1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[12pt]{aastex}
3: %\documentclass[apjl]{emulateapj}
4: %\usepackage{apjfonts}
5:
6: %\slugcomment{Submitted to ApJ, 12 July 2007; resubmitted 10 August 2007}
7: \slugcomment{to appear in ApJ}
8: \shortauthors{Hoard et al.}
9: \shorttitle{Mid-IR Spectrum of EF Eri}
10:
11: \begin{document}
12:
13: \title{The Mid-Infrared Spectrum of the Short Orbital Period Polar
14: EF~Eridani from the Spitzer Space Telescope}
15:
16: \author{D. W. Hoard\altaffilmark{1},
17: Steve B. Howell\altaffilmark{2},
18: Carolyn S. Brinkworth\altaffilmark{1},
19: David R. Ciardi\altaffilmark{3},
20: Stefanie Wachter\altaffilmark{1}
21: }
22: \altaffiltext{1}{Spitzer Science Center, California Institute of Technology,
23: Pasadena, CA 91125}
24: \altaffiltext{2}{WIYN Observatory and National Optical Astronomy Observatory,
25: Tucson, AZ 85719}
26: \altaffiltext{3}{Michelson Science Center, California Institute of Technology,
27: Pasadena, CA 91125}
28:
29:
30: \begin{abstract}
31: We present the first mid-infrared (5.5--14.5 $\mu$m) spectrum of
32: a highly magnetic cataclysmic variable, EF Eridani, obtained with
33: the Infrared Spectrograph on the {\em Spitzer Space Telescope}.
34: The spectrum displays a relatively flat, featureless continuum.
35: A spectral energy distribution model consisting of a 9500 K white
36: dwarf, L5 secondary star, cyclotron emission corresponding to a
37: $B\approx13$ MG white dwarf magnetic field, and an optically thin
38: circumbinary dust disk is in reasonable agreement with the extant
39: 2MASS, IRAC, and IRS observations of EF Eri. Cyclotron emission
40: is ruled out as a dominant contributor to the infrared flux
41: density at wavelengths $\gtrsim3$ $\mu$m. The spectral energy
42: distribution longward of $\sim5$ $\mu$m is dominated by dust
43: emission. Even longer wavelength observations would test the
44: model's prediction of a continuing gradual decline in the
45: circumbinary disk-dominated region of the spectral energy distribution.
46: \end{abstract}
47:
48: \keywords{stars: individual (EF Eri) --- novae, cataclysmic variables ---
49: stars: low-mass --- stars: brown dwarfs}
50:
51:
52: \section{Introduction}
53:
54: Cataclysmic variables (CVs) are interacting binary stars
55: containing a white dwarf (WD) primary and a low mass
56: secondary \citep{warner95}. The evolution of these
57: binaries is believed to proceed as follows (see
58: \citealt{howell01}):\ after a common envelope phase
59: following the post-main sequence evolution of the WD
60: progenitor, the low mass secondary star eventually
61: (over)fills its Roche lobe and the binary commences
62: mass transfer. Due to angular momentum losses, primarily
63: via magnetic braking and gravitational radiation, the
64: two component stars move closer together over time; that
65: is, their orbital period decreases. For the oldest CVs,
66: the orbital periods are very short (near 80 minutes) and
67: the secondary stars are very low mass, being
68: $\sim$0.06 M$_{\odot}$ stars or lower mass degenerate
69: brown dwarf-like objects.
70:
71: EF Eridani contains a strongly magnetic WD ($B\approx13$--14 MG;
72: \citealt{WR98,howell06b,beuermann07}), making it a member of the
73: polar class of CV (named for the highly polarized nature of
74: their emitted light). Unlike CVs containing non-magnetic WDs,
75: polars have no accretion disks (instead, accretion proceeds
76: directly from the inner Lagrangian point onto the magnetic
77: field lines of the WD) and generally do not undergo
78: dwarf-nova-type (i.e., disk instability) outbursts. They do,
79: however, experience periods of normal mass transfer from the
80: secondary star (high states) interspersed with times when
81: this accretion flow stops or is significantly decreased
82: (low states), possibly related to stellar activity on the
83: secondary star. EF Eri has been in a low state for the past
84: 10 years \citep{howell06b}. Interestingly, recent high energy
85: observations in the UV \citep{szkody06} and X-rays
86: \citep{schwope07} have shown that EF Eri still has a
87: $\sim20,000$~K hot spot remaining on the surface of its
88: $\sim10,000$~K WD even after a decade of very low mass
89: accretion ($\dot{M}\sim10^{-14} M_{\odot}$ yr$^{-1}$).
90: This hot region is presumably at or near the active accretion
91: pole, which is best modelled as a non-uniform spot covering
92: 10--20\% of one side of the WD \citep{beuermann07}.
93:
94: Our initial {\em Spitzer Space Telescope} \citep{werner04}
95: photometric observations of a small sample of polars using
96: the Infrared Array Camera (IRAC; \citealt{fazio04}) included
97: EF Eri as the brightest sample member, and revealed a nearly
98: flat mid-IR (3.6--8 $\mu$m) flux density level near 700 $\mu$Jy
99: \citep{howell06a,brinkworth07}. In \citet{howell06a}, we showed that
100: the IRAC spectral energy distributions (SEDs) of EF Eri and
101: three other polars with similar orbital periods
102: had flux density in excess of that produced by the two
103: component stars alone. It was concluded that the best
104: candidate for the excess emission was a circumbinary dust disk
105: with an inner edge temperature ($T_{\rm in}$) near 800~K.
106: In \citet{brinkworth07} (hereafter, B07), we
107: used these same data, as well as IRAC observations of two
108: additional polars (making a total of seven systems),
109: to examine a larger suite of more
110: sophisticated SED models. For EF Eri, we
111: found that the most plausible way to produce the observed
112: bright 8-$\mu$m flux density, without exceeding the observed
113: flux densities at shorter wavelengths, was via a geometrically
114: thin, optically thick circumbinary disk (CBD) with
115: T$_{in}\approx650$~K; however, we could not completely rule
116: out the possibility that at least some of the long wavelength
117: flux density in the SED is due to cyclotron emission.
118: The B07 model results were generally consistent with the need
119: for a prominent cyclotron emission component to explain the
120: near-IR (e.g., 2MASS) portion of the EF Eri SED.
121:
122: The {\em Spitzer} photometric observations alone could not
123: provide any further clarification as to the exact origin of
124: the 3.6--8-$\mu$m SED, as they had limited wavelength
125: resolution and coverage. Consequently, those data were not
126: able to strongly constrain models for the origin of the
127: observed mid-IR flux density of EF Eri (as described in more
128: detail in B07). Therefore, in an effort
129: to better understand the true nature and extent of the mid-IR
130: emitting source(s) in EF Eri, we obtained a spectrum spanning
131: 5.5--14.5 $\mu$m using the Infrared Spectrograph
132: (IRS; \citealt{houck04}) on {\em Spitzer}.
133:
134:
135: \section{Observations and Data Processing}
136: \subsection{Mid-IR Spectrum}
137:
138: Our spectroscopic observations of EF Eri were obtained using
139: the ``Short Low'' module of the IRS, which covers 5.2--8.7 $\mu$m
140: in second order (SL2) and 7.4--14.5 $\mu$m in first order (SL1)
141: at a resolution of $R\sim60$--120.
142: A third order (the SL3 ``bonus'' spectrum) is obtained with each
143: SL2 observation; it spans 7.4--8.6 $\mu$m and is primarily used
144: to ensure that the flux calibration from SL2 to SL1 is consistent
145: (in our case, all three orders were in excellent agreement in the
146: overlapping wavelength region, so we did not apply any offsets
147: between orders, and used the data from all three orders in
148: constructing a final spectrum -- see below).
149: We obtained four cycles of 240~s
150: each for both SL1 and SL2 (i.e., a total of sixteen individual
151: exposures after counting the two nod positions obtained in each cycle).
152: The corresponding AOR reqkey number is 17052928.
153:
154: We used the Spitzer Science Center post-BCD software SPICE
155: (v1.4.1)\footnote{See
156: \url{http://ssc.spitzer.caltech.edu/postbcd/spice.html}.} to extract
157: the IRS spectra from the two-dimensonal images.
158: The input images consisted of the four S15.3.0 pipeline-combined
159: post-BCD images (i.e., the $*$bksub.fits files), each of which is
160: constructed from four appropriately background-subtracted, masked,
161: and co-added sub-exposures to give two images each for the SL1 and
162: SL2+SL3 orders (i.e., one image per order at each nod position).
163: We extracted the spectra from these two-dimensional images using
164: the optimal extraction algorithm with the standard aperture width.
165: This results in two extracted spectra (one for each nod position)
166: spanning the three SL orders.
167:
168: Next, we performed a weighted average of the two (or three) points
169: at each wavelength from the two nod spectra to obtain a preliminary
170: average spectrum. We calculated the weighted mean,
171: $\langle f_{\rm w} \rangle$, and standard
172: deviation of the weighted mean, $\sigma_{\rm wavg}$,
173: of the flux density over the full
174: wavelength range for the preliminary average spectrum. Then, we
175: rejected any point in an individual nod spectrum that was more
176: than $3\sigma_{\rm wavg}$ away from $\langle f_{\rm w} \rangle$
177: when the corresponding point(s) at the same wavelength in the
178: other nod was not (i.e., an outlier rejection).
179: Finally, we recalculated the average spectrum using the
180: outlier-rejected nod spectra, via a weighted average when two or
181: more data points were available for a given wavelength
182: or using the remaining data point when
183: only one survived rejection.
184:
185: The IRS spectrum of EF Eri is shown in Figure \ref{f:spectrum}.
186: It is characterized by a generally flat continuum with a slight
187: dip at the short wavelength end. There are no obvious emission
188: features, and only a few potential absorption features. However,
189: we note that the IRS was designed to achieve high sensitivity
190: at the cost of reduced dynamic range; hence, great care must be
191: used in the interpretation of weak spectral features\footnote{See
192: the IRS chapter of the Spitzer Observer's Manual, available at
193: \url{http://ssc.spitzer.caltech.edu/documents/som/}.}, especially
194: in spectra of faint targets like EF Eri.
195: The spectrum of EF Eri used in this work has not been scaled
196: from its original flux calibration -- the excellent agreement
197: in flux density between the IRAC photometric points and the IRS
198: spectrum (see \S\ref{s:otherdata} and Figure \ref{f:bigspec}),
199: as well as the smooth transition from the SL2 to SL1 data sections
200: (at $\lambda\approx7.5$ $\mu$m), attests that the overall continuum
201: shape of the spectrum is reliable. Consequently, our analysis
202: presented here will rely on looking at the gross properties of the
203: spectrum as a whole (i.e., the continuum shape and flux density
204: level), rather than focusing on specific features with low
205: statistical probability of being real.
206:
207:
208: \subsection{Other Data}
209: \label{s:otherdata}
210:
211: We also utilized the 2MASS and IRAC photometry of EF Eri reported
212: in B07, and the near-IR spectrum shown in Figure 1 of \citet{harrison07}.
213: Figure \ref{f:bigspec} shows all of the spectroscopic and photometric
214: data plotted together. Other than the points from IRAC channels 1
215: and 2 (3.6 and 4.5 $\mu$m, for which there are no overlapping
216: spectroscopic data), all of the photometric data are in
217: excellent agreement with the spectroscopic data. A slight exception
218: to this is the 2MASS $K_{\rm s}$-band point, which is significantly
219: lower in flux density (by more than $1\sigma$) compared to the
220: near-IR spectrum. This is likely due to the $K$-band variability
221: noted by \citet{harrison04}, which has been linked to cyclotron
222: emission that varies with orbital phase.
223:
224:
225: \section{Spectral Energy Distribution Models}
226: \subsection{The Code}
227: \label{s:code}
228:
229: In B07, we introduced our IR SED modeling code for CVs, which we
230: applied to photometric data for several polars, including EF Eri.
231: We have now made a number of improvements to the modeling code
232: for use here. First, we have modified the code to allow SEDs to
233: be calculated at higher wavelength resolution, which is more
234: appropriate for comparing to spectral data. This modification
235: has little effect on the WD and CBD components, but provides
236: the necessary resolution to resolve cyclotron humps in the
237: cyclotron component. For the models presented here, we have
238: used a wavelength increment of 0.05 $\mu$m in the range 1--14.5 $\mu$m.
239:
240: Second, we have modified the handling of the secondary star to
241: accommodate data at longer wavelengths. In B07, the secondary
242: star was represented by average 2MASS and IRAC photometry for
243: spectal type templates from \citet{patten06}. We have now added
244: the capability to combine the 2MASS and IRAC photometry of a single,
245: representative spectral type star from \citet{patten06} with
246: the IRS SL spectrum of the same star from \citet{cushing06}.
247: For the EF Eri models presented here,
248: the secondary star model component (see Table \ref{t:shared-params})
249: is represented by
250: the L5 star 2MASS~J15074769$-$1627386 \citep{reid00} scaled to
251: a distance of $d=132$ pc (see \S\ref{s:models}).
252:
253: Third, we have made a number of minor improvements to the
254: calculation of the optically thin CBD component SED, which is
255: primarily used in this work instead of the optically thick CBD used in B07.
256: In general, the procedure remains as described in B07.
257: The SED of the CBD is obtained by summing the contribution of
258: 1000 annular rings, each of which has the same width and the
259: correct temperature for its average radial distance from the
260: center-of-mass of the CV
261: (based on the $T \propto [1/r]^{3/4}$ profile explained in
262: more detail in B07).
263: The volume of a ring increases as its radial
264: distance increases, and we assume equal mass in each ring, which
265: results in rings that are successively less dense at larger
266: radial distances.
267:
268: We have imposed an
269: upper limit of 1000 K for the temperature of the inner edge of
270: the CBD. Not only is it difficult to devise a mechanism that
271: would heat the inner edge of the CBD in EF Eri above 1000 K, but
272: dust sublimation (hence, destruction of the CBD) likely occurs
273: at temperatures above 1000--2000 K.
274: Based purely on the WD effective temperature, we would expect the
275: temperature at the radius of the inner edge of the CBD to be
276: $T_{\rm in}\approx250$ K; including the contribution from the
277: secondary star, this temperature increases to
278: $T_{\rm in}\approx400$--500 K.
279: However, as noted in B07, the upper limit
280: to this temperature is difficult to assess because of the
281: uncertainties in how to properly account for the
282: contributions from the secondary star and accretion luminosity;
283: in addition, the conversion from the ambient temperature of
284: irradiation at a particular radius to temperature of the CBD
285: material at that radius is uncertain and highly dependent on the
286: physical parameters of the CBD and the WD accretion spot(s).
287: Especially considering the evidence from recent UV and X-ray
288: observations \citep{szkody06,schwope07} that there is a
289: significant hot spot on the WD in EF Eri, the temperature at
290: the inner edge of the CBD could be up to
291: several hundreds of K higher. Effectively, then, we define an
292: ``allowed'' range of $T_{\rm in}=400$--1000 K.
293:
294: Finally, and most significantly, we have replaced the purely
295: morphological, template-based cyclotron component from B07 with
296: a physical model based on the formulation discussed and used in,
297: for example, \citet{chanmugam80,TC87,schwope90} and references therein.
298: The calculation of the cyclotron SED is influenced by several
299: parameters:\ the WD magnetic field strength ($B$), the viewing
300: angle ($\theta$), the electron temperature ($kT$), a dimensionless
301: ``size'' parameter ($\Lambda$), and a normalization factor ($A$).
302: None of these parameters is strictly independent from the
303: others -- the change in the cyclotron SED produced by adjusting
304: one parameter can usually be offset by adjusting one or more of the
305: other parameters. This makes it difficult to produce a unique
306: solution without having reliable, very narrow constraints for as
307: many of the parameters as possible. We will describe the constraints
308: used to narrow down the possible cyclotron SEDs for EF Eri
309: in \S\ref{s:models}. In the remainder of this section, we will
310: discuss, in general terms, the manner in which each parameter
311: influences the cyclotron SED.
312:
313: To first order, the WD magnetic field strength ($B$) determines
314: the wavelengths at which cyclotron humps appear in the SED for
315: successively higher harmonic numbers. Smaller values of $B$
316: produce more redshifted humps. However, the wavelengths of the
317: harmonics are also influenced by the electron temperature ($kT$)
318: and the viewing angle ($\theta$). Increasing $kT$ or decreasing
319: $\theta$ redshifts the humps.
320: The humps become broader with increasing $kT$, increasing size
321: parameter $\Lambda$ (however, see below for additional effects
322: of changing $\Lambda$), or decreasing $\theta$. At very high
323: values of these parameters (especially $kT$), the humps are so
324: broad as to be completely overlapping and blended together,
325: effectively forming a humpless pseudo-continuum. At very low
326: values, the cyclotron humps are very narrow, resulting in the
327: cyclotron spectrum consisting of multiple discrete hump profiles
328: with regions of zero intensity between them.
329: The normalization factor ($A$) is a simple scaling factor given
330: by the ratio of the effective emitting area of the cyclotron
331: component to the square of the distance to the CV.
332:
333: The size parameter $\Lambda$ deserves additional explanation.
334: Although its value is set in the cyclotron SED calculation as
335: a single number, this parameter actually incorporates several
336: other parameters. It is defined (see \citealt{schwope90}) as
337: \begin{equation}
338: \Lambda = 2.01\times10^{5} \left(\frac{s}{10^5\,{\rm [cm]}}\right) \left(\frac{N_{\rm e}}{10^{16}\,{\rm [cm^{-3}]}}\right) \left(\frac{3\times10^7\,{\rm [G]}}{B}\right),
339: \end{equation}
340: \noindent where $s$ is the geometric path length through the
341: cyclotron emitting region and $N_{\rm e}$ is the electron number
342: density. For a fixed value of $B$, the value of $\Lambda$ is
343: dependent on only the product of $s$ and $N_{\rm e}$.
344: Functionally, larger values of $\Lambda$ correspond to higher
345: optical depths, which influences the harmonic number at which
346: the cyclotron humps transition from being optically thick
347: (at lower harmonics) to optically thin (at higher harmonics).
348: Optically thick cyclotron humps have a truncated, flat-topped
349: appearance compared to the rounded optically thin humps.
350: The value of $\Lambda$ also has a large effect on the relative
351: amplitudes of the cyclotron humps. The lower harmonic, optically
352: thick humps at longer wavelengths rapidly decline in amplitude as
353: harmonic number decreases compared to the optically thin humps at
354: higher harmonics (shorter wavelengths).
355:
356:
357: \subsection{The Models}
358: \label{s:models}
359:
360: \subsubsection{Distance to EF Eri}
361:
362: The distance used to scale the model components in this work is
363: the $1\sigma$ upper limit of a trigonometric parallax-derived
364: distance from \citet{thor03}, and is approximately halfway between
365: the two possible nominal parallax distance values of 113 and 162 pc
366: reported in that work. We increased the distance from the 105 pc
367: used in B07 (which was based on non-parallax estimates) because
368: it is now clear that cyclotron emission makes a non-negligible
369: contribution in the $J$ band (see discussion of models below),
370: whereas at the smaller distance, the observed $J$-band flux
371: density was accounted for completely by the WD and secondary star
372: components. The value of 132 pc used here is the minimum distance
373: for which the summed WD, secondary star, and cyclotron emission
374: do not exceed the observed photometric and spectroscopic data
375: at $J$ band.
376:
377:
378: \subsubsection{Cyclotron Component Constraints}
379: \label{s:cyccon}
380:
381: As noted in \S\ref{s:code}, our revised cyclotron model component
382: is calculated using a number of non-independent parameters, so it
383: is helpful to constrain, as much as possible, the range of valid
384: parameter space. To that end, we describe in this section the
385: various constraints that we used to limit the cyclotron model
386: parameters.
387:
388: \begin{description}
389: \item[$B$:] The WD magnetic field in EF Eri is frequently derived
390: as $B=13$--$14$ MG (e.g., \citealt{WR98,howell06b,beuermann07}).
391: A value of $B$ in this range is consistent with the location of
392: the cyclotron humps in the near-IR spectrum of EF Eri
393: (\citealt{harrison07}; also see Figure \ref{f:bigspec}).
394: We do not consider the much more complex scenario in which the
395: cyclotron spectrum of EF Eri results from two or more magnetic
396: accretion regions with significantly different field strengths.
397:
398: \item[$kT$:] The near-IR spectrum of EF Eri also provides useful
399: constraints for this parameter. Even after subtracting the WD
400: and secondary star components, there is residual flux between the
401: cyclotron humps. This requires $kT\gtrsim5$ keV (at lower values
402: of $kT$ and for reasonable values of $\Lambda$ -- see below -- the
403: cyclotron spectrum is composed of discrete humps with zero flux
404: level between them).
405: In addition, for $kT\lesssim5$ keV, and for values of $\Lambda$
406: that reproduce the observed optically thick to thin transition
407: and relative hump peak amplitudes, the cyclotron humps are too
408: narrow compared to the observed near-IR spectrum.
409: The fact that discrete cyclotron humps are observed requires
410: $kT\lesssim20$ keV.
411:
412: \item[$\theta$:] In the absence of consistent information
413: constraining this parameter, we chose to set $\theta=75^{\circ}$
414: for all models. In any case, $\theta$ cannot be much lower or
415: higher than this, or the widths and relative amplitudes of the
416: optically thin humps will not match the observed near-IR spectrum.
417:
418: \item[$\Lambda$:] For any given values of the other parameters,
419: $\Lambda$ is the most constrained parameter. This is because
420: only a narrow range of its values will reproduce the observed
421: near-IR spectrum, which shows the $n=4$ harmonic as optically
422: thick, and higher harmonics as optically thin. The value of
423: $\Lambda$ is further fine-tuned by matching the relative peak
424: amplitudes of both the optically thick and thin cyclotron humps.
425:
426: \item[$A$:] The main constraint on the scaling parameter is that
427: for a given distance, $A$ should correspond to an effective
428: emitting area of the cyclotron radiation ($a_{\rm cyc}$) that
429: is (much) smaller than the projected surface area of the WD.
430: \end{description}
431:
432: Most of these constraints are derived from the near-IR spectrum
433: of EF Eri, which shows the cyclotron hump at the $n=4$ harmonic
434: to be flat-topped and much lower in peak amplitude than the humps
435: at the $n\geq5$ harmonics (see Figure \ref{f:bigspec}).
436: This indicates that the transition
437: from optically thick to thin cyclotron emission occurs in the
438: vicinity of the $n=4$ harmonic. In particular, satisfying the
439: observed transition from optically thick to thin humps results
440: in the mid-IR contribution of the cyclotron emission being
441: increasingly negligible at longer wavelengths. That is, the
442: cyclotron component is only important at relatively short IR
443: wavelengths, and does not contribute significantly to the mid-IR
444: region spanned by the Spitzer data.
445:
446: As expected from the description in \S\ref{s:code}, we found it
447: difficult to determine a unique, ``best'' solution for the
448: cyclotron model component. We determined the relative goodness
449: of different model cyclotron components by calculating the
450: $\chi^2$ and standard deviation of the residuals ($\sigma_{\rm res}$)
451: in the 1--2.5 $\mu$m wavelength region for the summed cyclotron,
452: WD, and secondary star model components compared to the observed
453: near-IR spectrum.
454: Table \ref{t:cycmodels} lists several ``best'' cyclotron models
455: determined in this way for a range of $kT$ values.
456: As $kT$ decreases, we had to decrease $B$ and increase $\Lambda$
457: in order for the cyclotron component to continue to match the
458: wavelength spacing, widths, and relative peak amplitudes of the
459: observed cyclotron humps.
460: The goodness of the model also tends to decrease with decreasing
461: $kT$, although none of the models has particularly poor agreement
462: with the observed SED.
463: In the end, for the purposes of this work, the specific cyclotron
464: component that we use is not particularly important since all of
465: them contribute negligibly at mid-IR wavelengths. We will use
466: the $kT=10$ keV cyclotron model in the remainder of this work.
467:
468: In B07, we utilized a ``sum-over-fields'' approach to calculating
469: the cyclotron component, which considered the summed contributions
470: from cyclotron emission of electrons encountering a successively
471: stronger magnetic field as they approached the WD. In part, this
472: was an attempt to set a ``worst case'' limit for the contribution
473: of cyclotron emission at long wavelengths (since the strength of
474: cyclotron emission at long wavelengths is increased relative to
475: short wavelengths through this approach). However, another effect
476: of the sum-over-fields approach is to smear out the individual
477: cyclotron humps, which is clearly inconsistent with the observed
478: near-IR spectrum of EF Eri. Consequently, we have not used that
479: approach here. In any case, even if we considered a two-part
480: cyclotron component consisting of the single-field {\em and}
481: summed-fields cases, the contribution of the latter would have
482: to be extremely small in order to not dilute the strong observed
483: single-field spectrum (and, regardless of strength, would not
484: contribute significantly at $\lambda\gtrsim4$ $\mu$m).
485:
486:
487: \subsubsection{Comparison with Brinkworth et al.\ (2007)}
488: \label{s:comparison}
489:
490: Figure \ref{f:models}a shows the observational data from
491: Figure \ref{f:bigspec} with a model containing an optically thick
492: CBD similar to the best optically thick CBD component from B07
493: (see Model 1 in Table 4 of that work). It has been adusted
494: slightly to account for the larger distance used here by changing
495: the inner edge temperature from 655 K to 755 K.
496: The criterion used in B07, of best reproducing the 8-$\mu$m IRAC
497: point without exceeding any shorter wavelength point, has been
498: preserved. This model has the wrong spectral shape and
499: significantly exceeds the {\em longer} wavelength SED of EF Eri
500: that is revealed by our IRS spectrum.
501:
502:
503: \subsubsection{New Model Results}
504: \label{s:newmodels}
505:
506: Figure \ref{f:models}b shows a new model that utilizes the
507: additional constraints on flux density at long wavelengths
508: provided by our IRS spectrum of EF Eri.
509: The parameters common to this model and the one discussed
510: below are listed in Table \ref{t:shared-params}, while parameters
511: specific to this model are listed in Table \ref{t:specific-params}
512: (Model 1). For this model, we have utilized an optically thin CBD
513: composed of spherical dust grains that radiate as blackbodies
514: according to the radial temperature profile calculated as for
515: the optically thick CBD case (see \S\ref{s:code} and B07).
516:
517: This model shows significant improvement over that shown in
518: Figure \ref{f:models}a, especially at the short and long wavelength
519: ends. However, the model flux density is too low at the middle
520: wavelengths (i.e., IRAC channels 1 and 2 at 3.6 and 4.5 $\mu$m).
521: The total mass of the CBD is $\approx 10^{21}$ g, consistent with
522: the finding from B07 that the masses of CBDs in magnetic CVs are
523: many orders of magnitude smaller than predicted to be required to
524: influence the angular momentum loss history of these systems \citep{taam03}.
525: For lower and higher inner edge CBD temperatures, the overall
526: flux density level of the observed SED can be matched by increasing
527: or decreasing, respectively, the total disk mass (i.e., the number
528: of radiating dust grains).
529: If the temperature of the inner edge of the CBD is decreased, then
530: the match between the model and observed SEDs becomes worse -- the
531: CBD profile does not reach peak flux density until an even longer
532: wavelength, which exacerbates the problem of missing flux density
533: at the IRAC wavelengths.
534: A higher temperature for the inner edge of the CBD produces a
535: better match at the middle wavelengths. However, the model is
536: then too faint at the long wavelength end, since the CBD SED peaks
537: and begins to decline at a shorter wavelength.
538: (See \S\ref{s:cbdcon} for more discussion of the constraints on CBD
539: model parameters.)
540:
541: We have explored a possible means of reconciling
542: this CBD model with the 3.6 and 4.5 $\mu$m data.
543: The model in Figure \ref{f:models}c is similar to that shown in
544: Figure \ref{f:models}b, but uses a blackbody component to account
545: for the ``missing'' flux at 3.6 and 4.5 $\mu$m. The model
546: parameters are listed in Tables \ref{t:shared-params} and
547: \ref{t:specific-params} (Model 2). This model has the advantage
548: that it requires a CBD with a low inner edge temperature that
549: could easily be produced by irradiation from the stellar
550: components in EF Eri.
551: On the other hand, it has the disadvantage that the physical
552: origin of the additional component is unclear. The required
553: equivalent emitting area (corresponding to a radius of $46R_{\rm wd}$)
554: is too large to be contained in the stellar Roche lobes, which
555: points to the CBD. The required temperature is 1000 K, which is
556: somewhat uncomfortably warm from considerations of both the
557: origin of the heating and potential destruction of the dust grains.
558: However, it might suggest that a more complex radial temperature
559: profile in the CBD could produce an SED shape that is more
560: consistent with the observations. We have not explored this
561: possibility in detail because it would introduce yet more free
562: parameters into our already barely constrained model.
563:
564:
565: \subsubsection{Uniqueness, Plausibility, and Constraints of the Circumbinary Disk Models}
566: \label{s:cbdcon}
567:
568: Much like the cyclotron emission component (see \S\ref{s:code}
569: and \S\ref{s:cyccon}), the model CBD SEDs shown in this work are not,
570: strictly speaking, unique solutions, in the sense that very similar
571: results can be achieved from somewhat different combinations of input
572: parameters. We have tried to minimize this as much as possible by
573: constraining plausible parameter ranges based on whatever other
574: information, observational data, and reasonable assumptions are available.
575: In this section, we describe in detail the justification for the
576: constraints that have been assumed in fixing the exponent in the
577: radial temperature profile calculation (see \S\ref{s:code}) and
578: the inner radius, $R_{\rm in}$, of the model CBD. We also expand
579: upon the discussion of the failure of the optically thick CBD
580: model first mentioned in \S\ref{s:comparison}.
581:
582: In general terms, the influence of the exponent in the radial
583: temperature profile can be described as follows:\
584: a larger (smaller) exponent in the CBD radial temperature profile
585: leads to a steeper (shallower) temperature gradient near the
586: inner edge of the disk and overall lower (higher) temperature
587: throughout the disk, which corresponds to large (small) dust grains
588: that do (do not) cool efficiently. In practice, we have found that
589: an exponent of 3/4, as used here and in B07, produces the most
590: viable results in comparison with our observations of CVs.
591: Smaller exponents produce CBD SEDs that rise too steeply and are,
592: overall, too bright to reproduce the observed SEDs without
593: arbitrarily increasing the distances to the CVs to many
594: hundreds or thousands of pc.
595: Larger exponents produce CBD SEDs that rise too shallowly at longer
596: wavelengths and are, overall, too faint to match the observed mid-IR
597: flux densities without arbitrarily increasing the temperature of
598: the disk's inner edge to unrealistic levels.
599: For example, we can obtain a model optically thin CBD SED that
600: is essentially indistinguishable from that in Model 1
601: (Figure \ref{f:models}b) by using a larger radial
602: temperature profile exponent (1 instead of 3/4; see \S\ref{s:code})
603: and smaller total disk mass ($4.41\times10^{20}$ g instead
604: of $9.55\times10^{20}$ g); however, one objection to this
605: approach is that departure from an exponent of 3/4 implies
606: dust grains that are not blackbodies \citep{FKR}, whereas this is
607: an implicit assumption of the CBD model calculations.
608:
609: The inner edge radius of the CBD models calculated in this work
610: (for both optically thick and thin cases) is fixed
611: at $R_{\rm in}=73R_{\rm wd}$,
612: which is a lower limit set by the tidal truncation radius of EF Eri.
613: Increasing the inner radius of the CBD from this
614: value worsens the agreement between the CBD models and the
615: observed SED, because the irradiation-induced temperature of
616: the inner edge will then be lower. This effectively removes flux
617: from the short IR wavelengths of the model SED, whereas the main
618: problem we have in reproducing the observations is that the models
619: already have too little flux at short IR wavelengths. (The outer
620: radius of the CBD is calculated to correspond to a temperature of
621: 20 K according to the radial temperature profile in use, but at
622: wavelengths shortward of 15 $\mu$m the resultant profile is
623: insensitive to increasing the temperature of the outer radius
624: cut-off by as much as an order of magnitude.)
625: Even if we arbitrarily (and unphysically) decrease the inner
626: edge radius of the CBD to $50R_{\rm wd}$, which is barely larger
627: than the distance from the CV's center-of-mass to the back
628: of the secondary star's Roche lobe ($49R_{\rm wd}$),
629: the resultant model CBD SED (for both
630: optically thick and thin cases) has only a few percent improvement
631: (increase) in flux density at IRAC channels 1 and 2. Similarly,
632: the agreement with the IRAC channels 3 and 4 and IRS data is not
633: significantly better than that achieved by the CBD component used
634: in Model 1 (Figure \ref{f:models}b).
635:
636: The optically thick CBD models are parameterized solely by the
637: radial temperature profile (including the value of $T_{\rm in}$)
638: and size (i.e., inner and outer radii). As such, the variety
639: of possible model optically thick CBD SEDs is rather more limited
640: than for the optically thin case. Changing either the radial
641: temperature profile exponent or $R_{\rm in}$ does not yield an
642: optically thick CBD model that reproduces the observed SED any
643: better than the model shown in Figure \ref{f:models}a
644: and discussed in \S\ref{s:comparison}.
645: For example, using an exponent of 1, we can produce an optically
646: thick CBD SED whose shape (i.e., relative intensity at each
647: wavelength) is almost indistinguishable from that of the optically
648: thin model CBD SED in Model 1. However, it is overall 15--30\%
649: fainter than the optically thin model, so produces much worse
650: agreement with the observations. Smaller exponents produce model
651: CBD SEDs that are far too bright, especially at longer wavelengths.
652: Increasing $T_{\rm in}$ for the optically thick CBD produces more
653: flux at shorter wavelengths, but also increases the steepness of
654: the SED and the excess flux at longer wavelengths. Decreasing
655: $T_{\rm in}$ begins to flatten the SED, but also makes it overall
656: too faint at all wavelengths, especially the short IR wavelengths.
657: Increasing $R_{\rm in}$ also fails, since (as described above)
658: this results in even less flux at the short IR wavelengths.
659: In light of the flat shape of the EF Eri SED at wavelengths
660: longer than the IRAC regime, we conclude from the behavior of
661: the optically thick CBD models described in this section that they
662: are much less likely than the more flexible optically thin CBD models
663: as viable explanations of the mid-IR SED of EF Eri.
664:
665:
666: \section{Conclusions}
667:
668: Our newly obtained mid-IR spectrum of EF Eri has allowed us to
669: further constrain and refine the SED model for this system first
670: presented in B07. Based on the B07 model, which was constrained
671: by only 2MASS and IRAC photometric data, we would have expected
672: the 8--14 $\mu$m SED of EF Eri to either rise
673: (if dominated by an optically thick CBD) or fall (if dominated
674: by short wavelength cyclotron emission). However, the observed
675: spectrum defied both of our expectations, by remaining almost
676: flat compared to the IRAC data. This does allow us to eliminate
677: cyclotron emission as a dominant component in the mid-IR SED of
678: EF Eri beyond $\lambda\approx3$ $\mu$m. At the same time, we
679: also show that a geometrically thin, but optically thick,
680: CBD is unlikely as a viable explanation for the spatial
681: distribution of dust in EF Eri. Instead, the
682: dust is more likely present as a geometrically and optically thin CBD.
683: In all cases, however, there are inconsistencies between our CBD
684: models and the observed SED of EF Eri at the IRAC channel 1 and 2
685: wavelengths (3.6 and 4.5 $\mu$m) that imply a more complex situation
686: is present than represented by our simple CBD models.
687: As also found in B07, the total mass of dust in the CBD is still
688: several orders of magnitude too small to strongly affect the
689: secular evolution of CVs in the context of current models of
690: angular momentum loss mechanisms in CVs.
691:
692: Based on our study of EF Eri, we can make several generalizations
693: regarding the mid-IR observational properties of CVs that contain dust.
694: First, EF Eri is a low-field polar. In polars with moderate to
695: strong WD magnetic fields of several tens of MG or more, cyclotron
696: emission will be shifted to even shorter wavelengths and be even
697: less important for understanding the mid-IR SED.
698: Second, EF Eri contains a very low mass, faint secondary star.
699: Longer orbital period systems will have correspondingly more
700: massive, brighter secondary stars. However, in the IRAC bands,
701: an M5 dwarf is only $\approx10$ times brighter than an L5 brown
702: dwarf \citep{patten06}. Although the observed 3.6 $\mu$m flux
703: density in EF Eri is comparable to that of an M5 star, at 8 $\mu$m
704: the observed flux density is $\approx4$ times that of an M5 star.
705: So, even in systems containing a more massive secondary star, dust
706: emission at a comparable level to that in EF Eri produces a mid-IR
707: SED far in excess of the stellar components.
708: Finally, EF Eri lacks an accretion disk. In non-magnetic CVs, the
709: hot accretion disk will appear in the IR as a Rayleigh-Jeans-like
710: tail similar in shape to the WD SED, but likely significantly
711: brighter than both stellar components. However, the maximum
712: possible emitting area for the accretion disk is limited by the
713: size of the WD Roche lobe. The CBD, on the other hand, can have
714: an emitting area many orders of magnitude larger. Hence, even in
715: the presence of an accretion disk, the system SED at the longest
716: mid-IR wavelengths could still be dominated by dust emission.
717:
718: An extrapolation of our current CBD model for EF Eri to even
719: longer wavelengths predicts a continued gradual decline in the
720: overall flux density, with (for example) the total flux density
721: of EF Eri at 24 $\mu$m being about 90\% of the 8-$\mu$m value.
722: Longer wavelength observations (e.g., at the {\em Spitzer}
723: Peak-up Imaging 22-$\mu$m or MIPS 24-$\mu$m bands) could test
724: this prediction.
725:
726:
727:
728: \acknowledgments
729:
730: This work is based in part on observations made with the
731: {\em Spitzer Space Telescope}, which is operated by the Jet
732: Propulsion Laboratory, California Institute of Technology,
733: under a contract with the National Aeronautics and Space
734: Administration (NASA).
735: Support for this work was provided by NASA.
736: We thank the Spitzer Science Center (SSC) Director for his
737: generous allocation of observing time for the NASA/NOAO/{\em Spitzer
738: Space Telescope} Observing Program for Students and Teachers.
739: The National Optical Astronomy Observatory (NOAO), which is
740: operated by the Association of Universities for Research
741: in Astronomy (AURA), Inc., under cooperative agreement with the
742: National Science Foundation (NSF), has provided many in kind
743: contributions for which SBH is grateful.
744: This work makes use of data products from the
745: Two Micron All Sky Survey, which is a joint project of the
746: University of Massachusetts and the Infrared Processing and
747: Analysis Center/Caltech, funded by NASA and the NSF.
748: CSB acknowledges support from the SSC Enhanced Science Fund
749: and NASA's Michelson Science Center.
750: DWH thanks Axel Schwope for helpful advice on calculating
751: cyclotron spectra.
752:
753:
754: \clearpage
755:
756: \begin{thebibliography}{99}
757:
758: \bibitem[Beuermann et al.(2007)]{beuermann07} Beuermann, K.,
759: Euchner, F., Reinsch, K., Jordan, S., G\"{a}nsicke, B.~T.\ 2007,
760: \aap, 463, 647
761:
762: \bibitem[Beuermann et al.(2000)]{beuermann00} Beuermann, K.,
763: Wheatley, P., Ramsay, G., Euchner, F., \& G\"ansicke, B.~T.\ 2000,
764: \aap, 354L, 49
765:
766: \bibitem[Brinkworth et al.(2007)]{brinkworth07} Brinkworth, C.~S.,
767: et al.\ 2007, \apj, 659, 1541 (B07) %%% 9 authors
768:
769: \bibitem[Chanmugam(1980)]{chanmugam80} Chanmugam, G.\ 1980, \apj, 241, 1122
770:
771: \bibitem[Cushing et al.(2006)]{cushing06} Cushing, M.~C., et al.\
772: 2006, \apj, 648, 614 %%% 11 authors
773:
774: \bibitem[Fazio et al.(2004)]{fazio04} Fazio, G., et al.\ 2004, \apjs, 154, 10
775:
776: \bibitem[Frank, King, \& Raine(2002)]{FKR} Frank, J., King, A., \&
777: Raine, D.\ 2002, in Accretion Power in Astrophysics (Cambridge:\
778: Cambridge University Press), ch.\ 5
779:
780: \bibitem[Harrison et al.(2007)]{harrison07} Harrison, T.~E.,
781: Campbell, R.~K., Howell, S.~B., Cordova, F.~A., \& Schwope, A.~D.\ 2007,
782: \apj, 656, 444
783:
784: \bibitem[Harrison et al.(2004)]{harrison04} Harrison, T.~E.,
785: Howell S.~B., Szkody, P., Homeier, D., Johnson, J., \& Osborne, H.\ 2004,
786: \apj, 614, 947
787:
788: \bibitem[Higdon et al.(2004)]{higdon04} Higdon, S.~J.~U., et al.\ 2004,
789: \pasp, 116, 975 %%% 11 authors
790:
791: \bibitem[Houck et al.(2004)]{houck04} Houck, J.~R., et al.\ 2004,
792: \apjs, 154, 18 %%% 35 authors
793:
794: \bibitem[Howell \& Ciardi(2001)]{HC01} Howell, S.~B., \& Ciardi, D.~R.\
795: 2001, \apjl, 550, L57
796:
797: \bibitem[Howell et al.(2001)]{howell01} Howell, S.~B., Nelson, L., \&
798: Rappaport, S.\ 2001, \apj, 550, 897
799:
800: \bibitem[Howell et al.(2006a)]{howell06a} Howell, S.~B., et al.\ 2006,
801: \apjl, 646, L65 %%% 11 authors
802:
803: \bibitem[Howell et al.(2006b)]{howell06b} Howell, S.~B., Walter, F.~M.,
804: Harrison, T.~E., Huber, M.~E., Becker, R.~H., \& White, R.~L.\ 2006,
805: \apj, 652, 709
806:
807: \bibitem[Patten et al.(2006)]{patten06} Patten, B.~M., et al.\ 2006,
808: \apj, 651, 502 %%% 12 authors
809:
810: \bibitem[Piirola et al.(1987)]{piirola87} Piirola, V., Coye, G.~V., \&
811: Reiz, A.\ 1987, \aap, 186, 120
812:
813: \bibitem[Reid et al.(2000)]{reid00} Reid, I.~N., Kirkpatrick, J.~D.,
814: Gizis, J.~E., Dahn, C.~C., Monet, D.~G., Williams, R.~J.,
815: Liebert, J., \& Burgasser, A.~J.\ 2000, \aj, 119, 369
816:
817: \bibitem[Schwope et al.(1990)]{schwope90} Schwope, A.~D.,
818: Beuermann, K., \& Thomas, H.-C.\ 1990, \aap, 230, 120
819:
820: \bibitem[Schwope et al.(2007)]{schwope07} Schwope, A.~D., Staude, A.,
821: Koester, D., \& Vogel, J.\ 2007, \aap, 469, 1027
822:
823: \bibitem[Sloan et al.(2003)]{sloan03} Sloan, G.~C., Kraemer, K.~E.,
824: Price, S.~D., \& Shipman, R.~F.\ 2003, \apjs, 147, 379
825:
826: \bibitem[Sloan et al.(2005)]{sloan05} Sloan, G.~C., et al.\ 2005,
827: \apj, 632, 956 %%% 19 authors
828:
829: \bibitem[Szkody et al.(2006)]{szkody06} Szkody, P., Harrison, T.~E.,
830: Plotkin, R.~M., Howell, S.~B., Seibert, M., \& Bianchi, L.\ 2006,
831: \apjl, 646, L147
832:
833: \bibitem[Taam et al.(2003)]{taam03} Taam, R.~E., Sandquist, E.~L., \&
834: Dubus, G.\ 2003, \apj, 592, 1124
835:
836: \bibitem[Thompson \& Cawthorne(1987)]{TC87} Thompson, A.~M., \&
837: Cawthorne, T.~V.\ 1987, \mnras, 224, 425
838:
839: \bibitem[Thorstensen(2003)]{thor03} Thorstensen, J.~R.\ 2003, \aj, 126, 3017
840:
841: \bibitem[Warner(1995)]{warner95} Warner, B.\ 1995, in Cataclysmic
842: Variable Stars (Cambridge:\ Cambridge University Press)
843:
844: \bibitem[Werner et al.(2004)]{werner04} Werner, M., et al.\ 2004,
845: \apjs, 154, 1
846:
847: \bibitem[Wheatley \& Ramsay(1998)]{WR98} Wheatley, P.~J., \&
848: Ramsay, G.\ 1998, Wild Stars in the Old West, 137, 446
849:
850: \end{thebibliography}
851:
852:
853: %%% START TABLES %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
854:
855: \clearpage
856:
857: \begin{deluxetable}{llllllll}
858: \tablewidth{0pt}
859: \tabletypesize{\footnotesize}
860: %\rotate
861: \tablecaption{Representative cyclotron models \label{t:cycmodels}}
862: \tablehead{
863: \colhead{$kT$} &
864: \colhead{$B$} &
865: \colhead{$\theta$} &
866: \colhead{$\log(\Lambda)$} &
867: \colhead{$\log(A)$} &
868: \colhead{$a_{\rm cyc}$} &
869: \colhead{$\chi^2$} &
870: \colhead{$\sigma_{\rm res}$} \\
871: \colhead{(keV)} &
872: \colhead{(MG)} &
873: \colhead{($^{\circ}$)} &
874: \colhead{ } &
875: \colhead{ } &
876: \colhead{($a_{\rm wd})$\tablenotemark{a}} &
877: \colhead{ } &
878: \colhead{(mJy)}
879: }
880: \startdata
881: 12 & 13.6 & 75 & 2.845 & $-25.178$ & 0.0145 & 0.202 & 0.0432 \\
882: 10 & 13.4 & 75 & 3.267 & $-25.106$ & 0.0172 & 0.224 & 0.0475 \\
883: 8 & 13.35 & 75 & 3.653 & $-25.044$ & 0.0198 & 0.424 & 0.0535 \\
884: 6 & 13.2 & 75 & 4.342 & $-24.919$ & 0.0264 & 0.728 & 0.0737
885: \enddata
886: \tablenotetext{a}{The cyclotron emitting area $a_{\rm cyc}$ determined
887: by the scale factor $A$ is given in units of the projected WD surface
888: area $a_{\rm wd}=\pi R_{\rm wd}^2$.}
889: \end{deluxetable}
890:
891:
892: \begin{deluxetable}{lll}
893: \tablewidth{0pt}
894: \tabletypesize{\footnotesize}
895: %\rotate
896: \tablecaption{Shared model parameters \label{t:shared-params}}
897: \tablehead{
898: \colhead{Component} &
899: \colhead{Parameter} &
900: \colhead{Value}
901: }
902: \startdata
903: System: & Orbital Period, $P_{\rm orb}$ (min) & 81.022932(8) [5] \\
904: & Inclination, $i$ ($^{\circ}$) & 55(5) [5] \\
905: & Distance, $d$ (pc) & 132 [6] \\
906: WD: & Temperature, $T_{\rm wd}$ (1000 K) & 9.5(0.5) [1] \\
907: & Mass, $M_{\rm wd}$ ($M_{\odot}$) & 0.6 [1] \\
908: & Radius, $R_{\rm wd}$ ($R_{\odot}$) & 0.0125 [2] \\
909: SS: & Template Star & 2MASS J15074769$-$1627386 \\
910: & Spectral Type & L5 [2,3] \\
911: & Temperature, $T_{2}$ (1000 K) & 1.7 [2] \\
912: & Mass, $M_2$ ($M_{\odot}$) & 0.055 [4] \\
913: & Radius, $R_{2}$ ($R_{\odot}$) & 0.1 [2] \\
914: CBD: & Optical Depth Prescription & Thin \\
915: & Temperature Profile Exponent & 0.75 \\
916: & Disk Annulus Treatment & Equal mass \\
917: & Constant Height, $h_{\rm cbd}$ ($R_{\rm wd}$) & 0.00167 \\
918: & Grain Density, $\rho_{\rm grain}$ (g cm$^{-3}$) & 3.0 \\
919: & Grain Radius, $r_{\rm grain}$ ($\mu$m) & 1
920: \enddata
921: \tablerefs{
922: [1] = \citet{beuermann00},
923: [2] = \citet{brinkworth07} (and references therein),
924: [3] = \citet{HC01},
925: [4] = \citet{howell06b},
926: [5] = \citet{piirola87}, and
927: [6] = \citet{thor03}.
928: }
929: \end{deluxetable}
930:
931:
932: \begin{deluxetable}{llll}
933: \tablewidth{0pt}
934: \tabletypesize{\footnotesize}
935: %\rotate
936: \tablecaption{Model-specific parameters \label{t:specific-params}}
937: \tablehead{
938: \colhead{Component} &
939: \colhead{Parameter} &
940: \multicolumn{2}{c}{Values:} \\
941: \colhead{ } &
942: \colhead{ } &
943: \colhead{Model 1} &
944: \colhead{Model 2}
945: }
946: \startdata
947: CYC: & WD Magnetic Field, $B$ (MG) & 13.4 & 13.4 \\
948: & Electron Temperature, $kT$ (keV) & 10 & 10 \\
949: & Angle to Magnetic Field, $\theta$ ($^{\circ}$) & 75 & 75 \\
950: & Size Parameter, $\log(\Lambda)$ & 3.267 & 3.301 \\
951: & Scale Factor, $\log(A)$ & $-25.106$ & $-25.220$ \\
952: & Emitting Area, $a_{\rm cyc}$ ($a_{\rm wd}$) & 0.0172 & 0.0132 \\
953: & Model Goodness, $\chi^2$ & 0.224 & 0.286 \\
954: & Model Residuals, $\sigma_{\rm res}$ (mJy) & 0.0475 & 0.0522 \\
955: CBD: & Inner Edge Temperature, $T_{\rm in}$ (K) & 830 & 450 \\
956: & Inner Edge Radius, $R_{\rm in}$ ($R_{\rm wd}$) & 73 & 73 \\
957: & Outer Edge Temperature, $T_{\rm in}$ (K) & 20 & 20 \\
958: & Outer Edge Radius, $R_{\rm in}$ ($R_{\rm wd}$) & 10482 & 4634 \\
959: & Total Mass, $M_{\rm cbd}$ ($10^{20}$ g) & 9.55 & 17.3 \\
960: BB: & Temperature, $T$ (K) & \nodata & 1000 \\
961: & Radius of emitting area, $R_{\rm bb}$ ($R_{\rm wd}$) & \nodata & 45.6
962: \enddata
963: \end{deluxetable}
964:
965:
966: %%% END TABLES %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
967:
968:
969: %%% START FIGURES %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
970:
971: \clearpage
972:
973: \begin{figure}
974: \epsscale{1.00}
975: \plotone{f1.eps}
976: %\plotone{f1color.eps}
977: \epsscale{1.00}
978: \caption{Mid-IR spectrum of EF Eri from the {\em Spitzer Space
979: Telescope} (with $1\sigma$ error bars).
980: \label{f:spectrum}}
981: \end{figure}
982:
983:
984: \begin{figure}
985: \epsscale{1.00}
986: \plotone{f2.eps}
987: %\plotone{f2color.eps}
988: \epsscale{1.00}
989: \caption{IR observations of EF Eri. The photometric data from
990: \citet{brinkworth07} are shown as filled circles; from short to
991: long wavelength, these are 2MASS $J$, $H$, and $K_{\rm s}$, and
992: IRAC channels 1--4. Error bars on all of the photometric points are
993: the $1\sigma$ photometric uncertainties in the y-direction (these
994: are smaller than the plotting symbols for IRAC channels 1 and 2),
995: and the widths of the photometric bands in the x-direction.
996: The near-IR spectrum (taken from Figure 1 of \citealt{harrison07})
997: and IRS mid-IR spectrum of EF Eri are shown as solid lines (error
998: bars have been omitted for clarity).
999: The vertical hashmarks show the cyclotron harmonic peaks for $n=2$--7
1000: (top panel, right to left) and the cyclotron fundamental (bottom panel)
1001: for the model shown in Figure \ref{f:models}b and
1002: discussed in \S\ref{s:newmodels}.
1003: Note the different wavelength scales in the two panels.
1004: \label{f:bigspec}}
1005: \end{figure}
1006:
1007:
1008: \begin{figure}
1009: \epsscale{1.00}
1010: %\plotone{f3.eps}
1011: \plotone{f3color.eps}
1012: \epsscale{1.00}
1013: \caption{Mid-IR SED of EF Eri (from Figure \ref{f:bigspec}) with
1014: models: (a) optically thick CBD model, (b) new optically thin CBD
1015: and cyclotron model, and (c) new optically thin CBD and cyclotron
1016: model with an additional blackbody component.
1017: The model components are: WD (blue dotted line), secondary star
1018: (orange short dashed line), circumbinary disk (red long dashed
1019: line), cyclotron emission (green dot-dash line), and blackbody
1020: (purple dot-dot-dot-dash line); the total combined model is shown
1021: as a thick solid line (black).
1022: See text for additional details of the models.
1023: \label{f:models}}
1024: \end{figure}
1025:
1026:
1027: %%% END FIGURES %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1028:
1029:
1030: \end{document}
1031:
1032: