1: \documentclass{aastex}
2: \usepackage{emulateapj5,apjfonts}
3:
4: \slugcomment{2000 April 4, Accepted for publication in The Astrophysical Journal Letters}
5:
6: \shorttitle{NARROW X-RAY ABSORPTION LINES FROM NGC\,3783}
7:
8: \shortauthors{KASPI ET AL.}
9:
10: \begin{document}
11:
12: \title{Discovery of Narrow X-ray Absorption Lines from NGC\,3783 \\
13: with the {\it Chandra} HETGS}
14: \author{
15: Shai Kaspi,\altaffilmark{1}
16: W. N. Brandt,\altaffilmark{1}
17: Hagai Netzer,\altaffilmark{2}
18: Rita Sambruna,\altaffilmark{1}
19: George Chartas,\altaffilmark{1} \\
20: Gordon P. Garmire,\altaffilmark{1}
21: and
22: John A. Nousek\altaffilmark{1}
23: }
24: \altaffiltext{1}{Department of Astronomy and Astrophysics, 525 Davey
25: Laboratory, The Pennsylvania State University, University Park, PA, 16802.}
26: \altaffiltext{2}{School of Physics and Astronomy and the Wise
27: Observatory, Raymond and Beverly Sackler Faculty of Exact Sciences, \\
28: \hglue 0.5cm Tel-Aviv University, Tel-Aviv 69978, Israel.}
29:
30: \begin{abstract}
31:
32: We present the first grating-resolution X-ray spectra of the Seyfert~1
33: galaxy NGC\,3783, obtained with the High Energy Transmission Grating
34: Spectrometer on the {\it Chandra} X-ray Observatory. These spectra
35: reveal many narrow absorption lines from the H-like and He-like ions of
36: O, Ne, Mg, Si, S and Ar, as well as \ion{Fe}{17}--\ion{Fe}{21} L-shell
37: lines. We have also identified several weak emission lines, mainly from
38: O and Ne. The absorption lines are blueshifted by a mean velocity of
39: $\approx 440\pm 200$~km\,s$^{-1}$ and are not resolved, indicating a
40: velocity dispersion within the absorbing gas of a few hundred
41: km\,s$^{-1}$ or less. We measure the lines' equivalent widths and
42: compare them with the predictions of photoionization models. The
43: best-fitting model has a microturbulence velocity of 150~km\,s$^{-1}$
44: and a hydrogen column density of $1.3\times 10^{22}$~cm$^{-2}$. The
45: measured blueshifts and inferred velocity dispersions of the X-ray
46: absorption lines are consistent with those of the strongest UV
47: absorption lines observed in this object. However, simple models that
48: propose to strictly unify the X-ray and UV absorbers have difficulty
49: explaining simultaneously the X-ray and UV absorption line strengths.
50:
51: \end{abstract}
52:
53: \keywords{
54: galaxies: active ---
55: galaxies: nuclei ---
56: galaxies: Seyfert ---
57: galaxies: individual (NGC\,3783) ---
58: X-rays: galaxies ---
59: techniques: spectroscopic}
60:
61: \section{Introduction}
62:
63: X-ray spectra of Active Galactic Nuclei (AGNs) often show evidence for
64: deep \ion{O}{7} (739~eV) and \ion{O}{8} (871~eV) absorption edges. The
65: ionized gas component creating these edges is named the `warm absorber'
66: and is seen in most ($\gtrsim$70\%) Seyfert~1s and some quasars.
67: Theoretical models for the warm absorber suggest it should also be a
68: source of X-ray emission (e.g., Netzer 1996) and absorption (e.g.,
69: Nicastro, Fiore, \& Matt 1999) lines. While the current X-ray data are
70: consistent with the presence of such lines (e.g., George et~al 1998a),
71: it has been impossible to study them in any detail.
72:
73: The bright Seyfert~1 galaxy NGC\,3783 ($V\approx13.5$ mag, $z=0.0097$)
74: has one of the strongest X-ray warm absorbers known with a column
75: density of ionized gas of $N_{\rm H}\approx$~(1--2)$\times
76: 10^{22}$~cm$^{-2}$. {\it ASCA} spectra also show warm absorber
77: variability and excess flux around 600~eV which has been interpreted as
78: emission lines from the warm absorber, particularly \ion{O}{7}~568~eV
79: (e.g., George et~al. 1998a, 1998b). UV spectra of NGC\,3783 show
80: intrinsic absorption features due to \ion{C}{4}, \ion{N}{5} and
81: \ion{H}{1}. Two kinematic absorption components are known with radial
82: velocities of $\approx -560$~km\,s$^{-1}$ and $\approx
83: -1420$~km\,s$^{-1}$ relative to the optical redshift from [\ion{O}{3}].
84: The \ion{C}{4} absorption shows extreme variability (e.g., Crenshaw
85: et~al. 1999 and references therein). It is not clear whether the UV and
86: X-ray absorbing gas are the same, although some relation seems likely
87: (e.g., Shields \& Hamann 1997).
88:
89: This letter describes the discovery of a large number of X-ray
90: absorption lines in the {\it Chandra} spectrum of NGC 3783. In \S~2 we
91: describe the observation and data analysis, and \S~3 gives the
92: preliminary interpretation of the results.
93: \vspace{1cm}
94:
95:
96: \section{Observation and Data Analysis}
97: \label{obser}
98:
99: NGC\,3783 was observed using the High Energy Transmission Grating
100: Spectrometer (HETGS; C. R. Canizares et al., in preparation) on the
101: {\it Chandra} X-ray Observatory\footnote{See {\it The Chandra
102: Proposers' Observatory Guide} at http://asc.harvard.edu/udocs/docs/} on
103: 2000 January 21. The detector was the Advanced CCD Imaging Spectrometer
104: (ACIS; G. P. Garmire et al., in preparation). The total integration
105: time was 56~ks, and the observation was continuous. The data were
106: reduced with the standard pipeline by the {\it Chandra} X-ray Center
107: (CXC) on 2000 January 26. We used the {\it Chandra} Interactive
108: Analysis of Observations (CIAO) software to repeat the reduction
109: process using an updated response matrix file. Our reduction produced
110: somewhat less noisy spectra than those produced by the pipeline.
111:
112: The HETGS produces a zeroth order X-ray spectrum at the aim point on
113: the CCD and higher order spectra which have much higher spectral
114: resolution along the ACIS-S array. The higher order spectra are from
115: two grating assemblies, the High Energy Grating (HEG) and Medium Energy
116: Grating (MEG). Each grating assembly produces two spectra for each
117: order. Order overlaps are discriminated by the intrinsic energy
118: resolution of ACIS. The zeroth-order spectrum of NGC~3783 shows
119: substantial photon pile up, and we leave the detailed analysis of these
120: data to a future paper. We briefly comment, however, on the presence of
121: a weak ($\approx 10$ photon) off-nuclear ($\approx 34\arcsec$) source
122: located at $\alpha_{2000}=$~11$^{\rm h}$39$^{\rm m}$01$\fs$1,
123: $\delta_{2000}=$~$-37\arcdeg$44$\arcmin$52$\farcs$9, coincident with
124: one of the galaxy's spiral arms. This source may be a luminous
125: ($L_{0.5-8~{\rm keV}}$ $\approx 2 \times 10^{39}$ ergs\,s$^{-1}$) X-ray
126: binary in NGC~3783, although it could also be an unrelated background
127: source.
128:
129: \begin{figure*}
130: \centerline{\epsfxsize=18.5cm\epsfbox{spectra.eps}}
131: \caption{MEG first-order smoothed spectrum of NGC\,3783. The two upper
132: panels are binned by 0.005~\AA , and the bottom panel is binned by
133: 0.02~\AA . Fluxes are corrected for Galactic absorption. Dots with
134: error bars show the typical 2$\sigma$ statistical error at various
135: wavelengths. The identified absorption features are marked. The few
136: identified emission lines are marked with (r) for resonance, (i) for
137: intercombination, and (f) for forbidden for the He-like ions, as well as
138: L$\alpha$ and L$\beta$ for the H-like ions.}
139: \label{megspec}
140: \end{figure*}
141:
142: The first-order spectra from the HEG and MEG have S/N ratios of
143: $\approx 5$ and $\approx 2.5$, respectively (at around 7~\AA ). The
144: higher order spectra had only a few ($\approx 3$) photons per bin and
145: will not be presented here. In order to flux calibrate the spectra we
146: used the CIAO software to produce Ancillary Response Files (ARFs). The
147: flux calibration error is estimated to be $\lesssim$10\% below 6~\AA\
148: and $\lesssim$20\% above 6~\AA\ (CXC, private communication). We also
149: corrected the spectra for the Galactic absorption of $N_{\rm
150: H}=8.7\times 10^{20}$~cm$^{-2}$ and the cosmological redshift.
151:
152: The two MEG first-order spectra agree well with each other. The
153: dispersions of the MEG spectra are 0.005~\AA\ per bin, and the spectral
154: resolution is $\approx 0.023$~\AA\ (approximately constant across the
155: entire HETGS band). The two HEG spectra also agree well with each other
156: and with the MEG spectra, though they have lower S/N. The HEG spectra
157: have dispersions of 0.0025~\AA\ per bin, and the spectral resolution is
158: about two times better than for the MEG spectra. In Fig.~\ref{megspec}
159: we present the mean of the two MEG spectra. This mean spectrum was
160: smoothed using a simple 3-bin boxcar filter. Since above
161: $\sim$16~\AA\ there are $\lesssim$4 counts per 0.005~\AA\ bin, the
162: spectrum presented in the bottom panel of Fig.~\ref{megspec} is binned
163: to 0.02~\AA .
164:
165: The overall shape and flux of the X-ray spectrum are consistent with
166: those found in previous X-ray observations of NGC\,3783 (e.g., George
167: et~al. 1998b). The {\it Chandra} fluxes in several energy bands are
168: also consistent with those from a simultaneous 4~ks {\it ASCA} spectrum
169: to within $\approx$~10\%. The total flux measured from the {\it
170: Chandra} spectrum is
171: (2.0$\pm$0.2)$\times$10$^{-11}$~ergs\,cm$^{-2}$\,s$^{-1}$ in the
172: 0.5--2~keV band and
173: (4.2$\pm$0.9)$\times$10$^{-11}$~ergs\,cm$^{-2}$\,s$^{-1}$ in the
174: 2--7~keV band. When plotting the modeled {\it ASCA} spectrum on top of
175: the {\it Chandra} spectrum, we find general consistency in the overall
176: shape as well as with the oxygen absorption edges.
177:
178: The most striking new features seen in the {\it Chandra} spectrum of
179: NGC\,3783 are narrow X-ray absorption lines. In order to measure these
180: spectral features we interpolated over all the absorption lines and
181: fitted the resulting continuum with a cubic spline which serves as the
182: continuum reference point for measuring the lines. In Table~1
183: we list the information about selected absorption and emission lines.
184: In Fig.~\ref{megspec} we identify some absorption lines which are not
185: listed in Table~1 due to brevity and lack of model
186: predictions. There are also some absorption and emission features which
187: appear to be present but are not identified yet. We used multiple
188: lines from the same ion to help with line identification. We checked
189: for the existence of these lines in each of the MEG spectra (and also
190: in the HEG spectra when possible) and measured them in the smoothed
191: average spectrum.
192:
193: The measured lines' Full Widths at Half Maximum (FWHMs) are consistent
194: with the MEG resolution which is $\approx 1300$~km\,s$^{-1}$ at
195: 5~\AA\ and $\approx 500$~km\,s$^{-1}$ at 15~\AA . Hence, the lines are
196: not clearly resolved (see discussion in \S~\ref{dinamic}). We were not
197: able to put better constraints on the velocity field with the HEG due
198: to the small S/N.
199:
200: \section{Discussion}
201:
202: \subsection{Absorption Lines, Column Density and Ionization Parameter}
203: \label{ionsec}
204:
205: The strongest observed absorption lines in the X-ray spectrum of
206: NGC\,3783 are from the H-like and He-like ions of O, Ne, Mg, Si, S and
207: Ar, as well as \ion{Fe}{17}--\ion{Fe}{21} L-shell lines (see
208: Table~1 but note that some features represent blends, and the
209: values given are the summed EWs).
210:
211: As explained above, the lines are not resolved, and the only firm
212: conclusion is that, given a Gaussian profile, their FWHMs do not exceed
213: a few hundred km\,s$^{-1}$. Given this uncertainty, we have taken an
214: approach that relies solely on the measured EWs of the lines. We
215: assumed a certain microturbulence velocity in a single line-of-sight
216: cloud and calculated a theoretical spectrum with the photoionization
217: code ION99, \ the 1999 \ version of the
218:
219: %\begin{deluxetable}{lcccc}
220: %\tablecolumns{5}
221: %\tabletypesize{\small}
222: %\tabletypesize{\footnotesize}
223: %\tabletypesize{\scriptsize}
224: %\tablewidth{0pt}
225: %\tablecaption{Absorption and Emission Lines From NGC\,3783
226: %\label{lines}}
227: %\tablehead{
228: %
229: \footnotesize
230: \begin{center}
231: {\sc TABLE 1\\
232: Absorption and Emission Lines From NGC\,3783}
233: \vskip 4pt
234: \begin{tabular}{lcccc}
235: \hline
236: \hline
237: {} &
238: {Predicted} &
239: {Velocity} &
240: {Predicted} &
241: {Measured} \\
242: {} &
243: {$\lambda$} &
244: {Shift} &
245: {EW} &
246: {EW} \\
247: {Ion} &
248: {(\AA)} &
249: {(km\,s$^{-1}$)} &
250: {(m\AA)} &
251: {(m\AA)} \\
252: \hline
253: \multicolumn{5}{c}{Absorption Lines} \\
254: \hline
255: \ion{S }{16} & 4.728 & $-$571$\pm$317 & 0.7 & 8.0$\pm$2.6 \\
256: \ion{S }{15} & 5.039 & $-$535$\pm$297 & 8.2 & 7.5$\pm$2.5 \\
257: \ion{Si}{14} & 5.217 & $-$920$\pm$287 & 1.4 & 10.1$\pm$3.5\phn \\
258: \ion{Si}{14} & 6.181 & $-$534$\pm$242 & 7.7 & 17.3$\pm$4.6\phn \\
259: \ion{Si}{13} & 6.648 & $-$586$\pm$225 & 18.4 & 14.7$\pm$4.8\phn \\
260: \ion{Mg}{12} & 7.106 & $-$379$\pm$211 & 6.6 & 6.2$\pm$2.2 \\
261: \ion{Mg}{11} & 7.473 & $-$280$\pm$200 & 8.6 & 8.0$\pm$3.6 \\
262: \ion{Mg}{11} & 7.851 & $-$382$\pm$191 & 15.6 & 9.7$\pm$3.1 \\
263: \ion{Mg}{12}\tablenotemark{a} & 8.420 & $-$463$\pm$178 & $>$19.9 & 25.6$\pm$6.6\phn \\
264: \ion{Mg}{11}\tablenotemark{a} & 9.170 & $-$261$\pm$163 & $>$25.9 & 34.5$\pm$10.2 \\
265: \ion{Ne}{10} & 9.362 & $-$384$\pm$320 & 6.2 & 16.4$\pm$4.5\phn \\
266: \ion{Ne}{10}\tablenotemark{b} & 9.481 & $-$474$\pm$158 & 12.0 & 18.6$\pm$4.9\phn \\
267: \ion{Ne}{10}\tablenotemark{a} & 9.709 & $-$216$\pm$154 & $>$15.7 & 29.7$\pm$8.2\phn \\
268: \ion{Ne}{10}\tablenotemark{a} & 10.239 & $-$410$\pm$146 & $>$24.7 & 19.8$\pm$5.5\phn \\
269: \ion{Ne}{ 9}\tablenotemark{a} & 11.001 & $-$518$\pm$190 & $>$20.1 & 50.4$\pm$12.2 \\
270: \ion{Fe}{17} & 11.250 & $-$186$\pm$133 & 23.3 & 15.7$\pm$4.3\phn \\
271: \ion{Fe}{18}\tablenotemark{a} & 11.435 & $-$629$\pm$131 & $>$20.5 & 21.0$\pm$14.2 \\
272: \ion{Ne}{ 9}\tablenotemark{a} & 11.547 & $-$701$\pm$129 & $>$49.3 & 27.2$\pm$7.4\phn \\
273: \ion{Ne}{10}\tablenotemark{b} & 12.132 & $-$667$\pm$123 & 69.8 & 49.2$\pm$19.8 \\
274: \ion{Fe}{17}\tablenotemark{b} & 12.263 & $-$391$\pm$122 & 43.1 & 51.9$\pm$12.8 \\
275: \ion{O }{ 8} & 14.635 & \phn$-$20$\pm$102 & 34.3 & 24.8$\pm$12.2 \\
276: \ion{O }{ 8} & 14.821 & $-$182$\pm$101 & 38.9 & 27.8$\pm$12.8 \\
277: \ion{O }{ 8} & 15.176 & \phn$-$79$\pm$197 & 42.8 & 35.8$\pm$11.2 \\
278: \ion{Fe}{17} & 15.264 & $-$471$\pm$196 & 42.7 & 30.6$\pm$8.4\phn \\
279: \ion{Fe}{18} & 15.628 & $-$624$\pm$249 & 18.2 & 27.7$\pm$7.9\phn \\
280: \ion{O }{ 8}\tablenotemark{a} & 16.007 & $-$346$\pm$121 & $>$52.3 & 43.9$\pm$14.2 \\
281: \ion{O }{ 7} & 18.628 & $-$515$\pm$322 & 52.1 & 41.9$\pm$14.8 \\
282: \ion{N }{ 7} & 19.120 & $-$510$\pm$109 & 11.2 & 18.8$\pm$6.8\phn \\
283: \hline
284: \multicolumn{5}{c}{Emission Lines} \\
285: \hline
286: \ion{Mg}{12} & 9.300 & \phm{$-$}306$\pm$225 & 15 & 11.0$\pm$6.8\phn \\
287: \ion{Ne}{10} & 12.132 & \phm{$-$}444$\pm$123 & 36 & 44$\pm$24 \\
288: \ion{Ne}{ 9} & 13.447 & \phm{$-$}\phn64$\pm$111 & 19 & 99$\pm$88 \\
289: \ion{Ne}{ 9} & 13.553 & \phm{$-$}464$\pm$110 & 13 & 163$\pm$140 \\
290: \ion{Ne}{ 9} & 13.699 & \phm{$-1$}26$\pm$175 & 41 & 245$\pm$215 \\
291: \ion{O }{ 8} & 18.967 & \phm{$-$}415$\pm$126 & 460 & 335$\pm$310 \\
292: \ion{N }{ 7} & 20.912 & \phm{$-$}161$\pm$100 & 28 & 118$\pm$92\phn \\
293: \ion{O }{ 7} & 21.602 & \phm{$-1$}54$\pm$69\phn & 99 & 197$\pm$169 \\
294: \ion{O }{ 7} & 21.807 & \phm{$-$}229$\pm$68\phn & 52 & 135$\pm$109 \\
295: \ion{O }{ 7} & 22.102 & \phm{$-$}154$\pm$67\phn & 201 & 631$\pm$583 \\
296: \hline
297: \end{tabular}
298: \vskip 2pt
299: \parbox{3.485in}{
300: \small\baselineskip 9pt
301: \footnotesize
302: \indent
303: $\rm ^a${Blended\,with\,Fe\,absorption lines which\,are\,not\,taken into
304: account in the model.} \\
305: $\rm ^b${Blended with Fe absorption lines which are taken into account
306: in the model.}
307: }
308: \end{center}
309: \setcounter{table}{1}
310: \normalsize
311:
312: \noindent code ION (Netzer 1996). We then
313: compared the results with the observations. The most important
314: ingredients of the modelare the microturbulence velocity, the column
315: density, the spectral energy distribution (SED), the ionization
316: parameter, and the gas composition.
317:
318: The atomic data required for calculating line optical depths are taken
319: from various sources. The oscillator strengths for the H-like and
320: He-like transitions of all metals are well known (e.g., Porquet \&
321: Dubau 2000 and references therein). This is not the case for the
322: various iron L-shell lines where there is no standard published data
323: set. In this paper we rely on various publications (e.g., Mason et al.
324: 1979; Bhatia \& Mason 1980; Cornille et al. 1992, 1994; Phillips et
325: al. 1996; Saba et al. 1999) as well as on the National Institute of
326: Standards and Technology (NIST) data set. These were combined with the
327: Phillips et~al. (1999) line lists, which are based on solar X-ray
328: spectra and served as our major source for the wavelengths of the iron
329: lines. Only 3--5 of the strongest iron lines, per ion, have been used
330: in the analysis.
331:
332: Our models are based on those in George et~al. (1998b) that gave
333: satisfactory fits to the 1993 and 1996 {\it ASCA} observations of the
334: source. They include a weak UV-bump AGN SED (the weak IR case in Netzer
335: 1996) with $\Gamma_{0.1-50~{\rm keV}}$=1.75, a single constant density
336: cloud with $N_{\rm H}=1.3\times 10^{22}$~cm$^{-2}$, solar com-
337: %
338: %\begin{figure}
339: \centerline{\epsfxsize=8.5cm\epsfbox{ewcomp.eps}}
340: \figcaption{Measured versus modeled absorption line EWs. Filled circles
341: are model EWs that are a sum of all lines contributing to the blend.
342: Right-pointing triangles are lower limits for model EWs (see
343: \S~\ref{obser} for details). The agreement between the model and the
344: measured points is noticeable (Spearman rank-order correlation
345: probability of $\approx 10^{-6}$). A line with a slope of unity is
346: drawn to guide the eye.
347: \label{ewcomp}}
348: %\end{figure}
349: \centerline{}
350:
351: %\begin{deluxetable}{lclc}
352: %\tablecolumns{4}
353: %\tabletypesize{\small}
354: %\tabletypesize{\footnotesize}
355: %\tabletypesize{\scriptsize}
356: %\tablewidth{0pt}
357: %\tablecaption{Best Model Column Densities
358: %\label{colden} }
359: %\tablehead{
360:
361: \footnotesize
362: \begin{center}
363: {\sc TABLE 2\\
364: Best Model Column Densities}
365: \vskip 4pt
366: \begin{tabular}{lclc}
367: \hline
368: \hline
369: {} &
370: {Column} &
371: {} &
372: {Column} \\
373: {Ion} &
374: {($10^{17}$ cm$^{-2}$)} &
375: {Ion} &
376: {($10^{17}$ cm$^{-2}$)} \\
377: \hline
378: \ion{O}{7} & 5.0 & \ion{S}{16} & 0.05 \\
379: \ion{O}{8}& 38 & \ion{Ar}{17} & 0.06 \\
380: \ion{Ne}{9}& 4.0 & \ion{Ar}{18} & 0.003 \\
381: \ion{Ne}{10}& 6.9 & \ion{Fe}{17} & 1.2 \\
382: \ion{Mg}{11}& 2.5 & \ion{Fe}{18} & 1.5 \\
383: \ion{Mg}{12} & 1.5 & \ion{Fe}{19} & 1.0 \\
384: \ion{Si}{13} & 2.2 & \ion{Fe}{20} & 0.4 \\
385: \ion{Si}{14} & 0.5 & \ion{Fe}{21} & 0.07 \\
386: \ion{S}{15} & 0.5 & \ion{Fe}{22} & 0.02 \\
387: \ion{C}{4} & 2.4$\times 10^{-5}$ & \ion{N}{5} & 4.7$\times 10^{-4}$ \\
388: \hline
389: \end{tabular}
390: \end{center}
391: \setcounter{table}{2}
392: \normalsize
393: \centerline{}
394:
395: \noindent position,
396: and a 0.1--10 keV ionization parameter of $U_{\rm x}=0.13$. The
397: standard ($E>13.6$~eV) ionization parameter for this continuum is 19.3
398: and the mean temperature is 1.9$\times 10^5$~K. We have assumed
399: Doppler profiles with $V^2=V^2_{\rm thermal}+V^2_{\rm turbulence}$ and
400: calculated absorption EWs for $100<V_{\rm
401: turbulence}<600$~km\,s$^{-1}$. The best agreement with observations is
402: for $V_{\rm turbulence}$=150~km\,s$^{-1}$. The model EWs are tabulated
403: in Table~1. For a few lines we list only the model's EW
404: lower-limit since the measurement includes blends for which we do not
405: yet have a modeled EW. In most cases, the blends are expected to
406: increase the modeled EW by $\approx$~10--20\%. The comparison between
407: the observed and calculated EWs is shown in Fig.~\ref{ewcomp}, and the
408: deduced column densities, for the ions producing the strongest lines,
409: are given in Table~2. All of the strong lines predicted by
410: the model are observed in the spectra. The microturbulence velocity of
411: our best fitted model is consistent with the velocity dispersion of the
412: UV absorption lines (161$\pm$22~km\,s$^{-1}$; Crenshaw et~al. 1999).
413:
414: We have carried out basic tests to verify that the deduced line widths
415: are in the right range. If the `true' widths are much larger than
416: assumed, the resonance lines in the H-like and He-like series will be
417: optically thin. The observed EWs show that this is not the case. On the
418: other hand, if the `true' line widths are much smaller than
419: 150~km\,s$^{-1}$, several of these lines will be saturated, giving very
420: similar EWs to all. This is not observed either. While these are not
421: strong conclusions given the S/N of the observations, it is encouraging
422: to see the good agreement between the observed and calculated EWs,
423: suggesting that the line optical depths are indeed in the intermediate
424: range. According to the model, the largest optical depth for a line
425: with $E>1$~keV is $\tau$(\ion{Ne}{10}~12.132~\AA)=35.
426:
427: \subsection{Dynamics of the Absorbing Gas}
428: \label{dinamic}
429:
430: All the identified absorption lines in the HETG spectrum of NGC\,3783
431: are blueshifted by a few hundred~km\,s$^{-1}$ with respect to the
432: systemic velocity. The unweighted mean blueshift is
433: 440$\pm$200~km\,s$^{-1}$. Kaastra et~al. (2000) have recently observed
434: the bright Seyfert~1 NGC\,5548 using the Low Energy Transmission
435: Grating Spectrometer (LETGS) on {\it Chandra}. The low-energy X-ray
436: lines reported in that paper indicate a similar level of ionization,
437: and the absorption lines are blueshifted by similar amounts. Judging by
438: these first two results, it seems that the warm X-ray material in
439: Seyfert~1 galaxies is outflowing with typical velocities of several
440: hundred km\,s$^{-1}$. This has important implications for the dynamics
441: of the absorbing gas as well as for the X-ray-UV connection.
442:
443: The acceleration of X-ray ionized gas by the strong central radiation
444: source in AGN has not been studied in detail so far (for some
445: discussion see Reynolds \& Fabian 1995). The recent work by Chelouche
446: \& Netzer (2000) is the first detailed study of this kind. This study
447: discusses the motion of discrete, X-ray ionized clouds in hydrostatic
448: equilibrium, under a large range of conditions. It shows that
449: acceleration to velocities of a few hundred to a few thousand
450: km\,s$^{-1}$ is a natural consequence of the AGN environment. The
451: terminal velocity is of the same order as the escape velocity in the
452: region where the flow originates. Using these results, and assuming
453: that the conditions in the nucleus of NGC\,3783 are similar to those
454: considered by Chelouche \& Netzer (2000), we conclude that the origin
455: of the warm absorbing gas in this source is outside the Broad Line
456: Region, where the escape velocity drops below $\approx
457: 1000$~km\,s$^{-1}$. Better line profile analysis will enable us to
458: constrain this location more accurately.
459:
460: Another important comparison is with the UV absorption lines observed
461: from this source. Several papers (e.g., Mathur, Elvis, \& Wilkes 1995;
462: Shields \& Hamann 1997 and references therein) suggest that the UV and
463: X-ray absorbers may originate in the same component. The Crenshaw et
464: al. (1999) observations reveal two UV absorption systems, the strongest
465: of which has a velocity consistent with the mean observed velocity of
466: the X-ray gas and a velocity dispersion consistent with our best fitted
467: model. Our model, which depends on the assumed SED, predicts column
468: densities for \ion{C}{4} and \ion{N}{5} (see Table~2) which
469: are 37 and 5 times smaller, respectively, than what has been calculated
470: by Crenshaw et al. (1999) using the 1993--1995 data. Hence the UV
471: absorption lines appear to be inconsistent with our X-ray absorber
472: model and are likely coming from a separate component.
473:
474: \subsection{Emission Lines and Covering Factor}
475:
476: The HETG spectra of NGC\,3783 revel several weak emission lines (see
477: Table~1) some of which show hints of P Cygni profiles. These
478: have a marginally significant redshift with an unweighted mean of
479: 230$\pm$170~km\,s$^{-1}$ relative to the systemic velocity measured
480: from the [\ion{O}{3}] line. Kaastra et~al. (2000) found similar
481: redshifts for the emission lines in the LETG spectrum of NGC\,5548.
482: While the EWs of some of the lines are well within the MEG and HEG
483: capability, the low S/N and bumpy continuum do not allow reliable EW
484: measurements. The values given in Table~1 are therefore
485: highly uncertain. However, even these rough measurements are good
486: enough to test the idea that the central source in NGC\,3783 is
487: surrounded by gas clouds with properties similar to those deduced for
488: the line-of-sight absorber. This is done by comparing the EWs
489: calculated by ION with those observed, for various assumed covering
490: factors.
491:
492: The calculated emission-line EWs, for an assumed geometry of full
493: covering, are listed in Table~1. Given these values, we find
494: that the mean deduced covering factor is close to unity, in agreement
495: with previous estimates (e.g., George et al. 1998b). This is the {\it
496: first direct constraint} on the covering factor of a warm absorber in
497: any AGN. Unfortunately, the uncertainty on the covering factor is
498: large, due to the uncertain EWs. Further discussion of this issue, as
499: well as detailed analysis of the continuum, emission features, and
500: absorption features, are deferred to a future paper.
501:
502: \acknowledgments
503: We thank all the members of the {\it Chandra} team for their enormous
504: efforts. We also thank Herman L. Marshall and an anonymous referee for
505: helpful comments. We gratefully acknowledge the financial support of
506: NASA grant NAS~8-38252 (G. P. G., PI), NASA LTSA grant NAG~5-8107 (S.
507: K., W. N. B.), and the Alfred P. Sloan Foundation (W. N. B.). H. N.
508: acknowledges support from the Israel Science Foundation and the Jack
509: Adler Chair for Extragalactic Astronomy.
510:
511: \begin{thebibliography}{}
512: \bibitem[Bhatia (1980)]{B80} Bhatia, A. K., \& Mason, H. E. 1980, \aap, 83, 380
513: \bibitem[Chelouche and Netzer (2000)]{CN00} Chelouche, D., \& Netzer, H.
514: 2000, ApJ, Submitted
515: \bibitem[Cornille, et al. (1992)]{1992A&A...259..669C} Cornille, M., Dubau,
516: J., Loulergue, M., Bely-Dubau, F., \& Faucher, P. 1992, \aap, 259, 669
517: \bibitem[Cornille, et al. (1994)]{1994A&AS..105...77C} Cornille, M., Dubau,
518: J., Faucher, P., Bely-Dubau, F., \& Blancard, C. 1994, \aaps, 105, 77
519: \bibitem[Crenshaw et al. (1999)]{1999ApJ...516..750C} Crenshaw, D. M. ,
520: Kraemer, S. B., Boggess, A. , Maran, S. P., Mushotzky, R. F., \& Wu, C. -C.
521: 1999, \apj, 516, 750
522: \bibitem[George et al. (1998)]{1998ApJS..114...73G} George, I. M.,
523: Turner, T. J., Netzer, H. , Nandra, K., Mushotzky, R. F., \& Yaqoob,
524: T. 1998a, \apjs, 114, 73
525: \bibitem[George et al. (1998)]{1998ApJ...503..174G} George, I. M., Turner, T.
526: J., Mushotzky, R., Nandra, K., \& Netzer, H. 1998b, \apj, 503, 174
527: \bibitem[Kaastra et~al. (2000)]{astro-ph..0002345} Kaastra, J. S., Mewe, R.,
528: Liedahl, D. A., Komossa, S., \& Brinkman, A. C. 2000, A\&A, 354, L83
529: \bibitem[Mason, Doschek, Feldman \& Bhatia (1979)]{1979A&A....73...74M}
530: Mason, H. E., Doschek, G. A., Feldman, U., \& Bhatia, A. K. 1979, \aap, 73, 74
531: \bibitem[Mathur, Elvis \& Wilkes (1995)]{1995ApJ...452..230M} Mathur, S.,
532: Elvis, M., \& Wilkes, B. 1995, \apj, 452, 230
533: \bibitem[Netzer (1996)]{1996ApJ...473..781N} Netzer, H. 1996, \apj, 473, 781
534: \bibitem[Nicastro Fiore \& Matt (1999)]{1999ApJ...517..108N} Nicastro, F.,
535: Fiore, F., \& Matt, G. 1999, \apj, 517, 108
536: \bibitem[Phillips (1996)]{P96} Phillips, K. J. H., Bhatia, A. K., Mason,
537: H. E., \& Zarro, D. M. 1996, \apj, 466, 549
538: \bibitem[Phillips (1999)]{P99} Phillips, K. J. H., Mewe, R.,
539: Harra-Murnion, L. K., Kaastra, J. S., Beiersdorfer, P., Brown, G. V.,
540: \& Liedahl, D. A. 1999, \apjs, 138, 381
541: \bibitem[Porquet (2000)]{P20} Porquet, D., \& Dubau, J. 2000, A\&AS, in
542: press (astro-ph/0002319)
543: \bibitem[Reynolds \& Fabian (1995)]{1995MNRAS.273.1167R} Reynolds, C.
544: S., \& Fabian, A. C. 1995, \mnras, 273, 1167
545: \bibitem[Saba (1999)]{S99} Saba, J. L. R., Schmelz, J. T., Bhatia, A. K., \&
546: Strong, K. T. 1999, \apj, 510, 1064
547: \bibitem[Shields \& Hamann (1997)]{1997ApJ...481..752S} Shields, J. C., \&
548: Hamann, F. 1997, \apj, 481, 752
549: \end{thebibliography}
550:
551: \end{document}
552: