1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3: \newcommand{\gtappeq}{\raisebox{-0.6ex}{$\,\stackrel
4: {\raisebox{-.2ex}{$\textstyle >$}}{\sim}\,$}}
5: \newcommand{\deltE}{\Delta\kern-1ptE}
6: \newcommand{\lam}{$\lambda$}
7: \def\ion#1#2{#1\,{\sc #2}}
8:
9: \input epsf
10:
11: \usepackage{longtable}
12: \usepackage{amssymb}
13: \usepackage{graphicx}
14:
15: \slugcomment{To be submitted to ApJ}
16: \shorttitle{High Energy 3D Photoionization Code}
17: \shortauthors{Ercolano et al.}
18: \begin{document}
19: \title{X-ray enabled MOCASSIN: a 3D code for photoionized media}
20: \author{Barbara Ercolano}
21: \affil{Harvard-Smithsonian Center for Astrophysics, MS-67
22: \\ 60 Garden Street, \\ Cambridge, MA 02138, USA}
23: \author{Peter R. Young}
24: \affil{STFC Rutherford Appleton Laboratory, \\Chilton, Didcot, \\
25: Oxfordshire, OX11 0QX, UK}
26: \author{Jeremy J. Drake \& John C. Raymond}
27: \affil{Harvard-Smithsonian Center for Astrophysics\\
28: 60 Garden Street, \\ Cambridge, MA 02138, USA}
29:
30: \begin{abstract}
31: We present a new version of the fully 3D
32: photoionization and dust radiative transfer code, {\sc mocassin}, that uses a Monte Carlo
33: approach for the transfer of radiation. The X-ray enabled {\sc mocassin} allows a fully
34: geometry independent description of low-density gaseous
35: environments strongly photoionized by a radiation field extending from
36: radio to gamma rays. The code has been thoroughly benchmarked against
37: other established codes routinely used in the literature, using simple
38: plane parallel models designed to test performance under standard
39: conditions. We show the results of our benchmarking exercise and
40: discuss applicability and limitations of the new code, which should be of guidance for
41: future astrophysical studies with {\sc mocassin}.
42:
43: %In the tradition of previous realizes of the {\sc mocassin} code, the
44: %X-ray version will also be made publicly available to the scientific
45: %community shortly after publication of this article.
46:
47: \end{abstract}
48:
49: \keywords{radiative transfer --- plasmas}
50:
51: \section{Introduction}\label{s:intro}
52:
53: Photoionized environments characterize a wide range of astrophysical
54: problems involving sources of X--radiation. With the advent of new
55: technology used for instruments on board of (e.g.) {\it XMM-Newton}
56: and {\it Chandra}, high resolution spectroscopy of such
57: environments has become a reality. Paerels et al. (2000), for instance,
58: observed the photoionized wind in Cygnus X--3 with {\it Chandra}
59: High Energy Transmission Grating Spectrometer (HETGS) showing the
60: discrete emission to be excited by recombination in a tenuous
61: X--ray--photoionized medium which is not symmetric with the source of
62: the wind. Other examples include
63: the detection of several recombination emission lines (from
64: Fe~{\sc xxvi} at 1.78~{\AA} to N~{\sc vi} at 29.08~{\AA}), by
65: Jimenez-Garate et al. (2005) in a 50~ks
66: observation of the bright X-ray binary Hercules X-1 with {\it
67: Chandra} HETGS. We also note the {\it Chandra} ACIS observations
68: of the deeply eclipsing cataclysmic variable DQ Herculis by Mukai et
69: al. (2003), who were able to pin down the origin of the soft X-rays
70: from this system as being due to scattering of the unseen
71: central X-ray source, probably in an accretion disk wind.
72:
73: A number of 1D photoionization codes, including G.~Ferland's {\sc cloudy}
74: (Ferland et al. 1998) and T.~Kallman's {\sc xstar} (Kallman \&
75: McCray 1980, Kallman \& Bautista 2001) continue to represent powerful analytical tools for
76: the analysis of astrophysical spectra from the X-ray to the infrared
77: regime. These codes are designed for diffuse, optically thin media, which may
78: also be irradiated by a non-thermal X-ray continuum, and have
79: been, for instance, applied to the modelling of Narrow Line Regions (NLRs) of Active
80: Galactic Nuclei (AGNs). Several other codes have been developed which specialize in emission
81: and reflection spectra from optically thick hot photoionized media,
82: irradiated by a non-thermal continuum extending to the hard X-ray
83: region, such as X--ray irradiated accretion disks (e.g. Ross \& Fabian,
84: 1993; Nayakshin et al. 2000; Dumont et al. 2000).
85:
86: To date, these and most other photoionization codes have numerically
87: solved the equations of radiative transfer (RT) under the assumption
88: of spherical symmetry or in plane parallel geometries, whereupon the problem is reduced
89: to a 1D calculation. While very few real X-ray sources are
90: spherically symmetric, this approximation has been driven by the
91: available computing power and the complexity of the multi-dimensional
92: case. Mauche et al. (2004) developed a Monte Carlo code to investigate
93: the radiation transfer of Ly$\alpha$, He$\alpha$, and recombination
94: continua photons of H- and He-like C, N, O and Ne produced in the
95: atmosphere of a relativistic black hole accretion disk. This code,
96: however, while accounting for Compton scattering and photoabsorption
97: followed by recombination, does not calculate the ionization state of
98: the plasma. To our knowledge, there are currently no general,
99: self-consistent and publicly available X-ray photoionization and dust RT codes capable of
100: working in 3D. Although the 1D codes mentioned above are powerful
101: tools for the analysis of the pan-chromatic spectra of numerous
102: astrophysical environments, their application is restricted to
103: rather simplified geometries.
104:
105: The computational demand of realistic 3D simulations has recently come
106: within the reach of low-cost clusters. Taking advantage of this,
107: the first self-consistent, 3D photoionization and dust RT code was
108: developed for the IR-UV regime using Monte Carlo
109: techniques (Ercolano et al. 2003a, 2005). The code, {\sc mocassin}
110: (MOnte CArlo SimulationS of Ionized Nebulae)
111: was designed to build realistic models of photoionized environments
112: of arbitrary geometry and density distributions, and can
113: simultaneously treat the dust RT. The code can also treat
114: illumination from multiple point- or arbitrarily extended
115: sources. The fully parallel version of {\sc mocassin} (documented and
116: publicly available) is well-tested for
117: classical nebulae, according to standard photoionization benchmarks
118: (P\'equignot et al., 2001), and has been successfully applied to the
119: modeling of H~{\sc ii} regions (e.g. Ercolano, Bastian \&
120: Stasi\'{n}ska, 2007) and planetary nebulae (e.g. Ercolano et
121: al. 2003b,c, 2004; Gon\c{c}alves et al., 2006; Schwarz \& Monteiro,
122: 2006; Wright, Ercolano \& Barlow, 2006). We now present a significantly
123: improved version of the {\sc mocassin} code (version 3.00)
124: which extends to the X-ray regime. In the
125: tradition of previous releases of the {\sc mocassin} code, the
126: X-ray version will also be made publicly available to the scientific
127: community shortly after publication of this article\footnote{Previous
128: versions are available upon request from BE.}.
129:
130: In Section~\ref{s:code} we
131: summarize the physical processes and atomic data added/modified
132: in the new implementation
133: and discuss the applicability and limitations of the code in its
134: current form. In Section~\ref{s:bench} we compare our code to
135: established 1D codes for benchmark tests, comprising emission line
136: spectra from model NLR and from three thin low-density slab models
137: illuminated by a hard continuum. A brief summary is given in Section~\ref{s:summary}.
138:
139:
140: \section{The X-ray enabled {\sc mocassin} code}\label{s:code}
141:
142: \subsection{Basic philosophy and underlying assumptions}
143: The original {\sc mocassin} code
144: was developed in order to provide a 3D modelling tool capable of
145: dealing with asymmetric and density and/or chemically inhomogeneous media, as well as,
146: if required, multiple, non-centrally located, point and extended sources of exciting
147: radiation. The code can self-consistently treat the transfer of
148: ionizing and non-ionizing radiation through a medium composed of gas
149: and/or dust.
150: The numerical techniques employed and the physical processes considered
151: by the code are described in detail by Ercolano et al. (2003a,
152: 2005). In brief, {\sc mocassin} locally simulates the processes of absorption,
153: re-emission and scattering of photons as they diffuse through a
154: gaseous/dusty medium. The radiation field is expressed in terms of
155: energy packets which are the calculation `quanta'. The energy packets
156: are created at the illuminating source(s), which may be placed
157: anywhere inside the grid and be of point-like or diffuse nature. Their trajectories through the nebula are
158: computed, as the packets suffer scatterings, absorptions and
159: re-emissions, according to the local
160: gas and dust opacities and emissivities. The packets' trajectories yield
161: a measure of the local radiation field, from which the local photoionization and
162: recombination rates, as well as the heating and cooling integrals, are
163: determined.
164:
165:
166: It is worth reminding the reader of the physical assumptions that both
167: the previous and current versions of the {\sc mocassin} code rely
168: upon, which also define its applicability and limitations.
169: \begin{itemize}
170: \item All physical processes affecting the gas are in steady
171: state. This implies that the atomic physics and heating and cooling
172: timescales are short
173: compared to those of gas-motion or to the rate of change of the
174: ionizing field.
175: \item In some cases large optical depths may occur in the core of
176: emission lines of astrophysically abundant ions. As described by
177: Ercolano et al. (2005), resonant scattering is accounted for via an
178: escape probability method. The underlying assumption is that
179: photons in the line centre will be scattered many times in a region
180: close to where they were originally emitted, until they
181: finally escape through the line wings or are destroyed through
182: continuum photo-absorption. While escape probability methods are
183: often used in photoionization models, it is well known that these
184: may lower the accuracy of calculations in high density environments
185: (e.g. Avrett \& Loeser, 1988; Dumont et al. 2003). Furthermore, our current
186: scheme assumes a static gas distribution, which would need to be modified to
187: treat winds or accretion disks with very large shear.
188: \item The contribution from recombination is included for
189: H--like, He--like and the lower Fe~{\sc xxiv} lines. Recombination,
190: however, can dominate for several other lines in some extreme cases;
191: in general, this process strengthens the 3s Fe L-lines compared
192: to the 3d (Liedahl et al., 1990). Nevertheless, in the most
193: commonly encountered cases, the contributions from recombination of
194: He-like and H-like species are strongest.
195: \item Very high density environments are problematic and solutions may
196: carry larger uncertainties. Apart from the line transfer
197: issue mentioned above, a higher density
198: limit of $\sim10^{13}$~cm$^{-3}$, at temperatures of $\sim$ 10$^4$~K, is imposed by our approximate treatment of three-body
199: recombination and collisional ionization, which become important at
200: higher densities (see also Ferland 2006). However, for highly
201: ionized, hotter gas, e.g. $\sim10^6$~K gas in the corona of an X-ray binary, the
202: high density limit is increased to $\sim10^{15}$~cm$^{-3}$.
203: \item A treatment for unresolved transition arrays (UTAs), observed in several
204: spectra of AGNs (e.g. Sako at al. 2001), is not
205: currently included in {\sc mocassin}. Calculations were
206: presented by Behar, Sako and Kahn (2001) for inner shell n=2-3
207: photoexcitation of the 16 iron charge states Fe~{\sc i} through Fe~{\sc
208: xvi} and their data will be included in a future version.
209:
210: \end{itemize}
211:
212: \subsection{X-ray extension}
213:
214: \subsubsection{Atomic Data}
215:
216: {\sc mocassin}'s atomic database was updated in order to include the
217: latest data releases. Details of {\sc mocassin's} complete
218: database for the original version are given in Appendix~A of Ercolano
219: et al. (2003a). Data updates (additions and
220: replacements) implemented for version 3.00 are as follows:
221: \begin{itemize}
222: \item Free-bound emission for H~{\sc i},
223: He~{\sc ii} and He~{\sc i} uses the data recently calculated by Ercolano \& Storey (2006).
224:
225: \item The radiative and dielectronic rates of Badnell et al. (2003), Badnell (2006a,b),
226: Zatsarinny et al. (2003, 2004a,b, 2006), Colgan et al. (2003,
227: 2004), Altun et al. (2004, 2006), Mitnik \& Badnell (2004) have been
228: included and are used as default. The data cover all elements up to
229: Zn including sequences up to Na-like electron target, all other
230: species are treated with the previously available data, except for
231: species included in the Nahar (1997,1999) data set, if chosen by the
232: user (see following item).
233:
234: \item The total (radiative and dielectronic) rates of Nahar
235: (1997, 1999) for recombination to all ions of carbon, nitrogen and oxygen and of Nahar (2000) for
236: recombination to Si~{\sc i}, S~{\sc iii}, Ar~{\sc v}, Ca~{\sc vii}
237: and Fe~{\sc xii} can now be included, if explicitly chosen. As well
238: as the recombination data set mentioned as the default in the
239: previous item, all data used in previous versions of {\sc mocassin} are still available for use and compatible with the X-ray version.
240:
241: \item The fits of Verner (1996) to Opacity Project (Seaton et al. 1993) data for the photoionization
242: cross-sections from all shells are used (the previous
243: versions used the same fitting procedures, but only allowed
244: photoionization from the outer valence electron shell).
245:
246: \item The data set for collision strengths, transition probabilities and energy
247: levels has been substantially updated using version 5.2 of the {\sc chianti}
248: atomic database (Landi et al. 2006, Dere et al., 1997). The updates are
249: described in more detail in Section~\ref{s:chianti}. Further updates
250: are planned for the near future when version 6.0 of the database
251: becomes available.
252:
253: \item Recombination lines from hydrogenic species are calculated using
254: fits to the line emissivities of Storey \& Hummer (1995). The data includes
255: species up to Z=8, and temperatures extending to 10$^5$~K for
256: Z\,$\geq$\,2. Species with Z\,$\geq$\,8 are calculated by scaling
257: the He~{\sc ii} data. The same hydrogenic scaling to He~{\sc ii} is also
258: used for species with 2\,$\leq$\,Z\,$\leq$\,8 when the local
259: electron temperatures exceed the highest tabulated temperature of
260: 10$^5$~K.
261:
262: \item Recombination lines from He-like species are calculated using
263: fits to recombination data that will become publicly available
264: in the next version of {\sc chianti} database.
265: The available total (direct+cascade) temperature-dependent
266: recombination coefficients to the individual levels are used in the
267: statistical equilibrium matrix that calculates the level
268: populations, such that the final emissivities automatically include
269: contributions from recombination. As well as the radiative-only
270: rates derived from the Mewe et al. (1985) calculations for He-like
271: iron, the dataset also includes more recent data of
272: Porquet \& Daubau (2000), which include dielectronic contributions,
273: when significant, for He-like ions with Z~=~6, 7, 8, 10, 12 and 14.
274:
275: \item Recombination contributions to Li-like Fe~{\sc xxiv} transitions
276: are calculated in a similar fashion to He-like contributions (see
277: item above), using the temperature-dependent recombination
278: coefficients of Gu et al. (2003). The data, however, extend only to
279: the 3d level, transitions originating from higher levels (e.g. Fe~{\sc
280: xxiv}~7.99~{\AA}) are calculated via hydrogenic scaling of the coefficients given in
281: Storey \& Hummer (1995).
282:
283: \item The ionization rate coefficients of atoms and ions by
284: electron impact now uses the fits presented by Voronov (1997).
285: The collisional ionization and heating effects of the suprathermal secondary electrons
286: following inner-shell photoionization are computed using
287: approximations from Shull \& van Steenberg (1985). Ionization by
288: suprathermal secondaries is generally only important for very low
289: ionization fractions.
290: \end{itemize}
291:
292:
293: \subsubsection{Inner-shell photoionization}
294: Previous versions of the {\sc mocassin} code were designed to
295: treat the transfer of low-energy ionizing radiation, typical of H~{\sc
296: ii} regions and PNe. In these environments, only ionization from the
297: outer valence electron shell needs to be considered. We have now
298: lifted this limitation by including inner shell photoionization using
299: the Auger yields of Kaastra \& Mewe (1993) and the photoionization
300: cross-sections given by Verner et al. (1996). The solution of the
301: ionization balance equations is rendered more complicated by the
302: coupling of non-adjacent states through the emission of multiple
303: electrons by high energy photons. We adopt a similar iteration scheme
304: as that described by Ferland (2006) to solve the resulting ionization
305: balance matrix.
306:
307: \subsubsection{Thomson/Compton scattering}
308:
309: The Compton cross-section is calculated via the Klein-Nishina
310: formula, which simplifies to the Thompson case at non-relativistic
311: energies. Compton heating and cooling contributions are calculated using standard formulae
312: (e.g. Rybicki \& Lightman, 1986). Bound Compton ionization and heating
313: from high energy photons ($\gtrsim$\,2.3 keV for hydrogen) are also taken into
314: account, using the formalism described by Ferland (2006, Eqn 294).
315:
316: The great advantage of a 3D Monte Carlo transfer technique is that
317: scattering events can be treated taking into account the real geometry
318: of the object. The re-processed directions of Compton-scattered
319: packets, for example, are calculated stochastically using
320: probability density functions based on the redistribution functions
321: obtained from the Klein-Nishina formulae. This is of particular
322: interest for the calculation of fluorescence spectra, as
323: described in Drake, Ercolano \& Swartz (2007), Drake \& Ercolano
324: (2007a) and Drake \& Ercolano (2007b), but it can in principle be applied to any
325: line or continuum energy packet.
326:
327: \subsection{Energy levels, collision strengths and radiative decay
328: rates}
329: \label{s:chianti}
330:
331: Energy levels, collision strengths and radiative decay rates have been
332: updated with
333: data from version 5.2 of the {\sc chianti} atomic database (Landi et al.,
334: 2006; Dere et al., 1997). Since highly excited levels are generally not significantly
335: populated in photoionized plasmas, not all data from {\sc chianti} were
336: converted into the {\sc mocassin} format. Thus, for example, while the
337: {\sc chianti} models for many of the boron-like ions consist of 125 levels,
338: only data for the lowest 10 levels are currently used by {\sc
339: mocassin}. The number of levels used for each iso-electronic sequence are
340: listed in Table~\ref{tbl.levels}. The
341: Maxwellian-averaged collision strengths, $\Upsilon$, are stored as
342: spline fits in {\sc chianti} \citep{dere97} which allow $\Upsilon$ to be
343: calculated for any temperature. For {\sc mocassin}, $\Upsilon$s
344: are calculated from the {\sc chianti} spline fits for the temperature range $2.0\le
345: \log\,T\le 6.4$ at 0.2~dex intervals.
346: % It is to be noted, however, that the
347: %{\sc chianti} spline fits are not necessarily accurate over the entire
348: %temperature range, either because the original collision strengths
349: %were not available over this temperature range, or because the {\sc chianti}
350: %spline fits do not accurately represent the original data. This latter
351: %point is particularly important given the different applications for
352: %{\sc chianti} and {\sc mocassin} and we discuss it further.
353:
354: Up until version 4 of {\sc chianti} all collision data were fitted with a five
355: point spline, following the method of \citet{burgess92}. For some
356: transitions, however, a five point spline was not able to accurately fit
357: the entire set of collision strengths and thus it was necessary to
358: omit some of the original data points for the fit. The choice of which
359: data points to omit was influenced by the major application of
360: {\sc chianti} --
361: the analysis of emission line spectra from \emph{collisionally-ionized}
362: plasmas such as the solar corona -- thus care was taken to ensure the
363: fits were accurate in the temperature range where the ion was most
364: abundant. \citep[Tables of these temperature ranges are given in][for example.]{mazzotta98}
365: %\citet{mazzotta98}, for example.)
366: Oftentimes this led to the collision
367: strength data at low temperatures being omitted from the fit. With
368: version 4 of {\sc chianti} \citep{young03} nine point spline fits to the
369: collision strength became available as an option, thus allowing
370: improved fits at low temperatures. As it is the low temperatures that
371: are important for \emph{photoionized} plasmas, the early {\sc chianti}
372: data-sets have been re-assessed to ensure that the low temperature data
373: points are being accurately reproduced. Poor fits were improved by
374: re-fitting the original data with nine point splines. An example of an improved fit to the
375: \ion{O}{iv} ground transition is shown in Fig.~\ref{fig.chianti-o4}.
376:
377: In the process of adapting the {\sc chianti} data files for {\sc mocassin}, the
378: existing data within {\sc mocassin} have been compared with those in
379: {\sc chianti}. In general, {\sc mocassin} used the same or older
380: atomic data; in the latter case {\sc chianti} data simply replaced the
381: {\sc mocassin} data (after the checks described in the previous paragraph had been
382: performed). In some cases, {\sc mocassin} made use of better atomic data
383: (either more recent calculations, or data more suited to photoionized
384: plasmas), and so these data were assessed and added to {\sc chianti}.
385: The comparison between {\sc chianti} and {\sc mocassin} also proved valuable for
386: identifying errors in the data files and a number of corrections have
387: been made to {\sc chianti}. These updates will appear in the next version of that
388: database.
389:
390: \section{Benchmark Tests} \label{s:bench}
391:
392: The complexity of the calculations involved in large scale
393: photoionization codes, where a number of coupled microphysical process
394: are at play, imposes the need for careful testing before predictions
395: from such codes may be trusted. In terms of energetics, care should be
396: taken that all ionization and recombination, heating and cooling
397: channels are correctly represented.
398:
399: The new code presented here has been thoroughly tested and compared, for the 1D case,
400: to the solutions to a number of benchmark models obtained by
401: independent codes routinely used in the literature.
402: The set of benchmarks used was presented by P\'{e}quignot et al. (2001)
403: and comprises a plane-parallel simulation for conditions typical of
404: NLRs of AGNs and three low-density thin slab models irradiated by
405: a strong broken power-law continuum radiation field, as described in
406: Table~10 of P\'{e}quignot et al. (2001), and reproduced here for
407: convenience (see Table~\ref{t:bpl}). For the NLR model
408: the illuminating spectrum is a power law in energy units with slope
409: 1.3, and the
410: ionization parameter, $U_{13.6}$~=~0.01 is defined as
411: $F_{13.6}$\,/\,($c\,\times\,n_{\rm H}$), where $F_{13.6}$ is the
412: incident photon flux above 13.6~eV.
413: The ionization parameters of the three X-ray slab models, X0.1, X1.0
414: and X10.0 are $U_{100}$~=~0.1, 1.0 and 10.0, respectively, where the
415: definition of $U_{100}$ is analogous to that of $U_{13.6}$ above.
416:
417: The plane-parallel geometry assumed by the 1D test, can be mimicked
418: with our 3D code by modelling a thin, elongated cuboidal grid with
419: plane parallel diffuse illumination
420: coming at normal incidence from one of the smaller sides; the energy packets are only
421: allowed to escape from the side opposed to the illuminating side and all other sides act
422: as mirrors (for a discussion of the {\it mirror} technique see
423: Ercolano et al., 2003b). The main input parameters for the benchmark tests
424: are given in Table~1 of P\'{e}quignot et al. (2001) and repeated in
425: our Table~\ref{t:benchinput} for convenience.
426:
427: We run a number of tests to estimate the random error due to Monte
428: Carlo sampling in order to establish an efficient, yet reliable,
429: combination of cell numbers and energy packets needed to sample them. We
430: adopted grids of 300 depth points and three by
431: three in the short axes, with the number of packets needed to achieve
432: convergence varying from one to eight million. Errors of the order of
433: 0.5~\% in computed line strengths are commmon with this setup, with larger errors ($<$6\%)
434: sometimes obtained for the less abundant ions.
435:
436: We note that radiative transfer effects are largely unimportant for
437: the lines listed in the X-ray slab models. Furthermore, the models are
438: optically thin to ionizing radiation and electron temperature and
439: density, and ionization structure variations through the slabs
440: are minimal.
441:
442: Before discussing the benchmark results in detail, we note that many
443: atomic data updates and code
444: developments have occurred in the past seven years (see Ercolano
445: 2005). In particular,
446: the radiative and dielectronic
447: recombination coefficients calculated by Badnell et al. (2003), Badnell (2006a,b),
448: Zatsarinny et al. (2003, 2004a,b, 2006), Colgan et al. (2003,
449: 2004), Altun et al. (2004, 2006), Mitnik \& Badnell (2004),
450: covering all elements up to Zn (plus Kr, Mo and Xe), including sequences
451: up to Na-like electron target\footnote{The data set is publicly available form
452: http://amdpp.phys.strath.ac.uk/tamoc/DATA/.}, represent a significant
453: improvement over earlier sets.
454:
455: \subsection{Narrow Line Region}
456:
457: We have organized the benchmark comparison in a similar fashion to
458: P\'{e}quignot et al. (2001); for the NLR model we check
459: that {\sc mocassin}'s prediction of each quantity falls within the
460: range of the values obtained by the other codes, listed in Table~10 of
461: P\'{e}quignot et al. (2001). In those cases where {\sc mocassin}'s
462: prediction lies outside the range, we compute the {\it isolation
463: factors}, defined as the ratio of the value given by {\sc mocassin}
464: to the upper limit of the range, if our value is above the range, or
465: the ratio of the lower limit of the range to the value given by {\sc
466: mocassin}, if our value is below the range.
467: Table~\ref{t:nlr} lists line fluxes, temperature and ionized helium
468: fraction as predicted by {\sc mocassin} (column 4) as well as the
469: minimum and maximum values (columns 2 and 3) predicted by the other
470: codes and given in Table~7 of P\'{e}quignot et al (2001). The
471: isolation factors computed for {\sc mocassin} are given in the last
472: column of our Table~\ref{t:nlr}, where we have added a minus or plus
473: sign to indicate
474: whether our prediction lies below or above the range predicted by the
475: other codes.
476:
477: In P\'{e}quignot et al. (2001) it is stated that isolation factors
478: larger that 1.3 should be considered ``indicative of a very
479: significant departure and possible problem''.
480: From Table~\ref{t:nlr} we find that only six
481: out of 48 quantities predicted carried an isolation factor with
482: absolute value larger that
483: 1.3, with only one of these being larger that 2.0. We judge this to be
484: indicative of a high degree of consistency between the results of {\sc
485: mocassin} and other codes, especially in the light of
486: the atomic physics updates that have occurred between the year 2001 and
487: today. Furthermore, the temperature and ionization structure of our
488: model is also in good agreement with those calculated by other codes
489: indicating that the differences in some line predictions can
490: probably be ascribed to atomic physics changes.
491:
492: The largest isolation factors were obtained for sulphur
493: ions. The values in Table~\ref{t:nlr} show that our predictions for the [S~{\sc iii}]
494: lines are generally lower than those predicted by other
495: codes, and we suspect that dielectronic recombination effects may be at play in this
496: case. Rates for third row elements are poorly
497: known. In the case of recombination of S~{\sc iv} into S~{\sc
498: iii} the rates calculated by Nahar (2000) are available. These
499: calculations used an ab-initio method to derive total (electron+ion)
500: recombination (Nahar \& Pradhan 1994, 1995) which enables the consideration of radiative and dielectronic
501: recombination processes in a unified manner. However not all modelers
502: choose to include
503: the total recombination coefficients due to the uncertainties inherent to the method
504: that relies on theoretical predictions of the resonance
505: positions. Here we have used the data of Colgan et al. (2003), which
506: were not available to the 2001 benchmark modelers.
507: Furthermore, no data was available in 2001 for S~{\sc
508: iii}~$\rightarrow$~S~{\sc ii} dielectronic recombination and estimates were
509: used instead (e.g. Ali et al., 1991). Here we used the data of Colgan
510: et al. (2004). It seems likely that some of the
511: scatter recorded for these lines is indeed due the different atomic
512: data available and assumptions made.
513:
514: As a final note we should add that the temperature at the inner edge
515: calculated by {\sc mocassin} is slightly lower than the values
516: obtained by the other codes. This is to be expected since {\sc
517: mocassin} uses a exact treatment of the radiative transfer,
518: including the diffuse component, while comparison codes approximate
519: the transfer of the diffuse component by either iterating along only
520: one direction or by adopting an ``outward-only'' approximation, but
521: along several directions. As noticed by P\'{e}quignot et al. (2001),
522: the kinetic temperature tends to be lower
523: in the innermost layers of models with exact transfer, as the ionizing
524: radiation field there is softer. Figure~\ref{f:tene} shows electron
525: temperatures, $T_{\rm e} [K]$, and densities, $n_{\rm e} [cm^{-3}]$,
526: calculated with {\sc mocassin} as a function of column density in the slab.
527:
528: \subsection{X-ray irradiated plane-parallel slabs}
529:
530: The P\'{e}quignot et al. (2001) benchmarks constitute
531: the first attempt to assess
532: the accuracy of photoionization codes in the X-ray regime. As well as
533: atomic data issues, already pointed out in the previous section, a
534: further complication for this set of models is posed by the fact
535: that, in general, for the
536: far-UV lines listed, the original benchmarks considered sums of
537: multiplet lines in a rather liberal sense, so that {\it energetics}
538: were privileged at the expense of accuracy of wavelengths (with the
539: exceptions of some fine-structure optical lines). It is therefore possible
540: that part of the scatter amongst the original benchmark values
541: arose because of the different way multiplets have been handled in
542: different model calculations (P\'equignot, priv. comm.). This problem persists and should be kept in mind when
543: looking at the comparisons performed in this
544: article. The original benchmark tests were quite
545: rough and their main
546: aim was to highlight gross discrepancies, particularly in derived temperatures.
547: A new benchmarking exercise has yet to appear for the high-energy regime;
548: however it is still useful for us to check our code against the 2001
549: benchmarks as major problems in the thermal and ionization balances
550: would be uncovered by such a comparison. Furthermore, we propose that
551: the solutions presented here be taken as an improved set, obtained
552: with up-to-date atomic data.
553:
554: Models of X-ray slabs are assessed by comparing, for each quantity,
555: {\sc mocassin}'s prediction with the minimum and maximum limits of
556: the range of values obtained by the other codes and listed in
557: Tables~11-13 of P\'{e}quignot et al. (2001).
558:
559: In Figures~\ref{f:x01}, \ref{f:x1} and \ref{f:x10} we use red
560: (striped) histograms to illustrate {\sc mocassin}'s results, while the filled and
561: empty black histograms respectively mark the lower and the higher limit of the
562: range of solutions obtained by the other codes in the original
563: exercise. All line fluxes are in units of $[erg\,s^{-1}]$ for a column
564: density of 10$^{16}\,cm^{-2}$, with the emissivities being summed over 4$\pi$.
565: In aid of future benchmarking efforts we provide in
566: Table~\ref{t:xlines} the individual line fluxes in $[erg\,s^{-1}]$
567: as predicted by this version (3.00) of {\sc mocassin}.
568:
569: In spite of all the complications listed above, the agreement of {\sc
570: mocassin}'s predictions with the benchmarks is reassuring and adds
571: confidence to current and future astrophysical applications of the
572: code. In particular, very good agreement is shown for the relative
573: strengths of lines from far infrared to X-ray wavelengths. Some
574: exceptions are discussed in more detail below.
575:
576: %\subsubsection{X-ray benchmark X01}
577: %\label{s:x01}
578: The agreement between {\sc mocassin} and the codes in P\'equignot et
579: al. (2001) for the benchmark X01 is good, although the electron temperature of the slab,
580: listed in Table~\ref{t:xlines}, carries an isolation factor of
581: 1.14. While this is small and not of concern {\it per se }
582: it may affect the strengths of some of the emission
583: lines.
584:
585: %\subsubsection{X-ray benchmark X1}
586: %\label{s:x1}
587: Figures~\ref{f:x1} and~\ref{f:x10} demonstrate a very good agreement between {\sc mocassin} and
588: the codes in P\'equignot et al. (2001) for most of the lines listed in
589: Table~\ref{t:xlines} for benchmarks X1 and X10.
590: The small discrepancies with some of the H-like recombination lines (e.g. He~{\sc ii}~303.8~{\AA} and H~{\sc
591: i}~1216~{\AA} in X1 and Ar~{\sc xviii}~20.20~{\AA},
592: N~{\sc vii}~24.78~{\AA}, C~{\sc vi}~28.47~{\AA} and 33.74~{\AA} and He~{\sc
593: ii}~303.8~{\AA} in X10) are most probably due to different
594: extrapolation techniques of the Storey \& Hummer (1995) data to the
595: high temperatures of these models.
596:
597: \section{Summary}\label{s:summary}
598:
599: We have presented a new version of the fully 3D
600: Monte Carlo photoionization code, {\sc mocassin}. The code was
601: extended to allow the modelling of plasma irradiated by a hard
602: continuum spanning from radio to gamma rays. The atomic data set of
603: the code was also significantly updated and it is now synchronized
604: with the latest release of the {\sc chianti} database.
605: The applicability and
606: limitations of the new code were discussed, and the results
607: of a thorough benchmarking exercise presented.
608:
609: No major problems were found by the benchmark tests, although some
610: minor differences have been found. We have
611: highlighted a number of significant improvements in the atomic
612: datasets available today compared to those available at the time the
613: original benchmark tables were compiled. We provide here updated
614: values and we emphasize the need of a new benchmarking exercise to be
615: undertaken by the plasma modelling community.
616:
617: The good performance of the {\sc mocassin} code in all benchmark
618: tests demonstrates that it is ready for application to real
619: astrophysical problems. The public version of the X-ray enhanced {\sc
620: mocassin} code is available on request from the author.
621:
622: \section*{Acknowledgments}
623:
624: We thank Daniel P\'{e}quignot for crucial guidance with the benchmark
625: comparisons. We also would like to thank Gary
626: Ferland for helpful discussion on solving the ionization balance
627: matrix for non-adjacent stages. Our thanks are also extended to Pete
628: Storey for help with the hydrogenic data set and dielectronic
629: recombination data. We thank the {\sc chianti} team for making the new
630: data available to us. Finally, we thank the anonymous referee for constructive
631: comments that helped to significantly improve the quality of this work.
632: JJD was supported by the Chandra X-ray Center NASA contract NAS8-39073
633: during the course of this research. BE was supported by {\it Chandra}
634: grants GO6-7008X and GO6-7009X.
635:
636:
637: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
638:
639:
640: \begin{table}
641: \caption{The number of energy levels used for each iso-electronic
642: sequence when compiling the {\sc mocassin} atomic data files.}
643: \begin{center}
644: \begin{tabular}{llll}
645: \noalign{\hrule}
646: \noalign{\smallskip}
647: %
648: Iso-electronic &Number &Iso-electronic &Number \\
649: Sequence & of levels &Sequence & of levels \\
650: \noalign{\hrule}
651: \noalign{\smallskip}
652: %
653: H & 9 & Na & 3 \\
654: He & 17 & Mg & 5 \\
655: Li & 15 & Al & 12 \\
656: Be & 10 & Si & 27 \\
657: B & 10 & P & 5 \\
658: C & 15 & S & 5 \\
659: N & 13 & Cl & 2 \\
660: O & 9 & Ar & -- \\
661: F & 2 & K & 2 \\
662: Ne & -- & Ca & 9 \\
663: & & Sc &19 \\
664: \noalign{\hrule}
665: \end{tabular}
666: \end{center}
667: \label{tbl.levels}
668: \end{table}
669:
670:
671: \begin{table}
672: \begin{center}
673: \caption{Input parameters for benchmark models.\tablenotemark{a}\label{t:benchinput}}
674: \tablenotetext{a}{ Plane parallel
675: geometry is assumed for all models. PL1: power law, slope 1.3; BPL:
676: Broken Power Law (Table~\ref{t:bpl}). All abundances are given by
677: number with respect to hydrogen.}
678: \begin{tabular}{lcccc}
679: \tableline
680: Model: & NLR & X$^{\rm thin}_{0.1}$ & X$^{\rm thin}_{1}$ &
681: X$^{\rm thin}_{10}$ \\
682: \tableline
683: Ionizing Spectrum & PL1 & BPL & BPL & BPL\\
684: $n_{\rm H}\,[$cm$^{-3}]$& 10$^4$ & 10$^5$ & 10$^5$ & 10$^5$ \\
685: $N{\rm (H)}^{\rm out} [$cm$^{-2}]$& 10$^{22}$ & 10$^{16}$ & 10$^{16}$
686: & 10$^{16}$ \\
687: U$_{13.6\,eV}$ & 0.01 & -- & -- & -- \\
688: U$_{100\,eV}$ & -- & 0.1 & 1 & 10 \\
689: He/H & 0.1 & 0.1 & 0.1 & 0.1 \\
690: C/H\,$\times$\,10$^5$ & 30.0 & 37.0 & 37.0 & 37.0 \\
691: N/H\,$\times$\,10$^5$ & 10.0 & 11.0 & 11.0 & 11.0 \\
692: O/H\,$\times$\,10$^5$ & 80.0 & 80.0 & 80.0 & 80.0 \\
693: Ne/H\,$\times$\,10$^5$ & 10.0 & 11.0 & 11.0 & 11.0 \\
694: Mg/H\,$\times$\,10$^5$ & 3.0 & 3.7 & 3.7 & 3.7 \\
695: Si/H\,$\times$\,10$^5$ & 3.0 & 3.5 & 3.5 & 3.5 \\
696: S/H\,$\times$\,10$^5$ & 1.5 & 1.6 & 1.6 & 1.6 \\
697: Ar/H\,$\times$\,10$^5$ & -- & 0.37 & 0.37 & 0.37 \\
698: Fe/H\,$\times$\,10$^5$ & -- & 4.0 & 4.0 & 4.0 \\
699: \tableline\noalign{\smallskip}
700: \end{tabular}
701: \end{center}
702: \end{table}
703:
704:
705: \begin{table}
706: \begin{center}
707: \caption{Broken power law for optically thin X-ray models.\label{t:bpl}}
708: \begin{tabular}{lcccccc}
709: \tableline
710: log h$\nu$ (eV) & -4.8663 & -0.8663 & 1.6108 & 2.0000 & 5.0000 &
711: 7.1337 \\
712: log F$_{\nu}$ (erg) & 1.0 & 14.5 & 12.7 & 10.6 & 7.6 & 1.0 \\
713: \tableline\noalign{\smallskip}
714: \end{tabular}
715: \end{center}
716: \end{table}
717:
718:
719: \begin{table}
720: \begin{center}
721: \caption{Standard Narrow Line Region (NLR). \tablenotemark{a}\label{t:nlr}}
722: \tablenotetext{a}{
723: The $Min$ and $Max$ values
724: of each $Quantity$ were taken from the range of values predicted by
725: the various codes listed in Table~7 of P\'{e}quignot et al. (2001).
726: When {\sc mocassin}'s prediction for a given quantity lies outside
727: the range $[Min,Max]$ the isolation factor is defined as the
728: positive ratio of {\sc mocassin}/$Max$ if {\sc mocassin} $>$ $Max$
729: or the negative ratio of $Min$/{\sc mocassin} if {\sc mocassin} $<$
730: $Min$. All line intensities are given relative to H$\beta$.}
731: \begin{tabular}{lcccc}
732: \tableline
733: Quantity & Min & Max & {\sc mocassin}& isolation \\
734: \tableline
735: H$\beta$ erg/s/cm$^2$ & 1.09 & 1.49 & 1.24 & 0 \\
736: H$\beta$ 4861 & 1.00 & 1.00 & 1.00 & 0 \\
737: He~{\sc i} 5876 & 0.100 & 0.139 & 0.111 & 0 \\
738: He~{\sc ii} 4686 & 0.226 & 0.260 & 0.226 & 0.\\
739: C~{\sc ii} $]$2325+ & 0.362 & 0.600 & 0.438 & 0 \\
740: CII1335 & 0.09 & 0.148 & -- & 0 \\
741: CIII$]$1907+1909 & 2.33 & 4.09 & 3.20 & 0. \\
742: CIV1549+ & 3.36 & 4.90 & 3.84 & 0. \\
743: $[$NI$]$5200+5198 & 0.034 & 0.230 & 0.033 & -1.03 \\
744: $[$NII$]$6584+6548 & 1.19 & 3.27 & 0.933 & -1.28 \\
745: $[$NII$]$5755 & 0.018 & 0.310 & 0.013 & -1.39 \\
746: NIII$]$1749+ & 0.177 & 0.265 & 0.157 & -1.12 \\
747: $[$NIII$]$57.3$\mu$m& 0.042 & 0.050 & 0.041 & -1.03 \\
748: NIV$]$1487+ & 0.203 & 0.250 & 0.343 & +1.37 \\
749: NV1240+ & 0.106 & 0.191 & 0.096 & -1.10 \\
750: $[$OI$]$63.1$\mu$m & 0.220 & 3.10 & 0.156 & -1.43 \\
751: $[$OI$]$6300+6363 & 1.37 & 2.17 & 1.44 & 0 \\
752: $[$OII$]$3726+3729 & 1.25 & 1.85 & 1.23 & -1.02 \\
753: $[$OIII$]$51.8$\mu$m & 0.667 & 0.858 & 0.919 & +1.07 \\
754: $[$OIII$]$88.3$\mu$m & 0.092 & 0.112 & 0.117 & +1.04 \\
755: $[$OIII$]$5007+4959 & 31.2 & 34.9 & 39.3 & +1.12 \\
756: $[$OIII$]$4363 & 0.286 & 0.348 & 0.273 & -1.04 \\
757: $[$OIV$]$25.9$\mu$m & 1.72 & 2.53 & 1.948 & 0 \\
758: OIV$]$1403+ & 0.311 & 0.510 & 0.374 & 0 \\
759: OV$]$1218+ & 0.120 & 0.201 & 0.148 & 0 \\
760: OVI1034+ & 0.030 & 0.059 & 0.041 & 0 \\
761: $[$NeII$]$12.8$\mu$m & 0.145 & 0.832& 0.165 & 0 \\
762: $[$NeIII$]$15.5$\mu$m & 1.38 & 2.86 & 1.44 & 0 \\
763: $[$NeIII$]$3869+68 & 1.64 & 2.46 & 1.92 & 0 \\
764: $[$NeIV$]$2423+ & 0.394 & 0.586 & 0.387 & -1.02 \\
765: $[$NeV$]$3426+3346 & 0.520 & 0.670 & 0.456 & -1.14 \\
766: $[$NeV$]$24.2$\mu$m & 0.162 & 0.450& 0.340 & 0 \\
767: $[$NeVI$]$7.63$\mu$m & 0.226 & 0.288& 0.243 & 0 \\
768: MgII~2798+ & 1.23 & 2.33 & 1.77 & 0. \\
769: $[$MgIV$]$4.49$\mu$m & 0.092 & 0.116 &0.060 & -1.53 \\
770: $[$MgV$]$5.61$\mu$m & 0.100 & 0.230 &0.170 & 0 \\
771: $[$SiII$]$34.8$\mu$m & 0.728 & 1.37 &0.664 & -1.09 \\
772: SiII$]$2335+ & 0.114 & 0.257 &0.177 & 0 \\
773: SiIII$]$1892+ & 0.089 & 0.312 &0.130 & 0 \\
774: SiIV1397+ & 0.078 & 0.121 &0.101 & 0 \\
775: $[$SII$]$6716+6731 & 0.750 & 1.86 &1.80 & 0 \\
776: $[$SII$]$4069+4076 & 0.082 & 0.180 &0.222 & +1.23 \\
777: $[$SIII$]$18.7$\mu$m & 0.560 & 0.968 &0.284 & -1.97 \\
778: $[$SIII$]$33.6$\mu$m & 0.302 & 0.488 &0.140 & -2.16 \\
779: $[$SIII$]$9532+9069 & 1.98 & 2.32 &1.30 & -1.54 \\
780: $[$SIV$]$10.5$\mu$m & 0.850 & 1.74 &1.57 & 0 \\
781: T$_{inner} [$K$]$ & 16840 & 17100 & 16250 & -1.04 \\
782: $<T[N_{\rm H^+}N_{\rm e}]> [$K$]$ & 11970 & 12920 & 12480 & 0 \\
783: $<$He$^+$/He$>$/$<$H$^+$/H$>$ & 0.729 & 0.766 & 0.736 & 0 \\
784: \tableline\noalign{\smallskip}
785: \end{tabular}
786: \end{center}
787: \end{table}
788:
789:
790:
791: \begin{table*}
792: \begin{center}
793: \caption{X-ray irradiated slabs. Emission line fluxes predicted by
794: {\sc mocassin}, given in units of $[erg/s]$. \label{t:xlines}
795: }
796: \begin{tabular}{lcc|lcc|lcc}
797: \tableline
798: \multicolumn{3}{c}{X01} &\multicolumn{3}{c}{X1} &\multicolumn{3}{c}{X10} \\
799: Ion & $\lambda\,[${\AA}$]$ & Flux & Ion & $\lambda\,[${\AA}$]$ & Flux & Ion & $\lambda\,[${\AA}$]$ & Flux \\
800: \tableline
801: Mg9 & 9.314 & 1.17E-04 & S16 & 4.792 & 7.05E-05 &Fe26 & 1.392 & 1.98E-05 \\
802: Ne10 & 12.13 & 1.30E-04 & S15 & 5.101 & 1.40E-04 &Fe26 & 1.425 & 4.06E-05 \\
803: Ne9 & 13.45 & 1.29E-04 & Sil14 & 6.182 & 2.30E-04 &Fe26 & 1.503 & 1.11E-04 \\
804: Ne9 & 13.55 & 1.53E-04 & Sil13 & 6.648 & 7.32E-05 &Fe25 & 1.573 & 7.75E-05 \\
805: Ne9 & 13.70 & 4.27E-04 & Sil13 & 6.688 & 1.08E-04 &Fe26 & 1.780 & 6.23E-04 \\
806: O8 & 15.18 & 1.19E-04 & Sil13 & 6.740 & 2.31E-04 &Fe25 & 1.851 & 5.72E-04 \\
807: O8 & 16.01 & 3.27E-04 & Mg12 & 7.106 & 4.91E-05 &Fe25 & 1.859 & 7.81E-04 \\
808: O8 & 18.97 & 1.84E-03 & Mg12 & 8.421 & 2.38E-04 &Fe25 & 1.868 & 1.29E-03 \\
809: N7 & 20.91 & 3.86E-05 & Mg11 & 9.314 & 8.98E-05 &Ar18 & 3.151 & 1.79E-05 \\
810: O7 & 21.60 & 4.57E-04 & Ne10 & 10.24 & 9.04E-05 &Ar18 & 3.731 & 8.75E-05 \\
811: O7 & 21.80 & 4.78E-04 & Ne10 & 12.13 & 4.14E-04 &S16 & 3.991 & 5.76E-05 \\
812: O7 & 22.10 & 1.58E-03 & Ne9 & 13.70 & 5.61E-05 &Ar17 & 3.994 & 2.23E-05 \\
813: N7 & 24.78 & 2.07E-04 & O8 & 15.18 & 1.15E-04 &S16 & 4.729 & 2.79E-04 \\
814: C6 & 28.47 & 9.70E-05 & O8 & 16.01 & 2.94E-04 &Si14 & 4.947 & 3.17E-05 \\
815: N6 & 29.54 & 1.24E-04 & O8 & 18.97 & 1.26E-03 &S15 & 5.101 & 3.54E-05 \\
816: C6 & 33.74 & 5.04E-04 & N7 & 24.78 & 9.64E-05 &Si14 & 5.217 & 8.27E-05 \\
817: C5 & 41.47 & 8.10E-05 & C6 & 33.74 & 1.68E-04 &Si14 & 6.182 & 3.78E-04 \\
818: O8 & 102.5 & 1.94E-04 & Fe19 & 101.5 & 3.23E-05 &Si13 & 6.740 & 2.34E-05 \\
819: C6 & 182.2 & 5.01E-05 & O8 & 102.5 & 1.05E-04 &Mg12 & 7.106 & 4.78E-05 \\
820: He2 & 256.0 & 2.32E-04 & Fe19 & 108.4 & 1.28E-04 &Mg12 & 8.421 & 2.10E-04 \\
821: Fe15 & 284.2 & 6.87E-04 & Fe22 & 101+17+36& 4.33E-04 &Fe26 & 9.652 & 6.60E-05 \\
822: He2 & 303.8 & 9.38E-04 & Fe20 & 121.8 & 3.03E-04 &Ne10 & 10.24 & 6.57E-05 \\
823: Si11 & 303.9 & 2.77E-04 & Fe21 & 128.0 & 6.14E-04 &Fe25 & 10.32 & 7.46E-07 \\
824: Fe13 & 316.0 & 8.33E-05 & Fe20 & 132.8 & 4.44E-04 &Fe24 & 10.63 & 1.05E-04 \\
825: Fe14 & 333.7 & 1.39E-04 & Fe23 & 132.8 & 3.64E-04 &Fe24 & 11.17 & 3.37E-04 \\
826: Fe16 & 335.4 & 1.52E-04 & Fe24 & 192.0 & 7.72E-05 &Ne10 & 12.13 & 2.81E-04 \\
827: Fe16 & 360.8 & 9.27E-05 & Fe24 & 255.1 & 3.90E-05 &O8 & 16.01 & 6.57E-05 \\
828: Mg9 & 368.1 & 5.32E-04 & He2 & 303.8 & 1.64E-04 &O8 & 18.97 & 2.81E-04 \\
829: Ne7 & 465.0 & 3.22E-04 & Ar16 & 389+ & 9.34E-05 &Ar18 & 20.20 & 2.34E-05 \\
830: Sil12 & 506+ & 3.05E-04 & S14 & 418+ & 1.58E-04 &N7 & 24.78 & 5.77E-05 \\
831: Si11 & 581.0 & 8.21E-05 & Si12 & 506+ & 7.30E-05 &S16 & 25.58 & 2.64E-05 \\
832: Mg10 & 615+23 & 1.04E-03 & Fe20 & 567.8 & 4.82E-05 &C6 & 28.47 & 2.41E-05 \\
833: O5 & 629.7 & 6.30E-05 & Fe20 & 721.4 & 9.57E-05 &Si14 & 33.42 & 3.38E-05 \\
834: Ne8 & 770+80 & 1.47E-03 & Fe22 & 845.4 & 8.52E-05 &C6 & 33.74 & 9.59E-05 \\
835: Ne7 &895.2 & 5.39E-05 & H1 & 1216 & 9.13E-05 &Mg12 & 45.51 & 1.81E-05 \\
836: O6 &1034 & 8.74E-04 & Fe21 & 1354 & 2.14E-04 &Ne10 & 65.56 & 2.34E-05 \\
837: H1 &1216 & 4.97E-04 & T/10$^5$K & & 6.99 &O8 & 102.5 & 5.89E-05 \\
838: Fe13 & 2579 & 6.12E-06 & & & &Fe23 & 132.8 & 1.42E-04 \\
839: Fe13 & 3388 & 5.60E-06 & & & &Fe24 & 192.0 & 3.09E-04 \\
840: Fe14 & 5303 & 9.21E-04 & & & &Fe24 & 255.1 & 1.38E-04 \\
841: S12 &7536 & 1.02E-04 & & & &He2 & 303.8 & 7.97E-05 \\
842: Fe13 & 10747 & 7.91E-05 & & & &T/10$^6$K& & 1.44 \\
843: Fe13 & 10798 & 5.22E-05 & & & & & &\\
844: Si10 & 14302 & 2.95E-04 & & & & & &\\
845: T/10$^4$K & & 12.2 & & & & & &\\
846: \tableline\noalign{\smallskip}
847: \end{tabular}
848: \end{center}
849: \end{table*}
850:
851: \begin{figure}
852: \plotone{f1.eps}
853: %\centerline{\epsfxsize=8cm\epsfbox{f1.eps}}
854: \caption{\footnotesize Plot demonstrating the difference between five and nine point
855: spline fits for the \ion{O}{iv} $^2P_{1/2}$ -- $^2P_{3/2}$ ground
856: transition. The original collision strength data are from
857: \citet{zhang94}. The five
858: point spline fit was optimized to the temperatures relevant to an
859: electron ionized plasma ($\sim 10^5$~K) and it fails to reproduce
860: the low-temperature region. The nine point spline gives a good fit to the low
861: temperature data points needed for photoionized plasmas.
862: \label{fig.chianti-o4}}
863: \end{figure}
864:
865:
866: \begin{figure}
867: \epsscale{1.0}
868: \plotone{f2.eps}
869: \caption{\footnotesize NLR benchmark; electron temperature, $T_{\rm e}
870: [K]$, and density, $n_{\rm e} [cm^{-3}]$, as a function of hydrogen
871: column density, $N_{\rm H} [cm^{-2}]$, through the slab.
872: \label{f:tene}
873: }
874: \end{figure}
875:
876:
877:
878:
879: \begin{figure}
880: \epsscale{1.0}
881: \plotone{f3.eps}
882: \caption{\footnotesize X-ray slab; U$_{\rm X}$=0.1 (X01). The red,
883: striped histograms illustrate {\sc mocassin's} results, while the
884: filled and empty histograms represent the lower and higher limits of the
885: range of predictions obtained by the codes in P\'equignot et al. (2001) .
886: \label{f:x01}
887: }
888: \end{figure}
889:
890: \begin{figure}
891: \epsscale{1.0}
892: \plotone{f4.eps}
893: \caption{\footnotesize X-ray slab; U$_{\rm X}$=1 (X1). The red,
894: striped histograms illustrate {\sc mocassin's} results, while the
895: filled and empty histograms represent the lower and higher limits of the
896: range of predictions obtained by the codes in P\'equignot et al. (2001).
897: \label{f:x1}
898: }
899: \end{figure}
900:
901: \begin{figure}
902: \epsscale{1.0}
903: \plotone{f5.eps}
904: \caption{\footnotesize X-ray slab; U$_{\rm X}$=10 (X10). The red,
905: striped histograms illustrate {\sc mocassin's} results, while the
906: filled and empty histograms represent the lower and higher limits of the
907: range of predictions obtained by the codes in P\'equignot et al. (2001) .
908: \label{f:x10}
909: }
910: \end{figure}
911:
912: \begin{thebibliography}{}
913:
914: \bibitem[Ali et al.(1991)]{1991PASP..103.1182A} Ali, B., Blum, R.~D.,
915: Bumgardner, T.~E., Cranmer, S.~R., Ferland, G.~J., Haefner, R.~I., \&
916: Tiede, G.~P.\ 1991, \pasp, 103, 1182
917:
918: \bibitem[Altun et al.(2004)]{2004AA...420..775A} Altun, Z., Yumak,
919: A., Badnell, N.~R., Colgan, J., \& Pindzola, M.~S.\ 2004, \aap, 420, 775
920:
921: \bibitem[Altun et al.(2006)]{2006AA...447.1165A} Altun, Z., Yumak,
922: A., Badnell, N.~R., Loch, S.~D., \& Pindzola, M.~S.\ 2006, \aap, 447, 1165
923:
924: \bibitem[Avrett \& Loeser(1988)]{1988ApJ...331..211A} Avrett, E.~H., \&
925: Loeser, R.\ 1988, \apj, 331, 211
926:
927: \bibitem[Badnell et al.(2003)]{2003AA...406.1151B} Badnell, N.~R., et
928: al.\ 2003, \aap, 406, 1151
929:
930: \bibitem[Badnell(2006)]{2006ApJS..167..334B} Badnell, N.~R.\ 2006a,
931: \apjs, 167, 334
932:
933: \bibitem[Badnell(2006)]{2006AA...447..389B} Badnell, N.~R.\ 2006b,
934: \aap, 447, 389
935:
936: \bibitem[Burgess \& Tully(1992)]{burgess92}
937: Burgess A., \& Tully J. A. 1992,
938: A\&A, 254, 436
939:
940: \bibitem[Colgan et al.(2003)]{2003AA...412..597C} Colgan, J.,
941: Pindzola, M.~S., Whiteford, A.~D., \& Badnell, N.~R.\ 2003, \aap, 412, 597
942:
943: \bibitem[Colgan et al.(2004)]{2004AA...417.1183C} Colgan, J.,
944: Pindzola, M.~S., \& Badnell, N.~R.\ 2004, \aap, 417, 1183
945:
946: \bibitem[Dere et al.(1997)]{dere97}
947: Dere, K. P., Landi, E., Mason, H. E., Monsignori-Fossi, B. F., \&
948: Young, P. R. 1997,
949: A\&AS, 125, 149
950:
951: \bibitem[Drake et al.(2007aa)]{sub1} Drake, J. J., Ercolano, B., Swartz,
952: D., 2007, \mnras, in press
953:
954: \bibitem[DrakeErcolano (2007a)]{sub2} Drake, J. J., Ercolano, B., 2007a,
955: \mnras, in press
956:
957: \bibitem[DrakeErcolano (2007b)]{sub2} Drake, J. J., Ercolano, B., 2007b,
958: \mnras, submitted
959:
960: \bibitem[Dumont et al.(2003)]{2003AA...407...13D} Dumont, A.-M., Collin,
961: S., Paletou, F., Coup{\'e}, S., Godet, O., \& Pelat, D.\ 2003, \aap, 407,
962: 13
963:
964: \bibitem[Dumont et al.(2000)]{2000AA...357..823D} Dumont, A.-M.,
965: Abrassart, A., \& Collin, S.\ 2000, \aap, 357, 823
966:
967: \bibitem[Ercolano et al.(2003a)]{2003MNRAS.340.1136E} Ercolano, B.,
968: Barlow, M.~J., Storey, P.~J., \& Liu, X.-W.\ 2003a, \mnras, 340, 1136
969:
970: \bibitem[Ercolano et al.(2003b)]{2003MNRAS.340.1153E} Ercolano, B.,
971: Morisset, C., Barlow, M.~J., Storey, P.~J., \& Liu, X.-W.\ 2003b, \mnras,
972: 340, 1153
973:
974: \bibitem[Ercolano et al.(2003c)]{2003MNRAS.344.1145E} Ercolano, B., Barlow,
975: M.~J., Storey, P.~J., Liu, X.-W., Rauch, T., \& Werner, K.\ 2003c, \mnras,
976: 344, 1145
977:
978: \bibitem[Ercolano et al.(2004)]{2004MNRAS.354..558E} Ercolano, B., Wesson,
979: R., Zhang, Y., Barlow, M.~J., De Marco, O., Rauch, T., \& Liu, X.-W.\ 2004,
980: \mnras, 354, 558
981:
982: \bibitem[Ercolano(2005)]{2005AIPC..804...35E} Ercolano, B.\ 2005, Planetary
983: Nebulae as Astronomical Tools, 804, 35
984:
985: \bibitem[Ercolano et al.(2005)]{2005MNRAS.362.1038E} Ercolano, B.,
986: Barlow, M.~J., \& Storey, P.~J.\ 2005, \mnras, 362, 1038
987:
988: \bibitem[Ercolano et al.(2007)]{2007arXiv0705.2726E} Ercolano, B., Bastian,
989: N., \& Stasinska, G.\ 2007, \mnras, in press, arXiv:0705.2726
990:
991: \bibitem[Ferland et al.(1998)]{1998PASP..110..761F} Ferland, G.~J.,
992: Korista, K.~T., Verner, D.~A., Ferguson, J.~W., Kingdon, J.~B., \&
993: Verner, E.~M.\ 1998, \pasp, 110, 761
994:
995: \bibitem[Ferland(2006)]{2006hbic.book.....F} Ferland, G.~J.\ 2006,
996: University of Kentucky Internal Report, 565 pages
997:
998: \bibitem[Gon{\c c}alves et al.(2006)]{2006MNRAS.365.1039G} Gon{\c c}alves,
999: D.~R., Ercolano, B., Carnero, A., Mampaso, A., \& Corradi, R.~L.~M.\ 2006,
1000: \mnras, 365, 1039
1001:
1002: \bibitem[Gu(2003)]{2003ApJ...582.1241G} Gu, M.~F.\ 2003, \apj, 582,
1003: 1241
1004:
1005: \bibitem[Kaastra \& Mewe(1993)]{1993AAS...97..443K} Kaastra, J.~S.,
1006: \& Mewe, R.\ 1993, \aaps, 97, 443
1007:
1008: \bibitem[Kallman \& McCray(1980)]{1980ApJ...242..615K} Kallman, T., \&
1009: McCray, R.\ 1980, \apj, 242, 615
1010:
1011: \bibitem[Kallman \& Bautista(2001)]{2001ApJS...133} Kallman, T., \&
1012: Bautista, M.\ 2001, \apj, 133, 221
1013:
1014: \bibitem[Landi et al.(2006)]{2006ApJS..162..261L} Landi, E., Del
1015: Zanna, G., Young, P.~R., Dere, K.~P., Mason, H.~E., \& Landini, M.\ 2006, \apjs,
1016: 162, 261
1017:
1018: \bibitem[Liedahl et al.(1990)]{1990ApJ...350L..37L} Liedahl, D.~A., Kahn,
1019: S.~M., Osterheld, A.~L., \& Goldstein, W.~H.\ 1990, \apjl, 350, L37
1020:
1021: \bibitem[Mauche et al.(2004)]{2004ApJ...606..168M} Mauche, C.~W., Liedahl,
1022: D.~A., Mathiesen, B.~F., Jimenez-Garate, M.~A., \& Raymond, J.~C.\ 2004,
1023: \apj, 606, 168
1024:
1025: \bibitem[Mazzotta et al.(1998)]{mazzotta98}
1026: Mazzotta, P., Mazzitelli, G., Colafrancesco, S., \& Vittorio,
1027: N. 1998,
1028: A\&AS, 133, 403
1029:
1030: \bibitem[Mewe et al.(1985)]{1985A&AS...62..197M} Mewe, R., Gronenschild,
1031: E.~H.~B.~M., \& van den Oord, G.~H.~J.\ 1985, \aaps, 62, 197
1032:
1033: \bibitem[Mitnik \& Badnell(2004)]{2004AA...425.1153M} Mitnik, D.~M.,
1034: \& Badnell, N.~R.\ 2004, \aap, 425, 1153
1035:
1036: \bibitem[Mukai et al.(2003)]{2003ApJ...594..428M} Mukai, K., Still, M., \&
1037: Ringwald, F.~A.\ 2003, \apj, 594, 428
1038:
1039: \bibitem[Nahar \& Pradhan(1994)]{1994PhRvA..49.1816N} Nahar, S.~N., \&
1040: Pradhan, A.~K.\ 1994, \pra, 49, 1816
1041:
1042: \bibitem[Nahar \& Pradhan(1995)]{1995ApJ...447..966N} Nahar, S.~N., \&
1043: Pradhan, A.~K.\ 1995, \apj, 447, 966
1044:
1045: \bibitem[Nahar \& Pradhan(1997)]{1997ApJS..111..339N} Nahar, S.~N., \&
1046: Pradhan, A.~K.\ 1997, \apjs, 111, 339
1047:
1048: \bibitem[Nahar(1999)]{1999ApJS..120..131N} Nahar, S.~N.\ 1999, \apjs,
1049: 120, 131
1050:
1051: \bibitem[Nahar(2000)]{2000ApJS..126..537N} Nahar, S.~N.\ 2000, \apjs,
1052: 126, 537
1053:
1054: \bibitem[Nayakshin et al.(2000)]{2000ApJ...537..833N} Nayakshin, S.,
1055: Kazanas, D., \& Kallman, T.~R.\ 2000, \apj, 537, 833
1056:
1057: \bibitem[Paerels et al.(2000)]{2000ApJ...533L.135P} Paerels, F., Cottam,
1058: J., Sako, M., Liedahl, D.~A., Brinkman, A.~C., van der Meer, R.~L.~J.,
1059: Kaastra, J.~S., \& Predehl, P.\ 2000, \apjl, 533, L135
1060:
1061: \bibitem[P{\'e}quignot et al.(2001)]{2001ASPC..247..533P}
1062: P{\'e}quignot, D., et al.\ 2001, Spectroscopic Challenges of Photoionized Plasmas,
1063: 247, 533
1064:
1065: \bibitem[Porquet \& Dubau(2000)]{2000AAS..143..495P} Porquet, D., \&
1066: Dubau, J.\ 2000, \aaps, 143, 495
1067:
1068: \bibitem[Rybicki \& Lightman(1986)]{1986rpa..book.....R} Rybicki,
1069: G.~B., \& Lightman, A.~P.\ 1986, Radiative Processes in Astrophysics, by George
1070: B.~Rybicki, Alan P.~Lightman, pp.~400.~ISBN 0-471-82759-2.~Wiley-VCH ,
1071: June 1986.,
1072:
1073: \bibitem[Ross \& Fabian(1993)]{1993MNRAS.261...74R} Ross, R.~R., \&
1074: Fabian, A.~C.\ 1993, \mnras, 261, 74
1075:
1076: \bibitem[Schwarz \& Monteiro(2006)]{2006ApJ...648..430S} Schwarz, H.~E., \&
1077: Monteiro, H.\ 2006, \apj, 648, 430
1078:
1079: \bibitem[Seaton et al.(1994)]{1994MNRAS.266..805S} Seaton, M.~J., Yan,
1080: Y., Mihalas, D., \& Pradhan, A.~K.\ 1994, \mnras, 266, 805
1081:
1082: \bibitem[Shull \& van Steenberg(1985)]{1985ApJ...298...268} Shull,
1083: J.~M., \& van Steenberg, M.\ 1985, \apj, 298, 268
1084:
1085: \bibitem[Storey \& Hummer(1995)]{1995MNRAS.272...41S} Storey, P.~J.,
1086: \& Hummer, D.~G.\ 1995, \mnras, 272, 41
1087:
1088: \bibitem[Verner et al.(1996)]{1996ApJ...465..487V} Verner, D.~A.,
1089: Ferland, G.~J., Korista, K.~T., \& Yakovlev, D.~G.\ 1996, \apj, 465, 487
1090:
1091: \bibitem[Voronov(1997)]{1997ADNDT..65....1V} Voronov, G.~S.\ 1997,
1092: Atomic Data and Nuclear Data Tables, 65, 1
1093:
1094: \bibitem[Wright et al.(2006)]{2006IAUS..234..545W} Wright, N., Ercolano,
1095: B., \& Barlow, M.~J.\ 2006, Planetary Nebulae in our Galaxy and Beyond,
1096: 234, 545
1097:
1098: \bibitem[Young et al.(2003)]{young03}
1099: Young, P. R., Del Zanna, G., Landi, E., et al. 2003,
1100: ApJS, 144, 135
1101:
1102: \bibitem[Zatsarinny et al.(2003)]{2003AA...412..587Z} Zatsarinny, O.,
1103: Gorczyca, T.~W., Korista, K.~T., Badnell, N.~R., \& Savin, D.~W.\
1104: 2003, \aap, 412, 587
1105:
1106: \bibitem[Zatsarinny et al.(2004)]{2004AA...426..699Z} Zatsarinny, O.,
1107: Gorczyca, T.~W., Korista, K., Badnell, N.~R., \& Savin, D.~W.\ 2004a,
1108: \aap, 426, 699
1109:
1110: \bibitem[Zatsarinny et al.(2004)]{2004AA...417.1173Z} Zatsarinny, O.,
1111: Gorczyca, T.~W., Korista, K.~T., Badnell, N.~R., \& Savin, D.~W.\
1112: 2004b, \aap, 417, 1173
1113:
1114: \bibitem[Zatsarinny et al.(2006)]{2006AA...447..379Z} Zatsarinny, O.,
1115: Gorczyca, T.~W., Fu, J., Korista, K.~T., Badnell, N.~R., \& Savin,
1116: D.~W.\ 2006, \aap, 447, 379
1117:
1118: \bibitem[Zhang et al.(1994)]{zhang94}
1119: Zhang, H. L., Graziani, M., \& Pradhan, A. K.,
1120: 1994, A\&A, 283, 319
1121:
1122:
1123: \end{thebibliography}
1124:
1125: \end{document}
1126:
1127: