1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass{emulateapj}
3:
4: \usepackage{lscape}
5:
6: \begin{document}
7:
8: \title{Observation of the first gravitational microlensing event in a sparse stellar field : the Tago event}
9:
10: \author{
11: A. Fukui\altaffilmark{1},
12: F. Abe\altaffilmark{1},
13: K. Ayani\altaffilmark{2},
14: M. Fujii\altaffilmark{4},
15: R. Iizuka\altaffilmark{5},
16: Y. Itow\altaffilmark{1},
17: K. Kabumoto\altaffilmark{3},
18: K. Kamiya\altaffilmark{1},
19: T. Kawabata\altaffilmark{1},
20: S. Kawanomoto\altaffilmark{6},
21: K. Kinugasa\altaffilmark{7},
22: R. A. Koff\altaffilmark{8},
23: T. Krajci\altaffilmark{9},
24: H. Naito\altaffilmark{5},
25: D. Nogami\altaffilmark{10},
26: S. Narusawa\altaffilmark{5},
27: N. Ohishi\altaffilmark{6},
28: K. Ohnishi\altaffilmark{11},
29: T. Sumi\altaffilmark{1},
30: F. Tsumuraya\altaffilmark{5}
31: }
32:
33: \altaffiltext{1}{Solar Terrestrial Environment Laboratory, Nagoya University, Nagoya, 464-8601, Japan.
34: e-mail:{\tt afukui,abe,itow,kawabata, kkamiya,sumi@stelab.nagoya-u.ac.jp}}
35: \altaffiltext{2}{Bisei Astronomical Observatory, 1723-70 Ohkura, Bisei, Okayama 714-1411, Japan}
36: \altaffiltext{3}{Okayama Astronomical Museum, 3037-5 Honjo, Kamogata, Asakuchi, Okayama 719-0232, Japan}
37: \altaffiltext{4}{ Fujii-Bisei Observatory, 4500 Kurosaki, Tamashima, Okayama 713-8126, Japan}
38: \altaffiltext{5}{Nishi-Harima Astronomical Observatory, Sayo-cho, Hyogo 679-5313, Japan}
39: \altaffiltext{6}{National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan}
40: \altaffiltext{7}{ Gunma Astronomical Observatory, 6860-86 Nakayama Takayama, Agatsuma, Gunma 377-0702, Japan}
41: \altaffiltext{8}{Antelope Hills Observatory, 980 Antelope Drive West Bennett, CO 80102, USA. }
42: %e-mail:{\tt bob@antelopehillsobservatory.org}
43: \altaffiltext{9}{Center for Backyard Astrophysics, Cloudcroft, NM 88317, USA.}
44: %e-mail:{\tt tom\_krajci@tularosa.net}
45: \altaffiltext{10}{Hida Observatory, Kyoto University, Kamitakara, Gifu 506-1317, Japan. e-mail:{\tt nogami@kwasan.kyoto-u.ac.jp}}
46: \altaffiltext{11}{Nagano National College of Technology, Tokuma 716, Nagano 381-8550, Japan. e-mail:{\tt kouji.ohnishi@nao.ac.jp}}
47:
48: %---------------------------------------------
49:
50: \begin{abstract}
51:
52: We report the observation of the first gravitational microlensing
53: event in a sparse stellar field, involving the brightest (V=11.4 mag) and closest ($\sim 1 $kpc) source star to date. This event was discovered by an amateur astronomer,
54: A. Tago, on 2006 October 31 as a transient brightening, by $\sim4.5$
55: mag during a $\sim$ 15 day period, of a normal A-type star
56: (GSC 3656-1328) in the Cassiopeia constellation.
57: Analysis of both spectroscopic
58: observations and the light curve indicates that this event was
59: caused by gravitational microlensing rather than
60: an intrinsically variable star. Discovery of this single event over a
61: 30 year period is roughly consistent with the expected microlensing
62: rate for the whole sky down to $V = 12$ mag stars. However, the
63: probability for finding events with such a high magnification ($\sim
64: 50)$ is much smaller, by a factor $\sim 1/50$, which implies that
65: the true event rate may be higher than expected. This discovery
66: indicates the potential of all sky variability surveys, employing
67: frequent sampling by telescopes with small apertures and wide fields
68: of view, for finding such rare transient events, and using the
69: observations to explore galactic disk structure and search for
70: exo-planets.
71: \end{abstract}
72:
73: \keywords{
74: Galaxy: disk -- stars: individual (GSC 3656-1328) -- gravitational lensing
75: %disk---Galaxy: disk -- stars: other
76: }
77:
78: %----------------------------------------------
79:
80: \section{Introduction}
81:
82: \label{sec:introduction}
83:
84: The idea that a star's gravity magnifies the light from a perfectly
85: aligned background source star, so-called gravitational
86: microlensing, was first presented by \cite{ein36} and developed by
87: \cite{ref64}. However, it was thought very difficult to observe such
88: an event due to its rareness. \cite{pac86} first estimated realistic
89: rates for microlensing toward crowded stellar fields, such as the
90: Large and Small Magellanic Clouds and the Galactic Bulge (GB,
91: \citealt{pac91}), to study Galactic dark matter, Galactic structure
92: and extrasolar planets. The first microlensing candidates were
93: reported by the MACHO (\citealt{alc93}), EROS (\citealt{aub93}) and OGLE
94: (\citealt{uda93}) collaborations. Following this, a few thousands of
95: microlensing events have been detected to date, mostly towards the GB.
96: Currently, the OGLE and MOA (\citealt{sum03}) collaborations are
97: detecting $\sim 600$ events every year.
98: All of these events have been found only in distant, dense stellar
99: fields. This is because the rate of microlensing is very small,
100: $\sim10^{-5}$ events/star/yr even toward the GB, which is the most
101: dense stellar field in the sky. The rate for the whole sky down to
102: $V=12$ mag for the source stars is estimated to be $\sim 0.05 - 0.2$ events/yr
103: (\citealt{nem98}, \citealt{han07}).
104:
105: Here we report the observation of the first microlensing event for a
106: very close ($\sim 1$ kpc), bright star as a microlensing source that is not within a dense
107: stellar field. We present spectroscopic follow-up data, which are used
108: to judge that this event is most likely due to microlensing rather
109: than an intrinsically variable star. We also present observed light
110: curves to demonstrate that the microlensing hypothesis is plausible.
111: In \S \ref{sec:discovery} we detail how the event was discovered. \S
112: \ref{sec:spectroscopic} contains the spectroscopic analysis, and \S
113: \ref{sec:photometric} is devoted to the photometric observations and
114: the light curve analysis. Discussion and conclusions are given in \S
115: \ref{sec:conclusion}.
116:
117: %----------------------------------------
118:
119: \section{Discovery}
120:
121: \label{sec:discovery}
122:
123: On 2006 October 31, an amateur astronomer A. Tago found that the star
124: GSC 3656-1328 at (R.A., Dec., J2000.0)$=$$(00h09m21.9948s,$
125: $+54^\circ39'43.832")$, i.e., ($l,b$)=$(116.8158^\circ, -07.7092^\circ)$, near
126: the Cassiopeia constellation, had brightened by $\sim 4.5$ mag.
127: He used a commercial digital camera with 70 mm f/3.2 optics, which has
128: a $\sim$400 $deg^2$ field of view. This discovery was rapidly
129: reported to the Central Bureau for Astronomical Telegrams
130: \footnotemark\footnotetext{\tt http://cfa-www.harvard.edu/iau/cbat.html}
131: (CBAT; CBET 711) and the Variable Star NETwork
132: (VSNET;\citealt{kat04}). Following this, a special alert was
133: announced by the American Association of Variable Star Observers
134: \footnotemark\footnotetext{\tt http://www.aavso.org/} (AAVSO; ASN
135: 22). Meanwhile, Y. Sakurai, an amateur astronomer, independently
136: discovered the same event.
137:
138: The source star is a normal A-type star with V=11.4 mag at 1 kpc (CBET
139: 711, 712). A. Tago had discovered this event during his systematic nova
140: survey along the Galactic disk, covering $\sim 4500$ $deg^2$ in
141: total down to $\sim$ 12 mag, and sampling once every other
142: night. Such surveys have been conducted for more than 30 years by
143: many amateur astronomers, including him. The object was found by
144: visually comparing successive images with a reference image using the
145: eye-blinking method.
146:
147: Spectroscopic and photometric follow-up observations were promptly
148: obtained, and as the spectra contained no known signatures of variable
149: stars, the possibility that the brightening was due to microlensing
150: was reported to the Astronomer's Telegram \footnotemark\footnotetext{\tt
151: http://www.astronomerstelegram.org/?read=931} (ATEL; ATEL 931)
152: web site.
153:
154: %-----------------------------------------------
155:
156: \section{Spectroscopic observations}
157:
158: \label{sec:spectroscopic}
159:
160: We started our spectroscopic campaign on this variable object 1
161: night after we received the report on CBET 711. Five sites were
162: employed: the Okayama Astrophysical Observatory (OAO), Bisei Astronomical
163: Observatory (BAO), Fujii Bisei Observatory (FBO), Gunma Astronomical
164: Observatory (GAO), and Nishi-Harima Astronomical Observatory (NHAO).
165: For the 2006 November/December period, spectra were obtained as
166: follows: 2006 November 1 (GAO, FBO, BAO, and OAO), November 4 (FBO and OAO),
167: November 8 (FBO and GAO), November 12 (BAO), December 2 (NHAO), and
168: December 11 (GAO). See Table \ref{tbl:telescope} for instrumental details.
169:
170: All frames were reduced in the standard manner using IRAF, CCDOPS, and
171: MaxIm. The latter two software packages were used for dark
172: subtraction and flat-field averaging in the case of the FBO data.
173: Flux calibration using a local standard star was applied to all
174: spectra except for those obtained at OAO with the HIgh Dispersion
175: Echelle Spectrograph (HIDES; \citealt{izu99}). The typical error in
176: the flux calibration is estimated to be 10\%-20 \%.
177:
178: Low-dispersion spectra were obtained in the period 2006 November 1 to
179: 2006 December 11. We continued the campaign for about 1 month after
180: the object returned to its quiescent state in order to check for any
181: spectral variation. All the spectra are displayed in Fig.
182: \ref{fig:low-dispersion}. The Vega spectrum shown at the top of this
183: figure was retrieved from the Medium resolution INT Library of
184: Empirical Spectra (MILES \footnotemark\footnotetext{\tt
185: http://www.ucm.es/info/Astrof/miles/miles.html}). It is clear that
186: all these spectra have the same blue trend in the continuum, which is
187: independent of the brightness of the object. This trend well matches
188: that in the Vega spectrum, and the wavelengths at the peak intensities
189: are also consistently close to that of Vega. These facts suggest that
190: the variable object has a spectrum of a normal early A-type dwarf, both
191: during the bright state and when in quiescence.
192:
193: The high-dispersion spectra normalized to a unity continuum value are
194: shown in Fig. \ref{fig:high-dispersion}. These two spectra were
195: obtained on 2006 November 1 and 4, when the object was about 6.4 and 1.9
196: times brighter in V compared with quiescence, respectively (see Fig. \ref{fig:lightcurve}).
197: While the object's brightnesses were
198: different, both spectra contain the same absorption-line features
199: (Balmer, Mg II and Ca II K) in terms of depth, width, equivalent
200: width, and so on. This is clearly demonstrated by the ratio of these
201: two spectra, displayed in the bottom panel of Fig.
202: \ref{fig:high-dispersion}. In addition, the absorption-line features
203: are quite close to those in the Vega spectrum, and the differences of
204: the widths and depths of these lines between the target and Vega
205: should be attributable to the difference between the projected rotation
206: velocities of the two objects. Note that some of the end effects
207: resulting from the diffraction orders could not be completely
208: eliminated during the analysis and have remained in the continuum.
209:
210: In summary, our observations revealed that the optical spectrum of the
211: variable object did not vary from when the object was $\sim$6.4 times
212: brighter than its quiescent state until it returned to this state, and
213: that the spectrum is almost the same as that of Vega. Apart from
214: microlensing, we are not aware of another large-amplitude brightening
215: phenomenon with a timescale of days that does not change the object's
216: physical properties, such as the temperature and density. For
217: example, stellar flares and nova explosions are accompanied by strong
218: enhancement of Balmer emission lines, and dwarf nova-type outbursts
219: show a decay of the Balmer emission lines. The invariance of the
220: variable's spectral features, and the fact that they are close to
221: Vega's spectrum, provide firm evidence that the source object is a
222: normal main-sequence star of late-B to early-A type, and that this
223: brightening event was caused by gravitational microlensing.
224:
225: %----------------------------------------
226:
227: \section{Photometric observations and light curve analysis}
228:
229: \label{sec:photometric}
230:
231: Photometric observations were made by various observers before and
232: after the alert, using both commercial digital cameras and
233: scientific CCDs. A. Tago, Y. Sakurai, and Y. Sugawara obtained 22, 9,
234: and 1 JPEG images, respectively, using digital cameras (digicam data).
235: The observation times of data posted by A. Tago on the CBAT were initially
236: found to be wrong, but were then corrected (A. Tago 2007, private
237: communication). R.A.K and T.K. both obtained CCD
238: images of the variable and posted them on the AAVSO website.
239: The former provided 3635 V-band frames, while the latter took
240: 3472 frames without a filter. I-band (13 CCD images) and V-band (24
241: CCD images) were taken by the north site (Hawaii) of ASAS
242: (\citealt{poj02}) by chance as part their survey of the north sky, and
243: not prompted by the alert.
244:
245: The scientific CCD images, in the FITS format, were reduced in the
246: standard manner, and photometry of the variable was extracted.
247: However, the JPEG images taken by commercial digital cameras were
248: reduced with special care because both dark and flat-field images were
249: not available, and the flux linearity was affected by the lossy
250: compression method used. We estimated the uncertainties due to these
251: effects from the scatter of a standard star as follows.
252:
253: First, using three colors we added all the flux in the three images to
254: produce a single JPEG image and then converted this into a FITS
255: formatted image. Then about 100 stars in a $5^\circ \times 7^\circ$
256: area around the event star in each image were cross-referenced with
257: stars in the Tycho-2 catalog (\citealt{hog00}) for comparison purposes.
258: These stars were also chosen to have a similar color, B-V = $0.10 \pm
259: 0.1$, and they ranged from 6.9 to 12 mag in $V_T$, defined by Tycho-2.
260: The gain uniformity of the digital cameras in this region was
261: determined by examining sky-flat images taken by the cameras, and it
262: was found that they were flat over the area to better than $3\%$. We
263: then performed aperture photometry on these stars using IRAF, and
264: determined the relation between the instrumental magnitudes and the $V_T$
265: magnitudes by fitting a quadratic function, taking errors in both
266: magnitudes into account. Since the IRAF errors are relative values,
267: we renormalized these errors by a factor so that $\chi^2 / dof$ was
268: unity. The IRAF error of the event star was also renormalized by the
269: same factor. Finally, the IRAF magnitudes and associated errors of the
270: event star were converted into $V_T$ magnitudes using these functions
271: with their errors.
272:
273: To test whether the photometric light curve was consistent with the
274: microlensing model or not, we fitted the combined light curve with a
275: simple point-lens and point-source model, characterized by the three
276: parameters $u_{0}$, $t_{0}$ and $t_{E}$ (\citealt{pac86}), where
277: $u_{0}$ is the minimum impact parameter corresponding to the
278: source-lens angular separation in units of the angular Einstein radius $\theta_{E}$, $t_{0}$ is the
279: time of maximum magnification, and $t_{E}$ is the event timescale, defined as
280: $\theta_{E} / \mu$ where $\mu$ is the relative proper motion between the source and lens stars.
281: The baseline flux of the source star and any blended flux in each data
282: set were also allowed to be free parameters in the fitting procedure.
283:
284: In Table \ref{tbl:param} we list the resultant parameters and the errors
285: that were obtained for the different
286: data subsets by checking the systematics within each one. The
287: uncertainties in the ASAS and AAVSO data were estimated to be equal to
288: the rms of their baseline values. These were determined
289: using only those observations separated by an interval of at least
290: $3 t_{E}$ from the peak magnification, where the amplitudes are
291: expected to be sufficiently low. The value for $t_{E}$ was determined
292: from the fitting parameters of the digicam data set. The $\chi^2/dof$ in
293: each data set is close to unity except for the AAVSO data. We found
294: that the AAVSO points were scattered about the fitted curve
295: significantly more than expected from the estimated errors, with a timescale of $<$ 1 day.
296: This is because the reduction and calibration of
297: the data sets were not optimal (e.g. no air mass corrections were
298: made), and the errors were also underestimated for some reason.
299: Other than this, one can see that the microlensing model provides a
300: satisfactory fit to all the data sets except for some digicam data points near the peak
301: (refer to Fig.\ref{fig:lightcurve}). We suspect this deviation is due to the systematic errors
302: from lossy compression, and we think this has an insignificant effect on the conclusion of this paper.
303: Consequently, we renormalized the errors of
304: the R.A.K and T.K. data sets so that $\chi^2/dof$ in each one was unity,
305: where the fitted model values were determined using all data sets,
306: including the unnormalized AAVSO data (see "*" in Table
307: \ref{tbl:param}). The blended flux parameter for each data subset is
308: nearly negligible in comparison with the source flux, except for the
309: ASAS data set. In this case there were contributions from an
310: unresolved neighboring star due to the low resolution.
311:
312: Fig. \ref{fig:lightcurve} shows the complete photometric light curve,
313: including the renormalized AAVSO data, and the best fit model to all
314: the data. As shown in Table \ref{tbl:param}, the differences in
315: parameters for the various data subsets are relatively small, e.g.,
316: $\sim 3$ days in $t_{\rm E}$ at most, which is roughly the size of the
317: statistical errors. Therefore we conclude that systematic biases in
318: all data sets are negligible, and the aggregate light curve can be
319: represented well by a point-lens and point-source microlensing model.
320:
321: %-----------------------------------------
322:
323: \section{Discussion \& Conclusion}
324:
325: \label{sec:conclusion}
326:
327: We have reported the observation of the first microlensing event in a
328: sparse stellar field, involving the brightest (V=11.4 mag) and closest
329: (1 kpc) source star to date. The spectra of the source star and the
330: overall light curve show that this event cannot have been caused by any known
331: mechanism for intrinsically variable stars, but it can be naturally
332: explained as a microlensing event.
333:
334: \cite{gau07} also reported the discovery of the same event, and also
335: concluded that this event is most likely due to microlensing. A subset
336: of the data used in this paper is in common with theirs, but our
337: analysis was carried out independently, and we have original data
338: taken by amateur astronomers, which includes coverage of the peak
339: magnification region.
340:
341: The rate for the microlensing event described here is estimated to be
342: very small at $\sim 0.05 - 0.2$ events/yr over the whole sky
343: (\citealt{nem98}, \citealt{han07}), so it seems reasonable for one such event to happen
344: in a 30 year period.
345: However, this event is a high amplification one, so the corresponding
346: rate is much smaller ($\sim 1/50$) than that given above. This may be
347: just good luck, or the true event rate may be higher than expected but low-
348: amplitude events may be missed by the current nova surveys with digital
349: cameras, where the detection efficiency is uncertain. Such a
350: possibility can be addressed by systematic all-sky surveys featuring
351: frequent sampling with small apertures and wide fields of view, such
352: as LOTIS (\citealt{par02}), ROTSE (\citealt{woz04}) and ASAS
353: (\citealt{poj02}). We can expect to detect some microlensing
354: events/yr for all-sky surveys down to $V = 15$ mag (\citealt{nem98}, \citealt{han07}).
355:
356: Searching for exo-planets (\citealt{mao91}) in very bright microlensing events
357: is also a worthwhile exercise.
358: Although the expected number of these events is small, e.g. 1 planet every 4 years if assuming 5 microlensing events observed each year by an all-sky survey and 5\% planet abundance,
359: the detection efficiency for planets in such bright events is much higher than for faint events,
360: because the photometric precision of a bright event is higher (Peale 2001).
361: In addition, such high precision can be achieved even by small-aperture telescopes distributed around the world.
362: Even if we can detect only one planet in such an all-sky survey in 2-10 years, it will be the first
363: planet discovered by microlensing in the solar neighborhood which can
364: be confirmed by other methods such as spectroscopy or direct imaging after the peak amplification phase has ended.
365: This will be very important to convince people that the planetary signal in
366: the microlensing that we are also seeing towards the bulge is actually caused by the planet.
367:
368: This discovery shows the potential of all-sky variability surveys
369: using cheap, small telescopes for finding rare transient events, which
370: in turn will provide information on disk structure and possible
371: exoplanets.
372:
373: %---------------------------------------
374:
375: \acknowledgments
376:
377: We would like to express our deep gratitude to A. Tago, Y. Sakurai,
378: and Y. Sugawara for providing the significant data we used for this
379: analysis, and acknowledge T. Fukashima, T. Izumi, K. Kadota, S.
380: Kaneko, S. Kiyota, K. Nakajima, H. Nishimura, H. Maehara, Y. Okamoto,
381: S. Tago, M. Yamamoto, J. Yokomichi, and many other observers for
382: providing their observations or valuable information. And, special
383: thanks to H. Yamaoka for providing support for the data collection. We also
384: would like to thank the many amateur observers who obtained valuable data and
385: made them publicly available by sending them to VSNET, AAVSO, and VSOLJ.
386: We acknowledge G. Pojmanski and D. Szczygiel for providing the data of
387: ASAS. This work was supported by Grant-in-Aids (15540240, 17340074,
388: 18253002, 18749004, and 18740119) from MEXT and JSPS.
389:
390: \begin{thebibliography}{}
391: \bibitem[Alcock et al.(1993)]{alc93}Alcock, C. et al. 1993 Nature 365, 621
392: \bibitem[Aubourg et al.(1999)]{aub93}Aubourg, E., et al. 1993, Nature, 365, 623
393: %\bibitem[Bond et al.(2004)]{bon04}Bond, I, et al. 2004, \apj, 606, L155
394: \bibitem[Einstein (1936)]{ein36} A. Einstein, 1936, Science, 84, 506
395: \bibitem[Gaudi et al.(2007)]{gau07}Gaudi, S. et al. 2007, \apj, submitted (astro-ph/0703125v2)
396: \bibitem[Han (2007)]{han07}Han, C. 2007, preprint (astro-ph/0708.1215v1)
397: \bibitem[H{\o}g et al.(2000)]{hog00}H{\o}g E. et al. 2000, A\&A, 355, L27
398: \bibitem[Izumiura (1999)]{izu99}Izumiura, H., 1999, in Proc. 4th East Asian Meeting on Astronomy, ed. P.S. Chen (Kunming: Yunnan Observatory), 77
399: \bibitem[Kato et al. (2004)]{kat04}Kato, T., Uemura, M., Ishioka, R., Nagami, D., Kunjaya, C., Baba, H., Yamaoka, H., 2004, PASJ, 56, S1
400: \bibitem[Mao \& Paczy\'{n}ski(1991)]{mao91}Mao, S. \& Paczy\'{n}ski, B. 1991, \apj, 374, L37
401: \bibitem[Marshall (1997)]{nem97}Marshall, S., 1997, BAAS, 191 (48.15)
402: \bibitem[Nemiroff (1998)]{nem98}Nemiroff, R.J., 1998, \apj, 509, 39
403: \bibitem[Paczy\'{n}ski(1986)]{pac86}Paczy\'{n}ski, B. 1986, \apj, 304,1
404: \bibitem[Paczy\'{n}ski(1991)]{pac91}Paczy\'{n}ski, B. 1991, \apj, 371, L63
405: \bibitem[Park et al. (2002)]{par02}Park, H. S. et al. 2002, \apj, 571, L131
406: \bibitem[Peale (2001)]{pea01}Peale, S. J. 2001, \apj, 552, 889
407: \bibitem[Pojma\'{n}ski (2002)]{poj02}Pojma\'{n}ski, 2002, Acta Astron., 52, 397
408: \bibitem[Refsdal (1964)]{ref64} Refsdal, S. 1964, MNRAS, 128, 295
409: \bibitem[Udalski et al.(1993)]{uda93}Udalski et al., 1993, Acta Astronomica, 43, 69
410: \bibitem[Sumi et al.(2003)]{sum03}Sumi, T. et al. 2003, ApJ, 591, 204
411: \bibitem[Wo\'{z}niak et al.(2004)]{woz04}Wo\'{z}niak, P. R. et al. 2004, \aj, 127, 2436
412: \end{thebibliography}
413:
414: %---------------------------------------------------
415:
416: \begin{deluxetable}{rrlr}
417: \tabletypesize{\scriptsize}
418: \tablecaption{Telescopes and instruments used for spectroscopic observations.
419: \label{tbl:telescope}}
420: \tablewidth{0pt}
421: \tablehead{
422: Site$^{a}$ & Tel. & Instruments & WR$^{b}$ (A) \\
423: }
424:
425: \startdata
426: OAO & 188cm & HIDES$^{c}$ & 0.07 \\
427: BAO & 101cm & Low Dispersion Spectrograph & 5 \\
428: FBO & 28cm & Low Dispersion Spectrograph & 10 \\
429: GAO & 150cm & Low Dispersion Spectrograph & 13-14 \\
430: NHAO& 200cm & MALLS$^{d}$ & 6 \\
431: \enddata
432:
433: \tablecomments{
434: $^{a}$ OAO: Okayama Astrophysical Observatory,
435: BAO: Bisei Astronomical Observatory,
436: FBO: Fujii Bisei Observatory,
437: GAO: Gunma Astronomical Observatory,
438: NHAO: Nishi-Harima Astronomical Observaotry,
439: $^{b}$ WR: Wavelength resolution,
440: $^{c}$ HIDES: HIgh Dispersion Echelle Spectrograph (Izumiura 1999),
441: $^{d}$ MALLS: Medium And Low-dispersion Long slit Spectrograph
442: }
443:
444: \end{deluxetable}
445:
446:
447: \begin{deluxetable}{lrrrrrr}
448: \tabletypesize{\scriptsize}
449: \tablecaption{Lensing fit parameters.
450: \label{tbl:param}}
451: \tablewidth{0pt}
452: \tablehead{
453: dataset & $t_0 (HJD - 2450000)$ & $u_{\rm 0}$ & $t_{\rm E} (day)$ & $N$ & $\chi^2/dof$ \\
454: }
455:
456: \startdata
457: digicam & 4039.93 (0.02) & 0.0197 (0.0134) & 4.70 (2.87) & 32 & 1.04\\
458: ASAS & 4039.99 (0.13) & 0.0221 (0.0129) & 6.20 (0.95) & 37 & 1.08\\
459: AAVSO & 4039.87 (0.006) & 0.0184 (0.0029) & 6.98 (0.02) & 7107 & 6.75\\
460: AAVSO$^*$ & 4039.98 (0.02) & 0.0367 (0.0038) & 7.74 (0.10) & 7107 & 0.97\\
461: ASAS+AAVSO$^*$ & 4039.92 (0.005) & 0.0229 (0.0014) & 7.53 (0.05) & 7144 & 0.98\\
462: digicam+AAVSO$^*$ & 4039.89 (0.004) & 0.0107 (0.0018) & 7.48 (0.004) & 7139 & 0.98\\
463: digicam+ASAS & 4039.98 (0.01) & 0.0148 (0.0020) & 7.24 (0.66) & 69 & 1.27\\
464: digicam+ASAS+AAVSO & 4039.88 (0.002) & 0.0199 (0.0010) & 6.99 (0.02) & 7176 & 6.71\\
465: digicam+ASAS+AAVSO$^*$ & 4039.91 (0.003) & 0.0189 (0.0011) & 7.53 (0.04) & 7176 & 0.99\\
466: \enddata
467:
468: \tablecomments{
469: digicam: Data from digital cameras. ASAS: ASAS I and V band. AAVSO: CCD data from R.A.K. and T.K.
470: *: Errors in R.A.K. and T.K. data sets are rescaled independently by multiplying the factors so that $\chi^2/dof$ in each data set was unity, calculated from a model fitted by using all data sets including the unnormalized AAVSO data. The numbers in parentheses represent 1-$\sigma$ statistical
471: errors which corresponding to the values where $\Delta \chi^{2}$ is 1.
472: }
473:
474: \end{deluxetable}
475:
476: %--------------------------------------------
477:
478: \begin{figure}[ht]
479:
480: \epsscale{0.75}
481: \includegraphics[angle=-90,scale=0.75,keepaspectratio]{f1.eps}
482: \caption{
483: Low-dispersion spectra obtained between 2006 November
484: 1, and 2006 December 11, and a spectrum of Vega (A0V).
485: The vertical axis is the relative flux using a logscale. The spectra
486: have been arbitrarily scaled in the vertical direction for visual purposes.
487: The Vega spectrum was retrieved from the Medium resolution INT
488: Library of Empirical Spectra (MILES).
489: (a), (f) and (i) were obtained at GAO. (b), (d) and (e) were taken at FBO.
490: (c) and (g) were taken at BAO and (h) was at NHAO.
491: The blue trend of the continuum did not change within the error of the flux
492: calibration, and is in good agreement with that of the Vega spectrum.
493: \label{fig:low-dispersion}}
494: \end{figure}
495:
496: \begin{figure}[ht]
497: \epsscale{0.5}
498: \includegraphics[angle=-90,scale=0.75,keepaspectratio]{f2.eps}
499: \caption{
500: High-dispersion spectra of the event star obtained on
501: 2006 November 1 and 4 at the OAO. We have also included a spectrum
502: of Vega for comparison purposes: it was obtained at Steward
503: Observatory using a 90 inch(2.3m) telescope, and is publicly available
504: on the Web (http://stellar.phys.appstate.edu/stdSO90.html).
505: These spectra have been normalized to a unity continuum value and
506: arbitrary shifted in the y axis direction for visual purposes.
507: The absorption features of H$\gamma, \delta,$ and H$\epsilon$,
508: Mg II and Ca II K did not change between the observations and are very
509: close to those of Vega. The bottom panel shows
510: the ratio of these two spectra ((j)/(k)).
511: \label{fig:high-dispersion}}
512: \end{figure}
513:
514: \begin{figure}[ht]
515: \epsscale{0.5}
516: \includegraphics[angle=-90,scale=0.6,keepaspectratio]{f3.eps}
517: \caption{
518: The light curve of the Tago event. The solid line indicates the best fit
519: point-lens and point-source model. Errors in the AAVSO data
520: have been re-normalized as noted in the text. The arrows (a) - (k)
521: indicate the epochs of the spectroscopic measurements depicted in
522: Fig \ref{fig:low-dispersion} and Fig \ref{fig:high-dispersion}.
523: \label{fig:lightcurve}}
524: \end{figure}
525:
526: \end{document}
527:
528: