, 293, 889
712: \bibitem[Rees et al.(1982)]{r1982} Rees, M.~J., Begelman, M.~C., Blandford, R.~D., \& Phinney, E.~S.\ 1982, \nat, 295, 17
713: \bibitem[Reig et al.(2004)]{rks2004} Reig, P., van Straaten, S., \& van der Klis, M.\ 2004, \apj, 602, 918
714: \bibitem[Robinson et al.(2001)]{rwy2001} Robinson, E.~L., Welsh, W.~F., \& Young, P.\ 2001, AIP Conf.~Proc.~599: X-ray Astronomy: Stellar Endpoints, AGN, and the Diffuse X-ray Background, 599, 902
715: \bibitem[Rodriguez et al.(2006)]{rodriguez2006} Rodriguez, J., Shaw, S.~E., \& Corbel, S.\ 2006, \aap, 451, 1045
716: \bibitem[Rupen et al.(2004)]{atel286} Rupen, M.~P., Mioduszewski, A.~J., \& Dhawan, V.\ 2004, The Astronomer's Telegram, 286, 1
717: \bibitem[Russell et al.(2006)]{r2006} Russell, D.~M., Fender, R.~P., Hynes, R.~I., Brocksopp, C., Homan, J., Jonker, P.~G., \& Buxton, M.~M.\ 2006, \mnras, 371, 1334
718: \bibitem[Russell et al.(2007)]{r2007a} Russell, D.~M., Fender, R.~P., \& Jonker, P.~G.\ 2007, \mnras, 379, 1108 
719: %\bibitem[Russell et al.(2007)]{r2007b} Russell, D.~M., Maccarone, T.~J., K{\"o}rding, E.~G., \& Homan, J.\ 2007, \mnras, 379, 1401 
720: \bibitem[Rutledge et al.(2001)]{r2001} Rutledge, R.~E., Bildsten, L., Brown, E.~F., Pavlov, G.~G., \& Zavlin, V.~E.\ 2001, \apj, 559, 1054
721: \bibitem[Shakura \& Sunyaev(1973)]{ss1973} Shakura, N.~I., \& Sunyaev, R.~A.\ 1973, \aap, 24, 337
722: \bibitem[{\v S}imon(2002)]{simon2002} {\v S}imon, V.\ 2002, \aap, 381, 151
723: \bibitem[Smith et al.(2002)]{s2002} Smith, D.~A., et al.\ 2002, \apjs, 141, 415
724: \bibitem[Sturner \& Shrader(2005)]{ss2005} Sturner, S.~J., \& Shrader, C.~R.\ 2005, \apj, 625, 923
725: \bibitem[van Paradijs \& McClintock(1994)]{vPM1994} van Paradijs, J., \& McClintock, J.~E.\ 1994, \aap, 290, 133
726: \bibitem[Vrtilek et al.(1990)]{v1990} Vrtilek, S.~D., Raymond, J.~C., Garcia, M.~R., Verbunt, F., Hasinger, G., \& Kurster, M.\ 1990, \aap, 235, 162
727: \bibitem[Wachter et al.(2002)]{w2002} Wachter, S., Hoard, D.~W., Bailyn, C.~D., Corbel, S., \& Kaaret, P.\ 2002, \apj, 568, 901
728: \bibitem[Welsh et al.(2000)]{wry2000} Welsh, W.~F., Robinson, E.~L., \& Young, P.\ 2000, \aj, 120, 943
729: \bibitem[van der Klis(2006)]{vdk2006} van der Klis, M.\ 2006, In: Compact stellar X-ray sources. Edited by Walter Lewin \& Michiel van der Klis. Cambridge Astrophysics Series, No. 39. Cambridge, UK: Cambridge University Press, 39
730: \bibitem[Yu \& Dolence(2007)]{y2007} Yu, W., \& Dolence, J.\ 2007, \apj, 667, 1043
731: 
732: \end{thebibliography}
733: \clearpage
734: 
735: %% Table:1
736: \begin{deluxetable}{ccclc}
737: \tablecolumns{5}
738: \tablewidth{0pc}
739: \tablecaption{Date of peak flux during the outbursts of \aql\ between
740: 	MJD 50984 -- MJD 54411 (1998 June 20 -- 2007 Nov 7).
741: 	\label{tab:outburst_log}}
742: \tablehead{
743: 	\colhead{R-band peak} & \colhead{J-band peak} & \colhead{ASM peak} &
744: 	\colhead{Outburst type: Evolution of} & \colhead{Major Outburst} \\
745: 	\colhead{(MJD)} & \colhead{(MJD)} & \colhead{(MJD)} &
746: 	\colhead{light curve morphology} & \colhead{Number}
747: }
748: 
749: \startdata
750: %  (1) R-band peak (2) J-band peak (3) ASM peak (4) Type (5) # Major
751:  51076 & --    & --    & Mini				 & -- \\
752:  51320 & --    & 51321 & Major: FRED $\rightarrow$ LIS	 & 1  \\
753:  51676 & --    & --    & Mini				 & -- \\
754:  51730 & --    & --    & Mini				 & -- \\
755:  51831 & 51831 & 51837 & Major: FRED			 & 2  \\
756:  52086 & 52088 & 52087 & Major: FRED $\rightarrow$ LIS	 & 3  \\
757:  52192 & 52192 & --    & Mini				 & -- \\
758:  52226 & 52226 & --    & Mini				 & -- \\
759:  52338 & --    & 52336 & Major: FRED			 & 4  \\
760:  52707 & 52711 & 52705 & Major: FRED			 & 5  \\
761:  52923 & 52923 & --    & Mini				 & -- \\
762:  53150 & 53150 & 53167 & Major: LIS  $\rightarrow$ FRED  & 6  \\
763:  53268 & 53268 & --    & Mini				 & -- \\
764:  53473 & 53473 & 53476 & Major: FRED $\rightarrow$ LIS	 & 7  \\
765:  53579 & 53580 & --    & Mini				 & -- \\
766:  53950 & 53951 & 53952 & Major: FRED $\rightarrow$ LIS	 & 8  \\
767:  54242 & 54245 & 54253 & Major: FRED $\rightarrow$ LIS	 & 9  \\
768:  54363 & 54367 & 54365 & Major: FRED $\rightarrow$ LIS	 & 10
769: \enddata
770: \end{deluxetable}
771: 
772: %% Figures.
773: 
774: \clearpage
775: \begin{figure}
776: \begin{center}
777: %%\plotone{J-R-ASM.eps}
778: \plotone{f1.eps}
779: \caption[Multi-wavelength long-term monitoring of \aql]
780: {Multi-wavelength long-term monitoring of \aql\ between 
781: MJD 50984 -- MJD 54411 (1998 June 20 -- 2007 Nov 7).
782: %%
783: Top panel: Near-IR J-band light curve. 
784: %%
785: Middle panel: Optical R-band light curve. The numbers in
786: parentheses refer to the serial number assigned to the major outbursts in 
787: Table ~\ref{tab:outburst_log}.
788: %%
789: Bottom panel: Logarithm of soft X-ray (\about 1.5-12 keV) count-rate as 
790: measured by the All-Sky Monitor onboard the RXTE. Since the ASM does not
791: detect the source when in quiescence, only those ASM data are plotted for which
792: the signal-to-noise ratio is $>4$.
793: %%
794: The R and J-band fluxes include the flux from \aql\ as well as from
795: the contaminating star. During periods of quiescence, the soft X-ray flux is below the
796: detection limit of ASM but the combined flux from \aql+contaminating star remains 
797: detectable in both R and J bands. Note that a major outburst which occurred 
798: during late 2005 -- early 2006 could be seen only with the ASM but not the
799: ground based telescopes due to close angular proximity to the Sun. Therefore
800: we have not labelled this outburst in the middle panel.
801: \label{longterm}}
802: \end{center}
803: \end{figure}
804: 
805: \begin{figure}
806: \begin{center}
807: %\includegraphics{R-ASM.ob1.eps}
808: %\includegraphics{R-ASM.ob2.eps}
809: %\includegraphics{R-ASM.ob3.eps}
810: \includegraphics{f2a.eps}
811: \includegraphics{f2b.eps}
812: \includegraphics{f2c.eps}
813: \caption[Light curve during outbursts of \aql]
814: {Zooming into the R-band and ASM light curves shown in Fig.~\ref{longterm},
815: showing the detailed light curve morphology during major outbursts 1--3. 
816: FRED outbursts are marked by filled circles, LIS by open circles. 
817: In this figure we are using a linear scale for the ASM count-rates to 
818: emphasize the difference between LIS and FRED.
819: \label{zoom_lc_1-3}}
820: \end{center}
821: \end{figure}
822: 
823: \begin{figure}
824: \begin{center}
825: %\includegraphics{R-ASM.ob4.eps}
826: %\includegraphics{R-ASM.ob5.eps}
827: %\includegraphics{R-ASM.ob6.eps}
828: \includegraphics{f3a.eps}
829: \includegraphics{f3b.eps}
830: \includegraphics{f3c.eps}
831: \caption[Light curve during outbursts of \aql]
832: {Same as Fig.~\ref{zoom_lc_1-3}, but for outbursts 4--6.
833: \label{zoom_lc_4-6}}
834: \end{center}
835: \end{figure}
836: 
837: 
838: \begin{figure}
839: \begin{center}
840: %\includegraphics{R-ASM.ob7.eps}
841: %\includegraphics{R-ASM.ob8.eps}
842: %\includegraphics{R-ASM.ob9.eps}
843: \includegraphics{f4a.eps}
844: \includegraphics{f4b.eps}
845: \includegraphics{f4c.eps}
846: \caption[Light curve during outbursts of \aql]
847: {Same as Fig.~\ref{zoom_lc_1-3}, but for outbursts 7--9.
848: \label{zoom_lc_7-9}}
849: \end{center}
850: \end{figure}
851: 
852: \begin{figure}
853: \begin{center}
854: %\includegraphics{R-ASM.ob10.eps}
855: \includegraphics{f5.eps}
856: \caption[Light curve during outbursts of \aql]
857: {Same as Fig.~\ref{zoom_lc_1-3}, but for outburst 10.
858: \label{zoom_lc_10}}
859: \end{center}
860: \end{figure}
861: 
862: 	
863: \begin{figure}
864: \begin{center}
865: %\plotone{fluxes_all.eps}
866: \plotone{f6.eps}
867: \caption[Optical--soft-X-ray correlation: All outbursts combined]
868: {Optical--soft-X-ray correlation in \aql. Data from all the major outbursts
869: from MJD 50984 -- MJD 54411 (1998 June 20 -- 2007 Nov 7) 
870: for which the X-ray state could be determined from contemporaneous 
871: RXTE observations \citep[][also Manuel Linares, priv. comm.]
872: {mc2003a,mc2003b,rks2004,mb2004,rodriguez2006,y2007}. 
873: The dashed lines L1 and L2 which follow the equations $f_R=1.25f_{ASM} - 75$ 
874: and $f_R=1.5f_{ASM} + 270$ respectively, roughly form an envelope containing 
875: most of the FRED points.
876: The dotted lines L3 and L4 which follow the equations 
877: $f_R = 9f_{ASM} - 240$ and $f_R = 12f_{ASM} + 240$ respectively, roughly form
878: an envelope containing most of the LIS points.
879: Filled circles represent data taken during {\em soft} X-ray state and open
880: squares are those during {\em hard} X-ray state.
881: \label{allfluxes}}
882: \end{center}
883: \end{figure}
884: 
885: \begin{figure} 
886: \begin{center}
887: %\includegraphics[width=0.90\textwidth, angle=0]{opt-nir_zoom1.eps}
888: \includegraphics[width=0.90\textwidth, angle=0]{f7.eps}
889: \caption[Evolution of J-R color during various outbursts of \aql]
890: {Evolution of J-R color during outbursts 2, 3, 5 and 6 of \aql. The
891: increase in brightness from quiescence is shown in the top panels, with the 
892: R-band data shown by filled circles and J-band data shown by open
893: squares. The outburst serial number (Table.~\ref{tab:outburst_log})  is
894: given in parentheses. Variation of corresponding J-R color is shown in the
895: bottom panels. Epochs of LIS as inferred from 
896: Figs.~\ref{longterm}--\ref{zoom_lc_10} are labelled. Unless labelled as a LIS,
897: the OIR/soft-X-ray light curve morphology resembles that of FRED outbursts.
898: \label{r-j_1}} 
899: \end{center}
900: \end{figure}
901: \begin{figure} 
902: \begin{center}
903: %\includegraphics[width=0.90\textwidth, angle=0]{opt-nir_zoom2.eps}
904: \includegraphics[width=0.90\textwidth, angle=0]{f8.eps}
905: \caption[Evolution of J-R color during various outbursts of \aql]
906: {Same as Fig.~\ref{r-j_1}, but for outbursts 7--10.
907: \label{r-j_2}} 
908: \end{center}
909: \end{figure}
910: 
911: \begin{figure}
912: \begin{center}
913: %\includegraphics[width=0.73\textwidth, angle=-90]{cmd1.ps}
914: \includegraphics[width=0.73\textwidth, angle=-90]{f9.ps}
915: \caption[Evolution of the outbursts on OIR \cmd]
916: {Evolution of outbursts 2, 3, 5 and 6 of \aql\ on the OIR \cmd. 
917: Observations when the source was in hard/power law state during the rise 
918: of an outburst are shown by ($\bigtriangleup$). The decaying hard state, 
919: are shown by ($\bigtriangledown$). Outbursts where state transitions were 
920: not reported are shown by ($\sq$). Observations during soft/thermal 
921: dominated state are shown by ($\circ$). The quiescent level is shown by the 
922: dashed line. 
923: A typical error bar is shown in the top left panel near (1.4, 17.75). 
924: The solid, 
925: curve from bottom-right to top-left in each panel shows the 
926: color and brightness expected by heating a disk-shaped \bb\ of radius 5 
927: light seconds, inclined to the observer at $45\degr$ and at a distance of 3
928: kpc. The quantity $f$ represents a multiplicative area correction factor.
929: The expected color and magnitude of the disk between temperatures of 
930: 4000 --6000 K are labelled. 
931: \label{aql-cmd1}} 
932: \end{center}
933: \end{figure}
934: \begin{figure}
935: \begin{center}
936: %\includegraphics[width=0.73\textwidth, angle=-90]{cmd2.ps}
937: \includegraphics[width=0.73\textwidth, angle=-90]{f10.ps}
938: \caption[Evolution of the outbursts on OIR \cmd]
939: {Same as Fig.~\ref{aql-cmd1}, but for outbursts 7, 8, 9 and 10.
940: \label{aql-cmd2}} 
941: \end{center}
942: \end{figure}
943: 
944: \end{document}
945: