1: \documentclass[twocolumn]{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: %\documentclass[manuscript]{aastex}
4: %\usepackage{emulateapj5}
5: %\documentclass{emulateapj}
6: %% preprint2 produces a double-column, single-spaced document:
7: %\documentclass[preprint2]{aastex}
8:
9: \usepackage{natbib}
10: \usepackage{epsfig}
11: \usepackage{lscape}
12: \usepackage{graphicx}
13: \bibliographystyle{apj}
14:
15: %% You can insert a short comment on the title page using the command below.
16: \slugcomment{To Appear in ApJ}
17:
18: \shorttitle{{\it Spitzer} 24$\mu$m Time-Series Observations of GU Boo}
19:
20: \shortauthors{von Braun et al.}
21:
22: %----------------------------------------------------------------------
23:
24: \begin{document}
25:
26: \title{{\it Spitzer} 24\micron\ Time-Series Observations of
27: the Eclipsing M-dwarf Binary GU Bo\"otis }
28:
29: \author{Kaspar von Braun\altaffilmark{1}}
30: \author{Gerard T. van Belle\altaffilmark{1,4}}
31: \author{David R. Ciardi\altaffilmark{1}}
32: \author{Mercedes L\'{o}pez-Morales\altaffilmark{2}}
33: \author{D.W. Hoard\altaffilmark{3}}
34: \author{Stefanie Wachter\altaffilmark{3}}
35:
36: \altaffiltext{1}{Michelson Science Center, California Institute of
37: Technology, MC 100-22, Pasadena, CA 91125;
38: kaspar, gerard, ciardi@ipac.caltech.edu}
39:
40: \altaffiltext{2}{Carnegie Fellow, Department of Terrestrial Magnetism,
41: Carnegie Institution of Washington, 5241 Broad Branch Rd. NW,
42: Washington, DC 20015; mercedes@dtm.ciw.edu}
43:
44: \altaffiltext{3}{{\it Spitzer} Science Center, California Institute of
45: Technology, MC 220-6, Pasadena, CA 91125;
46: hoard, wachter@ipac.caltech.edu}
47:
48: \altaffiltext{4}{European Southern Observatory, Karl-Schwarzschild-Str. 2,
49: 85748 Garching, Germany; gerard.van.belle@eso.org}
50:
51: %----------------------------------------------------------------------
52:
53: \begin{abstract}
54:
55: We present a set of {\it Spitzer} 24$\mu$m MIPS time series observations of
56: the M-dwarf eclipsing binary star GU Bo\"otis. Our data cover three secondary
57: eclipses of the system: two consecutive events and an additional eclipse six
58: weeks later. The study's main purpose is the long wavelength (and thus limb
59: darkening-independent) characterization of GU Boo's light curve, allowing for
60: independent verification of the results of previous optical studies. Our
61: results confirm previously obtained system parameters. We further compare GU
62: Boo's measured 24$\mu$m flux density to the value predicted by spectral
63: fitting and find no evidence for circumstellar dust. In addition to GU Boo, we
64: characterize (and show examples of) light curves of other objects in the field
65: of view. Analysis of these light curves serves to characterize the photometric
66: stability and repeatability of {\it Spitzer's} MIPS 24\micron\ array over
67: short (days) and long (weeks) timescales at flux densities between
68: approximately 300--2,000$\mu$Jy. We find that the light curve root mean square
69: about the median level falls into the 1--4\% range for flux densities higher
70: than 1mJy. Finally, we comment on the fluctuations of the 24\micron\ background
71: on short and long timescales.
72:
73: \end{abstract}
74:
75: %----------------------------------------------------------------------
76:
77: \keywords{techniques: photometric, infrared: stars, stars: fundamental
78: parameters, stars: binaries: eclipsing, stars: individual: GU Boo,
79: circumstellar matter, dust}
80:
81: %----------------------------------------------------------------------
82:
83: \section{Introduction}\label{introduction}
84:
85: GU Bo\"otis is a nearby, low-mass eclipsing binary system, consisting of two
86: nearly equal mass M-dwarfs \citep{lmr05}. It is one of currently very few
87: ($\sim$5) known nearby ($<$ 200 pc) double-lined, detached eclipsing binary
88: (DEB) systems composed of two low-mass stars \citep{lm07}. Eclipsing binaries
89: can be used as tools to constrain fundamental stellar properties such as mass,
90: radius, and effective temperature. Given the fact that over 70\% of the stars
91: in the Milky Way are low-mass objects with $M< 1 M_\odot$ \citep{hik97},
92: coupled with the considerable uncertainty over the mass-radius relation for
93: low-mass stars, objects such as GU Boo are of particular interest in exploring
94: the low-mass end of the Hertzsprung-Russell diagram.
95:
96: While simultaneous analysis of DEB light curves and radial velocity (RV)
97: curves provides insight into the component masses and physical sizes, an
98: estimate of their intrinsic luminosities can only be made with the knowledge
99: of the amount and properties of dust along the line of sight. As such, it is
100: important to understand whether low-mass DEB systems used to constrain stellar
101: models contain dust which, in turn, may lead to an underestimate of their
102: surface temperatures and thus luminosities. This problem has been documented
103: in \citet{dfm99,mlg01,tr02,ribas03}. In particular, \citet{ribas03} states
104: that the most likely explanation for the temperature discrepancy between
105: observations and models for the low-mass DEB CU Cancri is the presence of
106: either circumstellar or circumbinary dust. The detection of dust in any
107: system such as GU Boo would therefore additionally shed insight into formation
108: and evolution of the low-mass DEBs.
109:
110: The characterization of the effects of limb darkening and star spots
111: introduces additional free parameters and thus statistical uncertainty in the
112: calculation of the stellar radii and masses. Using the {\it Spitzer Space
113: Telescope}, we obtained 24$\mu$m time series observations of three separate
114: instances of GU Boo's secondary eclipse (see \S \ref{observations}) to create
115: a light curve far enough in the infrared to not be contaminated by the effects
116: of limb darkening and star spots. We purposely timed the observations such
117: that each secondary eclipse event is preceded by a sufficient length of time
118: to establish GU Boo's out-of-eclipse flux density in order to detect any
119: infrared excess possibly caused by thermal dust emission.
120:
121: A further goal of our study is to characterize the photometric stability of
122: the Multiband Imaging Photometer (MIPS) on {\it Spitzer} at 24$\mu$m over
123: short and long time scales, similar to what was done for bright objects in \S
124: 5 of \citet{rye04}. Time-series observing is atypical (albeit increasingly
125: common) for {\it Spitzer}, which is the reason why there are very few
126: published photometric light curves based on {\it Spitzer} observations. The
127: recent spectacular observations of primary and secondary eclipses of
128: transiting planets are notable exceptions \citep[see for
129: instance][]{cam05,dsr05,cac07,dhl07,gdb07,kca07}. Of these, the \citet{dsr05}
130: study was performed at 24$\mu$m. We therefore observed two consecutive
131: secondary eclipses of GU Boo ($\sim$ 12 hours apart), and then a third event
132: about six weeks later (see Table \ref{table1_AORs}).
133: %In addition to providing the scientific return outlined above, GU Boo was
134: %chosen as a target due to the fact that it is of similar brightness in the
135: %near infrared ($K_{s} \sim 10.22$) as TrES-1 ($K_{s} \sim 9.8$) \citep{csd03}.
136:
137: We describe our observations and data reduction methods in \S
138: \ref{obs_red} and discuss our findings with respect to {\it Spitzer's}
139: photometric stability in \S \ref{repeatability}. The analysis of GU Boo's
140: light curve is described in \S \ref{analysis}. We probe for the existence of
141: an infrared excess in GU Boo's spectral energy distribution in \S
142: \ref{flux_comparison}. In \S \ref{other_lightcurves}, we show light curves of
143: other well sampled objects in the field along with a brief summary of their
144: respective properties, and we summarize and conclude in \S \ref{conclusions}.
145:
146: %----------------------------------------------------------------------
147:
148: \section{{\it Spitzer} Observations and Data Reduction}\label{obs_red}
149:
150: %----------------------------------------------------------------------
151:
152: \subsection{Observations}\label{observations}
153:
154: We used the MIPS 24\micron\ array aboard the {\it Spitzer Space Telescope}
155: \citep{wrl04} to observe GU Boo in February and April of 2006, as
156: outlined in Table \ref{table1_AORs}. The MIPS 24\micron\ array (MIPS-24), is a
157: Si:As detector with 128 $\times$ 128 pixels, an image scale of 2.55"
158: pixel$^{-1}$, and a field of view of 5.4' $\times$ 5.4' \citep{rye04}. Our
159: exposures were obtained using the standard MIPS 24$\mu$m small field
160: photometry pattern, which consists of four cardinal dither positions located
161: approximately in a square with 2 arcmin on a side. For two of these cardinal
162: positions, there are four smaller subposition dithers (offset by $\sim$ 10
163: arcsec), and for the other two cardinal positions, there are three such
164: subpositions. This results in a dither pattern in which {\it Spitzer} places
165: the star at 14 different positions on the array.
166:
167: Our goal was to observe three independent secondary eclipses of GU Boo: two
168: consecutive ones and another one several weeks after the first two
169: \citep{bbc07}. Of our total of nine of {\it Spitzer's} Astronomical
170: Observation Requests (AORs), three were used for each secondary eclipse event
171: (see Table \ref{table1_AORs}). Each AOR contained eight
172: exposures\footnote{The term ``exposure'' here can be thought of as a cycle of
173: observations, but since the word ``cycle'' is reserved for another unit of
174: {\it Spitzer} data collection, it carries the somewhat misleading name
175: ``exposure''.} with 36 individual Basic Calibrated Data (BCD) frames each. The
176: first BCD in each exposure is 9s long, the subsequent 35 are 10s long. The
177: first two BCDs of every exposure were discarded due to a ``first frames
178: effect''. This procedure left 34 BCDs per exposure, 272 BCDs per AOR, 816 BCDs
179: per secondary eclipse event, and 2448 BCDs for the entire project (all 10s
180: exposure time).
181:
182: For background information on {\it Spitzer} and MIPS, we refer the reader to
183: the Spitzer Observer's Manual, obtainable at
184: \url{http://ssc.spitzer.caltech.edu/documents/som/}. For information
185: specifically related to MIPS data reduction, please consult the MIPS Data
186: Handbook (MDH -- \url{http://ssc.spitzer.caltech.edu/mips/dh/}) and
187: \citet{gre05}.
188:
189: %----------------------------------------------------------------------
190:
191: \input{tab1.tex}
192:
193: %----------------------------------------------------------------------
194:
195: \subsection{Data Processing and \textsf{mopex} $/$ \textsf{apex}
196: Reduction} \label{reductions}
197:
198: %----------------------------------------------------------------------
199:
200: \subsubsection{Mosaicing}
201: \label{mosaicing}
202:
203: The MIPS-24 data are provided by the {\it Spitzer Archive} in the
204: (flatfielded) BCD format. We applied further post-processing to these data in
205: order to correct for small scale artifacts, in particular using
206: IRAF's\footnote{IRAF is distributed by the National Optical Astronomy
207: Observatory, which is operated by the Association of Universities for Research
208: in Astronomy, Inc, under cooperative agreement with the National Science
209: Foundation.} CCDRED package to remove the weak ``jailbar'' features in the
210: images (as described in the MDH).
211:
212: The {\it Spitzer} software package \textsf{mopex} \citep{mk05,mm05} was used
213: to co-add the individual MIPS BCD frames into mosaics of 17 frames, using overlap
214: correction and outlier rejection in the process. The choice of 17 frames was
215: made to balance three aspects:
216:
217: \begin{enumerate}
218: \item We want to obtain sufficient signal-to-noise ratio (SNR) for measured
219: flux densities in the combined images and subsequent data points in the light
220: curves (SNR $>$ 10 for GU Boo; see Table \ref{tab3_obj_stats}).
221: \item We need to maintain a sufficiently high effective observing
222: cadence to temporally resolve elements of GU Boo's light curve for fitting
223: purposes.
224: \item We do not want to be forced to combine frames from different exposures
225: into a single light curve data point (see \S \ref{observations}).
226: \end{enumerate}
227: The interpolated, remapped mosaics have a pixel scale of 2.45''
228: pixel$^{-1}$. We show in Figure \ref{mosaic_AOR} the MIPS-24 field of view of
229: GU Boo.
230: \begin{figure}
231: \plotone{f1.eps}
232: \caption{\label{mosaic_AOR} A {\it Spitzer} MIPS 24$\mu$m mosaic of
233: GU Boo (marked with a circle at the center of the image). This mosaic was
234: created using all 272 frames in one AOR and is about 8 arcmin on a side. North
235: is up, east is to the left. The change in noise structure as a function of
236: position is due to different effective exposure times (only the inner $\sim$
237: 3' $\times$ 3' were covered by all 272 BCD frames). The white specks in the NW
238: corner are flatfielding residuals, caused by a fleck of paint or dust grain on
239: the pickoff mirror (from {\it Spitzer's} launch), imaged at the four cardinal
240: dither positions (see \S \ref{observations}).}
241: \end{figure}
242:
243: %----------------------------------------------------------------------
244:
245: \subsubsection{Photometry}
246:
247: For photometric reductions of the mosaiced images, we utilized the
248: \textsf{apex} component of \textsf{mopex} to perform point-source extraction
249: as described in \citet{mm05}\footnote{Also see information on \textsf{apex} at
250: \url{http://ssc.spitzer.caltech.edu/postbcd/apex.html} and the User's Guide
251: at \url{http://ssc.spitzer.caltech.edu/postbcd/doc/apex.pdf}}. This step
252: included background subtraction of the images, and the fitting of a resampled
253: point response function (PRF). In order to match the PRF centroid as closely
254: as possible to the centroid of the stellar profile, the first Airy ring is
255: initially subtracted from the stellar profiles, and the source detection
256: happens on the resulting image. Photometry of the detected sources is then
257: performed on the original images.
258:
259: For single frame photometry on mosaiced images, \textsf{apex} provides the
260: option of using a template model PRF produced by the analysis of $\sim$ 20
261: bright, isolated stars in the {\it Spitzer Archive}, or using one's own data
262: to create a PRF. It furthermore allows for a resampling factor in both x and y
263: directions. The scatter in our light curves was minimized when using the model
264: PRF provided by the {\it Spitzer Science Center}, oversampled by a factor of 4
265: in both x and y directions. Using the PRF created from our own data or
266: sampling any PRF to a higher resolution resulted in noticeably larger root
267: mean square (rms) dispersion in our light curves, most likely due to
268: systematic errors introduced in the low SNR regime of our data (see Table
269: \ref{tab3_obj_stats}).
270:
271: We note that, currently, \textsf{apex} only provides the option of using a
272: synthetic PRF \citep[Tiny
273: Tim\footnote{\url{http://ssc.spitzer.caltech.edu/archanaly/contributed/stinytim.}};][]{krist93}
274: for photometry on individual BCD frames (i.e., not mosaiced), as applied by
275: \citet{dsr05} and \citet{rhs06}. Since the flux density of our target star
276: ($\sim$ 600$\mu$Jy) is so much lower than HD 209458 \citep[$\sim$22mJy;][]{dsr05}, our SNR regime did not allow for performing photometry on
277: single BCD frames.
278:
279: %----------------------------------------------------------------------
280:
281: \subsubsection{Background Fluctuations}
282: \label{background_fluctuations}
283:
284: The \textsf{apex} error analysis is described in the \textsf{apex} User's
285: Guide, and parts of it can be found in \citet{mml02,ml05,mm05}. We briefly
286: summarize the general idea here. Errors in the photometry are dominated by the
287: statistical background fluctuations in the images. These fluctuations are
288: calculated per pixel by estimating the Gaussian noise inside a sliding window
289: whose size is defined by the user (45 $\times$ 45 interpolated pixels in this
290: case). Thus, \textsf{apex} produces ``noise tiles'' for the computation of the
291: SNR of the point sources in the corresponding mosaiced image tiles (see column
292: 7 in Table \ref{tab3_obj_stats}).
293:
294: To provide an estimate of the background fluctuations from image to image, we
295: show in Fig. \ref{background} the surface brightness for every image in the
296: three observing sets. These estimates were obtained by calculating the median
297: surface brightness level for the inner 90\% of the image (in area). The error
298: bars correspond to the standard deviation about this median over the same
299: area. Note that the fluctuations of the background within observing sets are
300: very small, but they are different for the temporally offset observing set 3
301: (see Table \ref{table1_AORs}). Surface brightness values are given in the {\it
302: Spitzer} native units of MJy per steradian. The surface brightness values in
303: the cores of the brightest objects in the field are typically 1--1.5 MJy/sr
304: above the background level (23--25 MJy/sr). A linear fit (weighted by the
305: standard deviation values of the data points) to observing sets 1 \& 2 returns
306: a slope of $-0.033 \pm 0.023$ MJy/sr/day. The same fit for all three observing
307: sets produced a statistically consistent slope of $-0.030 \pm 0.001$
308: MJy/sr/day, indicating a smoothly decreasing background level over the course
309: of our observations (Table \ref{table1_AORs}).
310:
311: The {\it Spitzer} tool {\textsf
312: Spot}\footnote{\url{http://ssc.spitzer.caltech.edu/propkit/spot/index.html}}
313: predicts the surface brightness of the Zodiacal background in {\it Spitzer}
314: images. \citet{hhl06}, for instance, used this model to correct for ostensible
315: variations in instrument sensitivity over a period of around four days. We
316: compared our background estimates to the predictions in {\textsf Spot} and
317: found that the model underestimates our measurements by 3.5--4.0 MJy/sr. One
318: potential reason for any offset between observed and estimated backgrounds is
319: that {\textsf Spot} calculates a monochromatic background, whereas the
320: measured background is integrated over the wavelength passband and sensitivity
321: function of the MIPS-24 detector. Using {\textsf Spot}, we calculated
322: background estimates in 6 hour increments from 2006 Feb 20 00:00:00 to 2006
323: Feb 22 00:00:00, and from 2006 Apr 01 00:00:00 to 2006 Apr 02 00:00:00. The
324: slopes of the Feb backgrounds and the Feb--Apr backgrounds are -0.022 and
325: -0.032 MJy/sr/day, respectively; {\textsf Spot} does not provide error
326: estimates in its predictions. It thus appears that, within statistical
327: uncertainties, the behavior of the {\textsf Spot} background model is
328: consistent with our empirical results (at least for the time scales of our
329: observations), lending further justification to the approach by \citet{hhl06}.
330:
331: %findings for observing sets 1 and 2 by
332: %3.46 MJy/sr (i.e., it predicts a surface brightness of 21.008 MJy/sr), and it
333: %underestimates our results by 4.0 MJy/sr for observing set 3. One potential
334: %reason for any offset between observed and estimated backgrounds is that
335: %{\textsf Spot} calculates a monochromatic background, whereas the measured
336: %background is integrated over the wavelength passband and sensitivity function
337: %of the MIPS-24 detector.
338:
339: If, however, the difference in slope between the Spot estimates for just Feb
340: and Feb--Apr is indeed real but simply not detectable at our temporal
341: resolution (indicating that the background change with time is not a simple
342: linear function), any resulting discrepancy between calculated and observed
343: background values may be attributable to the fact that the model is calculated
344: for Earth, whereas Spitzer is in an Earth-trailing orbit and thus looking
345: through different amounts of Zodiacal dust \citep{hhl06}.
346:
347: \begin{figure}
348: \includegraphics[angle=-90, width=\linewidth]{f2.eps}
349: \caption{\label{background} Temporal fluctuations of the image background
350: measured in surface brightness units of MJy/sr for the three observing sets
351: (\ref{table1_AORs}). The lower, middle, and upper panel correspond to
352: observing set 1, 2, and 3, respectively. Note that the background level is
353: similar between observing sets 1 and 2 ($24.48\pm0.009$ MJy/sr and
354: $24.46\pm0.009$ MJy/sr, respectively), but lower for observing set 3
355: ($23.28\pm0.008$ MJy/sr). The typical fluctuations within a given image
356: (represented by the error bars) are around 0.05--0.06 MJy/sr. The large error
357: bar for the one data point in observing set 3 is caused by a cosmic ray.}
358: \end{figure}
359:
360: %----------------------------------------------------------------------
361:
362: %\subsection{Heliocentric Julian Dates of Observations}\label{dates}
363: %
364: %Observation timing information currently reported in the Spitzer FITS headers
365: %are local spacecraft time (SCLK\_OBS), universal time (UTCS\_OBS), and a
366: %geocentric modified Julian date (MJD\_OBS), but not the heliocentric Julian
367: %date (HJD)\footnote{Future implementations of the Spitzer pipeline (the S16
368: %software release) will implement a heliocentric Julian date field (HMJD\_OBS),
369: %but unfortunately the data available for this study did not have this feature
370: %available.}. From the Spitzer Helpdesk, we were able to obtain spacecraft
371: %coordinates relative to the Sun for the observation dates of our MIPS
372: %observing campaigns. From those relative coordinates one can compute time
373: %differences between Spitzer-centric modified Julian date (MJD\_Spitzer) and
374: %MJD\_OBS throughout the observing campaigns -- essentially the GU Boo light
375: %time-of-arrival lag between Spitzer and the earth. A HJD relative to MJD\_OBS
376: %could then be extracted for proper orbit phasing of the GU Boo data. Offsets
377: %between the MJD\_Spitzer and the MJD\_OBS values were approximately 95.5
378: %seconds for 20-21 Feb 2006 observations, and 60.2 seconds for 1 Apr 2006
379: %observations. Offsets between the HJD and the MJD\_OBS values were 143 to 146
380: %seconds for 20-21 Feb 2006 observations, and 294 seconds for 1 Apr 2006
381: %observations. An additional complication is that the MJD\_OBS time is
382: %currently reported in days with a precision of only 3 decimal places, which
383: %was insufficiently granual for our needs. In cross-referencing MJD\_OBS
384: %against UTCS\_OBS, we were able to interpolate the MJD\_OBS values to fraction
385: %of a second precision.
386:
387: %----------------------------------------------------------------------
388:
389: \section{Precision and Repeatability of the {\it Spitzer} Photometry}
390: \label{repeatability}
391:
392: %----------------------------------------------------------------------
393:
394: \subsection{PRF Fitting Versus Aperture Photometry}\label{prf_apphot}
395:
396: To create GU Boo's 24\micron\ light curve, we performed both PRF photometry as
397: described in \S \ref{reductions} and additionally utilized \textsf{apex's}
398: option of simultaneously obtaining aperture photometry. Figure
399: \ref{guboo_prf_apphot_comp} shows the agreement between PRF and aperture
400: photometry based on an aperture radius of 6'' (with 20--32'' background
401: annulus) which minimized the rms in the flat part of GU Boo's
402: phased\footnote{We use the period calculated by \citet{lmr05} (0.488728 days;
403: see Table \ref{table2_basic_params}) for our phasing throughout this paper.}
404: light curve. Using information from the MDH, we applied a multiplicative
405: aperture correction of 1.699 to the photometry. The median flux density
406: obtained by PRF photometry for the flat part of the phased light curve is 614
407: $\pm$ 49$\mu$Jy compared to 608 $\pm$ 59$\mu$Jy for the aperture photometry.
408: Thus, the absolute median flux density values agree very well for the two
409: different photometry approaches, but the PRF photometry exhibits smaller
410: scatter around the median magnitude. We note that the current version of
411: \textsf{apex} does not calculate photometry errors in the aperture correction,
412: and the principal reason why we performed aperture correction is to verify the
413: absolute flux density level of our sources as calculated by PRF fitting.
414:
415: \begin{figure}
416: \plotone{f3.eps}
417: \caption{\label{guboo_prf_apphot_comp} Comparison between
418: PRF photometry (blue crosses) and aperture photometry (red squares) for GU
419: Boo's phased light curve. Photometric error bars are omitted for the sake of
420: clarity. The scatter in the flat part of the light curve is smaller for PRF
421: photometry (49$\mu$Jy) than aperture photometry (59$\mu$Jy), but the median
422: flux density level of the flat part of the light curve is identical within the
423: errors for both photometry approaches.}
424: \end{figure}
425:
426: %----------------------------------------------------------------------
427:
428: \subsection{Absolute Versus Relative Photometry}\label{abs_relphot}
429:
430: In order to remove statistically correlated noise from GU Boo's light curve,
431: we performed relative photometry as described in equations 2 and 3 of
432: \citet{eh01}. We picked comparison objects based on the number of
433: observational epochs: in order to obtain a relative offset per photometric
434: data point in GU Boo's light curve (all data points are treated independently
435: of each other), it is advantageous to use stars with (at least) as many data
436: points as GU Boo itself. Four objects out of Table \ref{tab3_obj_stats}
437: fulfill this criterion: numbers 18, 31, 58, and 66 (see Figures
438: \ref{18_lc}--\ref{66_lc} for their light curves). The cross-referencing in
439: Table \ref{tab3_obj_stats} shows that objects 31 and 66 are stars, and objects
440: 18 and 58 are galaxies (as are all other objects in the field that we were
441: able to cross-reference). Note, however, that object 31's closest match in
442: SDSS \citep{sdss07} and 2MASS \citep{2mass03,2mass06} catalogs is 9'' away
443: whereas for object 66, the distance to the closest SDSS match was only
444: 0.15''. We originally presumed that, despite the fact that they are galaxies,
445: objects 18 and 58 would be unresolved at our large pixel size (\S
446: \ref{observations}) and tested that hypothesis by comparing the flux density
447: obtained by PRF photometry to that obtained by aperture photometry (6''
448: aperture; \S \ref{prf_apphot}). Figures \ref{stars_comp} and
449: \ref{galaxies_comp} shows that our presumption does not hold true for object
450: 58, and we discarded it from our relative photometry procedure.
451:
452: We find that, by performing differential photometry as outlined above, the
453: scatter in the flat part of GU Boo's phased light curve reduces by 9.4\% over
454: the PRF photometry (see Fig. \ref{guboo_prf_apphot_comp}) to 45$\mu$Jy. Light
455: curve fitting as described in \S \ref{analysis} was performed on the
456: differential photometry.
457:
458: \begin{figure}
459: \includegraphics[angle=-90, width=\linewidth]{f4.eps}
460: \caption{\label{stars_comp} Comparison between PRF photometry (blue crosses)
461: and aperture photometry (red squares) for objects 31 (left) and 66 (right).
462: Photometric error bars are omitted for the sake of clarity. The light curves
463: are phased to the period of GU Boo for the sake of comparison. The ordinate
464: scale is the same as Figures \ref{guboo_prf_apphot_comp}, \ref{galaxies_comp},
465: and \ref{18_lc}--\ref{66_lc}. The flux density values (PRF vs aperture
466: photometry) are identical within the errors for both objects.}
467: \end{figure}
468:
469: \begin{figure}
470: \includegraphics[angle=-90, width=\linewidth]{f5.eps}
471: \caption{\label{galaxies_comp} Comparison between PRF photometry (blue
472: crosses) and aperture photometry (red squares) for objects 18 (left) and 58
473: (right). Photometric error bars are omitted for the sake of clarity. The light
474: curves are phased to the period of GU Boo for the sake of comparison. The
475: ordinate scale is the same as Figures \ref{guboo_prf_apphot_comp},
476: \ref{stars_comp}, and \ref{18_lc}--\ref{66_lc}. The flux density values (PRF vs
477: aperture photometry) agree within the errors for object 18, but are discrepant
478: for object 58.}
479: \end{figure}
480:
481: %----------------------------------------------------------------------
482:
483: \subsection{Intra-Set Versus Inter-Set Photometric Stability}
484:
485: Figure \ref{global_rms} shows fractional rms values versus median flux
486: densities for all objects in Table \ref{tab3_obj_stats}. For every object, we
487: plot fractional rms for each individual observing set as well as for the
488: three sets combined. Observing sets 1 and 2 were obtained during the
489: MIPS006500 campaign, observing set 3 during MIPS006700 (Table
490: \ref{table1_AORs}). Consistent with the results in \citet{rye04}, we find
491: that inter-set repeatability of {\it Spitzer's} MIPS-24 is comparable to the
492: intra-set repeatability, both in terms of median flux density and the rms
493: scatter of the light curves, despite varying background levels (see \S
494: \ref{background_fluctuations}). For objects with a flux density in excess of
495: 1mJy, the rms scatter approaches 1--2 \%, similar to the scatter found for the
496: brightest sources observed with MIPS-24 in \citet{rye04}. Because of the
497: intrinsic variability produced by the stellar eclipse, GU Boo has the largest
498: fractional rms ($\sim$ 0.2). However, when we subtract the fit (see \S
499: \ref{analysis}) from GU Boo's light curve, the fractional rms falls to 0.081,
500: consistent with stars of similar median brightness. We show GU Boo's light
501: curve for the three individual observing sets in Fig. \ref{lc_guboo} and the
502: phased light curve along with the fit in Fig. \ref{lc_fit}.
503:
504: %Figure \ref{global_rms} shows the fractional rms around median values for all
505: %objects in Table \ref{tab3_obj_stats} for each individual observing set as
506: %well as for the 3 sets combined. Observing sets 1 and 2 were obtained during
507: %the MIPS006500 campaign, observing set 3 during MIPS006700 (Table
508: %\ref{table1_AORs}). We find that inter-set repeatability of {\it
509: %Spitzer's} MIPS-24 is comparable to its repeatability within a set, both in
510: %terms of median flux density level as well as rms values, even when the
511: %mid-infrared background levels vary between sets (see \S
512: %\ref{background_fluctuations}). For the objects with a flux density in excess
513: %of 1mJy, the rms values approach the 1--2 \% level. Both findings are
514: %consistent with what \citet{rye04} find for the brightest objects.
515:
516: %GU Boo is the point with the largest rms, due to its intrinsic variability. If
517: %we subtract the fit (see \S \ref{analysis}) from GU Boo's light curve, the
518: %fractional rms falls to 0.081, right into the middle of data points with
519: %similar median brightnesses. We show GU Boo's light curve for the three
520: %individual observing sets in Fig. \ref{lc_guboo} and the phased light curve
521: %along with the fit in Fig. \ref{lc_fit}.
522:
523: In order to compare our rms values to background-limited noise values, we used
524: {\it Spitzer's} {\textsf
525: SENS-PET}\footnote{\url{http://ssc.spitzer.caltech.edu/tools/senspet}} to
526: predict the MIPS-24 sensitivity ($1 \sigma$ above background for 170 seconds
527: integration time; see \S \ref{mosaicing}) for low and medium background levels
528: (solid and dashed line in Fig. \ref{global_rms}, respectively). The
529: mid-infrared background at the time of observations of GU Boo is 23--24.5
530: MJy/sr, which is between the typical low and medium background levels used by
531: {\textsf SENS-PET} (see also \S \ref{background_fluctuations}). We find that
532: the {\textsf SENS-PET} predictions are consistent with our empirically
533: determined error estimates. Except for the variable GU Boo, typical values
534: for the rms scatter of the light curves (Table \ref{tab3_obj_stats}) are
535: approximately equal to average photometric measurement uncertainties of
536: individual data points (see Figs. \ref{18_lc}, \ref{31_lc}, and \ref{66_lc}).
537:
538: \begin{figure}
539: \epsscale{0.7}\plotone{f6.eps}
540: \caption{\label{global_rms}
541: Median flux density versus fractional rms for the 24 objects that have
542: photometry for more than 72 out of 144 observational epochs. Triangles,
543: squares, and crosses represent the data from observings sets 1, 2, and 3,
544: respectively (see Table \ref{table1_AORs}), to illustrate the repeatability of
545: {\it Spitzer}/MIPS-24 within individual observing sets. Circles mark the data
546: points from the combination of all 3 observing sets (to show the inter-set
547: stability). The data point with the highest fractional rms is GU Boo, due to
548: its intrinsic variability. When subtracting our fit from its light curve (see
549: Fig. \ref{lc_fit}), GU Boo's fractional rms falls to 0.081. The solid and
550: dashed lines indicate MIPS-24's sensitivity for our exposure times as a
551: function of flux density for low and medium background levels, respectively.}
552: \end{figure}
553:
554: %\begin{figure}
555: %\plotone{figures/campaigns_rms.eps}
556: %\caption{\label{global_rms}
557: %A plot of median flux density versus fractional rms for the 24 stars that have
558: %photometry for more than 72 out of 144 observational epochs. Shown are the
559: %individual 3 observing sets (see Table \ref{table1_AORs}) to illustrate the
560: %repeatability of {\it Spitzer}/MIPS-24 within individual observing sets,
561: %as well as a plot of all 3 sets combined (to show the inter-set
562: %stability). The data point with the highest fractional rms is GU Boo. }
563: %\end{figure}
564:
565: \begin{figure}
566: \includegraphics[angle=-90, width=\linewidth]{f7.eps}
567: %\includegraphics[angle=270]{figures/66_lc.eps}
568: \caption{\label{lc_guboo} GU Boo's 24$\micron$ light curve, based on absolute
569: PRF photometry (see \S \ref{prf_apphot} and \S \ref{abs_relphot}). The three
570: panels represent the three MIPS-24 observing sets during which the individual
571: secondary eclipse events were observed. GU Boo is object number 51 in our
572: numbering system (see Table \ref{tab3_obj_stats}).}
573: \end{figure}
574:
575: %\input{tables/tab3_tab3_obj_stats.tex}\label{tab3_obj_stats}
576:
577: %----------------------------------------------------------------------
578:
579: %\subsection{Light Curve Production}\label{lightcurves}
580:
581: %----------------------------------------------------------------------
582:
583: \section{Analysis of GU Boo's Photometric Light Curve}\label{analysis}
584:
585: We modeled the secondary eclipse observations of GU Boo using the JKTEBOP
586: code \citep{sms04,szm04}. JKTEBOP is based on the original EBOP code
587: \citep{pe81,e81}, but with the addition of the Levenberg-Marquardt
588: optimization algorithm \citep{NR92} to find the best fitting model, and also
589: the implementation of a Monte Carlo simulation algorithm to determine robust
590: uncertainties in the fitted parameters \citep{ssm05}.
591:
592: The orbital period and initial epoch of the primary eclipse were set to the
593: values given in the ephemeris derived by \citet{lmr05}. We further fixed the
594: mass ratio and the radius ratio of the stars, as well as the eccentricity of
595: the system ($e$=0) to the values obtained in that work. We assumed no limb
596: darkening effects in the light curves, as expected for observations this far
597: into the infrared \citep[][and references therein]{cdg95,rhs06,cbb07,s07}, and
598: no significant gravitational darkening or reflection effects, based on the
599: spherical shape of the stars and the similarity in effective temperatures. All
600: these are reasonable assumptions, based on the results of the study of GU Boo
601: at visible wavelengths, and they are, in fact, hard to test in detail, given
602: the photometric precision of the {\it Spitzer} light curve at this flux
603: density level.
604:
605: In the absence of primary eclipse observations, to calculate the luminosity
606: ratio of the system, we place a further constraint to the fit by fixing the
607: value of the surface brightness ratio of the stars to J=J2/J1=0.9795. This
608: value, combined with the adopted radius ratio and the no limb darkening
609: assumption, gives a luminosity ratio of of $L_{2}/L_{1}$ = 0.9697, which is
610: consistent with the expected value for GU Boo at 24$\mu$m.
611:
612: The parameters initially left free in the models were: (1) the fractional sum
613: of the radii, i. e., $(R_{1} + R_{2})/a$, where $R_{1}$ and $R_{2}$ are the
614: component radii, and $a$ is the orbital separation, computed from the stellar
615: masses and the orbital period of the system, (2) the inclination of the orbit
616: $i$, (3) the amount of third light $L_{3}$, and (4) a phase offset parameter
617: $\phi$ (to account for small errors in the ephemeris).
618:
619: %and a light scale factor to adjust the out of eclipse baseline of the models
620: %to the baseline of the normalized light curve data. This last parameters was
621: %set to 0.991, after the first few model trials proved to give basically
622: %identical results.
623:
624: Our best model solution is illustrated in Figure \ref{lc_fit}, with a reduced
625: $\chi^2$ of 1.7, and a mean fractional error per data point of 9.5\%
626: (cf. Fig. \ref{global_rms}). Formal errors in the fitted parameters were
627: derived using the Monte Carlo algorithm implementation in JKTEBOP for a total
628: of 1000 iterations. We obtain a radius for the secondary component of $R_{2}$
629: = 0.66 $\pm$ 0.02 $R_{\sun}$. Our value of the orbital inclination is $i$ =
630: 89.3 $\pm$ 0.8 degrees. Both values are slightly larger than the ones
631: obtained by \citet{lmr05} at optical wavelengths, $R_{2}$ = 0.62 $\pm$ 0.02
632: $R_{\sun}$ and $i$ = 87.6 $\pm$ 0.2 degrees. The two secondary radius
633: estimates agree to within random statistical errors (1.4$\sigma$). In the
634: case of the inclination, our value is not as well constrained as in the
635: optical, since we lack a full light curve that includes a primary eclipse. We
636: show our estimates for GU Boo's system parameters in Table
637: \ref{table2_basic_params}. For the tested third light contribution, we obtain
638: a value of $L_{3}$ = -0.04 $\pm$ 0.07, consistent with $L_{3}$ = 0.
639:
640: Finally, we find a phase shift of $\Delta\phi$ = -0.014 $\pm$ 0.001. This
641: phase shift is 1.5 times larger than expected from the \citet{lmr05}
642: ($|\Delta\phi|$ = 0.009), but can still be attributed to uncertainties in the
643: original period estimation. The \citet{lmr05} observations were conducted in
644: 2003, near JD=2452733. The number of elapsed periods inbetween those
645: observations and our {\it Spitzer} AORs is about 2150. The 1-$\sigma$ error
646: in the \citet{lmr05} period estimate is $2\times 10^{-6}$ days, which
647: accumulates to 0.0043 days, about 0.009 in phase, in 2150 periods. Thus, the
648: discrepancy we find corresponds to about 1.5$\sigma$ from the \citet{lmr05}
649: ephemeris predictions. We estimate that this offset is based on normal
650: statistiscal errors. An alternative explanation would be that a third body
651: orbiting the system could cause this shift, but since (1) we show in \S
652: \ref{flux_comparison} that GU Boo's flux is consistent with its modeled
653: spectral energy distribution, and (2) we calculate the third light component
654: to be $L_{3}$ = 0, any such claim would be unsubstantiable with our data.
655:
656: Equation \ref{ephemeris} shows the updated ephemeris equation of GU Boo by
657: combining the seven minima in table 5 of \citet{lmr05} with the three new
658: minima presented in this work.
659:
660: \begin{equation}
661: \label{ephemeris}
662: T(\mbox{Min I}) = \mbox{HJD}2452723.981327(1)+ 0.4887247(8) \cdot E.
663: \end{equation}
664:
665: Uncertainty digits are given in parentheses. $E$ represents the number of
666: elapsed periods since the initial epoch, $T(\mbox{Min I})$ the time of primary
667: eclipse minimum.
668:
669: %----------------------------------------------------------------------
670:
671: \input{tab2.tex}\label{basic_params}
672:
673: %----------------------------------------------------------------------
674:
675: \begin{figure}
676: \plotone{f8.eps}
677: \caption{\label{lc_fit} Our best fit overlaid on top of the phased 24$\micron$
678: data of GU Boo derived from our relative photometry of the system. The fit has
679: a reduced $\chi^2 = 1.7 \pm$ 0.1. The bottom panel shows the residuals around
680: the fit. The calculated system parameters, which agree well with the results
681: from the optical study in \citet{lmr05}, are shown in Table
682: \ref{table2_basic_params}. }
683: \end{figure}
684:
685: %----------------------------------------------------------------------
686:
687: \section{Comparison Between Expected and Measured 24\micron\ Flux Density of
688: GU Boo}\label{flux_comparison}
689:
690: In addition to the relative photometry of GU Boo (Fig. \ref{lc_fit}), we also
691: performed absolute photometry as reported in \S \ref{prf_apphot}. The
692: 24\micron\ flux density of $614 \pm 49\mu$Jy was determined from the median
693: flux level outside of eclipse. To test the accuracy of the absolute flux
694: density level, we show in this section a spectral energy distribution (SED)
695: model between 0.11 and 35$\micron$, scaled to the optical and near-infrared
696: (NIR) magnitudes of GU Boo \citep{lmr05}.
697:
698: The GU Boo system components are two M stars of nearly identical mass,
699: temperature and radius. For our SED model, we assumed both stellar components
700: to be M1V stars with effective temperatures of 3800 K
701: \citep[e.g.,][]{lmr05}. The model SED was constructed from the M1V
702: $0.11-2.5$\micron\ optical-NIR templates of \citet{p98} and the {\it Spitzer}
703: $5-35$\micron\ Infrared Spectrograph \citep[IRS; ][]{hrc04} spectra of
704: GL~229A, an M1V (3800 K) star \citep{crm06}. To build the SED model
705: (Fig. \ref{guboo_sed}), the M1V optical-NIR template was scaled to GU Boo's
706: optical-NIR flux densities based on table 1 in \citet{lmr05}. To connect the
707: {\it Spitzer} IRS spectrum to the optical-NIR template, we fit a power law of
708: the form $F_\nu \propto \nu^{n}$ (see dashed line in Fig. \ref{guboo_sed}) to
709: the IRS spectrum. We found the best-fit exponent to the power law to be
710: $n=1.9$. The IRS spectrum and the power law (extrapolated to 2.4$\micron$)
711: were then scaled to the red edge of the optical-NIR template. The slope of the
712: power law was maintained to ensure a continuous transition between the
713: optical-NIR template and the IRS spectrum (see Fig. \ref{guboo_sed}). Note
714: that only the optical and NIR flux densities were used to scale the SED model;
715: i.e., the scaling does not utilize the 24\micron\ data point.
716:
717: The SED model predicts a mid-infrared flux density for GU Boo of
718: $F_{\nu}(24\mu{\rm m}) \approx 650\mu$Jy. The measured 24\micron\ flux
719: density of GU Boo ($614 \pm 49\mu$Jy) is within 1$\sigma$ of the predicted
720: flux density, agreeing remarkably well with the simple SED model presented
721: here. We conclude that the stellar components are solely responsible for the
722: mid-infrared emission of GU Boo.
723:
724: The few M- and K-dwarf DEB systems studied to date (GU Boo included) reveal
725: that many of the binary components have larger radii (by 10-20\%) and cooler
726: effective temperatures (by 100 K to hundreds of K) than predicted by stellar
727: evolutionary models \citep[e.g.,][]{tr02,ribas03,lmr05,lm07}. Magnetic
728: activity and metallicity can account for the radius discrepancy
729: \citep{lm07} and, in principle, also for the temperature discrepancy. An
730: alternative explanation for the temperature discrepancy, however, is the
731: presence of dusty material around the systems. The excellent agreement of our
732: observed mid-infrared flux density with the model SED suggests that there is
733: little, if any, (warm) circumstellar dust in GU Boo, likely ruling out
734: circumstellar dust as a viable explanation for discrepancies with the stellar
735: evolutionary models.
736:
737: %----------------------------------------------------------------------
738:
739: \begin{figure}
740: \includegraphics[angle=90, width=\linewidth]{f9.eps}
741: \caption{\label{guboo_sed} The SED of GU Boo, based on M1V $0.11-2.5$\micron\
742: optical-NIR templates of \citet{p98} and {\it Spitzer} $5-35$\micron\ IRS
743: spectra of GL~229A (spectral type M1V; $T_{eff}$ = 3800 K). The dashed line
744: respresents the interpolation between template and spectra. For details, see
745: \S \ref{flux_comparison}.}
746: \end{figure}
747:
748: %----------------------------------------------------------------------
749:
750: \section{Light Curves of Selected Objects in the Field of GU Boo}
751: \label{other_lightcurves}
752:
753: In this Section, we present a brief summary of selected other light curves in
754: the field of GU Boo, along with basic determination of spectral types of the
755: objects identified as stars (see \S \ref{abs_relphot}). We limit our selection
756: to the three objects that were used to perform the relative photometry (see \S
757: \ref{abs_relphot}). Figures \ref{18_lc}--\ref{66_lc} display these light
758: curves. They are all on the same scale with different zeropoints. Parameters
759: for all objects with at least 72 out of the 144 epochs are listed in Table
760: \ref{tab3_obj_stats}. We do not show light curves for the rest of the field
761: objects since they can essentially be described as flat lines with some
762: scatter around the median magnitude, which is characterized by the values in
763: Table \ref{tab3_obj_stats}.
764:
765: \begin{figure}
766: \includegraphics[angle=-90, width=\linewidth]{f10.eps}
767: \caption{\label{18_lc} The MIPS-24 light curves of object 18 (a galaxy) in
768: the field of GU Boo. The three panels represent the three MIPS-24 observing
769: sets. For parameters, see Table \ref{tab3_obj_stats}.}
770: %\caption{\label{18_lc} The MIPS-24 light curve of object 18.
771: %The three panels represent the three MIPS-24 observing
772: %campaigns. For parameters, please see Table \ref{tab3_obj_stats}.}
773: \end{figure}
774:
775: \begin{figure}
776: \includegraphics[angle=-90, width=\linewidth]{f11.eps}
777: \caption{\label{31_lc}The MIPS-24 light curves of object 31 (possibly a star) in
778: the field of GU Boo. The three panels represent the three MIPS-24 observing
779: sets. For parameters, see Table \ref{tab3_obj_stats}. If the cross
780: referencing in \S \ref{abs_relphot} is correct, object 31 is an M3III giant
781: (see text in \S \ref{other_lightcurves}).}
782: %\caption{\label{31_lc} The MIPS-24 light curve of object 31.
783: %The three panels represent the three MIPS-24 observing
784: %campaigns. For parameters, please see Table \ref{tab3_obj_stats}.}
785: \end{figure}
786:
787: \begin{figure}
788: \includegraphics[angle=-90, width=\linewidth]{f12.eps}
789: \caption{\label{66_lc}The MIPS-24 light curves of object 66 (a star) in
790: the field of GU Boo. The three panels represent the three MIPS-24 observing
791: sets. For parameters, see Table \ref{tab3_obj_stats}. Our SED fitting
792: indicates this to be an A2V dwarf (see text in \S \ref{other_lightcurves}).}
793: \end{figure}
794:
795: Spectral typing for the two stars (objects 31 and 66) was attempted by means
796: of SED fitting of photometry available in the literature: both objects have
797: Sloan DSS $ugriz$ \citep{sdss07} data points, and star 31 additionally has
798: 2MASS $JHK_s$ \citep{2mass03,csd03} and Johnson $RI$ \citep{mlc03} magnitudes
799: available for it. SED fits were performed using the {\tt sedFit} program
800: discussed in \S 3.1 of \citet{bcb07}. The best SED match for star 31 is an
801: M3III giant ($\chi^2_{reduced} \sim 1.6$), whereas star 66's SED was found to
802: be consistent with an A2V dwarf ($\chi^2_{reduced} \sim 0.9$). Note that one
803: assumption we make here is that the cross referencing for object 31 is
804: correct, despite the large distance from its closest matches in the SDSS and
805: 2MASS catalogs (Table \ref{tab3_obj_stats}).
806:
807: %Luminosity classification is not well constrained directly for such an
808: %approach, but a choice between main sequence and giant star luminosity class
809: %was made based upon the reddening reported by the SED fit. As seen in Table
810: %***, stars 31 and 66 have similar Sloan $r$ band brightnesses, although star
811: %31 has substantially higher overall flux as evidenced by the rising magnitudes
812: %towards the red end in the Sloan filters, confirmed by the Johnson and 2MASS
813: %data. For star 31, a M-type fit was most appropriate, while an earlier A-type
814: %fit was more appropriate for star 66. Given the apparent brightness of star
815: %31, a giant star's luminosity class was appropriate, compared to a main
816: %sequence class for star 66.
817:
818: %More exotic relative
819: %luminosity classification (e.g., a supergiant M-type versus an A-type giant) is
820: %expected to be sufficiently unlikley that it was not considered.
821:
822: %Given that expectation for luminosity class, the resulting best SED fits are
823: %found at the bottom of Table TBD, including $\chi^2$ per degree of freedom fit
824: %values, spectral type, and expected extinction as characterized by reddening
825: %value $A_V$; plots of the fits are seen in Figures TBD and TBD. Star 31 is
826: %expected to be a M3III, while star 66 is an A2V object. Both light curves are
827: %statistically consistent with non-variable flux levels at $F_{24\mu m} = TBD
828: %\pm TBD$ and $TBD \pm TBD$, respectively.
829:
830: %An interesting aspect of the fits is that for star 66 is that extrapolation of
831: %the Rayleigh-Jeans tail to 24$\mu$m implies an expected flux level of less
832: %than 1$\mu$Jy; for star 31, the expected flux level is roughly 2$\mu$Jy.
833: %Thus our initial expectation here is that both objects exhibit significant
834: %mid-infrared excess, which is consistent with detection of such optically
835: %faint stars by Spitzer. Furthermore, it is consistent with previous
836: %detections of debris disk around A-type main sequence stars (ref TBD - Rieke
837: %2005?) and dusty mass loss from M giants (ref TBD).
838:
839:
840: %----------------------------------------------------------------------
841:
842: \section{Summary and Conclusions}\label{conclusions}
843:
844: We used MIPS-24 onboard the {\it Spitzer Space Telescope} to obtain
845: time-series photometry of the M-dwarf DEB GU Boo. Our observations cover three
846: secondary eclipse events, two consecutive ones and an additional event six
847: weeks later. Analysis of the photometry shows that the flux density values for
848: aperture photometry and PRF photometry agree, and that the PRF photometry
849: produces smaller scatter in the light curve. This scatter can be further
850: reduced by performing relative photometry based on three comparison objects in
851: the field. We find that the repeatability of MIPS-24 photometry is consistent
852: over all temporal scales we sampled: within an observing set and on time
853: scales of 24 hours and six weeks.
854:
855: Our mid-IR analysis of GU Boo's light curve is less affected by stellar
856: surface features than its optical counterpart. The results we produce show
857: very good agreement with the previously obtained system parameters based on
858: optical and near-IR work. A comparison between GU Boo's flux density and its
859: model SED based on stellar templates and IRS spectra shows no IR excess,
860: leading us to the conclusion that no warm circumstellar dust is present in the
861: system.
862:
863: Finally, light curves of other objects in the field indicate that the
864: photometric stability of {\it Spitzer's} MIPS-24 is comparable over short
865: (hours to days) and long (weeks) time scales, despite fluctuations in the
866: image mid-IR background on time scales of weeks.
867:
868: %----------------------------------------------------------------------
869:
870: \acknowledgements
871:
872: We gratefully acknowledge the allocation of {\it Spitzer} Director's
873: Discretionary Time (DDT) for this project. We furthermore thank D. Frayer,
874: S. Carey, J. Colbert, and P. Lowrance for providing valuable insight into the
875: mysterious world of \textsf{mopex} and \textsf{apex}, as well as the anonymous
876: referee for the thorough study of the manuscript and some very insightful
877: comments and suggestions that significantly improved the quality of this
878: publication. Thanks also to J. Southworth for useful clarifications on the use
879: of JKTEBOP, and to M. Cushing for providing the IRS spectrum of GL 229A in
880: electronic format. This research has made use of NASA's Astrophysics Data
881: System. This publication makes use of data products from the Two Micron All
882: Sky Survey, which is a joint project of the University of Massachusetts and
883: the Infrared Processing and Analysis Center/California Institute of
884: Technology, funded by NASA and the NSF. Funding for the SDSS and SDSS-II has
885: been provided by the Alfred P. Sloan Foundation, the Participating
886: Institutions, the National Science Foundation, the U.S. Department of Energy,
887: NASA, the Japanese Monbukagakusho, the Max Planck Society, and the Higher
888: Education Funding Council for England. The SDSS Web Site is
889: \url{http://www.sdss.org/}. The SDSS is managed by the Astrophysical Research
890: Consortium for the Participating Institutions. The Participating Institutions
891: are the American Museum of Natural History, Astrophysical Institute Potsdam,
892: University of Basel, University of Cambridge, Case Western Reserve University,
893: University of Chicago, Drexel University, Fermilab, the Institute for Advanced
894: Study, the Japan Participation Group, Johns Hopkins University, the Joint
895: Institute for Nuclear Astrophysics, the Kavli Institute for Particle
896: Astrophysics and Cosmology, the Korean Scientist Group, the Chinese Academy of
897: Sciences (LAMOST), Los Alamos National Laboratory, the Max-Planck-Institute
898: for Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New
899: Mexico State University, Ohio State University, University of Pittsburgh,
900: University of Portsmouth, Princeton University, the United States Naval
901: Observatory, and the University of Washington.
902:
903: %----------------------------------------------------------------------
904:
905: \bibliography{spitzer}
906:
907: %----------------------------------------------------------------------
908:
909: \clearpage
910: %\LongTables
911:
912: %\begin{rotate}
913: \begin{landscape}
914:
915: \input{tab3.tex}
916:
917: \clearpage
918:
919: \end{landscape}
920: %\end{rotate}
921:
922: %----------------------------------------------------------------------
923:
924: \end{document}