1: \documentclass[12pt,preprint]{aastex}
2: \newcommand{\mrk}{Mrk~279}
3: \newcommand{\kms}{km~s$^{-1}$}
4: \newcommand{\lam}{$\lambda$}
5: \newcommand{\lya}{Ly$\alpha$}
6: \newcommand{\fuse}{{\it FUSE}}
7: \newcommand{\stis}{STIS}
8: \newcommand{\chand}{{\it Chandra}}
9: \newcommand{\xmm}{{\it XMM-Newton}}
10:
11: \begin{document}
12:
13: \title{Variable Intrinsic Absorption in Mrk~279}
14:
15: \author{Jennifer E.\ Scott\altaffilmark{1,2},
16: Nahum Arav\altaffilmark{3,4},
17: Jack R.\ Gabel\altaffilmark{3,5},
18: Gerard A.\ Kriss\altaffilmark{6,7},
19: Jessica Kim Quijano\altaffilmark{6},
20: Jelle S.\ Kaastra\altaffilmark{8},
21: Elisa Costantini\altaffilmark{8}
22: Kirk Korista\altaffilmark{9}}
23:
24: \altaffiltext{1}{
25: Observational Cosmology Laboratory,
26: National Aeronautics and Space Administration,
27: Goddard Space Flight Center,
28: Greenbelt, MD 20771 USA}
29:
30: \altaffiltext{2}{
31: present address: Department of Physics, Astronomy, and Geosciences, Towson University,
32: Towson, MD 21252 USA; jescott@towson.edu}
33:
34: \altaffiltext{3}{Center for Astrophysics and Space Astronomy, University
35: of Colorado, 389 UCB, Boulder, CO 80309 USA}
36:
37: \altaffiltext{4}{
38: present address: Department of Physics, Virginia Polytechnic Institute \& State University,
39: Blacksburg, VA 24061 USA;
40: arav@vt.edu}
41:
42: \altaffiltext{5}{
43: present address: Department of Physics,
44: Creighton University,
45: Omaha, NE 68178 USA;
46: jackgabel@creighton.edu}
47:
48: \altaffiltext{6}{Space Telescope Science Institute, 3700 San Martin Drive,
49: Baltimore, MD 21218 USA; [gak,jkim]@stsci.edu}
50:
51: \altaffiltext{7}{Center for Astrophysical Sciences, Department of
52: Physics and Astronomy,
53: The Johns Hopkins University, Baltimore, MD 21218 USA}
54:
55: \altaffiltext{8}{SRON National Institute for Space Research,
56: Sorbonnelaan 2, 3584 CA Utrecht, The Netherlands; [J.S.Kaastra,
57: e.costantini]@sron.nl}
58:
59: \altaffiltext{9}{Department of Physics, Western Michigan University, Kalamazoo, MI 49008 USA;
60: korista@wmich.edu}
61:
62: \addtocounter{footnote}{-7}
63:
64: \begin{abstract}
65: We examine the variability in the
66: intrinsic absorption in the Seyfert 1 galaxy \mrk\
67: using three epochs of observations
68: from the {\it Far Ultraviolet Spectroscopic Explorer (FUSE)} and two
69: epochs of observations with the Space Telescope Imaging Spectrograph
70: on the {\it Hubble Space Telescope}.
71: Rather than finding simple photoionization responses of the absorbing gas to changes in the
72: underlying continuum, the observed changes in the absorption profiles
73: can be understood more clearly
74: if the effective covering fraction of the gas in all emission components,
75: continuum and broad and intermediate velocity width emission lines, is
76: accounted for.
77: While we do not uniquely solve for all of these separate covering fractions
78: and the ionic column densities using the spectral data, we examine the parameter space
79: using previously well-constrained solutions for continuum and single emission
80: component covering fractions. Assuming full coverage of the continuum, we find
81: that of the two velocity components of the Mrk~279 absorption most likely associated with
82: its outflow,
83: one likely has zero coverage of the intermediate line region while the
84: other does not. For each component, however, the broad line region is more
85: fully covered than the intermediate line region.
86: Changes in the \ion{O}{6} column densities are unconstrained due to
87: saturation, but
88: we show that small changes in the nonsaturated \ion{C}{4} and \ion{N}{5}
89: column densities
90: are consistent with the outflow gas having zero or partial
91: covering of the intermediate line region and an ionization
92: parameter changing from $\sim$0.01 to $\sim$0.1 from 2002 to 2003
93: as the UV continuum flux increased by a factor
94: of $\sim$8.
95: The absence of a change in the \ion{C}{3}
96: absorbing column density is attributed to this species arising outside
97: the Mrk279 outflow.
98: \end{abstract}
99:
100: \keywords{galaxies: active --- galaxies: individual (\mrk) ---
101: galaxies: Seyfert --- quasars: absorption lines --- ultraviolet: galaxies ---
102: X-ray: galaxies}
103:
104: \section{Introduction}
105:
106: Mass outflows from low-redshift active galactic nuclei (AGNs)
107: are inferred from blueshifted, variable X-ray and UV absorption
108: which often shows evidence for partial coverage of the contintuum source
109: in nearly one half of all nearby Seyfert galaxies
110: (see review in Crenshaw, Kraemer, \& George 2003 and recent results in Dunn et al.\ 2007).
111: In addition to possibly playing a strong role in the
112: accretion process
113: (Blandford \& Begelman 1999, 2004),
114: AGN outflows likely have substantial impacts on
115: their environments, providing feedback energy
116: to regulate the formation of galaxies and clusters of
117: galaxies
118: (Granato et al.\ 2004, Scannapieco \& Oh 2004, Scannapieco, Silk, \& Bouwens 2007);
119: to set the luminosity-temperature relation in galaxy
120: clusters (Cavaliere et al.\ 2002);
121: to distribute metals in galaxy clusters (Moll et al.\ 2007) and into the intergalactic
122: medium (Adelberger et al.\ 2003).
123:
124: Theoretical models of AGN outflows, from accretion disk winds
125: (K\"{o}nigl \& Kartje 1994, Murray et al.\ 1995, Proga 2000, 2003,
126: Proga, Stone, \& Kallman 2000, Proga \& Kallman 2004),
127: or from ablation from the obscuring torus
128: (Krolik \& Kriss 1995, 2001)
129: place the absorbing material at different distances from
130: the central engine and with different orientations with respect to the
131: broad line region (BLR).
132: A large sample of high resolution X-ray and UV spectra
133: of ouflows in nearby Seyferts from
134: {\it Chandra}, \xmm,
135: the {\it Far Ultraviolet Spectroscopic Explorer (FUSE)}
136: and from
137: the Space Telescope Imaging Spectrograph (STIS) onboard
138: the {\it Hubble Space Telescope}
139: has been building
140: in the literature over the last decade,
141: providing much insight into the physical conditions in
142: these ouflows.
143: However, a broad understanding of these outflows has yet to emerge, and
144: several outstanding questions remain, namely:
145: What is the relationship between
146: the X-ray- and UV- absorbing gas?;
147: What is the location and structure of the absorbing material?;
148: Is the observed variability in some absorption profiles
149: a response to flux variations in
150: the central engine
151: (Krolik \& Kriss 1995,
152: Shields \& Hamann 1997, Crenshaw et al.\ 2000, Kraemer et
153: al.\ 2002)
154: or to bulk motions of the absorbing gas (Crenshaw \& Kraemer 1999,
155: Kraemer et al.\ 2001, Gabel et al.\ 2003b)?
156:
157: In general, it is difficult to associate a particular UV absorption component
158: with the X-ray absorber unambiguously, in part due to relatively low spectral resolution
159: in X-ray data and in part because the ionization modeling
160: is sensitive to the shape of the UV-to-X-ray continuum (Kaspi et al.\ 2001).
161: However, with the
162: exception of campaigns on Mrk~279, using \fuse, \stis, and \chand\ with
163: both the {\it High Energy Transmission Grating (HETG)}
164: (Scott et al.\ 2004, Gabel et al.\ 2005a, S04 and G05 hereafter) and the
165: {\it Low Energy Transmission Grating}
166: (Costantini et al.\ 2007), on NGC~7469 (Scott et al.\ 2005)
167: using \fuse, \stis, and \chand/{\it HETG},
168: the long-term observing campaign on NGC~3783, using
169: \fuse, \stis, and \chand/{\it HETG}
170: (Kaspi et al.\ 2002, Netzer et al.\ 2003, Gabel et al.\ 2003a, 2005b),
171: and the campaign on NGC~4151 using \fuse, \stis, and \chand/{\it HETG}
172: (Kraemer et al.\ 2005, Kraemer et al. 2006, Crenshaw \& Kraemer 2007),
173: UV and X-ray observations are generally obtained at different times.
174:
175: Several observations of the absorption profiles taken over
176: a period of time
177: are particularly useful for addressing the last two questions
178: mentioned above,
179: as they provide contraints on the response time of the gas
180: to variability in the central engine and thus limits on the
181: distance of the absorber from the continuum-emitting region.
182: Mrk~279 ($z_{em}=0.0305$) has been the subject of an intensive
183: campaign of simultaneous observations in the UV and X-ray
184: (S04, Kaastra et al.\ 2004, Arav et al.\ 2005, 2007,
185: G05, Costantini et al.\ 2007).
186: In this paper, we use several epochs of UV spectral obsevations
187: of Mrk~279 over 1999 to 2003 from
188: \fuse\ and \stis\ presented in S04 and G05
189: to investigate the variability of the
190: column density and/or covering fraction of the outflow over time.
191: We choose to concentrate on the UV data because the
192: 2002 epoch of \chand\ observations coincided with a low
193: flux state. The grating data therefore have signal-to-noise
194: (S/N) too low to perform detailed analyses (S04).
195:
196: The spectral coverage of
197: \fuse\ provides information on \ion{H}{1} absorption in Ly~$\beta$
198: and higher order
199: Lyman transitions, and \ion{C}{3}~\lam977,
200: while
201: the \stis\ observations cover Ly$\alpha$ and the \ion{N}{5},
202: \ion{Si}{4}, and \ion{C}{4} doublets.
203: The different epochs permit us to investigate the
204: effects of UV continuum variability on the intrinsic absorption.
205: The analysis presented
206: here will also take into account changes in absorption profiles that
207: may be a result of differing contributions of the various emission
208: sources, i.e. continuum versus broad lines, in different observation epochs.
209:
210: \section{Continuum and Emission Lines}
211: \label{sec-cont}
212:
213: Table~\ref{table-data} summarizes the data used in this analysis.
214: See S04 and G05 for full presentations of
215: the \fuse\ and \stis\ spectra.
216: The 1999 \fuse\ spectrum and the 2002 and 2003 \fuse\ + \stis\
217: spectra are shown in Figure~\ref{fig:specall}.
218:
219: We fit a power law continuum of the form
220: $f_{\lambda} = f_{1000} \left(\frac{\lambda}{1000 \; {\rm \AA}}\right)^{-\alpha}$ to
221: the 2003 \fuse\ and \stis\ spectra simultaneously, where
222: the normalization, $f_{1000}$, is the flux at 1000 \AA, and we
223: include
224: the Galactic extinction
225: law of Cardelli, Clayton \& Mathis (1989) with R$_{V}=3.1$ and
226: E~(~B~-~V~)~=~0.016 (Schlegel, Finkbeiner, \& Davis 1998) in the continuum fits.
227: The best fit continuum slope and normalization are $\alpha=1.261\pm0.005$ and
228: $f_{1000}=1.230\pm0.002 \times 10^{-13}$ ergs s$^{-1}$ cm$^{-2}$ \AA$^{-1}$.
229: For the 1999- and 2002-epoch data we found $(\alpha,f_{1000})=
230: (1.60\pm0.02,1.325\pm0.004\times10^{-13})$ and
231: $(0.86\pm0.02,0.133\pm0.001\times10^{-13})$, respectively.
232: These results are shown graphically in Figure~\ref{fig:varflux}.
233: The 1999 and 2003 epochs are high flux states, and the
234: continuum becomes bluer as it brightens.
235:
236: We fit Gaussian profiles to the emission lines in the combined \fuse+\stis\
237: spectrum.
238: The emission lines we identify and fit are listed in Table~\ref{table-emspec}.
239: See Table~3 of S04 for the fits to the 1999- and 2002-epoch spectra.
240: An F-test confirms that the most prominent features require contributions from both a broad velocity
241: component (BLR, FWHM$\sim7500-10000$ \kms) and an intermediate one (ILR, FWHM$\sim3000$ \kms).
242: The resulting values of the reduced chi-squared, $\chi^2_\nu=1.6$ and $1.5$, for 2002 and
243: 2003 data, respectively, are significantly improved from the values of $\sim$2 achieved
244: with no ILR component. We do not attempt to add any more components to the fit to
245: improve the $\chi^2_\nu$ because there are no remaining systematics in the fit
246: residuals and because
247: these added components would not be well-motivated physically.
248: The full spectra with fits are shown in Figure~\ref{fig:specall}.
249:
250: We note that some of the fitted emission lines show velocity shifts of hundreds of
251: \kms\ from their 2002 positions. We performed an emission line fit for the 2003
252: data with the line centroids tied to their 2002 values. This led to a larger
253: $\chi^2_\nu$ in the final fit (1.6 versus 1.5), but did not change the results of
254: the absorption line analysis presented below. Since we will treat our
255: emission model as phenomenological and because S04 did not constrain the
256: line centroids in this way (e.g. in fitting the 1999, 2000, and 2002 epoch spectra),
257: we will use the emission model with the emission line velocity shifts in place.
258:
259: We show the emission profiles for the 2002 and 2003 spectra in
260: Figures~\ref{fig:o6_em}-\ref{fig:c4_em}.
261: For all these ions, the continuum flux is clearly stronger relative
262: to the emission lines in the high flux states in 1999 and 2003.
263: The \ion{O}{6} intermediate line emission is stronger relative to the
264: broad line in 2002 than in 1999 or in 2003, when the continuum fluxes
265: were higher. In fact, the ILR component is very weak in the 2003 spectrum.
266: The same is true for the Ly$\alpha$ and \ion{N}{5} intermediate lines. They too are
267: stronger relative to their
268: respective broad lines in the 2002 low flux state than in 2003, the only two
269: epochs we can compare with \stis\ data. For \ion{C}{4}, the difference between the
270: low and high flux states is less pronounced.
271: Figure~\ref{fig:c4_em} shows that
272: the intermediate width line is only slightly more prominent relative
273: to the broad line in 2002 than in 2003.
274:
275: \section{Intrinsic Absorption}
276:
277: S04 identified several velocity components to the absorption in the \mrk\ spectrum
278: extending over -600 km s$^{-1}$ - +100 km s$^{-1}$ with respect to the systemic
279: redshift of the AGN, taken to be $z=0.0305\pm0.0003$. These components, based on
280: the velocity structure in the Ly$\beta$ profile, were labeled 1, 2, 2a, 2b, 3, 4, 4a, and 5.
281: G05 added a component 2c to this,
282: based on additional velocity structure in the \ion{C}{4} and \ion{N}{5} absorption.
283: S04 concluded that Components 1 and 4 ($v \sim +90$ km s$^{-1}$ and $-450$ km s$^{-1}$
284: in Ly$\beta$, respectively)
285: are likely associated with the disk and possibly the halo of the host galaxy of the AGN and/or
286: its companion, MCG +12-13-024,
287: while the origins of Components 3 and 5 ($v \sim -385$ km s$^{-1}$ and $-540$ km s$^{-1}$
288: in Ly$\beta$, respectively)
289: are ambiguous. The absorption from
290: highly ionized species over the entire velocity range
291: is taken to be intrinsic by G05, as is the \ion{H}{1} absorption that is uncontaminated by any
292: possible contribution from low-ionization components seen in \ion{C}{2}, \ion{C}{3}, \ion{N}{3},
293: \ion{Si}{2}, and \ion{Si}{3}, i.e. $v \approx 200-300$ km s$^{-1}$.
294: We will follow that convention in this paper, referring to Components 2 and 4 in the
295: highly ionized species to be the absorption complexes centered at $v \approx -300$
296: and $v \approx -480$ respectively.
297:
298: The most general form of the intrinsic absorption profile for a particular species
299: accounts not only for partial covering, but also for multiple emission sources:
300: \begin{equation}
301: \label{equ:int}
302: I_j = \sum_i R^i_j (C^i_j e^{-\tau_j} + 1 - C^i_j)
303: \end{equation}
304: where $I$ is the line intensity at velocity bin $j$, $R$ is the ratio of
305: source $i$ to the total flux at $j$,
306: $\tau$ is the
307: optical depth of the absorber at $j$, and $C$ is the effective covering fraction
308: of the absorber with respect to source $i$ at velocity bin $j$ (S04, G05).
309:
310: The effective covering fraction at bin $j$
311: is the weighted sum of individual covering fractions summed over
312: all emission sources (G05):
313: \begin{equation}
314: C_j = \sum_i C^i_j R^i_j
315: \end{equation}
316: G05 found that changes in the Ly$\alpha$ absorption profile
317: between the 2002 and 2003 epochs could be explained by the change in the
318: effective covering fraction arising from the change in the relative
319: contribution of the continuum and broad line region to the total flux
320: in each epoch. No changes in the individual emission component covering fractions
321: or in the \ion{H}{1} column density
322: are necessary to produce the observed reduction in depth of the absorption trough
323: in Components 2-2a between the 2002 and 2003 epochs.
324:
325: However, we found that this method cannot account for the changes in the absorption
326: profiles of \ion{O}{6}, \ion{N}{5}, and \ion{C}{4} between the 2002 and 2003
327: epochs if only the continuum and
328: emission line sources are considered separately.
329: Below we will examine this ionic absorption variability
330: while treating the intermediate
331: width emission line components as a separate emission source.
332: For \ion{O}{6}, we also include the 1999 epoch for which we have additional \fuse\ data.
333: We will use the well-constrained covering fraction and column density solutions
334: derived by G05 from the high flux state 2003 data shown in Figure~\ref{fig:cno2003}.
335: These solutions are the global CNO fits, the simultaneous fits to the
336: \ion{O}{6}, \ion{N}{5}, and \ion{C}{4} doublets described by G05.
337: The continuum covering fraction is effectively unity over the entire velocity
338: range of the absorption while the total emission line covering fraction varies.
339: For the remainder of this paper, however, we shall set
340: $C^{\rm cont}=1$ (see also Arav et al.\ 2007).
341:
342: First, however, in the following section
343: we will describe the general observed variation of the intrinsic absorption.
344: For each species, in each epoch, we show and discuss the normalized absorption
345: profiles in the following figures and sections
346: under the assumption
347: $C^{\rm ILR}=C^{\rm BLR}=C^{\rm cont}=1$.
348: We know this
349: assumption is not strictly true, given the covering
350: fraction and column density solutions shown in Figure~\ref{fig:cno2003}; but we
351: use this as a heuristic model and a starting point for further analysis.
352: In Section~\ref{sec-photo}, we discuss photoionization models for this outflow,
353: and in Section~\ref{sec-cvsp} we discuss the importance of these
354: photoionization effects relative to changes in the effective covering fraction
355: of the absorption resulting from variations in the underlying emission sources,
356: primarily the broad and intermediate velocity width emission lines.
357:
358: \subsection{Description of Normalized Absorption Profiles}
359: \subsubsection{\ion{O}{6}}
360: Figure~\ref{fig:o6doub}
361: shows the \ion{O}{6} profile.
362: Given the equal depths of the red and blue doublet components, it is clear that
363: the \ion{O}{6} absorption in the core of Component 4 is completely saturated
364: in the 2002 and 2003 epochs, and less so in the 1999 high flux state.
365: If $C^{\rm ILR}=1$ as assumed here,
366: the increase in absorption depth of the saturated core between
367: 2002 and 2003 indicates the total
368: effective covering fraction increased.
369: This is a contradiction to our assumption here, i.e. unity covering
370: fractions for all emission components for all epochs, confirming
371: what we know from Figure~\ref{fig:cno2003}, namely that
372: $C^{\rm BLR}$ and possibly also $C^{\rm ILR}$ are not unity.
373:
374: The \ion{O}{6} absorption is saturated in Component 2 in all three
375: epochs. For $C^{\rm ILR}=1$, the top panel of Figure~\ref{fig:o6doub} shows that
376: the absorption depth is largest
377: in 2003 and smallest in 2002. This is true of Component 4 as well.
378: These are the epochs of the
379: smallest and largest ILR/BLR flux, respectively, indicating
380: that the profile change is caused by the change in this emission line ratio,
381: contributing to a change in the total effective covering fraction.
382:
383: \subsubsection{\ion{N}{5}}
384: In Component 4, the entire \ion{N}{5} profile is marginally saturated in 2002
385: and unsaturated in 2003. For $C^{\rm cont}=C^{\rm BLR}=C^{\rm ILR}=1$,
386: as shown in Figure~\ref{fig:n5doub},
387: the absorption
388: depth is similar but nonzero in both epochs.
389: This indicates once again that the emission line covering fractions are
390: nonunity, but also that the change in the effective line covering fraction
391: between the two epochs is not as dramatic as for \ion{O}{6}.
392:
393: For Component 2, the \ion{N}{5} profile is
394: saturated in 2002, but in 2003 it is saturated in the core and in the red
395: wing of the profile, while the blue wing is unsaturated.
396: The top panels of Figure~\ref{fig:n5doub} show
397: the absorption depth in 2003 is greater than in 2002,
398: particularly in the consistently saturated regions.
399: This could be the result of the changing ILR/BLR flux ratio and the
400: resulting variation in the effective line covering fraction, as described above
401: for the \ion{O}{6} profile.
402:
403: \subsubsection{\ion{C}{4}}
404: In 2002, Component 4 of the \ion{C}{4} profile is possibly saturated in the
405: red wing and unsaturated elsewhere, while in 2003
406: it is unsaturated over most of the profile.
407: The absoprtion depth decreases from 2002 to 2003, indicating
408: that the \ion{C}{4} column density decreased as the continuum
409: flux increased by a factor of $\sim8$ between 2002 and 2003.
410: The profile depth change between 2002 and 2003
411: is less than expected for a simple column density response
412: to this flux increase, leading to the
413: conclusion that the total effective covering fraction decreased over this time.
414: Although there is only a small change in the ILR/BLR emission flux ratio
415: (Fig.~\ref{fig:c4_em}) the continuum level relative to the emission lines rises
416: significantly between 2002 and 2003.
417: The effective covering fraction thus rises since $C^{\rm cont}=1$.
418:
419: In Component 2, the \ion{C}{4} profile is unsaturated in both 2002 and 2003, and
420: the absorption depth is unchanged between these two epochs, most likely also due to
421: ionization changes balancing an increase in the effective covering fraction.
422:
423: \subsubsection{Lyman series}
424: For \ion{H}{1}, the only part of the absorption complex that can
425: be reliably attributed to intrinsic gas is Component 2
426: ($v \approx 200-300$ km s$^{-1}$).
427: In this velocity component, the \lya\ absorption depth
428: is noticeably weaker in 2002 than in 2003. As pointed out
429: by G05, this is explained by a lower effective
430: covering fraction in 2002 due to a larger relative contribution
431: by the BLR relative to the continuum and the fact that $C^{\rm cont}=1$ while
432: $C^{\rm BLR}<1$.
433: For our assumption $C^{\rm ILR}=C^{\rm BLR}=1$, shown in Figure~\ref{fig:lyseries},
434: we see this reported
435: change in the \lya\ profile, and little change in the Ly$\beta$ profile, which we
436: would expect to be more influenced by the changes in the \ion{O}{6} emission components.
437:
438: For other lines further along the Lyman series from
439: Ly$\beta$,
440: there is little or no underlying
441: emission line flux beneath these profiles.
442: Examining Ly$\gamma$, a Lyman series line unlikely to be affected by saturation,
443: we find no discernable difference in the profiles
444: from 1999 to 2002, but the profile does appear less deep in 2003.
445: However there is no difference in Ly$\delta$ or Ly$\epsilon$ between
446: the two high flux states, 1999 and 2003.
447:
448: \section{Photoionization}
449: \label{sec-photo}
450: To investigate a wider range of possiblities for the various emission source covering
451: fractions, and to
452: disentangle the effects of photoionization from those of covering fraction and emission
453: source variation on the various species considered above,
454: we computed photoionzation models similar to those of
455: Krolik \& Kriss
456: (1995, 2001) using the XSTAR code (Kallman 2000).
457: We calculate the photoionization state
458: of the gas using the supersolar abundance found by Arav et al.\ (2007)
459: spectral energy distributions (SEDs) appropriate to the 2002 and 2003
460: \fuse\ and \chand\
461: observations shown in Figure~\ref{fig:sed}.
462: Both SEDs are normalized to the measured \fuse\ and \chand\
463: spectra over the spectral range of these instruments. The parameters of the \fuse\ spectra are
464: discussed in
465: Section~\ref{sec-cont}.
466: We connect the 2002 \fuse\ and \chand\ measurements with
467: $\alpha_{\rm UV-X-ray}=0.57$; and we
468: approximate the 2002 \chand\ double power law spectrum (S04)
469: with a single power law with $\alpha=1.47$ from 0.5 to 50 keV and we extend the SED to $\sim$150 keV with
470: $\alpha=0$.
471: We approximate the 2003 \chand\ power law + modified blackbody continuum (Costantini et al.\ 2007)
472: with a smooth power law: $\alpha =0.5$ connecting the \fuse\ and \chand\ data to 0.7 keV
473: and $\alpha=0.88$
474: to $\sim$150 keV (Costantini et al.\ 2007, Arav et al.\ 2007).
475: For energies lower than the \stis\ bandpass and higher than the \chand\ bandpass, we use
476: cut-offs for a standard
477: AGN spectrum (cf.\ Costantini et al.\ 2007).
478: We executed the model with grids in input column density and ionization parameter
479: from $\log N = 18$ to $\log N = 21$
480: and from $U=0.01$ to $U=10$, respectively.
481:
482: In Figure~\ref{fig:ratio}, we show results from the $\log N = 20$ run.
483: We plot the ratio of the column densities of
484: \ion{C}{3}, \ion{C}{4}, \ion{N}{5}, and \ion{O}{6} in the XSTAR simulations
485: using SED$_{2002}$ to the values from the XSTAR simulations using SED$_{2003}$
486: as a function of the ionization parameter in the SED$_{2002}$ simulations.
487: To take this ratio, we compare the column densities in each simulation at
488: equivalent ionization parameters given the different SEDs. The number of
489: ionizing photons supplied by SED$_{2002}$ is about one-tenth the number
490: supplied by SED$_{2003}$. Therefore, we take the ratio of the column
491: densities of each ion at $U_{2002}$ and $U_{2003}=10U_{2002}$. These
492: are plotted against $U_{2002}$ in Figure~\ref{fig:ratio}.
493:
494: \section{Effective Covering Fraction versus Photoionization}
495: \label{sec-cvsp}
496: We now seek to establish whether
497: the changes in the absorption profiles of the
498: highly ionized species \ion{C}{4}, \ion{N}{5}, and \ion{O}{6}
499: between
500: the 2002 and 2003 epochs can be explained by partial covering of some portion of the
501: line emission combined with changes in the relative contributions of the component broad and
502: intermediate line
503: fluxes to the total emission profile, or if some change in the column density in response
504: to changes in the continuum flux is required by the data.
505: To address this, we use the well-constrained 2003 covering fractions and column
506: densities derived by G05 and attempt to reproduce the 2002 absorption profiles
507: in two ways: (1) by allowing the covering fraction of the ILR to be different from
508: that of the BLR; and (2) by varying the column densities according to changes in the
509: ionizing flux as predicted by our photoionization models.
510: From Figure~\ref{fig:ratio}, we see how we expect the column densities of
511: each ion to change under photoionization alone.
512:
513: From the emission line fits, we consider three emission sources,
514: the continuum, the BLR, and the ILR. Thus,
515: Equation~\ref{equ:int} becomes
516: \begin{equation}
517: I_j = R^{\rm cont}_j (C^{\rm cont}_j e^{-\tau_j} + 1 - C^{\rm cont}_j) +
518: R^{\rm BLR}_j (C^{\rm BLR}_j e^{-\tau_j} + 1 - C^{\rm BLR}_j) +
519: R^{\rm ILR}_j (C^{\rm ILR}_j e^{-\tau_j} + 1 - C^{\rm ILR}_j).
520: \end{equation}
521: We assume $C^{\rm cont}=1$, consistent with G05.
522: For doublets, however, the resulting system of equations is still underconstrained.
523: Therefore,
524: we assume three possibilities for $C^{\rm ILR}$:
525: (1) $C^{\rm ILR}=C^{\rm BLR}$; (2) $C^{\rm ILR}=0.5 C^{\rm BLR}$; and
526: (3) $C^{\rm ILR}=0$.
527:
528: The overall goal of this experiment is to reproduce the 2002 spectrum from
529: the column densities derived from the higher quality 2003 data. We
530: attempt to do this by varying both the changes in the ionic column densities
531: using our photoionization modeling as a guide, and by varying the covering fractions
532: of the intermediate line region using the three cases above.
533: Here is a summary of our method:
534: \begin{itemize}
535:
536: \item{We begin with the well-constrained solutions for the
537: the emission line covering fractions from the 2003 data presented by G05.
538: In that solution, the line covering fractions of \ion{C}{4}, \ion{N}{5}, and
539: \ion{O}{6} were tied to the same value, 0.64 and 0.76 for the centers of Components 2 and 4,
540: respectively.
541: }
542:
543: \item{We use
544: the new emission line fits to the 2003 data presented in Section~\ref{sec-cont},
545: including intermediate line components,
546: to derive $C^{\rm BLR}$ and $C^{\rm ILR}$ from the values of $R^{\rm BLR}$ and $R^{\rm ILR}$ at each
547: velocity bin, from the total line covering fraction given by G05.
548: Note that in doing this, we are formally calculating {\it different} covering fractions
549: for \ion{C}{4}, \ion{N}{5}, and \ion{O}{6} because the emission line flux underlying
550: each set of doublets is different. We will return to this point in the discussion in
551: Section~\ref{sec-disc}.}
552:
553: \item{We
554: apply these to the 2002 epoch given the values of $R^{\rm BLR}$ and $R^{\rm ILR}$ derived
555: from that spectrum to examine how the profiles may have changed from changes in these
556: ratios alone.}
557:
558: \item{We also examine how the profiles change if the column densities
559: of the various ions have changed as well, by reasonable factors gleaned from our
560: photoionization modeling as shown in Figure~\ref{fig:ratio}.}
561:
562: \item{We center the x-axis of
563: Figure~\ref{fig:ratio} and concentrate our attention on the case $\log(U_{2003})\sim-1$
564: (and $\log(U_{2002})\sim-2$)
565: since this is the result of the detailed photoionization modeling and abundance analysis
566: of Arav et al.\ (2007).}
567:
568: \end{itemize}
569:
570: \section{Discussion}
571: \label{sec-disc}
572: The results for \ion{O}{6}, \ion{N}{5}, and \ion{C}{4}
573: are shown in Figures~\ref{fig:o6land}-\ref{fig:c4land}. In each lower left panel, we show
574: the blue and red components of the doublet profiles from the 2002 data.
575: In the other three quadrants, we
576: show the blue component of the 2002 profile in black. Superimposed on this are the attempts
577: to reproduce this profile from the parameter constraints from the high
578: signal-to-noise 2003 data described above,
579: under variations in both the column densities
580: and the covering fractions.
581: The column density changes from 2002 to 2003 increase
582: clockwise from top left to bottom right, with one panel
583: for each ion showing how the profiles would change if {\it no} change in the ion
584: column density occured between 2002 and 2003.
585:
586: Figure~\ref{fig:o6land} shows that because the
587: \ion{O}{6} profile is saturated, the column density
588: is unconstrained. For Component 2, regardless of the column density the only value
589: of $C^{\rm ILR}$ consistent with the profile is zero, while for Component 4,
590: $C^{\rm ILR}=0.5 C^{\rm BLR}$. The photoionization model predicts an increase
591: of about a factor of four in the \ion{O}{6} column density from 2002 to 2003,
592: given the change in the continuum level over that period.
593:
594: For \ion{N}{5}, shown in Figure~\ref{fig:n5land},
595: the best fit solution for both
596: Components 2 and 4 is
597: a column density decrease of
598: $\sim$30\% from 2002 to 2003 and
599: $C^{\rm ILR}=0$, although $C^{\rm ILR}=0.5C^{\rm BLR}$ also provides an acceptable
600: fit for Component 4.
601: Our pure photoionization model would predict
602: a decrease of about one half that amount in the \ion{N}{5} column density between 2002 and 2003.
603:
604: For \ion{C}{4}, Figure~\ref{fig:c4land} shows
605: that a decrease in the column density from 2002 to 2003 is required to
606: match the profiles in both components, regardless of the value of
607: $C^{\rm ILR}$. A change in the column density by larger than a factor of
608: $\sim$2, however, is not consistent with the observed absorption depths.
609: Our photoionization modeling indicates that we would expect
610: a larger decrease in the \ion{C}{4} column
611: density from photoionization alone, a factor of nearly six.
612: This covering fraction is relatively poorly constrained because
613: the relative contributions of the BLR and ILR change very little
614: between the 2002 and 2003 epochs, as noted above.
615: Of our models, the best fits to the 2002 spectrum come from
616: $N$(\ion{C}{4})$_{2002}$=$2 \times N$(\ion{C}{4})$_{2003}$
617: and $C^{\rm ILR}=0$ for Component 2 and $C^{\rm ILR}=C^{\rm BLR}$ for
618: Component 4. This ILR covering fraction solution for Component 4, however, is
619: not significantly different from $C^{\rm ILR}=0.5C^{\rm BLR}$.
620:
621: In the formal solution for these covering fractions we assumed
622: that the effective line covering fractions
623: for \ion{O}{6}, \ion{N}{5}, and \ion{C}{4} are all the same in the 2003 epoch
624: but that the resulting individual BLR and ILR covering
625: fractions are not necessarily equal, and that those individual BLR and ILR covering
626: fractions did not change between 2002 and 2003.
627: However, from the success of the G05 global CNO covering fraction fits,
628: the covering fractions for each of these highly ionized species are likely to be comparable.
629: Thus, we summarize the results detailed above in this context and find that
630: (1) the BLR covering fraction of \ion{C}{4}, \ion{N}{5}, and \ion{O}{6}
631: in the cores of Components 2 and 4 are $\sim$0.8-0.9;
632: (2) the ILR covering fractions
633: are zero for Component 2 and $\sim$0.4 for Component 4; and
634: (3) the column density in \ion{O}{6} decreased by an indetermiate factor while
635: the column density in \ion{N}{5} decreased by one third, and
636: that of \ion{C}{4} decreased by a factor of $\sim$2
637: between the two epochs.
638: Should we choose to constrain
639: the individual broad and intermediate emission line covering fractions of
640: \ion{C}{4}, \ion{N}{5}, and \ion{O}{6}
641: to be equal (like G05, but unlike S04), we could use this set of three
642: doublets to solve for six variables, the continuum, broad line, and emission line
643: covering fractions and the column densities of \ion{O}{6}, \ion{N}{5}, and \ion{C}{4}
644: as a function of velocity across the absorption profiles.
645: We leave this for future work.
646:
647: The \ion{C}{3} profile presents somewhat of a puzzle in light of the
648: results we have outlined thusfar. We expect the
649: column density of this species to be quite sensitive to the factor of
650: $\sim$8 increase in the ionizing flux between the 2002 and 2003 observations.
651: Figure~\ref{fig:ratio} shows that for $U_{2002}=0.01$,
652: the \ion{C}{3} column density should decrease by a factor of about 25-30 from 2002 to 2003.
653: In Figure~\ref{fig:c3}, we demonstrate that we see no significant change in the \ion{C}{3} profile
654: in either Component 4 or in Component 2.
655:
656: One way to resolve this problem is to invoke a long photoionization response time for
657: the gas.
658: To test this possibility, we follow an equilibrium photoionization model of the gas
659: over time. We began with a model that matches the well-constrained 2003 column densities
660: and dropped the photoionizing flux by a factor of 8 to evolve the absorber back in time to
661: the 2002 epoch. Following the trend of ionic abundances,
662: we find that the \ion{C}{3} column
663: density should decrease by a factor of two in about one hour for
664: $n_{\rm e} = 3 \times 10^4$ cm$^{-3}$,
665: in about one day for $n_{\rm e} = 1500$ cm$^{-3}$ and in about one year for
666: $n_{\rm e} = 5$ cm$^{-3}$.
667: Therefore, for a very small change in the \ion{C}{3} column density, a small total density
668: is required if the absorbing gas is in equilibrium.
669: This density, along with $\log U = -1$ and the luminosity of Mrk~279, require that the gas lie at
670: $\sim$2 kpc from the ionizing source if it is part of the Mrk~279 outflow.
671: This would certainly present us with a conundrum given the nonunity covering fractions
672: we have consistently found for the emission lines in absorption Components 2 and 4.
673:
674: The stable \ion{C}{3} profiles are not likely solely the result of problems with
675: the normalization of the spectra. We confirm this
676: by comparing the profiles of two nearby ISM lines in the
677: three \fuse\ epochs.
678: In Figure~\ref{fig:ism} we show two prominent but unsaturated
679: ISM lines, \ion{S}{3}~\lam1012 and \ion{Ar}{1}~\lam1048,
680: in all three \fuse\ observation epochs.
681: We do see some normalization problems, particularly in the \ion{Ar}{1}
682: region, likely attributable mainly to the low S/N in the 2002 data.
683: The absorption depth in the 2002 \ion{Ar}{1}
684: profile is higher than in the other epochs when it is not expected to vary at all.
685: This may be a contributor to the apparent lack of variability in the
686: \ion{C}{3} profiles.
687: We note, however, that the deepest absorption in the \ion{C}{3} profile peaks at
688: -540 \kms, -460 \kms, and at -385 \kms. As discussed by S04,
689: the low ionization absorption at these velocities may not be part of
690: the Mrk~279 outflow but may in fact be attributable to either the host galaxy
691: of the AGN or to a companion galaxy, MCG +12-13-024.
692:
693: Photoionization models of quasar line emission
694: indicate that intermediate line region lies at approximately 1 pc
695: from the central ionizing source, a factor of $\sim$10 further than
696: than the broad line region
697: (Brotherton et al.\ 1994).
698: However, the UV and X-ray observations of the Seyfert galaxy NGC~4151 place the D+E subcomponents of its
699: outflow absorption in an accretion disk wind which is the ILR and which lies
700: at only 0.1 pc from the continuum source (Kraemer et al.\ 2006,
701: Crenshaw \& Kraemer 2007, see also Corbin 1997).
702: For Mrk~279, we have found ILR covering fractions that are consistently smaller than
703: those of the BLR and continuum, and in some cases $C^{\rm ILR}=0$. This
704: would place the absorbing gas within or interior to the ILR. Assuming that
705: the width of the intermediate lines, $\sim$2900~\kms, is produced gravitationally,
706: a black hole mass of $10^{7.54}$ M$_{\sun}$ (Peterson et al.\ 2004) gives a radius
707: of $\sim$0.02 pc for the ILR, an estimate that is only slightly larger than
708: measured constraints on the size of the BLR, $\sim$0.01 pc (Maoz et al.\ 1990).
709: We note, however, that this estimate of the ILR size is influenced by the
710: relatively broad ILR emission features, certainly as compared with NGC~4151.
711:
712: \section{Summary}
713: We find that some of the variability, or lack thereof, in the absorption
714: profiles of the various species present in the \fuse\ and \stis\ data for Mrk~279 can
715: be understood in terms of changes in the effective covering fraction when all emission
716: components are included in the absorption model.
717: Accounting for individual covering fractions in the continuum, broad lines, and intermediate
718: velocity width lines, we find less dramatic changes in the
719: column densities of the highly ionized species are required by the absorption data than those
720: that are predicted by photoionization modeling alone. In particular, the contribution of
721: the covering fraction of the intermediate velocity width emission lines to the overall
722: effective covering fraction for a given ion and the changes in the relative contribution
723: of this component to the overall emission between the two epochs can account for a significant portion
724: of the changes observed in the absorption profiles.
725:
726: If the \ion{C}{3} absorption is intrinsic to the outflow, the UV continuum flux variation between
727: 2002 and 2003 is not commensurate with an absorber of very high ionization
728: parameter or of very large density, placing this absorbing structure at a substantial
729: distance from the ionizing source. We consider this unlikely, however, given the
730: nonunity broad and intermediate emission line covering fractions derived for the intrinsic
731: absorption.
732:
733: \acknowledgements
734: J.\ E.\ S.\ acknowledges the support of a National Research Council Associateship held at
735: NASA Goddard Space Flight Center and the support of the Jess and Mildred Fisher
736: Endowed Chair in the Biological and Physical Sciences, held at Towson University.
737:
738: \begin{thebibliography}
739:
740: \bibitem[Adelberger, Steidel, Shapley, \&
741: Pettini(2003)]{2003ApJ...584...45A} Adelberger, K.~L., Steidel, C.~C.,
742: Shapley, A.~E., \& Pettini, M.\ 2003, \apj, 584, 45
743:
744: \bibitem[Arav et al.]{arav2005} Arav,~N., Kaastra,~J., Kriss,~G.~A.,
745: Korista,~K.~T., Gabel,~J., \& Proga,~D. 2005, \apj, 620, 665
746:
747: \bibitem[Arav et al. 2007]{arav2007} Arav,~N., et al. 2007, \apj, 658, 829
748:
749: \bibitem[Blandford \& Begelman(1999)]{1999MNRAS.303L...1B} Blandford,
750: R.~D., \& Begelman, M.~C.\ 1999, \mnras, 303, L1
751:
752: \bibitem[Blandford \& Begelman(2004)]{2004MNRAS.349...68B} Blandford,
753: R.~D., \& Begelman, M.~C.\ 2004, \mnras, 349, 68
754:
755: \bibitem[Brotherton et al.(1994)]{1994ApJ...430..495B} Brotherton, M.~S.,
756: Wills, B.~J., Francis, P.~J., \& Steidel, C.~C.\ 1994, \apj, 430, 495
757:
758: \bibitem[CCM]{ccm89} Cardelli, J. A., Clayton, G. C., \& Mathis,
759: J. S. 1989, \apj, 345, 245
760:
761: \bibitem[Cavaliere, Lapi, \& Menci(2002)]{2002ApJ...581L...1C} Cavaliere,
762: A., Lapi, A., \& Menci, N.\ 2002, \apjl, 581, L1
763:
764: \bibitem[Corbin(1997)]{1997ApJS..113..245C} Corbin, M.~R.\ 1997, \apjs,
765: 113, 245
766:
767: \bibitem[Costantini et al. (2007)]{cost2007} Costantini, E., et al. 2007, \aap, 461, 121
768:
769: \bibitem[C]{ck1999} Crenshaw, D. M. \& Kraemer, S. B. 1999, \apj, 521, 572
770:
771: \bibitem[Crenshaw
772: \& Kraemer(2007)]{2007ApJ...659..250C} Crenshaw, D.~M., \& Kraemer, S.~B.\ 2007, \apj, 659, 250
773:
774: \bibitem[Crenshaw et al.\ 1999]{cren1999}
775: Crenshaw, D. M., Kraemer, S. B., Boggess, A., Maran, S. P., Mushotzky,
776: R. F., \& Wu, C. 1999, \apj, 516, 750
777:
778: \bibitem[Crenshaw, Kraemer, \& George(2003)]{2003ARA&A..41..117C} Crenshaw,
779: D.~M., Kraemer, S.~B., \& George, I.~M.\ 2003, \araa, 41, 117
780:
781: \bibitem[Crenshaw et al.\ 2000]{cren2000}
782: Crenshaw, D. M., Kraemer, S. B., Hutchings, J. B.,
783: Danks, A. C., Gull, T. R., Kaiser, M. E.,
784: Nelson, C. H., \& Weistrop, D. 2000, \apj, 545, L27
785:
786: \bibitem[dl90]{dl90} Dickey, J. M. \& Lockman F. J. 1990, \araa, 28, 215
787:
788: \bibitem[Dunn et al.(2007)]{2007AJ....134.1061D} Dunn, J.~P., Crenshaw,
789: D.~M., Kraemer, S.~B., \& Gabel, J.~R.\ 2007, \aj, 134, 1061
790:
791: \bibitem[Gabel et al.\ 2005a]{gabel2005a} Gabel, J. R. et al.\ 2005a, \apj, 623,
792: 85 (G05)
793:
794: \bibitem[Gabel et al.\ 2005b]{gabel2005b} Gabel, J. R. et al.\ 2005b, \apj, 631,
795: 741
796:
797: \bibitem[Gabel et al.\ 2003a]{gabel2003a} Gabel, J. R. et al.\ 2003a, \apj, 595, 120
798:
799: \bibitem[Gabel et al.\ 2003b]{gabel2003b} Gabel, J. R. et al.\ 2003b, \apj, 583, 178
800:
801: \bibitem[G98]{george98} George, I. M., Turner, T. J., Netzer, H.,
802: Nandra, K., Mushotzky, R. F., \& Yaqoob, T. 1998, \apjs, 114, 73
803:
804: \bibitem[Granato et al.(2004)]{2004ApJ...600..580G} Granato, G.~L., De
805: Zotti, G., Silva, L., Bressan, A., \& Danese, L.\ 2004, \apj, 600, 580
806:
807: \bibitem[Kaastra et al.(2004)]{2004A&A...428...57K} Kaastra, J.~S., et al.\
808: 2004, \aap, 428, 57
809:
810: \bibitem[Kallman2000]{k2000} Kallman, T. R. 2000, XSTAR User's Guide (Greenbelt: NASA GSFC)
811:
812: \bibitem[K01]{k01} Kaspi, S., et al. 2001, \apj, 554, 216
813:
814: \bibitem[K02]{ks02} Kaspi, S., et al. 2002, \apj, 574, 643
815:
816: \bibitem[KK94]{kk1994} K\"{o}nigl, A. K. \& Kartje, J. F. 1994, \apj, 434, 446
817:
818: \bibitem[K2001a]{kraemer2001a} Kraemer, S. B. et al. 2001, \apj, 551, 671
819:
820: \bibitem[Kraemer et al.(2005)]{2005ApJ...633..693K} Kraemer, S.~B., et al.\
821: 2005, \apj, 633, 693
822:
823: \bibitem[Kraemer et al.(2006)]{2006ApJS..167..161K} Kraemer, S.~B., et al.\
824: 2006, \apjs, 167, 161
825:
826: \bibitem[K2002a]{kraemer2002} Kraemer, S. B., Crenshaw, D. M.,
827: George, I. M., Netzer, H.,
828: Turner, T. J., \& Gabel, J. R. 2002, \apj, 577, 98
829:
830: \bibitem[]{KK95} Krolik, J.~H. \& Kriss, G.~A. 1995, \apj, 447, 512
831:
832: \bibitem[]{KK02} Krolik, J.~H. \& Kriss, G.~A. 2001, \apj, 561, 684
833:
834: \bibitem[Maoz et al.(1990)]{1990ApJ...351...75M} Maoz, D., et al.\ 1990,
835: \apj, 351, 75
836:
837: \bibitem[Moll et al.(2007)]{2007A&A...463..513M} Moll, R., et al.\ 2007,
838: \aap, 463, 513
839:
840: \bibitem[Murr]{murray1995} Murray, N., Chiang, J., Grossman, S. A., \&
841: Voit, G. M. 1995, \apj, 451, 498
842:
843: \bibitem[Netzer et al.(2003)]{2003ApJ...599..933N} Netzer, H., et al.\
844: 2003, \apj, 599, 933
845:
846: \bibitem[Peterson et al.(2004)]{2004ApJ...613..682P} Peterson, B.~M., et
847: al.\ 2004, \apj, 613, 682
848:
849: \bibitem[Proga(2000)]{2000ApJ...538..684P} Proga, D.\ 2000, \apj, 538, 684
850:
851: \bibitem[Proga(2003)]{2003ApJ...585..406P} Proga, D.\ 2003, \apj, 585, 406
852:
853: \bibitem[Proga \& Kallman(2004)]{2004ApJ...616..688P} Proga, D., \&
854: Kallman, T.~R.\ 2004, \apj, 616, 688
855:
856: \bibitem[Proga et al.(2000)]{2000ApJ...543..686P} Proga, D., Stone, J.~M.,
857: \& Kallman, T.~R.\ 2000, \apj, 543, 686
858:
859: \bibitem[Scannapieco \& Oh(2004)]{2004ApJ...608...62S} Scannapieco, E.~\&
860: Oh, S.~P.\ 2004, \apj, 608, 62
861:
862: \bibitem[Scannapieco et al.(2005)]{2005ApJ...635L..13S} Scannapieco, E.,
863: Silk, J., \& Bouwens, R.\ 2005, \apjl, 635, L13
864:
865: \bibitem[Scott et al.(2004)]{2004ApJS..152....1S} Scott, J.~E., et al.\
866: 2004, \apjs, 152, 1 (S04)
867:
868: \bibitem[Scott et al.(2005)]{2005Scott} Scott, J.~E., et al.\
869: 2005, \apj, 634, 193
870:
871: \bibitem[S]{shields1997} Shields, J. C. \& Hamann, F. 1997, \apj, 481, 752
872:
873: \end{thebibliography}
874:
875: \include{tab1}
876: \include{tab2}
877:
878: \clearpage
879: \begin{figure}
880: \epsscale{0.8}
881: \plotone{f1.eps}
882: \caption{1999 \fuse\ spectrum (top) and 2002 and 2003
883: \fuse\ + \stis\ spectra (middle and bottom) of \mrk\ with
884: errors (dotted line) and continuum and emission line fits (red)
885: \label{fig:specall}}
886: \end{figure}
887:
888: \clearpage
889: \begin{figure}
890: \epsscale{0.9}
891: \plotone{f2.eps}
892: \caption{Variation in continuum flux and UV spectral index 1999-2003
893: \label{fig:varflux}}
894: \end{figure}
895:
896: \clearpage
897: \begin{figure}
898: \epsscale{0.8}
899: \plotone{f3.eps}
900: \caption{\ion{O}{6} emission line profile in 1999, 2002, and 2003 \fuse\ spectra (cyan) with
901: individual broad line (red) and intermediate line (blue) components and continuum (dotted line)
902: shown. Note the expanded scale used to display the 2002 spectrum (middle panel).
903: All fitted emission lines are listed in Table~\ref{table-emspec}.
904: \label{fig:o6_em}}
905: \end{figure}
906:
907: \clearpage
908: \begin{figure}
909: \epsscale{0.8}
910: \plotone{f4.eps}
911: \caption{Ly$\alpha$ emission line profile in 2002 and 2003 \stis\ spectra (cyan) with
912: individual broad line (red) and intermediate line (blue) components and continuum (dotted line)
913: shown. The narrow velocity component of the fit (FWHM=697 km s$^{-1}$, fixed to the 2002 value)
914: is not shown separately
915: here.
916: \label{fig:lya_em}}
917: \end{figure}
918:
919: \clearpage
920: \begin{figure}
921: \epsscale{0.8}
922: \plotone{f5.eps}
923: \caption{\ion{N}{5} emission line profile in 2002 and 2003 \stis\ spectra (cyan) with
924: individual broad line (red) and intermediate line (blue) components and continuum (dotted line) shown.
925: \label{fig:n5_em}}
926: \end{figure}
927:
928:
929: \clearpage
930: \begin{figure}
931: \epsscale{0.8}
932: \plotone{f6.eps}
933: \caption{\ion{C}{4} emission line profile in 2002 and 2003 \stis\ spectra (cyan) with
934: individual broad line (red) and intermediate line (blue) components and continuum (dotted line) shown.
935: \label{fig:c4_em}}
936: \end{figure}
937:
938: \clearpage
939: \begin{figure}
940: \epsscale{0.8}
941: \plotone{f7.eps}
942: \caption{
943: Global CNO doublet fit from Gabel et al.\ (2005)
944: to line and continuum covering fractions (top panel) and
945: \ion{O}{6}, \ion{N}{5}, and \ion{C}{4} column densities (bottom panel).
946: Positions of Components 2 and 4 are
947: labeled at top of figure.
948: \label{fig:cno2003}}
949: \end{figure}
950:
951: \clearpage
952: \begin{figure}
953: \epsscale{0.8}
954: \plotone{f8.eps}
955: \caption{Normalized \ion{O}{6} doublet over all observation epochs for
956: $C_{\rm ILR}= C_{\rm BLR}=C_{\rm cont}=1$. Positions of Components 2 and 4 are
957: labeled at top. Top panel shows the blue doublet components
958: in the 1999, 2002, and 2003 epochs together, while the second panel shows this
959: for the red components. Bottom three panels show the blue and red doublet components
960: together in each of the three individual observation epochs.
961: \label{fig:o6doub}}
962: \end{figure}
963:
964: \clearpage
965: \begin{figure}
966: \epsscale{0.8}
967: \plotone{f9.eps}
968: \caption{Normalized \ion{N}{5} doublet over all observation epochs for
969: $C_{\rm ILR}= C_{\rm BLR}=C_{\rm cont}=1$. Positions of Components 2 and 4 are
970: labeled at top. Top panel shows the blue doublet components
971: in both the 2002 and 2003 epochs together, while the second panel shows this
972: for the red components. Bottom three panels show the blue and red doublet components
973: together in each individual observation epoch.
974: \label{fig:n5doub}}
975: \end{figure}
976:
977: \clearpage
978: \begin{figure}
979: \epsscale{0.8}
980: \plotone{f10.eps}
981: \caption{Normalized \ion{C}{4} doublet over all observation epochs for
982: $C_{\rm ILR}= C_{\rm BLR}=C_{\rm cont}=1$. Positions of Components 2 and 4 are
983: labeled at top. Top panel shows the blue doublet components
984: in both the 2002 and 2003 epochs together, while the second panel shows this
985: for the red components. Bottom three panels show the blue and red doublet components
986: together in each individual observation epoch.
987: \label{fig:c4doub}}
988: \end{figure}
989:
990: \clearpage
991: \begin{figure}
992: \epsscale{0.8}
993: \plotone{f11.eps}
994: \caption{Normalized Lyman series over all observation epochs for
995: $C_{\rm ILR}= C_{\rm BLR}=C_{\rm cont}=1$. Positions of Components 2 and 4 are
996: labeled at top. Each panel shows an individual component of the Lyman series in all
997: observation epochs: 2002 and 2003 only for Ly$\alpha$; 1999, 2002, and 2003 for
998: Ly$\beta$ and Ly$\gamma$; and 1999 and 2003 only for Ly$\delta$ and Ly$\epsilon$.
999: \label{fig:lyseries}}
1000: \end{figure}
1001:
1002: \clearpage
1003: \begin{figure}
1004: \epsscale{0.9}
1005: \plotone{f12.eps}
1006: \caption{Spectral energy distribution of Mrk 279 in 2002 and 2003, based
1007: on \fuse\ and \chand\ observations.
1008: \label{fig:sed}}
1009: \end{figure}
1010:
1011: \clearpage
1012: \begin{figure}
1013: \epsscale{0.9}
1014: \plotone{f13.eps}
1015: \caption{
1016: Ratio of the column density of
1017: \ion{C}{3}, \ion{C}{4}, \ion{N}{5}, and \ion{O}{6} in the XSTAR
1018: simulations using SED$_{2002}$ to those using SED$_{2003}$ as a function of the 2002 ionization paramter.
1019: The total input column density for all is $\log N = 20$.
1020: \label{fig:ratio}}
1021: \end{figure}
1022:
1023: \clearpage
1024: \begin{figure}
1025: \epsscale{0.8}
1026: \plotone{f14.eps}
1027: \caption{Blue and red components of \ion{O}{6} profile from 2002 \fuse\ data (lower left); blue component
1028: is repeated in other windows with profiles expected from \ion{O}{6} column density measured from 2003 \fuse\
1029: data and different ILR covering fractions: $C_{\rm ILR}=0$ (red), $C_{\rm ILR}=0.5C_{\rm BLR}$ (blue),
1030: $C_{\rm ILR}=C_{\rm BLR}$ (green). Positions of Components 2 and 4 are
1031: labeled at top.
1032: \label{fig:o6land}}
1033: \end{figure}
1034:
1035: \clearpage
1036: \begin{figure}
1037: \epsscale{0.8}
1038: \plotone{f15.eps}
1039: \caption{Blue and red components of \ion{N}{5} profile from 2002 \stis\ data (lower left); blue component
1040: is repeated in other windows with profiles expected from \ion{N}{5} column density measured from 2003 \stis\
1041: data and different ILR covering fractions: $C_{\rm ILR}=0$ (red), $C_{\rm ILR}=0.5C_{\rm BLR}$ (blue), $C_{\rm ILR}=C_{\rm BLR}$ (green). Positions of Components 2 and 4 are
1042: labeled at top.
1043: \label{fig:n5land}}
1044: \end{figure}
1045:
1046: \clearpage
1047: \begin{figure}
1048: \epsscale{0.8}
1049: \plotone{f16.eps}
1050: \caption{Blue and red components of \ion{C}{4} profile from 2002 \stis\ data (lower left); blue component
1051: is repeated in other windows with profiles expected from \ion{C}{4} column density measured from 2003 \stis\
1052: data and different ILR covering fractions: $C_{\rm ILR}=0$ (red), $C_{\rm ILR}=0.5C_{\rm BLR}$ (blue), $C_{\rm ILR}=C_{\rm BLR}$ (green). Positions of Components 2 and 4 are
1053: labeled at top.
1054: \label{fig:c4land}}
1055: \end{figure}
1056:
1057: \clearpage
1058: \begin{figure}
1059: \epsscale{0.9}
1060: \plotone{f17.eps}
1061: \caption{The \ion{C}{3}~\lam977 profile in the 1999, 2002, and 2003 \fuse\ observations.
1062: \label{fig:c3}}
1063: \end{figure}
1064:
1065: \clearpage
1066: \begin{figure}
1067: \epsscale{0.9}
1068: \plotone{f18.eps}
1069: \caption{Absorption profiles of interstellar species \ion{S}{3} and \ion{Ar}{1} in the
1070: 1999, 2002, and 2003 epochs of \fuse\ data.
1071: \label{fig:ism}}
1072: \end{figure}
1073:
1074: \end{document}
1075: