1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[referee]{aa}
3:
4:
5: \begin{document}
6: \title{
7: Super-solar N/C in the NLS1 Galaxy Markarian\,1044
8: }
9: \author{Dale L. Fields\altaffilmark{1}, Smita Mathur\altaffilmark{1}, Richard W. Pogge\altaffilmark{1}, Fabrizio Nicastro\altaffilmark{2} \& Stefanie Komossa\altaffilmark{3}}
10: \altaffiltext{1}
11: {Department of Astronomy, the Ohio State University, 140 West 18th
12: Avenue, Columbus, OH 43210, USA}
13: \altaffiltext{2}
14: {SAO,
15: 60 Garden Street, 02138, Cambridge, MA, USA}
16: \altaffiltext{3}
17: {Max-Planck-Institut f\"ur extraterrestrische Physik,
18: Giessenbachstrasse 1, D-85748 Garching, Germany}
19:
20:
21: \email{fields@astronomy.ohio-state.edu}
22:
23: \begin{abstract}
24:
25: Narrow-Line Seyfert 1s (NLS1s) are known have extreme values of a number
26: of properties compared to Active Galactic Nuclei (AGN) as a class. In
27: particular, previous emission-line studies have suggested that NLS1s are
28: unusually metal rich compared to broad-line AGN of comparable
29: luminosity. We present low- and medium-resolution spectroscopic
30: observations of the NLS1 Markarian\,1044 with the Hubble Space Telescope
31: Imaging Spectrometer (STIS). We identify two blueshifted intrinsic
32: absorption systems at $-1145$ and $-295$\,km\,s$^{-1}$ relative to the
33: systemic velocity of the galaxy. Using a simple photoionization model
34: of the absorbing gas, we find that the strongest and best-measured of
35: the absorption systems has $N/C\approx6.96\,[N/C]_\sun$. We also report
36: on the discovery of three new Ly$\alpha$ forest lines with $\log N_{H
37: I}\ge12.77$. This number is consistent with the 2.6 expected in the
38: path length to the source redshift of Mrk\,1044.
39:
40: \end{abstract}
41:
42: \section{Introduction} \label{sec:intro}
43:
44: The Narrow-Line Seyfert 1 (NLS1) class has come a long way since being
45: identified by \citet{OP}. There, they were defined as Seyfert 1s with
46: relatively narrow permitted lines ($\le2000$\,km\,s$^{-1}$) and a weak
47: [\ion{O}{3}]/H$\beta$\ ratio ($\le3$). The principal component analysis
48: of \citet{BG} found that NLS1s lie at one extreme end of the eigenvector
49: with the most variation in AGN spectral properties. In addition, NLS1s
50: often have very steep X-ray spectra, placing them on one end of the
51: anti-correlation between soft X-ray spectral slope and H$\beta$\ FWHM
52: seen in Seyfert 1s \citep{BBF} and quasars \citep{L97}.
53:
54: While work continues to identify the physical driver of eigenvector 1,
55: the leading contender is accretion rate expressed as a fraction of the
56: Eddington accretion rate ($\rm \dot{m}=\dot{M}/\dot{M}_{Edd}$) as
57: suggested by \citet{P95}. This would imply that NLS1s and some quasars
58: are accreting at a large fraction of the Eddington rate. This receives
59: support from evidence that NLS1s appear to have smaller supermassive
60: black holes than many other Seyferts, and yet have comparable
61: luminosities \citep[and references therein]{G04} and references therein.
62: Another type of AGN suspected to have a large fractional accretion rate
63: are the high-redshift quasars. These have much larger luminosities, and
64: correspondingly larger black hole masses than other AGN.
65:
66: One other property that NLS1s and high-redshift quasars appear to share
67: is super-solar gas-phase metallicity. The NLS1 PG1404+226 has unusually
68: strong \ion{N}{5} absorption lines relative to Ly$\alpha$, \ion{C}{4}
69: and X-ray absorption features, and this can only be explained by
70: super-solar metallicity \citep{M2Kb}. \citet{W99} found that the
71: strengths of the \ion{N}{5}\,$\lambda1240$ emission lines systematically
72: stronger than average, and the \ion{C}{4}\,$\lambda1549$ were
73: systematically weaker in AGN with narrow permitted lines. This is
74: similar to what appears in high redshift quasars \citep{HF,O94,SH}. In
75: addition, the strength of the fluorescent Fe-K$\alpha$ emission line in
76: some NLS1s is also thought to indicate super-solar abundances
77: \citep{CJ}. Finally, \citet{KM01} noted that the steep X-ray spectra
78: seen in NLS1s eliminates the possibility of having a multi-phase medium
79: with solar abundances. Only if the gas is super-solar can it
80: successfully form a pressure-confined broad-line region.
81:
82: Metallicity, and more importantly, abundance ratios reflect the star
83: formation history of the nuclear gas. This in turn might tell us about
84: the particular circumstances leading to the activation of the AGN.
85: There are theories that link the many similarities between NLS1s and
86: quasars into a scenario stating that these objects are ``young,'' that
87: is, the black holes have recently begun (re-)accreting gas \citep{M2Ka}.
88: In the case of Narrow-Line Seyfert 1s, this is very likely, as the mass
89: e-folding time is short enough to conclude that if that black hole had
90: any similarly major accretion events in its past, it would not have the
91: relatively low mass it has today. There are even suggestions that some
92: NLS1s do not fall directly on the $M_{BH}-\sigma$ relation
93: \citep{M01,GM04}. Because NLS1s could be part of an evolutionary track
94: (as opposed to the Unified Model), there is no a priori reason why NLS1s
95: should have the same star formation history, and thus metal abundance
96: ratios, as normal Seyfert galaxies. Indeed, \citet{SN} find that NLS1s
97: have significantly higher \ion{N}{5}/\ion{C}{4} ratios than standard
98: Seyferts. This result, however, is not confirmed in
99: \ion{N}{4}/\ion{C}{4}. In addition, data on this ratio is based upon
100: emission line ratios. A single emission line can originate in multiple
101: locations, each with different physical conditions, complicating the
102: conversion of line flux to column densities to abundance. If one wishes
103: to get an unambiguous result, one has to measure the absorption lines.
104: With this in mind, the emission lines results indicate super-solar
105: metallicity \citep{HF}, but require confirmation.
106:
107: We seek to determine the metal abundance of a NLS1 using absorption
108: lines. The largest contributor to the uncertainty in the abundances is
109: the uncertainty in the photoionization correction. This must be well
110: determined if there is to be any confidence in our results.
111: Historically, studies have contained few transitions and few (sometimes
112: only one) species per element due to limited wavelength coverage
113: (e.g. only UV or only X-ray) and thus do not have the leverage to
114: determine the correct ionization model. Our program consists of
115: near-simultaneous observations of the NLS1 Markarian\,1044 with the HST,
116: FUSE, and Chandra observatories. This long wavelength baseline ensures
117: that we will have the range of species and elements to accurately model
118: the ionization correction. In this paper, we present observations of
119: Mrk\,1044 with the FUV-MAMA instrument onboard the Hubble Space
120: Telescope. Future papers will detail the FUSE and Chandra observations
121: and the self-consistent model derived from all three's observations. In
122: \S\ref{sec:data} we detail the the FUV-MAMA observation and data. We
123: then follow in \S\ref{sec:analysis} with the analysis. We discuss the
124: results in \S\ref{sec:discuss} and give concluding remarks in
125: \S\ref{sec:conclude}.
126:
127: \section{Observations and Data Reduction} \label{sec:data}
128:
129: Mrk\,1044 was observed on UTC 2003 Jun 28 using the Hubble Space
130: Telescope Imaging Spectrometer (STIS) with the FUV-MAMA detector. All
131: observations were acquired during a single 5-orbit visit that ran from
132: MJD 52818.19912 through MJD 52818.49349. A journal of observations is
133: given in Table~\ref{tab:obs}. The 52X0.2 aperture and a clear filter
134: were used for all observations, centering the aperture on the bright
135: nucleus of Mrk\,1044. Two low-dispersion UV spectra with the G140L
136: grating (central wavelength of $\lambda1426$\AA) were acquired during
137: the first orbit, with a cumulative exposure time of 2311\,seconds.
138: Medium-dispersion spectra were acquired with the G140M grating during
139: the remaining 4 orbits at central wavelengths of $\lambda1222$,
140: $\lambda1272$, and $\lambda1567$ covering the emission lines of
141: Ly$\alpha$, \ion{N}{5}, and \ion{C}{4}, respectively. The Ly$\alpha$
142: and \ion{C}{4} regions were observed for 1 orbit each, acquiring 2
143: spectra per orbit for cumulative exposure times of 2734\,seconds each,
144: and \ion{N}{5} was observed for 2 orbits (4 spectra) for a cumulative
145: exposure time of 5583\,seconds.
146:
147: The standard STSDAS\footnote{STSDAS is a product of the Space Telescope
148: Science Institute, which is operated by AURA for NASA.} {\tt calstis}
149: pipeline was used to reduce the raw data to wavelength- and
150: flux-calibrated 2D spectra. The nuclear spectra of Mrk\,1044 were
151: extracted from the linearized 2D spectra (STIS ``x2d'' science frames)
152: using a 9 pixel (0.45 arcsec) aperture using the {\tt spectroid}
153: centroiding aperture-extraction task in XVista\footnote{XVista is the
154: current incarnation of Lick Observatory Vista, maintained in the public
155: domain at http://ganymede.nmsu.edu/holtz/xvista.}. Corresponding error
156: spectra were also extracted from the STIS error arrays for each spectrum
157: and co-added in quadrature. These 1D spectra form the basis of our
158: subsequent analysis.
159:
160: The G140L spectrum provides us with a reliable internal relative
161: intensity calibration for the three non-overlapping G140M spectra, and
162: provides information on the overall UV spectral slope. The G140L
163: spectrum with the flux-calibrated G140M spectra superimposed is shown in
164: Figure~\ref{fig:masterspec}. The individual flux-calibrated G140M
165: spectra are shown in Figure~\ref{fig:threespec}.
166:
167: \section{Analysis} \label{sec:analysis}
168:
169: As can be seen in Figure~\ref{fig:masterspec}, the UV spectrum of
170: Mrk\,1044 has a flat UV continuum with a slope of $\alpha=0.0$ for
171: $f_{\nu}\propto\nu^{-\alpha}$. Its emission line features are similar
172: to those of many other AGN. Most prominent are the Ly$\alpha$,
173: \ion{N}{5}, and \ion{C}{4} emission lines. Less prominent, but
174: certainly present are the \ion{Si}{2}\,$\lambda1265$,
175: \ion{O}{1}\,$\lambda1305$, \ion{C}{2}\,$\lambda1335$,
176: \ion{Si}{4}\,$\lambda\lambda1394$,$1403$+\ion{O}{4}]\,$\lambda1402$,
177: \ion{N}{4}]\,$\lambda1486$, \ion{He}{2}\,$\lambda1640$, and
178: \ion{O}{3}]\,$\lambda1663$ lines. Emission lines in general have a
179: narrow ($<1500$\,km\,s$^{-1}$) component and a broad
180: ($>3000$\,km\,s$^{-1}$) base. The presence of this base causes
181: significant blending of Ly$\alpha$ and \ion{N}{5}.
182:
183: The three G140M spectra show a total of 26 possible absorption lines,
184: all but four of which have good identifications. Line identification
185: proceeded from three assumptions: 1) lines near the rest-frame
186: wavelength of a well-known and strong transition are likely due to
187: Galactic ISM, 2) any \ion{N}{5} or \ion{C}{4} absorption line systems
188: intrinsic to Mrk\,1044 will also appear at the same velocity in
189: Ly$\alpha$, and 3) any remaining unidentified absorption lines blueward
190: of Mrk\,1044's Ly$\alpha$ emission line, but redward of 1216\AA\ could
191: be the signature of the low redshift Ly$\alpha$ forest.
192:
193: Line doublets of \ion{N}{5} and \ion{C}{4} were identified by
194: superposing velocity maps for each of the doublets. From the NASA/IPAC
195: Extragalactic Database (NED), we adopt a systemic velocity of Mrk\,1044
196: of 4932\,km\,s$^{-1}$ (z=0.01645). Velocity maps are shown in
197: Figure~\ref{fig:velmap} for Ly$\alpha$, \ion{N}{5}, and \ion{C}{4}.
198: From these, we identify two absorption systems at $-1145$\,km\,s$^{-1}$
199: and $-295$\,km\,s$^{-1}$. We find that the two candidate absorption
200: systems in \ion{N}{5} at $-2100$\,km\,s$^{-1}$ and $+75$\,km\,s$^{-1}$
201: are actually Galactic absorption from
202: \ion{S}{2}\,$\lambda\lambda1251$,$1254$ in the case of the former and
203: \ion{S}{2}\,$\lambda1260$ and an unknown absorption line superposed in
204: velocity space in the case of the latter. In both cases, no
205: corresponding absorption system is detected in \ion{C}{4} or in
206: Ly$\alpha$.
207:
208: We use two methods (each with slightly varying analysis paths) to
209: measure the column densities of the absorbing material. The first is
210: the standard curve of growth. More details of this process may be found
211: in \citet{Spitzer}. To determine the velocity spread parameter, we
212: calculate values of the column density for each line for 'b' values of
213: 10, 20, 40, and 80\,km\,s$^{-1}$. We then assume the velocity spread
214: that gives the most consistent result between the line pair. The second
215: method is the apparent optical depth method described by \citet{SS91}.
216: This method is usable in well-resolved systems and thus it will provide
217: an independent method with which to estimate the column densities.
218:
219: Lines in each of the spectra were first measured approximately with the
220: analysis package LINER \citep{LINER}. Best fit parameters for the
221: emission lines were then used as initial guesses into the STSDAS
222: analysis package SPECFIT \citep{SPECFIT}. To determine the absorption
223: line parameters via the curve of growth method, LINER was used with a
224: local continuum parameterized by fitting a fifth order polynomial over a
225: limited wavelength range (e.g. only on one side of an emission line). A
226: polynomial of order five was found to be a good balance between fitting
227: the intrinsic structure of the continuum and letting noise introduce
228: unphysical structure into the continuum fitting. In this manner, the
229: local continuum and then two or three absorption lines would be fit and
230: the process repeated for the next absorption feature. Again, these
231: results were used as inputs into SPECFIT for final determination of the
232: equivalent widths.
233:
234: The analysis of the lines using the apparent optical depth method
235: proceeded differently. Because estimating the continuum is the single
236: largest contributor to the systematic error budget, we performed nine
237: by-eye spline fits to the continuum around each group of lines. All
238: nine were done as reasonable fits, but several were purposefully fit to
239: the outer envelope of the noise. A mean was then taken of the nine
240: resultant column densities with the scatter functioning as our
241: measurement of error.
242:
243: \subsection{Mrk\,1044 Emission Lines} \label{sec:mrkemi}
244:
245: Most of the metallicity studies of AGN have been restricted to using the
246: emission lines only. Therefore, we give measurements of our emission
247: lines to place the results of our absorption line study within the
248: proper context. For each ion Table~\ref{tab:emission} gives the rest
249: wavelength, the observed equivalent width, and the FWHM of the fit
250: components. For most emission lines we fit both a narrow and a broad
251: component. Both the \ion{N}{5} and \ion{C}{4} emission lines exhibited
252: an excess of red-ward flux which we parameterized with an additional
253: Gaussian. The \ion{Si}{4}+\ion{O}{4}] complex around 1400\AA\ can be
254: parameterized with several narrow components and one very broad one, but
255: these are so poorly constrained individually that we only report the
256: gross properties of the complex.
257:
258: The literature contains many measures of AGN metallicity. \citet{SN}'s
259: Figure 2 gives \ion{N}{5}/\ion{C}{4} and \ion{N}{5}/\ion{He}{2} as a
260: function of $\nu L_{\nu}$ at 1450\AA. Mrk\,1044 has ratios of 0.6 and
261: 3.1, respectively, at $1.3\times10^{12}L_\sun$. These values place
262: Mrk\,1044 right in the middle of the NLS1 metallicity distributions,
263: though this object is slightly more luminous than the rest of the NLS1
264: sample considered by \citet{SN}. Nevertheless, Mrk\,1044 remains at or
265: above the mean metallicity-luminosity correlation. We can also place
266: Mrk\,1044 alongside the ultraviolet measurements of NLS1s in
267: \citet{W00}. This object's \ion{C}{4}/Ly$\alpha$ value of 0.59 is
268: slightly higher than any of the NLS1s shown in their Figure 2.
269: Mrk\,1044's (\ion{Si}{4}+\ion{O}{4}])/Ly$\alpha$ value of 0.17 is at the
270: upper value bounded by their NLS1 sample. Finally, we can compare this
271: object against the ``nitrogen-rich'' quasar Q0353-383 \citep{O80} and
272: its recently discovered fraternity \citep{BO}. One of the metallicity
273: indicator used in these studies is $f_{CIV}/(f_{\rm Ly\alpha}+f_{NV})$,
274: of which Q0353-383 has the lowest value found to date: 0.07. By
275: comparison, Mrk\,1044 has a value of 0.45, a factor of six difference.
276: It should be noted that \citep{O80} found a value of
277: $N/C\approx10$-$20\,[N/C]_\sun$, which is only a factor of 2-3 higher
278: than the value we will discuss below in \S\ref{sec:discuss}.
279:
280: \subsection{Mrk\,1044 Absorption Line Systems} \label{sec:mrkline}
281:
282: Of the 24 absorption lines found in our UV spectra, 10 of these appear
283: to be intrinsic to Mrk\,1044, organized into two distinct absorption
284: systems. System 1 at $-1145$\,km\,s$^{-1}$ contains strong lines of
285: Ly$\alpha$, \ion{N}{5}, and \ion{C}{4}. System 2 at
286: $-295$\,km\,s$^{-1}$ contains strong lines of Ly$\alpha$ and \ion{C}{4},
287: but has weak \ion{N}{5}. Due to their inherent weakness, estimations of
288: the strengths of the \ion{N}{5} lines in System 2 suffer accordingly.
289: Table~\ref{tab:mrk1044} gives the measured and derived parameters for
290: these systems: the observed wavelength, the observed equivalent width,
291: the column densities derived both the curve of growth method and the
292: optical depth integration, and the velocity offset from the systemic.
293: Equivalent widths are typically around 0.2\AA\ in the case of System 1
294: and 0.02\AA\ in the case of System 2. The errors in the equivalent
295: widths were determined by using estimates of the photon noise in each
296: pixel. The errors in the velocity are set at 0.5 pixels, as given by
297: the STIS Instrument Handbook for Cycle 13. Centroiding errors are much
298: smaller and therefore ignored.
299:
300: When we calculate the column densities using the curve of growth method,
301: we find that a velocity spread parameter of 20\,km\,s$^{-1}$ gives the
302: most consistent result for both pairs of \ion{N}{5} and \ion{C}{4}
303: lines. If we assume that this contributes the greatest to the width of
304: the line, we can determine if these lines are fully resolved. The STIS
305: Instrument Handbook gives a nominal spectral resolution of 1.4 pixels.
306: This corresponds to between 18 and 14\,km\,s$^{-1}$ in width at the
307: lines of \ion{N}{5} and \ion{C}{4}, respectively. This is approximately
308: similar to the preferred curve of growth width of 20\,km\,s$^{-1}$ so we
309: will use that as our measure of intrinsic width. In the case of System
310: 1, the FWHM of the \ion{N}{5} and \ion{C}{4} line pairs are 71, 71, 50
311: and 47\,km\,s$^{-1}$, from blue to red. All of these lines are fully
312: resolved. In the case of System 2, the FWHM of the lines are 90, 100,
313: 43 and 38\,km\,s$^{-1}$. The \ion{C}{4} lines are close to being
314: resolved. While the best-fit values of the \ion{N}{5} line widths
315: suggest these two lines are resolved, the large uncertainties prevent
316: this from being definite.
317:
318: We calculate the degree of saturation of these two systems by measuring
319: the ratios of the equivalent widths of each line doublet. Both
320: \ion{N}{5} and \ion{C}{4} have natural ratios of 2.0. This can be used
321: to derive a covering fraction ($C_f$) for each system following
322: \citet{H97}. System 1 has a consistent value of $C_f=0.80$ determined
323: from both the \ion{N}{5} and \ion{C}{4} lines. System 2 has less
324: consistent values with $C_f$ of 0.97 (\ion{N}{5}) and 0.94 (\ion{C}{4}),
325: but this is not unexpected given the more marginal nature of System 2's
326: \ion{N}{5} doublet, the $\lambda1243$\AA\ line of which we only detect at
327: about the 2.5$\sigma$ level.
328:
329: We calculate \ion{N}{5}/\ion{C}{4} column ratios for Systems 1 and 2.
330: We use the weighted mean for each species from each doublet. As
331: Table~\ref{tab:mrk1044} shows, both methods find very similar results
332: for the \ion{N}{5} abundance, but the curve of growth method finds a
333: systematically smaller \ion{C}{4} abundance than the apparent optical
334: depth method. This does not affect the \ion{N}{5}/\ion{C}{4} column
335: ratios at the 3$\sigma$ level. In System 1, we find
336: \ion{N}{5}/\ion{C}{4} ratios of $2.7\pm7$\% and $2.0\pm10$\% for the
337: curve-of-growth and optical depth methods, respectively. For System 2
338: we find ratios of $1.0\pm20$\% and $0.7\pm30$\% for the two methods
339: respectively.
340:
341: \subsection{Galactic Absorption Lines} \label{sec:galline}
342:
343: With good certainty, we detect absorption lines of \ion{N}{1},
344: Ly$\alpha$, \ion{Si}{2}, \ion{Si}{3}, \ion{S}{2}, and \ion{C}{4} in our
345: medium-resolution spectra and Ly$\alpha$, \ion{O}{1}, \ion{C}{2}, and
346: \ion{Si}{2} in our low-resolution spectrum from the ISM of our galaxy.
347: To determine the velocity spread parameter $b$ for the curve of growth
348: method, we look for the most consistent result between lines of the same
349: species. Species for which we have only a single identified line we
350: report a range of densities corresponding to a range of $b$ of
351: 20-80\,km\,s$^{-1}$. We also report column densities using the apparent
352: optical depth method for all lines. We give the results in
353: Table~\ref{tab:galactic} in the form of observed wavelength, likely
354: identification, the candidate's rest wavelength, the velocity width, the
355: observed equivalent width, the derived column densities for both
356: methods, and the velocity relative to Mrk\,1044's systematic velocity.
357: The velocities of the absorption components are scattered about zero
358: and, in 12 of 14 cases, within three standard deviations of zero. It
359: should be noted that there are systematic velocity shifts with respect
360: to the parent spectrum. Two of the G140M spectra (Ly$\alpha$ and
361: \ion{C}{4}) have all their Galactic absorption lines shifted to the
362: blue, while the other G140M spectrum (\ion{N}{5}) and the G140L spectrum
363: have their Galactic absorption lines shifted to the red. This can also
364: be seen, though not as clearly, in the lines given in
365: Table~\ref{tab:mrk1044}. These shifts are not statistically significant
366: if one accepts the absolute wavelength calibration error of 0.5 pixels.
367: The errors in the equivalent widths are estimated from the photon noise.
368:
369: Most of the FWHMs of the lines in the medium-resolution spectra sit
370: between 50 and 70\,km\,s$^{-1}$ with a few lines
371: $\gtrsim100\,km\,s^{-1}$ (the Si lines and \ion{C}{4}\,$\lambda$1548).
372: The uncertainty in the width is not taken from any SPECFIT fit (which
373: claims a formal precision on the order of one-tenth to one percent error
374: on any one measurement) but instead from the scatter around the mean
375: width measurement over many continuum fittings. This uncertainty is
376: approximately 5\,km\,s$^{-1}$ for almost all lines. The \ion{S}{2}
377: triplet lines have the best agreement amongst themselves, all coming
378: within one standard deviation of their mean width, 61\,km\,s$^{-1}$.
379: The \ion{N}{1} lines have a mean width of 70\,km\,s$^{-1}$ and agree to
380: within one standard deviation. The Si lines have a mean width of
381: 126\,km\,s$^{-1}$ and agree to within one standard deviation. The only
382: real disagreement is from the \ion{C}{4} lines which disagree at greater
383: than three standard deviations.
384:
385: The Galactic \ion{C}{4} lines suffer from another problem. The
386: velocities (i.e. centroids) of these lines are in great disagreement
387: with each other. The uncertainty in the velocity given is that induced
388: by the absolute wavelength calibration. Wavelengths relative to one
389: another in the same spectrum have their uncertainties smaller by a
390: factor of 2.5 to 5. This creates a disagreement in the centroids of the
391: \ion{C}{4} lines of more than $6\sigma$. In addition, as can be seen
392: from visual inspection of the spectrum (Figure~\ref{fig:threespec}), the
393: profiles of these lines are dissimilar. It is possible that weak
394: absorption from High Velocity Clouds (HVCs) is preferentially altering
395: the shape of the stronger \ion{C}{4}\,$\lambda1548$ line. This could
396: also explain the large differences in the FWHM of these lines. This
397: hypothesis, however, cannot be tested with the current data and must
398: wait for FUSE confirmations of the existence of HVCs through the
399: detection (or not) of \ion{O}{6}.
400:
401: The three absorption lines in the G140L spectrum (\ion{O}{1},
402: \ion{C}{2}, and \ion{Si}{2}\,$\lambda1527$) have large widths of 920,
403: 560, and 390\,km\,s$^{-1}$, respectively. All are fully resolved. It
404: is possible that what we identify as simply \ion{O}{1} could also be
405: blended with \ion{S}{1}\,$\lambda1303$. This would explain the large
406: width and would eliminate \ion{O}{1}'s large inferred velocity.
407: Unfortunately, deblending does not return a clean result, so we cannot
408: make this correction with any certainty and instead elect to omit it.
409: The large values of inferred widths in the G140L spectrum is somewhat
410: surprising, given that the low ionization lines seen in the higher
411: resolution spectra are all so narrow. The cause of this is not yet
412: known and we caution anyone from trusting the properties of these lines
413: at this time.
414:
415: The four unidentified absorption lines fail at least one of our
416: criteria. To be identified, a line must come from a strong transition
417: of an abundant element without having a sufficiently large velocity
418: difference. For example, \ion{Si}{2}\,$\lambda1264.7$\AA\ could be
419: identified with the line at 1263.5\AA\ but for the fact that this would
420: give it a relative velocity of $-286$\,km\,s$^{-1}$. We also use
421: additional information such as the fact that the line in question has a
422: width of 52\,km\,s$^{-1}$ which makes it unlikely to come from Si whose
423: lines average 123\,km\,s$^{-1}$. It is possible that the absorption
424: lines near the Ly$\alpha$ emission line of Mrk\,1044 are Ly$\alpha$
425: whose systems simply do not have the column to appear in
426: \ion{N}{5} or \ion{C}{4}.
427:
428: \subsection{Intervening Ly$\alpha$ Absorption?} \label{sec:lyaline}
429:
430: The two unidentified absorption lines blueward of the Ly$\alpha$
431: emission line have velocities relative to Mrk\,1044's systemic velocity
432: of about 2000 and 3000\,km\,s$^{-1}$. While this does not particularly
433: favor an origin as material intrinsic to Mrk\,1044, it is not
434: inconsistent with that hypothesis. If these absorption lines are due to
435: intrinsic Ly$\alpha$, the inferred column densities are intermediate
436: between Systems 1 and 2. Unlike those two systems, however, these
437: unidentified systems do not show absorption in either \ion{N}{5} or
438: \ion{C}{4}. An alternative hypothesis is that these two systems are due
439: to the low-redshift Ly$\alpha$-forest. To test this, we calculate the
440: number of expected absorption systems from the results of \citet{PSS}.
441: They parameterize the number of systems per unit log column density and
442: per unit redshift $\partial^2 N/\partial N_{H I} \partial z$ as $C_{H I}
443: N_{H I}^{-\beta}$ with $C_{H I}=10^{10.3}$ and $\beta=1.65$ between
444: $\log N_{H I}$ of 12.3 and 14.5. We calculate the $3\sigma$ detection
445: limit of Ly$\alpha$ absorption to be $\log N_{H I}=12.77$. Because this
446: is almost a factor of ten lower in column density than the weaker of our
447: systems, we reanalyze the spectrum for weak lines. Our best continuum
448: fit (flattest continuum-divided spectrum) can be seen in
449: Figure~\ref{fig:lyaforest}. While any one continuum fit may produce
450: several well-detected line candidates, we require any line to be
451: well-detected in the majority of our nine continuum fits. Only one
452: additional line passes this test, being detected, on average, at exactly
453: $3\sigma$. The wavelengths, velocity widths, equivalent widths,
454: calculated column densities and relative velocities of the two strong
455: Ly$\alpha$ systems and the weak, newly found system are given in
456: Table~\ref{tab:lyaforest}. For the decade above our detection limit,
457: and for a path length $\Delta z=z_{\rm Mrk\,1044}=0.01645$, 2.0 systems
458: are expected, and two systems (within our column density errors) are
459: detected. The expected number of systems at all higher column densities
460: ($\log N_{H I}>13.77$) and the same path length is 0.6 and we detect
461: one. In this manner, the attribution of these absorption lines to
462: intergalactic Ly$\alpha$ absorption is perfectly reasonable.
463:
464: \section{Discussion} \label{sec:discuss}
465:
466: The \ion{N}{5}/\ion{C}{4} ratio does not tell us the N/C abundance ratio
467: until we apply an ionization correction for each species. This
468: ionization correction is the greatest source of systematic uncertainty
469: in the determination of abundances. An accurate ionization correction
470: requires measurements of multiple species of a single element, and
471: preferably as many additional elements as possible. For this reason, we
472: cannot create an accurate ionization model with just this HST data. We
473: are able, however, to set a lower limit on the N/C abundance ratio by
474: calculating the \ion{N}{5}/\ion{C}{4} ratio in a region with physical
475: parameters optimized such that the maximum amount of N is in \ion{N}{5}.
476: We invert the equation
477:
478: \begin{equation}
479: \label{eqn:optimal}
480: \frac{\mbox{\ion{N}{5}}}{\mbox{\ion{C}{4}}}=\frac{A(N)}{A(C)}\times
481: \frac{\mbox{(\ion{N}{5}/N)}}{\mbox{(\ion{C}{4}/C)}}
482: \end{equation}
483:
484: \noindent where A(N)/A(C) is the total Nitrogen to total Carbon
485: abundance ratio (henceforth N/C), the quantity we want to measure.
486: Following \citet{HF} (their Figure 10), we find that \ion{N}{5} has a
487: maximum ionization fraction of 0.40 at $\log U$ of $-1.55$. The
488: fraction of carbon in \ion{C}{4} at the same ionization parameter is
489: 0.31. Using our observed ratio of \ion{N}{5}/\ion{C}{4}=2.67 from
490: System 1, this gives a N/C ratio of 2.07, or 6.96\,$[N/C]_\sun$
491: ($\pm7$\%). Doing the same analysis with the weaker absorption lines of
492: System 2 gives a super-solar N/C ratio of only 2.3\,$[N/C]_\sun$
493: ($\pm20$\%). The errors quoted only incorporate uncertainties due to
494: measurements derived from the spectra.
495:
496: To demonstrate the discrepancies caused by improper ionization
497: corrections, we compute N/C for two alternative models. We do this by
498: running the photoionization-equilibrium code Cloudy\,94\footnote{Cloudy
499: version C94.00, obtained from the Cloudy webpage
500: http://www.nublado.org/} \citep{cloudy} with two differing spectral
501: energy distributions (SEDs): a power law SED with $\alpha=-1.5$ and
502: Cloudy's standard AGN SED. We find that System 1 has an abundance ratio
503: of 4.3\,$[N/C]_\sun$ in the case of the simple power law SED and
504: 2.9\,$[N/C]_\sun$ in the case of the standard AGN template SED. This
505: can be compared with value of 6.96 using \citet{HF} as above. The
506: estimated error due to photon noise is only $\sim7$\%, which
507: demonstrates that our result is dominated by systematic uncertainty in
508: the ionization corrections, and more accurate results await data at
509: other wavebands that should permit calculation of a detailed
510: photoionization model for these spectra. Until such a model is
511: constructed, we are limited to stating only that Mrk\,1044 has line
512: strengths consistent with a super-solar N/C ratio.
513:
514: \section{Conclusions} \label{sec:conclude}
515:
516: We have analyzed FUV-MAMA spectra of the Narrow-Line Seyfert 1 galaxy
517: Mrk\,1044. We find two absorption systems blueshifted with respect to
518: the systemic velocity of the galaxy. We find absorption lines from
519: Galactic ISM in both the low-resolution and medium-resolution spectra.
520: We find super-solar N/C in the outflow systems of Mrk\,1044 utilizing a
521: very simple ionization model. We defer constructing a more detailed
522: photoionization model until after analyzing the FUSE and Chandra data so
523: as to avoid the problems that have historically plagued models derived
524: from single wavelength regions.
525:
526: \begin{acknowledgements}
527: The authors wish to thank Mike Crenshaw for his assistance with the
528: apparent optical depth method. This research has made use of the
529: NASA/IPAC Extragalactic Database (NED) which is operated by the Jet
530: Propulsion Laboratory, California Institute of Technology, under
531: contract with the National Aeronautics and Space Administration.
532: Primary support for this work was provided by NASA grant HST-GO-09687.O2
533: from the Space Telescope Science Institute, which is operated by the
534: Association of Universities for Research in Astronomy, Inc., under NASA
535: contract NAS 5-26555.
536: \end{acknowledgements}
537:
538: \begin{thebibliography}{}
539:
540: \bibitem[Bentz \& Osmer(2004)]{BO}
541: Bentz, M. \& Osmer, P. 2004, \aj, 127, 576
542:
543: \bibitem[Boller, Brandt \& Fink(1996)]{BBF}
544: Boller, Th., Brandt, N., \& Fink, H. 1996, A\&A, 305, 53
545:
546: \bibitem[Boroson \& Green(1992)]{BG}
547: Boroson, T. \& Green, R. 1992, \apjs, 80, 109
548:
549: \bibitem[Brandt \& Boller(1998)]{BB}
550: Brandt, N. \& Boller, Th. 1998, A\&A, 319, 7
551:
552: \bibitem[Collin \& Joly(2000)]{CJ}
553: Collin, S. \& Joly, M. 2000 \nar, 44, 531
554:
555: \bibitem[Ferland et al.(1998)]{cloudy}
556: Ferland, G. J., Korista, K. T., Verner, D. A., Ferguson, J. W., Kingdon, J. B., Verner, E. M. 1998, PASP, 110, 761.
557:
558: \bibitem[Grupe \& Mathur(2004)]{GM04}
559: Grupe, D. \& Mathur, S. 2004, \apj, 606L, 41
560:
561: \bibitem[Grupe et al.(2004)]{G04}
562: Grupe, D., Wills, B.J., Leighly, K.M., \& Meusinger, H. 2004 \aj, 127, 156
563:
564: \bibitem[Hamann et al.(1997)]{H97}
565: Hamann, F., Barlow, T.A., Junkkarinen, V., Burbidge, E.M. 1997, \apj, 478, 80
566:
567: \bibitem[Hamann \& Ferland(1999)]{HF}
568: Hamann, F. \& Ferland, G. 1991, \araa, 37, 487
569:
570: \bibitem[Komossa \& Mathur(2001)]{KM01}
571: Komossa, S. \& Mathur, S. 2001, A\&A, 974, 914
572:
573: \bibitem[Kriss(1994)]{SPECFIT}
574: Kriss, G.A. 1994 in Astronomical Data Analysis Software \& Systems III, A.S.P. Conf. Series, Vol. 61, ed. D. R. Crabtree, R. J. Hanisch, \& J. Barnes (Astronomical Society of the Pacific: San Francisco), p. 437.
575:
576: \bibitem[Laor et al.(1997)]{L97}
577: Laor, A., Fiore, F., Elvis, M., Wilkes, B., McDowell, J. 1997, \apj, 477, 93
578:
579: \bibitem[Mathur(2000a)]{M2Ka}
580: Mathur, S. 2000, \mnras, 314, L17
581:
582: \bibitem[Mathur(2000b)]{M2Kb}
583: Mathur, S. 2000, New Astronomy Reviews, 44, 469
584:
585: \bibitem[Mathur et al.(2001)]{M01}
586: Mathur, S., Matt, G., Green, P.J., Elvis, M., Singh, K.P. 2001, \apj, 551L, 13
587:
588: \bibitem[Osmer et al.(1994)]{O94}
589: Osmer, P., Porter, A., Green, R. 1994, \apj, 436, 678
590:
591: \bibitem[Osmer(1980)]{O80}
592: Osmer, P. 1980, \apj, 237, 666
593:
594: \bibitem[Osterbrock \& Pogge(1985)]{OP}
595: Osterbrock, D.E. \& Pogge, R.W. 1985, \apj, 297, 166
596:
597: \bibitem[Penton et al.(2004)]{PSS}
598: Penton, S.V., Stocke, J.T., \& Shull, J.M. 2004, \apjs, 152, 29
599:
600: \bibitem[Pogge \& Owen(1993)]{LINER}
601: Pogge, R.W., \& Owen, J.M. 1993, LINER; An Interactive Spectral Line Analysis Program, OSU Internal Report 93-01
602:
603: \bibitem[Pounds, Done \& Osborne(1995)]{P95}
604: Pounds, K., Done, C., \& Osborne, J. 1995, \mnras, 277, L5
605:
606: %\bibitem[De Robertis, Yee \& Hayhoe (1998)]{RYH}
607: %De Robertis, M. M., Yee, H.K.C., \& Hayhoe, K. 1998, \apj, 496, 93
608:
609: \bibitem[Savage \& Sembach(1991)]{SS91}
610: Savage, B.D., \& Sembach, K.R. 1991, \apj, 379, 245
611:
612: \bibitem[Shemmer \& Netzer(2002)]{SN}
613: Shemmer, O. \& Netzer, H. 2002, \apj, 567, L22
614:
615: \bibitem[Shields \& Hamann(1997)]{SH}
616: Shields, J. \& Hamann, F. 1997, RMxAC, 6, 221
617:
618: \bibitem[Spitzer(1978)]{Spitzer}
619: Spitzer, L. 1987, Physical Processes in the Interstellar Medium (New
620: York: Wiley-Interscience), 46
621:
622: %\bibitem[Virani \& De Robertis (2000)]{VR}
623: %Virani, S.N. \& De Robertis, M.M. 2000, \aj, 120, 1739
624:
625: \bibitem[Wills et al.(1999)]{W99}
626: Wills, B. et al. 1999 in ``Quasars and Cosmology'', Ed. G. Ferland \&
627: J. Baldwin
628:
629: \bibitem[Wills et al.(2000)]{W00}
630: Wills, B.J., Shang, Z. \& Yuan, J.M. 2000, \nar, 44, 511
631:
632: \end{thebibliography}
633: \clearpage
634:
635: %\newpage
636: \begin{deluxetable}{ccccc}
637: %\rotate
638: \tablecaption{Journal of Observations\label{tab:obs}}
639: \tablehead{
640: \colhead{}
641: &\colhead{Central}
642: &\colhead{UTC Date}
643: &\colhead{Exposure}
644: &\colhead{}\\
645: \colhead{Grating}
646: &\colhead{Wavelength [\AA]}
647: &\colhead{(start)}
648: &\colhead{Times [s]}
649: &\colhead{Datasets}}
650: \startdata
651: G140L&1425&2003-06-28T04:46:49&1200 \& 1011&O8K401010-20 \\
652: G140M&1567&2003-06-28T06:14:08&1294 \& 1440&O8K040130-40 \\
653: &1222&2003-06-28T07:50:11&1294 \& 1440&O8K040150-60 \\
654: &1272&2003-06-28T09:26:15&1294, 1409 \& 2$\times$1440&O8K040170-a0 \\
655: \enddata
656: \end{deluxetable}
657:
658: %\newpage
659: \begin{deluxetable}{ccccc}
660: %\rotate
661: \tablecaption{Emission Lines
662: \label{tab:emission}}
663: \tablehead{
664: \colhead{Ion}
665: &\colhead{$\lambda_{Rest}$}
666: &\colhead{Equivalent}
667: &\colhead{FWHM-Narrow}
668: &\colhead{FWHM-Broad}\\
669: \colhead{}
670: &\colhead{[\AA]}
671: &\colhead{Width [\AA]}
672: &\colhead{[\,km\,s$^{-1}$]}
673: &\colhead{[\,km\,s$^{-1}$]}}
674: \startdata
675: \ion{H}{1}&$1216$&$113.6\pm0.1$&$1120.6\pm0.3$&$3406\pm3$ \\
676: \ion{N}{5}&$1239$,$1243$&$40.2\pm0.4$&$1070\pm40$&$5530\pm50$\,\tablenotemark{a} \\
677: \ion{O}{1}&$1302$&$4.9\pm0.7$&$1350\pm7$&$3600\pm200$ \\
678: \ion{C}{2}&$1336$&$2.94\pm0.04$&$1020\pm20$&$3110\pm50$ \\
679: \ion{Si}{4}+\ion{O}{4}]&$1394$,$1403$+$1402$&$19.19\pm0.03$&$\sim2000$&$\sim10000$ \\
680: \ion{N}{4}]&$1486$&$0.984\pm0.005$&$1262\pm4$&\nodata \\
681: \ion{C}{4}&$1548$,$1551$&$67.0\pm0.3$&$1317\pm3$&$3400\pm30$\,\tablenotemark{b} \\
682: \ion{He}{2}&$1640$&$12.9\pm0.5$&$1440\pm50$&$7500\pm200$ \\
683: \ion{O}{3}]&$1663$&$5.2\pm0.4$&\nodata&$3400\pm200$ \\
684: \enddata
685: \tablenotetext{a}
686: {\mbox{Line asymmetry requires a third (red) Gaussian with a width of $3450\pm20$}}
687: \tablenotetext{b}
688: {\mbox{Line asymmetry requires a third (red) Gaussian with a width of $7320\pm30$}}
689: \end{deluxetable}
690:
691: %\newpage
692: \begin{deluxetable}{ccccccc}
693: %\rotate
694: \tablecaption{Measured and Calculated Parameters of Mrk\,1044
695: \label{tab:mrk1044}}
696: \tablehead{
697: \colhead{Line}
698: &\colhead{Observed}
699: &\colhead{FWHM}
700: &\colhead{Equivalent}
701: &\colhead{log(Column)}
702: &\colhead{log(Column)}
703: &\colhead{Velocity}\\
704: \colhead{}
705: &\colhead{Wavelength [\AA]}
706: &\colhead{[\,km\,s$^{-1}$]}
707: &\colhead{Width [m\AA]}
708: &\colhead{[$cm^{-2}$]\,\tablenotemark{a}}
709: &\colhead{[$cm^{-2}$]\,\tablenotemark{b}}
710: &\colhead{[\,km\,s$^{-1}$]}}
711: \startdata
712: System 1&&&&& \\
713: Ly$\alpha$&$1230.9819$&$92\pm7$&$341\pm5$&$15.07^{+0.05}_{-0.05}$&$14.11^{+0.04}_{-0.04}$&$-1156\pm7$ \\
714: \ion{N}{5}$1239$&$1254.4633$&$71\pm3$&$209\pm8$&$14.38^{+0.05}_{-0.04}$&$14.33^{+0.03}_{-0.03}$&$-1143\pm6$ \\
715: \ion{N}{5}$1243$&$1258.4999$&$71\pm3$&$162\pm6$&$14.44^{+0.03}_{-0.03}$&$14.45^{+0.02}_{-0.02}$&$-1143\pm6$ \\
716: \ion{C}{4}$1549$&$1567.7329$&$50\pm6$&$204\pm5$&$13.94^{+0.02}_{-0.03}$&$14.01^{+0.06}_{-0.07}$&$-1147\pm5$ \\
717: \ion{C}{4}$1551$&$1570.3602$&$47\pm5$&$156\pm4$&$14.04^{+0.02}_{-0.02}$&$14.33^{+0.05}_{-0.06}$&$-1143\pm5$ \\
718: System 2&&&&& \\
719: Ly$\alpha$&$1234.4395$&$50\pm8$&$16\pm3$&$12.48^{+0.07}_{-0.08}$&$12.63^{+0.08}_{-0.10}$&$-303\pm6$ \\
720: \ion{N}{5}$1239$&$1257.9923$&$90\pm20$&$23\pm5$&$13.07^{+0.09}_{-0.11}$&$13.18^{+0.11}_{-0.14}$&$-290\pm6$ \\
721: \ion{N}{5}$1243$&$1262.0425$&$100\pm50$&$10\pm4$&$13.00^{+0.15}_{-0.21}$&$12.99^{+0.25}_{-0.63}$&$-289\pm6$ \\
722: \ion{C}{4}$1548$&$1572.1322$&$43\pm10$&$42\pm4$&$13.03^{+0.04}_{-0.04}$&$13.23^{+0.10}_{-0.13}$&$-295\pm5$ \\
723: \ion{C}{4}$1551$&$1574.7397$&$38\pm10$&$29\pm2$&$13.16^{+0.03}_{-0.03}$&$13.44^{+0.11}_{-0.14}$&$-296\pm5$ \\
724: \enddata
725: \tablenotetext{a}
726: {Derived from curve of growth arguments}
727: \tablenotetext{b}
728: {Derived from optical depth integration}
729: \end{deluxetable}
730:
731: %\newpage
732: \begin{deluxetable}{cccccccc}
733: %\rotate
734: \tablecaption{Galactic and Unidentified Absorption Lines
735: \label{tab:galactic}}
736: \tablehead{
737: \colhead{$\lambda_{Obs}$}
738: &\colhead{Ion}
739: &\colhead{$\lambda_{Rest}$}
740: &\colhead{FWHM}
741: &\colhead{Equivalent}
742: &\colhead{log(Column)}
743: &\colhead{log(Column)}
744: &\colhead{Velocity}\\
745: \colhead{[\AA]}
746: &\colhead{}
747: &\colhead{[\AA]}
748: &\colhead{[\,km\,s$^{-1}$]}
749: &\colhead{Width [m\AA]}
750: &\colhead{[$cm^{-2}$]\,\tablenotemark{a}}
751: &\colhead{[$cm^{-2}$]\,\tablenotemark{b}}
752: &\colhead{[\,km\,s$^{-1}$]}}
753: \startdata
754: $1199.5251$&\ion{N}{1}&$1199.550$&$76\pm7$&$210\pm10$&$14.50^{+0.46}_{-0.34}$&$14.25^{+0.06}_{-0.07}$&$-6\pm7$ \\
755: $1200.1924$&\ion{N}{1}&$1200.223$&$65\pm5$&$160\pm10$&$14.41^{+0.36}_{-0.35}$&$14.30^{+0.05}_{-0.06}$&$-8\pm7$ \\
756: $1200.6842$&\ion{N}{1}&$1200.710$&$90\pm20$&$170\pm10$&$14.76^{+0.40}_{-0.36}$&$14.50^{+0.05}_{-0.06}$&$-6\pm7$ \\
757: $1206.4296$&\ion{Si}{3}&$1206.500$&$140\pm10$&$500\pm20$&13.53-16.72&$13.63^{+0.03}_{-0.03}$&$-18\pm7$ \\
758: $1224.3067$\,\tablenotemark{c}&\nodata&$$&$88\pm4$&$251\pm9$&$$&$$&$$ \\
759: $1227.6204$\,\tablenotemark{c}&\nodata&$$&$122\pm10$&$175\pm7$&$$&$$&$$ \\
760: $1235.6051$&\nodata&$$&$82\pm9$&$42\pm3$&$$&$$&$$ \\
761: $1250.6094$&\ion{S}{2}&$1250.584$&$60\pm6$&$93\pm8$&$15.15^{+0.04}_{-0.04}$&$15.18^{+0.05}_{-0.05}$&$+6\pm6$ \\
762: $1253.8245$&\ion{S}{2}&$1253.811$&$60\pm5$&$129\pm7$&$15.05^{+0.03}_{-0.03}$&$15.09^{+0.03}_{-0.04}$&$+3\pm6$ \\
763: $1259.5310$&\ion{S}{2}&$1259.519$&$64\pm5$&$169\pm5$&$15.10^{+0.02}_{-0.02}$&$15.13^{+0.02}_{-0.03}$&$+3\pm6$ \\
764: $1260.4310$&\ion{Si}{2}&$1260.422$&$125\pm3$&$498\pm8$&13.74-16.62&$13.24^{+0.02}_{-0.02}$&$+2\pm6$ \\
765: $1263.5315$&\nodata&$$&$49\pm9$&$33\pm6$&$$&$$&$$ \\
766: $1303.0942$\,\tablenotemark{d}&\ion{O}{1}&$1302.168$&$883\pm7$&$553\pm5$&15.04-18.45&$14.80^{+0.08}_{-0.09}$&$+210\pm70$ \\
767: $1334.9145$\,\tablenotemark{d}&\ion{C}{2}&$1334.532$&$560\pm10$&$720\pm30$&14.82-20.81&$14.18^{+0.21}_{-0.42}$&$+90\pm70$ \\
768: $1527.0709$\,\tablenotemark{d}&\ion{Si}{2}&$1526.707$&$380\pm60$&$280\pm50$&14.09-14.50&$14.09^{+0.08}_{-0.10}$&$+70\pm60$ \\
769: $1548.1517$&\ion{C}{4}&$1548.187$&$97\pm8$&$240\pm20$&$13.83^{+0.04}_{-0.05}$&$13.79^{+0.09}_{-0.11}$&$-7\pm5$ \\
770: $1550.6375$&\ion{C}{4}&$1550.772$&$50\pm10$&$120\pm20$&$13.80^{+0.05}_{-0.06}$&$13.76^{+0.08}_{-0.10}$&$-26\pm5$ \\
771: \enddata
772: \tablenotetext{a}
773: {Derived from curve of growth arguments}
774: \tablenotetext{b}
775: {Derived from optical depth integration}
776: \tablenotetext{c}
777: {Possible Ly$\alpha$ forest lines. See Table~\ref{tab:lyaforest}.}
778: \tablenotetext{d}
779: {From the low-resolution spectrum. The unusually large FWHMs indicate that these values should be used with caution.}
780: \end{deluxetable}
781:
782: %\newpage
783: \begin{deluxetable}{ccccccc}
784: %\rotate
785: \tablecaption{Likely Ly$\alpha$ Forest Lines
786: \label{tab:lyaforest}}
787: \tablehead{
788: \colhead{$\lambda_{Obs}$}
789: &\colhead{FWHM}
790: &\colhead{Equivalent}
791: &\colhead{log(Column)}
792: &\colhead{log(Column)}
793: &\colhead{Rest Frame}
794: &\colhead{Mrk\,1044}\\
795: \colhead{[\AA]}
796: &\colhead{[\,km\,s$^{-1}$]}
797: &\colhead{Width [m\AA]}
798: &\colhead{[$cm^{-2}$]\,\tablenotemark{a}}
799: &\colhead{[$cm^{-2}$]\,\tablenotemark{b}}
800: &\colhead{Velocity [\,km\,s$^{-1}$]}
801: &\colhead{Velocity [\,km\,s$^{-1}$]}}
802: \startdata
803: $1224.3067$&$88\pm4$&$251\pm9$&$14.27^{+0.07}_{-0.07}$&$13.80^{+0.02}_{-0.03}$&$2130\pm7$&$-2801\pm7$ \\
804: $1227.6204$&$120\pm10$&$175\pm7$&$13.79^{+0.04}_{-0.04}$&$13.56^{+0.05}_{-0.06}$&$2937\pm7$&$-1994\pm7$ \\
805: $1228.3089$&$70\pm20$&$45\pm15$&$12.96^{+0.14}_{-0.19}$&$12.77^{+0.10}_{-0.13}$&$3117\pm7$&$-1814\pm7$ \\
806: \enddata
807: \tablenotetext{a}
808: {Derived from curve of growth arguments}
809: \tablenotetext{b}
810: {Derived from optical depth integration}
811: \end{deluxetable}
812:
813: %\newpage
814: \begin{figure}
815: \epsscale{0.9}
816: \plotone{f1.eps}
817: \caption{\label{fig:masterspec}
818: The three G140M spectra superposed over the G140L spectrum. The
819: necessity of obtaining high-resolution spectra to determine the
820: metallicity can be seen very well here. The \ion{C}{4} features are
821: completely wiped out in the G140L spectrum. The \ion{N}{5} feature
822: visible in the G140L spectrum is, after consulting the G140M spectrum,
823: seen mostly to be caused by Galactic absorption. The emission lines
824: originating from Mrk\,1044 are marked above the spectrum. The
825: absorption lines caused by Galactic ISM are marked below the
826: spectrum. }\end{figure}
827: %\clearpage
828:
829: %\newpage
830: \begin{figure}
831: \epsscale{0.95}
832: \plotone{f2.eps}
833: \caption{\label{fig:threespec}
834: The three G140M spectra each shown separately. The absorption systems
835: intrinsic to Mrk\,1044 are marked above the spectrum and the identified
836: Galactic lines are marked below the spectrum. Unidentified lines are
837: marked as such.}\end{figure}
838: %\clearpage
839:
840: %\newpage
841: \begin{figure}
842: \epsscale{1.0}
843: \plotone{f3.eps}
844: \caption{\label{fig:velmap}
845: Velocity maps around Systems 1 (left) and 2 (right). In the case of
846: \ion{C}{4} and \ion{N}{5}, the dotted lines signify the shorter
847: wavelength line of the pair, while the solid lines signify the longer
848: wavelength line. System 1 is very secure with all three ions showing
849: absorption at that velocity. System 2 is less obvious with the
850: \ion{N}{5} lines showing little contrast. Also visible at
851: $\sim+75$\,km\,s$^{-1}$ is the chance alignment of a Galactic \ion{S}{2}
852: line with a yet unidentified line. Evidence against this candidate
853: being a true system is the lack of corroborating absorption at that
854: velocity in \ion{C}{4} and Ly$\alpha$. }\end{figure}
855: %\clearpage
856:
857: %\newpage
858: \begin{figure}
859: \epsscale{0.8}
860: \plotone{f4.eps}
861: \caption{\label{fig:lyaforest}
862: The normalized (continuum divided) spectrum blueward of the Ly$\alpha$
863: emission line of Mrk\,1044. This is one of nine by-eye spline fits to
864: the local continuum (i.e. the Ly$\alpha$ emission line) as described in
865: \S\ref{sec:analysis}. The absorption feature at 1228.3\AA\ is a
866: $3\sigma$ detection. Other possible absorption features such as those
867: blueward of 1224\AA\ or at 1233.5\AA\ have significant detections in
868: only some of the nine continuum fits. For example, the absorption
869: feature at 1233.5\AA\ is located near the steepest part of the
870: Ly$\alpha$ emission line. Small changes in the perceived ``continuum''
871: level severely affect its properties. This also affects the absorption
872: feature of System 2 to a lesser extent, though System 2 has the
873: advantage of being confirmed by the presence of \ion{N}{5} and
874: \ion{C}{4} at the same velocity (see Figure~\ref{fig:velmap}).
875: }\end{figure}
876: %\clearpage
877: \end{document}
878: