0712.2203/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: %\slugcomment{Not to appear in Nonlearned J., 45.}
3: 
4: \shorttitle{Off-limb formation of the He {\sc i} 10830 \AA\ and D$_3$ multiplets}
5: \shortauthors{Centeno et al.}
6: 
7: \begin{document}
8: 
9: \title{The influence of coronal EUV irradiance on the emission in the He {\sc i} 10830 \AA\ and D$_3$ multiplets}
10: 
11: \author{R. Centeno\altaffilmark{1,2}, J. Trujillo Bueno\altaffilmark{1,3}, H. Uitenbroek\altaffilmark{4} and M. Collados\altaffilmark{1}}
12: 
13: \altaffiltext{1}{Instituto de Astrof\'{\i}sica de Canarias, 38205 La Laguna, Tenerife, Spain}
14: \altaffiltext{2}{High Altitude Observatory (NCAR), Boulder CO 80301, USA}
15: \altaffiltext{3}{Consejo Superior de Investigaciones Cient\'{\i}ficas (Spain)}
16: \altaffiltext{4}{National Solar Observatory, Sac. Peak, NM, USA}
17: \email{rce@ucar.edu, jtb@iac.es, huitenbr@nso.edu, mcv@iac.es}
18: 
19: 
20: \begin{abstract}
21: Two of the most attractive spectral windows for spectropolarimetric 
22: investigations of the physical properties of the plasma structures in the 
23: solar chromosphere and corona are the ones provided by the spectral 
24: lines of the He {\sc i} 10830 \AA\ and 5876 \AA\ (or D$_3$) multiplets, 
25: whose polarization signals are sensitive to the Hanle and Zeeman effects.
26: However, in order to be able to carry out reliable diagnostics, it is crucial 
27: to have a good physical understanding of the sensitivity of the observed 
28: spectral line radiation to the various competing driving mechanisms.
29: Here we report a series of off-the-limb non-LTE 
30: calculations of the He {\sc i} D$_3$ and  10830 \AA\ emission profiles, 
31: focusing our investigation on their sensitivity to the EUV coronal 
32: irradiation and the model atmosphere used in the calculations. 
33: We show in particular that the intensity ratio of the blue to the red 
34: components in
35: the emission profiles of the He {\sc i} 10830 \AA\ multiplet turns out to be 
36: a good candidate as a diagnostic 
37: tool for the coronal irradiance. Measurements of this observable as a function 
38: of the distance to the limb and its confrontation with radiative transfer 
39: modeling might give us valuable information on the physical properties of 
40: the solar atmosphere and on the amount of EUV radiation at relevant wavelengths
41: penetrating the chromosphere from above.
42: 
43: \end{abstract}
44: 
45: 
46: \keywords{Radiative transfer, line: formation, Sun: chromosphere, Sun: UV radiation}
47: 
48: 
49: 
50: \section{Introduction}
51: 
52: The spectral lines of the multiplet of neutral helium at 10830 \AA\ are of 
53: great diagnostic value for investigating the dynamic and magnetic properties 
54: of plasma structures in the solar chromosphere and corona. This is because 
55: their polarization is sensitive to the presence of atomic level polarization 
56: and to the joint action of the Hanle and Zeeman effects
57: (e.g. Trujillo Bueno \& Asensio Ramos, 2007), which makes these lines 
58: especially sensitive to the strongest fields found in solar active regions 
59: (e.g., Harvey \& Hall 1977; R\"uedi et al. 1995; Centeno et al. 2006) and 
60: also to the weaker fields encountered in a variety of plasma 
61: structures such as filaments (e.g., Lin et al. 1998; Trujillo 
62: Bueno et al. 2002), regions of emerging magnetic flux (e.g., Lagg et al. 
63: 2004), chromospheric spicules (e.g., Trujillo Bueno et al. 2005; 
64: Socas-Navarro et al. 2005) 
65: and prominences (e.g., Merenda et al. 2006). Obviously, in order to be able 
66: to obtain reliable inferences from the observed spectral line radiation it 
67: is necessary to reach a rigorous physical understanding of the key 
68: mechanisms that are responsible for its intensity and polarization. The 
69: same applies to the lines of the He {\sc i} D$_3$ multiplet at 5876 \AA\, 
70: whose observed polarization has been used for investigating 
71: the magnetic field vector in solar prominences and spicules (e.g., Landi 
72: Degl'Innocenti 1982; Leroy, Bommier \& Sahal-Br\'echot, 1983; Casini et al. 
73: 2003; L\'opez Ariste \& Casini 2005; Bianda et al. 2006; Ramelli et al. 
74: 2006 a,b).
75: 
76: While the modeling of the spectral line intensity has been done by solving 
77: the standard non-LTE radiation transfer problem in one-dimensional (1-D)
78: semi-empirical models of the solar atmosphere (e.g., Pozhalova 1988; 
79: Avrett et al. 1994; 
80: Andretta \& Jones 1997), the physical interpretation of the observed spectral 
81: line polarization has been carried out taking into account the joint 
82: action of the Hanle and Zeeman effects within the framework of the quantum
83: theory of polarization, but  simplifying the radiative 
84: transfer problem by using the optically thin approximation (e.g., Landi
85: Degl'Innocenti, 1982; Leroy et al, 1983;
86: Casini et al 2004) or by solving the Stokes-vector transfer equation in
87: a slab of constant physical properties (Trujillo Bueno et al. 2002; 2005). 
88: 
89: In this slab model for the interpretation of polarization 
90: observations, the slab's optical thickness is simply chosen as a free 
91: parameter to fit the observed intensity profile. This ``cloud" model 
92: for the interpretation of polarization observations simplifies 
93: the forward modeling and inversion problem (see Trujillo Bueno \& Asensio 
94: Ramos, 2007; Asensio Ramos \& Trujillo 
95: Bueno 2007), but it does not give us 
96: information on the possible errors we make by not considering explicitly the 
97: physical mechanisms that are thought to be responsible for producing some 
98: optical thickness in such lines of neutral helium. 
99: 
100: \noindent According to the 
101: above-mentioned non-LTE radiative transfer investigations in 1D 
102: semi-empirical models of the solar atmosphere there are two candidates 
103: for producing a significant optical thickness in the lines of the 
104: 10830 \AA\ and D$_3$ multiplets, given that the collisional excitations 
105: at typical solar chromospheric temperatures ($<$20000 K) are unable to 
106: significantly populate the triplet levels of He {\sc i}. The first 
107: mechanism states that the EUV radiation irradiating the chromosphere 
108: ionizes part of the substantial amount of neutral helium 
109: atoms that are found in the ground singlet state at the typical chromospheric 
110: temperatures, which then lead to recombinations that produce an enhanced 
111: population of the triplet states (e.g., Pozhalova 1988; Avrett et al. 1994). 
112: The second one 
113: deals with the possibility of collisional excitation by a significant amount of 
114: transition region material at temperatures greater than 20000 K (e.g., 
115: Andretta \& Jones 1997). In our opinion, as we shall try to show 
116: in this paper, detailed off-limb observations of the intensity and 
117: polarization profiles of the He {\sc i} 10830 \AA\ multiplet should 
118: help us determine which is the dominating mechanism, or if both 
119: are playing an equally significant role. 
120: 
121: The mechanism of collisional excitation at temperatures greater than 20000 K 
122: seems to require (or is helped by) the existence of a temperature ``plateau" 
123: located in the lower transition region (TR) of the Sun, such as that included 
124: ad-hoc by Vernazza et al. (1981) in their semi-empirical models in order to 
125: be able to model the observed Ly$_{\alpha}$ radiation. Interestingly enough, 
126: the consideration of ambipolar diffusion effecs in such semi-empirical 
127: models led to a new series of models by Fontenla et al. (1990; 1991; 1993) 
128: that are able to explain the observed Ly$_{\alpha}$ intensity observations 
129: without the need of such a TR ``plateau". In this new series of semi-empirical 
130: models (FAL models), an enhanced population in the triplet states 
131: of He {\sc i} can only be produced by the above mentioned 
132: Photoionization-Recombination process (hereafter: the PR mechanism). 
133: 
134: The main aim of this paper is to investigate the sensitivity of the 
135: off-the-limb emission profiles of the 10830 \AA\ and D$_3$ lines to the 
136: amount of coronal EUV radiation. To this end, we have chosen the 
137: above-mentioned FAL-like models and have solved the RT problem in 
138: spherical geometry for increasing values of the EUV coronal irradiance. 
139: As we shall see, this has allowed us to demonstrate that the intensity 
140: ratio of the blue and red components of the 10830 \AA\ multiplet is a 
141: sensitive function of the amount of coronal EUV illumination, and to 
142: establish this very observable as an ideal candidate for mapping 
143: the amount of coronal EUV irradiance at more or less active 
144: off-limb locations and for determining whether or not the PR mechanism 
145: is truly the dominating one. We also show the variation 
146: of this intensity ratio with the atmospheric height when using an atmospheric 
147: model with a TR ``plateau" in the model's temperature profile. In our 
148: opinion, the comparison of this type of simulations with observations 
149: should allow us to determine whether there is any off-limb region of the 
150: Sun (e.g., above plages?) where collisional excitation plays a 
151: significant role in addition to the PR mechanism.
152: 
153: 
154: The outline of this paper is as follows. The formulation of the problem 
155: is presented in Sect. 2, where we describe our approach to model the off-limb emission profiles 
156: of the aforementioned multiplets. Section 3 focuses on the sensitivity
157: of these profiles to the EUV coronal irradiation and points out the 
158: existence of an interesting observable (i.e., the above-mentioned
159: intensity ratio) that can serve as diagnostic tool
160: for this physical quantity. Finally, in section 4 we summarize the highlights
161: and discuss possible lines of future research.
162: 
163: \section{Formulation of the problem}
164: 
165: For understanding the formation of these helium lines
166: we carried out a series of non-LTE radiative transfer calculations of the 
167: off-limb intensity profiles emerging from various semi-empirical 
168: atmospheric models using a 1D spherically symmetric geometry version of 
169: the RH code (see {\em http://www.nso.edu/staff/uitenbr/rh.html}). 
170: This non-LTE radiative transfer code, based on the Multi-level Accelerated
171: Lambda Iteration (MALI) scheme by Rybicki \& Hummer (1991, 1992), solves
172: the combined equations of statistical equilibrium and radiative transfer
173: for a multilevel atom in a given stellar atmospheric model
174: (see also Socas-Navarro and Trujillo Bueno, 1997). The code
175: deals with overlapping transitions, and besides from the opacities and
176: emissivities produced by the transitions in the {\em active} atom 
177: (the one that is considered in full non-LTE regime)
178: it also accounts for background radiation due to other
179: atoms, molecules and relevant continuum processes.
180: 
181: 
182: \begin{figure}[!t]
183: \begin{center}
184: \includegraphics[scale=.50]{f1.eps}
185: \caption{Schematic atomic model of He {\sc i}, with 46 energy levels and
186: the transitions between each pair of levels. Although
187: the multiple transitions among triplet levels are not represented
188: in the figure, they were taken into account in the calculations. 
189: The individual He {\sc i} 
190: 10830 \AA\ and D$_3$ transitions are highlighted in the inset.}
191: \label{fig:atomic-model}
192: \end{center}
193: \end{figure}
194: 
195: 
196: The calculations of the emergent emission spectra were done in the 
197: context of the FAL model atmospheres (Fontenla et al, 1990,
198: 1991, 1993), which require the above-mentioned PR mechanism for producing
199: significant off-limb emission in the He {\sc i} 10830 \AA\ and $D_3$ 
200: multiplets. We have used these semi-empirical models as tabulated (i.e.
201: without considering any possible feedback effect of the EUV irradiance
202: on the models themselves). The main analysis is carried out with the 
203: FAL-C model of the 
204: average quiet Sun, although a comparison of the results among  
205: FAL-C, FAL-P (representative of a plage region) and the M$_{\rm CO}$ model
206: of Avrett (i.e., the relatively cool M$_{\rm CO}$ model atmosphere illustrative of 
207: the quiet Sun that Avrett (1995) proposed to account for the 
208: observed molecular CO absorption at 4.6 $\mu$m) is described at the end 
209: of section 3. In addition, we show also a comparison with the results
210: obtained in the VAL-C model of Vernazza et al. (1981), whose temperature
211: plateau (with $T>20000$ K) in its TR to the coronal gas, gives rise to a 
212: localized collisional enhancement of the triplet states populations.
213: 
214: \noindent The atomic model of He that we used 
215: (see Fig. \ref{fig:atomic-model}) accounts for 165 possible
216: bound-bound transitions taking place among 53 energy levels (46 of 
217: He {\sc i}, 6 of He {\sc ii} and 1 of He {\sc iii}). The fine structure
218: of the triplet energy levels was taken into account, together with 
219: all the possible combinations of transitions between them.
220: The radiative transition parameters were obtained from the 
221: NIST database\footnote{National Institute for Standards and Technology, see
222: {\em http://physics.nist.gov/PhysRefData/ASD/index.html}} and 
223: the collisional rate coefficients were calculated using the Seaton (1962) 
224: impact approximation for neutral bound-bound and the Van Regemorter's (1962) 
225: approximation for treating bound-bound transitions caused by collisions
226: with electrons.
227: Both photoionization and collisional ionization bound-free cross-sections are
228: also tabulated in the atomic model file and were extracted from 
229: Landolt-B\"ornstein (1982).
230: 
231: \begin{figure}[!t]
232: \begin{center}
233: \includegraphics[scale=.60]{f2.eps}
234: \caption{Ionization-recombination scheme for the He atom triggered by the EUV coronal irradiance (CI). In the absence of CI (panel 1) nearly all the population of He is in the ground state of the singlet system. The photoionization-recombination process (panels 2 and 3) is able to populate the triplet system significantly.}\label{fig:recombination}
235: \end{center}
236: \end{figure}
237: 
238: 
239: Despite its simplicity from the atomic point of view, the atom of Helium
240: shows quite a complicated spectrum with two different spectral series: 
241: one of singlets (Para-Helium) involving singlet terms ($^1S$, 
242: $^1P$, $^1D$..) and the other one of triplets (Ortho-Helium) involving
243: triplet terms ($^3S$, $^3P$, $^3D$..).
244: In the electric dipole approximation, quantum selection rules do not
245: allow radiative transitions between singlet and triplet terms, so the
246: populations of the two sets of levels are not radiatively coupled. For this
247: reason, the lower energy level in the triplet system is a metastable one.
248: At typical solar chromospheric conditions, the bulk of the Helium population is
249: mainly concentrated in the $^1$S atomic ground level since the temperature is
250: not high enough for populating the rest of the levels significantly.
251: Thus, the insufficent population of the triplet system cannot account for the 
252: observed He {\sc i} 10830 \AA\ and 5876 \AA\ features.
253: It has been pointed out that the properties of the He {\sc i} 10830 \AA\ and 
254: $D_3$ multiplets depend mainly on the density and thickness of the 
255: chromosphere as well as on the incoming 
256: EUV irradiation from the corona that incides on the chromosphere 
257: (see Pozhalova 1988; Avrett et al. 1994; and earlier references therein). 
258: This EUV spectrum with $\lambda < 504$ \AA\ ionizes 
259: Helium increasing the He {\sc ii} number density, which subsequently recombine
260: with the free electrons populating in this way both singlet and triplet systems
261: (Fig. \ref{fig:recombination} describes this process in 3 schematic steps).
262: The excess population in the triplet energy levels is responsible for 
263: strengthening the He {\sc i} 10830 \AA\ and $D_3$ features.
264: 
265: 
266: \begin{figure}[!t]
267: \begin{center}
268: \includegraphics[scale=.50]{f3.ps}
269: \caption{EUV flux incident on the chromosphere based on Tobiska (1991).}
270: \label{fig:tobiska}
271: \end{center}
272: \end{figure}
273: 
274: 
275: The inclusion of coronal irradiance incident on the chromosphere 
276: was implemented in the RH code using the EUV flux given
277: by Tobiska (1991). This author compiled data -provided by six different 
278: satellites-
279: of the EUV coronal irradiance that reached the Earth in the period
280: between 1962 and 1991. 
281: These data were used to evaluate the chromospheric illumination due to coronal
282: sources, which is shown in Fig. \ref{fig:tobiska}.
283: 
284: 
285: 
286: \section{Results of the RT modeling calculations}
287: 
288: Following the strategy described in the last section
289: we carried out a thorough analysis of the sensitivity of the He {\sc i}
290: 10830 \AA\ and D$_3$ emission profiles to the ionizing coronal irradiance.
291: We will show, in particular, how off-limb observations of the He {\sc i}
292: 10830 \AA\ multiplet can serve as a diagnostic tool for this EUV 
293: irradiance and discuss 
294: their use as a constraint for future atmospheric models.
295: 
296: \subsection{He {\sc i} 10830 \AA\ and D3 emission profiles}
297: 
298: In a first approach to understand the off-limb emission of the lines of the 
299: 10830 \AA\ and D$_3$ multiplets we compute the emergent profiles from the 
300: FAL-C model atmosphere.
301: In order to calculate the non-LTE populations of He {\sc i} and the emergent 
302: profiles we used a version of the RH code in 1D spherically symmetric geometry.
303: We solved consistently the RT in spherical geometry
304: and the SE equations to determine the populations at each height, and
305: computed the emergent intensity profiles on a set of rays tangent to
306: each spherical shell of the model atmosphere.
307: 
308: Fig. \ref{fig:offlimb} represents the He {\sc i} 10830 \AA\ (left) and
309: 5876 \AA\ (right) profiles with increasing distance to the base of the 
310: photosphere\footnote{The height reference
311: is set to be the level at which the optical depth -in the vertical direction- 
312: of the continuum at 
313: 500 nm equals unity, $\tau_{500}=1$} when the incident coronal
314: irradiation takes its nominal value (see Fig. \ref{fig:tobiska}). 
315: The lower height
316: corresponds to the first profile seen in emission off the limb.
317: In the outermost layers of the FAL-C model atmosphere, both multiplet
318: transitions produce very little emission. As the distance to the limb 
319: decreases, the amplitude of the emergent profiles grows until it reaches
320: a maximum around 2000 km. Below this point, the emission
321: starts decreasing slowly until it turns into absorption inside the limb.
322: The interpretation is as follows: the density in the outer layers of the 
323: atmosphere (at large distances from the limb) is so low that it cannot 
324: produce measurable emission profiles. As we go deeper into the atmosphere,
325: the density increases quickly and so does the emission in the multiplets,
326: until it reaches a maximum at a height of $\sim 2000$ km. The extinction
327: of the EUV radiation as it travels inwards through the chromosphere 
328: leads to a reduction in the number of ionizations in the inner layers. Thus,
329: the emission in the spectral lines of the He {\sc i} 10830 \AA\ and D$_3$ 
330: multiplets starts
331: decreasing again because the PR mechanism cannot sustain the populations 
332: of the triplet system.
333: 
334: 
335: \begin{figure}[t!]
336: \centering
337: \includegraphics[width=7cm]{f4a.ps}
338: \includegraphics[width=7cm]{f4b.ps}
339: \protect\caption[ ]{Off-the-limb He {\sc i} 10830 \AA\ (left) and $D_3$ 
340: 5876 \AA\ (right) emission profiles at different distances to the limb
341: of the FAL-C semi-empirical model, when the nominal coronal irradiance
342: has been applied to the system. 
343: Heights are given relative to the reference $\tau_{500} = 1$. 
344: \label{fig:offlimb}}
345: \end{figure}
346: 
347: \subsection{Dependence with coronal irradiance}
348: 
349: 
350: One of the purposes of this investigation is to study the sensitivity
351: of the emission profiles to the amount of coronal irradiance (CI).
352: 
353: The top panels of Fig. \ref{fig:emission-profiles} show the 10830 \AA\ (left)
354: and $D_3$ (right) emergent profiles at the height of maximum emission (this 
355: is, 2023 km above $\tau_{500}=1$) for four different values of the CI
356: (0, 2, 5, and 10 times the nominal EUV spectrum). The tendency is 
357: clear:
358: the stronger the CI, the greater the emission. As we shall see below,
359: this is just a consequence of the increase in the population of the triplet
360: levels with the CI. 
361: The lower panels of Fig. \ref{fig:emission-profiles}
362: represent the same intensity profiles when each of them is normalized to
363: its maximum emission. The feature that strikes our attention most is the 
364: strong sensitivity of the relative amplitude of the blue component ($I_B$) of 
365: the He {\sc i} 10830 \AA\ multiplet to the strength of the incoming CI
366: (lower left panel of Fig. \ref{fig:emission-profiles}). The 
367: ratio $\cal R$ $= I_B / I_R$ of the amplitudes of the
368: blue to red components of the multiplet changes substantially with the
369: amount of ionizing radiation. This is an extremely interesting feature
370: since it could provide a useful diagnostic tool for the incoming coronal 
371: irradiance in off-limb observations of the He {\sc i} triplet. On the 
372: other hand, $D_3$ shows no observable relative changes among its spectral 
373: components in response to the ionizing radiaton. We will see further on
374: that this is just a natural consequence of the D$_3$ line not reaching 
375: enough optical thickness in these models to show a
376: differential saturation behavior.
377: 
378: 
379: In order to understand the changes in the 10830 profiles shown in
380: Fig. \ref{fig:emission-profiles} let us see how the atomic populations
381: behave in the absence/presence of the EUV ionizing spectrum.
382: Fig. \ref{fig:populations} shows the populations of the four $J$ levels 
383: involved in the 10830 \AA\ transitions ($^3S_1$, $^3P_2$, $^3P_1$ and 
384: $^3P_0$
385: from left to right and top to bottom) as a function of height in the
386: atmosphere, for five different values of the incoming CI (0, 1, 2, 5, and 10 
387: times the nominal EUV flux).
388: These plots clearly illustrate how the sole presence of CI dramatically
389: changes the populations of these levels. Furthermore, the stronger the
390: CI, the bigger this change. Note, however, that this sensitivity of 
391: the populations to the CI is only effective starting at a certain height
392: ($\sim 1000$ km) and that it keeps growing until it reaches the sharp
393: density decrease that characterizes the transition region in the FAL-C
394: model atmosphere. Below $1000$ km, the extinction of the CI is such that it 
395: has no effect on the populations. It is interesting to notice that
396: non-LTE effects play a significant role even in the absence of any coronal
397: irradiance (compare the solid lines with the LTE results), but the
398: ensuing overpopulation of the triplet levels above $\sim 1500$ km is
399: insufficient to produce any significant optical thickness in the 10830 \AA\
400: lines.
401: 
402: 
403: \begin{figure}[t!]
404: \centering
405: \includegraphics[width=7cm]{f5a.ps}
406: \includegraphics[width=7cm]{f5b.ps}\\
407: \includegraphics[width=7cm]{f5c.ps}
408: \includegraphics[width=7cm]{f5d.ps}
409: \protect\caption[ ]{{\em Top:} Emission intensity profiles of He {\sc i} 
410: 10830 \AA\ ({\em left}) and $D_3$ ({\em right}) at a height of 2023 km above
411: the $\tau_{500}=1$ level for 4 different values of the incident coronal 
412: irradiance on the FAL-C atmospheric model. {\em Bottom:} Same profiles but 
413: normalized to their maximum.
414: Note the sensitivity of the relative amplitudes of the blue and red components
415: of He 10830 \AA\ to the incoming CI.
416: \label{fig:emission-profiles}}
417: \end{figure}
418: 
419: \begin{figure}[t!]
420: \centering
421: \includegraphics[width=7cm]{f6a.ps}
422: \includegraphics[width=7cm]{f6b.ps}\\
423: \includegraphics[width=7cm]{f6c.ps}
424: \includegraphics[width=7cm]{f6d.ps}
425: \protect\caption[ ]{Populations of the energy levels involved in the 
426: 10830 \AA\ transitions (from left to right and top to bottom: $^3S$, $^3P_2$, 
427: $^3P_1$ and $^3P_0$) as a function of height in the FAL-C model
428: atmosphere (the reference
429: level is $\tau_{500}=1$). The populations were computed for 5 different
430: values of the CI (0, 1, 2, 5 and 10 times its nominal value). Note that
431: the sensitivity to the ionizing radiation starts at 1000 km above the base
432: of the photosphere. The solid line with thick dots represents the 
433: populations in LTE, that, by definition, do not change with the coronal 
434: irradiance.
435: \label{fig:populations}}
436: \end{figure}
437: 
438: \begin{figure}[t!]
439: \centering
440: \includegraphics[width=7cm]{f7a.ps}
441: \includegraphics[width=7cm]{f7b.ps}
442: \protect\caption[ ]{{\em Left}: Number of photoionizations in the FAL-C 
443: model atmosphere as a function of height. Black and grey show, respectively, 
444: the photoionizations from the ground and the metastable levels of the 
445: He {\sc i} atom for 4 different values of the coronal irradiance.
446: {\em Right}: Number of bound-bound transitions from the metastable level
447: $2^3S$ to the three upper levels of the 10830 \AA\ multiplet (represented with
448: different line thickness) as a function of height and for 0 ({\em solid}) and
449: 5 ({\em dashed}) times the nominal coronal irradiance.  
450: \label{fig:photoionizations}}
451: \end{figure}
452: 
453: 
454: The left panel of Figure \ref{fig:photoionizations} shows the number of
455: photoionizations per unit volume and time for increasing values of the 
456: coronal irradiance in the FAL-C model. Notice
457: that the number of photoionizations that take place from the ground level 
458: decreases with height, proportionally to the population of this level
459: - which follows the monotonical decrease of the atmospheric density profile.
460: On the contrary, the amount of photoionizations taking place from the 
461: metastable level increases with height, due to the fact that the population 
462: of this level increases with height between about 600 and 2100 km, even
463: in the LTE case (see Figure \ref{fig:populations}).
464:  The effect of the CI is such that it increases
465: the population of the triplet levels at the expense of the ground level. This
466: explains why, while the number of photoionizations from the metastable
467: level grows with increasing CI, the number of photoionizations of
468: the ground level becomes slightly smaller (due to the decrease in ground level
469: population).
470: 
471: \noindent It is of some interest to point out that even with 10 times the 
472: nominal coronal 
473: value of the EUV flux, the number of photoionizations that take place from
474: the ground level is ten orders of magnitude larger than that 
475: taking place from the metastable $^3S_1$ level. Even more interesting
476: is to keep in mind what the right panel of Figure \ref{fig:photoionizations}
477: illustrates - this is, that the number of bound-bound 10830 \AA\ transitions
478: per unit volume and time is way larger than the number of photoionizations 
479: from the same $^3S_1$ level. The same happens concerning the number of $D_3$
480: line transitions when compared to the number of photoionizations from their
481: lower and upper J-levels.
482: 
483: 
484: \subsection{Interpretation in terms of the optical thickness of the line formation region}
485: 
486: In a recent investigation, Trujillo Bueno et al. (2005) analyzed
487: a set of spectropolarimetric observations of quiet Sun chromospheric
488: spicules at $2''.5$ off the
489: visible limb in the He {\sc i} 10830 \AA\ multiplet. In their first attempt, 
490: the authors tried to fit the emergent Stokes profiles ignoring the effects 
491: of radiative transfer (i.e., using the optically thin approximation). 
492: However, they found 
493: that they could not reproduce the emission of the blue component of the 
494: multiplet under this assumption. In a more realistic approach they were able 
495: to fit the emission profiles assuming a plane-parallel slab of constant
496: physical properties and optical thickness $\tau$, that accounts for the 
497: collective effect of having several spicules interposed along the line of sight.
498: With a value of $\tau_{R}=3$ (of the optical thickness at the line center 
499: of the red blended component) the authors were able to reproduce the 
500: intensity ratio $\cal R$ of the blue and red components of the multiplet, 
501: proving that radiative transfer effects are non-negligible.
502: 
503: \noindent The solid line in the right panel of Fig. \ref{fig:ratio-vs-tau} 
504: illustrates
505: how the optical thickness of a slab of constant properties modifies the 
506: emission ratio $\cal R$ of the two components of the 10830 \AA\ multiplet. 
507: In the optically thin regime ($\tau_{R} < 1$), $\cal R$$=I_B/I_R$
508: takes a value around $\sim 0.12$, which corresponds to the nominal ratio
509: of oscillator strengths. For larger values of $\tau_{R}$, this ratio $\cal R$
510: increases with optical thickness until it reaches a saturation value of
511: $\sim 1$ for $\tau_{R}\sim 30$.
512: In this simple modeling, which is similar to that of Trujillo Bueno
513: et al. (2005), the ratio $\cal R$ is a function of $\tau_{R}$
514: in a wide range of optical thickness values. But what is the meaning
515: and physical origin of this parametrization? Qualitatively, the relation
516: between $\tau$ and the EUV coronal irradiation is easy to
517: understand. The ionizing radiation modifies the population of the metastable
518: level $2^3S$, which in time is responsible for the optical thickness of the
519: transition. But quantitatively, is this $\tau_{R}$ value chosen ad hoc
520: by Trujillo Bueno et al (2005) to fit the profiles representative of the
521: physical conditions of the solar corona and of the real integrated optical
522: thickness along the line of sight?
523: 
524: 
525: To address this question we synthesized the He {\sc i} 10830 \AA\ profiles
526: at a height of 2023 km above the base of the photosphere in the FAL-C
527: model atmosphere, for different values of the incident coronal irradiance. 
528: Then, we computed the integrated optical depths along the 
529: line-of-sight (LOS) at the line center 
530: wavelength of the red blended component of the multiplet for each case. 
531: The left panel of Fig. \ref{fig:ratio-vs-tau} relates the measured
532: ratio $\cal R$ to the amount of inward 
533: ionizing radiation for a fixed distance to the limb. As expected, the stronger
534: the CI, the greater the value of $\cal R$, although this relation
535: is far from being linear and tends to saturate for very large values of
536: the coronal irradiance.
537: Asterisks superposed to the right panel of Fig. \ref{fig:ratio-vs-tau} show 
538: these computed values of $\cal R$ as a function of the integrated optical 
539: depth along the line-of-sight. The excellent agreement with the 
540: constant-property slab modeling of Trujillo Bueno et al (2005) evidences that 
541: the ratio $\cal R$ depends only on the integrated $\tau$ along the LOS, rather
542: than on the spatial distribution of opacities. 
543: 
544: When $\cal R$ achieves values larger than the ratio of the oscillator 
545: strengths it means that the transitions are not in the optically thin regime
546: anymore. In Figure \ref{fig:emission-profiles} the reader can see that, while
547: the normalized emission intensity profiles of the He {\sc i} 10830 \AA\ 
548: multiplet show a strong sensitivity to the
549: incoming CI, He D$_3$, on the other hand, does not. This is a consequence of
550: the distribution of populations in the triplet system of the He atom. The
551: lower energy level of the 10830 \AA\ transition is metastable, thus implying 
552: that its population is orders of magnitude larger than that of the rest 
553: of the triplet levels (see Figure \ref{fig:populations}). This produces 
554: a larger opacity in the 10830 \AA\ multiplet that in turn affects the 
555: ratio $\cal R$. This is an interesting feature because, given the density 
556: profile of the atmosphere, this ratio should be a reliable tool for 
557: diagnosing the ionizing coronal irradiation, and viceversa.
558: 
559: 
560: \begin{figure}[t!]
561: \centering
562: \includegraphics[width=7cm]{f8a.ps}
563: \includegraphics[width=7cm]{f8b.ps}
564: \protect\caption[ ]{Influence of the coronal irradiance on the optical 
565: thickness and the ratio $\cal R$ in the FAL-C
566: model atmosphere. {\em Left}: $\cal R$ as a function of 
567: the number of times the nominal CI that was applied in the calculations,
568: for the emergent profiles at 2023 km above $\tau_{500}=1$.
569: {\em Right}: The solid line represents $\cal R$ as a function of the 
570: optical depth at the line center of the red component of the multiplet as
571: predicted by the optically thick slab modeling of Trujillo Bueno et al. (2005).
572: Asterisks superposed to the solid line are the values obtained from the
573: profiles simulated in the FAL-C model at the 
574: aforementioned height.\label{fig:ratio-vs-tau}}
575: \end{figure}
576: 
577: 
578: 
579: \subsection{Change of $\cal R$ with height in various model atmospheres}
580: 
581: 
582: \begin{figure}[t!]
583: \centering
584: \includegraphics[width=7cm]{f9a.ps}
585: \includegraphics[width=7cm]{f9b.ps}
586: \includegraphics[width=7cm]{f9c.ps}
587: \includegraphics[width=7cm]{f9d.ps}
588: \protect\caption[ ]{Normalized off-the-limb He {\sc i} 10830 \AA\ intensity 
589: profiles computed in FAL-C model atmosphere at different distances to the
590: limb (from left to right and top to bottom, 2137 km, 2023 km, 1444 km and 
591: 698 km above $\tau_{500}=1$).  Each panel shows the intensity spectra for
592: 0, 2, 5, and 10 times the nominal coronal irradiance). 
593: \label{fig:intensity-vs-distance}}
594: \end{figure}
595: 
596: 
597: The population of the metastable level changes with height, thus, the 
598: optical thickness in the 10830 \AA\ transitions (and the relative
599: amplitudes of the blue-to-red components of the multiplet, $\cal R$) should 
600: behave accordingly.
601: 
602: Fig. \ref{fig:intensity-vs-distance} illustrates the normalized emergent 
603: intensity profiles in the FAL-C model atmosphere at different 
604: heights above the limb
605: (from left to right and top to bottom: 2137 km, 2023 km, 1444 km and 698 km 
606: above $\tau_{500}=1$). Each panel shows the emission spectra for four values 
607: of the CI (0, 2, 5 and 10 times its nominal value). This figure proves how
608: $\cal R$ changes, not only with the amount of CI inciding on the chromosphere,
609: but also with height in the atmosphere. This is an obvious consequence 
610: of the dependence of $\cal R$ with the integrated optical depth along the
611: ray path, and thus with the density.
612: The top left panel of Fig. \ref{fig:intensity-vs-distance} corresponds to
613: the emergent profiles from one of the outer-most layers of the FAL-C model. 
614: The three components of the multiplet appear completely overlapped due to the
615: large values of the microturbulent velocity ($\sim 8 - 10$ kms$^{-1}$) at
616: these heights. The widening effect that this contribution produces on the
617: profiles is larger than the separation between the spectral components.
618: The closer we get to the limb, the smaller the microturbulence and the
619: easier it becomes to distinguish the red and blue components of the triplet.
620: In the FAL-C model, within this range of CIs, the maximum emission (and
621: ratio $\cal R$) is found at $\sim 2023$ km above $\tau_{500}=1$, although
622: in general, the stronger the CI, the lower in the atmosphere this would
623: take place.
624: 
625: 
626: \begin{figure}[t!]
627: \centering
628: \includegraphics[width=7cm]{f10a.ps}
629: \includegraphics[width=7cm]{f10b.ps}
630: \protect\caption[ ]{Comparison in 4 model atmospheres: FAL-C, FAL-P, 
631: VAL-C and the M$_{CO}$ model of Avrett (1995).
632: {\em Left}: Ratio $\cal R$ of the blue-to-red components of the emergent
633: He {\sc i} 10830 \AA\ intensity profiles as a function of height in the 
634: atmosphere when the applied CI takes 5 times its nominal value. 
635: {\em Right}: Hydrogen number density and temperature profiles as a 
636: function of height in the four model atmospheres. 
637: \label{fig:r-with-height}}
638: \end{figure}
639: 
640: 
641: The top left panel of Fig. \ref{fig:r-with-height} shows the ratio $\cal R$ 
642: as a function of height in four different standard model atmospheres 
643: (FAL-C, FAL-P, VAL-C and the M$_{CO}$ model of Avrett 1995) when the 
644: applied CI flux takes five times its nominal value.
645: The amplitudes of the blue and red components were extracted from a double 
646: gaussian fitting to the emergent intensity spectra. The higher layers of the
647: atmosphere, where the overlap between the two components due to the 
648: microturbulence did not allow an unambiguous determination of $\cal R$, 
649: were excluded from the calculation (this is, above 1700 km in FAL-P and 
650: over 2100 km in VAL-C, FAL-C and the M$_{CO}$ model of Avrett 1995).
651: The very different behavior of the four curves is a consequence of the 
652: differences between the density profiles and the vertical extent of 
653: the model atmospheres. 
654: The right panel of Fig. \ref{fig:r-with-height} illustrates the density 
655: ({\em gray}) and temperature ({\em black}) stratifications in these four 
656: atmospheres. While FAL-P has a lower transition region (TR) boundary
657: and a shallower chromosphere, all the other model atmospheres extend much 
658: higher and
659: have thicker chromospheres. The lower height of the TR boundary in the 
660: FAL-P atmosphere combined with the effect of the strong ionizing radiation 
661: coming downwards from the corona, shifts the region of formation of the He triplet 
662: towards lower heights in this model. Higher up, the medium becomes 
663: too rarified to produce enough opacity at the 10830 \AA\ wavelength. 
664: The density profile of the M$_{CO}$ model of Avrett (1995) in the transition region is not 
665: as steep as in FAL-C and FAL-P. For this reason, $\cal R$ does not decrease 
666: as fast in the highest layers of the atmosphere.
667: 
668: \noindent The case for VAL-C is somewhat different. The presence of a 
669: temperature ``plateau'' at a height of 2100 km that defines a very thick transition region (of a 
670: couple hundred km) produces an increment of the population of the metastable
671: level in this layer, even with no help from the ionizing radiation (cf. 
672: Andretta \& Jones 1997).
673: This is easy to see in Figure \ref{fig:r-with-height-0ci} where we 
674: represent the behavior of $\cal R$ as a 
675: function of height in the four model atmospheres for the case of no coronal 
676: irradiance inciding on the chromosphere. The temperature plateau and
677: the thickness of the hot TR of VAL-C make possible a significant number of
678: collisional excitations that lead to an over-population of 
679: the triplet system enough to produce a non-negligible optical thickness in 
680: the 10830 \AA\ transition, but only for off-limb line-of-sights
681: around 2100 km. Also FAL-P shows a very localized peak
682: produced by the sudden increase in the temperature stratification at
683: 1700 km above the level $\tau_{500}=1$.
684: 
685: 
686: \begin{figure}[t!]
687: \centering
688: \includegraphics[width=7cm]{f11.ps}
689: \protect\caption[ ]{Comparison in 4 model atmospheres (FAL-C, FAL-P,
690: VAL-C and the M$_{CO}$ model of Avrett 1995) of the ratio $\cal R$ as a 
691: function of height for the case with
692: no coronal irradiance inciding on the chromosphere. Note the effect of
693: the temperature ``plateau'' of the VAL-C model in the region between 
694: 1900 and 2100 km. Also FAL-P shows a very localized peak produced by
695: its sudden temperature increase at 1700 km above $\tau_{500}=1$.
696: \label{fig:r-with-height-0ci}}
697: \end{figure}
698: 
699: 
700: \section{Concluding comment: a diagnostic tool of EUV coronal irradiance}
701: 
702: 
703: Figure \ref{fig:r-with-height} shows the change of $\cal R$ with height
704: in four different model atmospheres. This variation is a very interesting 
705: feature since it sets a constraint on the thermodynamical structure of
706: the chromosphere and the amount of ionizing coronal irradiance inciding on it. 
707: $\cal R$ is an observable than can easily be determined
708: doing spectroscopic measurements of the He {\sc i} 10830 \AA\ multiplet.
709: In fact, S\'anchez-Andrade Nu\~no et al. (2007) made use of
710: this theoretical prediction and have recently presented a series of
711: observations of off-the-limb He {\sc i} 10830 \AA\ 
712: spectropolarimetric profiles aimed at the determination of $\cal R$. When
713: comparing the observations with our radiative transfer modeling, they find 
714: that the theoretical behavior
715: of the ratio $\cal R$ agrees qualitatively with the observed one, although 
716: a quantitative comparison shows poor agreement. 
717: The density stratification and the limited vertical extent of the 
718: semiempirical model atmospheres considered in this paper are not adequate 
719: for spicule modeling. 
720: The fact that the solar chromosphere is very inhomogeneous on many scales and
721: that spicules are small-scale intrusions of chromospheric matter into the 
722: hot corona, make it difficult for 1D models to reproduce off-limb 
723: observations.
724: 
725: 
726: 
727: The next step in this study should be to include in our 
728: RT modeling the spectral line polarization caused by the joint
729: action of the Hanle and Zeeman effects, as done by Trujillo Bueno \&
730: Asensio Ramos (2007) within the framework of the above-mentioned slab
731: model. New spectropolarimetric observations, at different positions above
732: the solar limb with different levels of ionizing radiation (i.e. active 
733: regions and coronal holes) should be compared 
734: to the theoretical calculations to help constrain the atmospheric models
735: and the physical processes that induce the formation of the He {\sc i}
736: 10830 \AA\ and D$_3$ multiplets.
737: 
738: 
739: 
740: \acknowledgments
741: 
742: This research was partially funded by the Spanish Ministerio de Educaci\'on 
743: y Ciencia through project AYA2007-63881.
744: The National Center for Atmospheric Research (NCAR) is sponsored by the 
745: National Science Foundation.
746: 
747: 
748: \begin{thebibliography}{}
749: \bibitem[Andretta \& Jones, 1997]{andretta07} Andretta, V., Jones, H.P., 1997, ApJ, {\bf 489}, 375.
750: \bibitem[Asensio Ramos \& Trujillo Bueno, 2007]{asensio07} Asensio Ramos, A., Trujillo Bueno, J., 2007, in The Physics of Chromospheric Plasmas, edited by P. Heinzel, I. Dorotovic and J. Rutten, ASP Conf. Series Vol {\bf 368}, 163.
751: \bibitem[Avrett et al, 1994]{avrett94} Avrett, E.H., Fontenla, J.M., Loeser, R., 1994, IAUS, {\bf 154}, 35.
752: \bibitem[Avrett, 1995]{avrett95} Avrett, E.H., 1995, in ``Infrared tools for Solar Astrophysics: what's next'', ed. J.R. Kuhn and M.J. Penn (Singapore: World Scientific), 303.
753: \bibitem[Bianda et al, 2005]{bianda05} Bianda, M., Ramelli, R., Trujillo Bueno, J., Stenflo, J., 2006, in ``Solar Polarization Workshop 4'', eds. R. Casini and B.W. Lites; ASP Conf. Series Vol {\bf 358}, 454.
754: \bibitem[Casini et al, 2004]{casini04} Casini, R., L\'opez Ariste, A., Tomczyk, S., Lites, B., 2003, ApJ, {\bf 598}, L67. 
755: \bibitem[Centeno et al, 2006]{centeno06} Centeno, R., Collados, M., Trujillo Bueno, J., 2006, ApJ, {\bf 640}, 1153.
756: 
757: \bibitem[Fontenla et al, 1990]{fontenla90} Fontenla, J.M., Avrett, E.H., Loeser,R., 1990, ApJ, {\bf 355}, 700.
758: \bibitem[Fontenla et al, 1992]{fontenla92} Fontenla, J.M., Avrett, E.H., Loeser,R., 1991, ApJ, {\bf 377}, 712.
759: \bibitem[Fontenla et al, 1993]{fontenla93} Fontenla, J.M., Avrett, E.H., Loeser,R., 1993, ApJ, {\bf 406}, 319.
760: \bibitem[Lagg et al 2004]{lagg04} Lagg, A., Woch, J., Krupp, N., Solanki, S.K., 2004, A\&A, {\bf 414}, 1109.
761: \bibitem[Landi Degl'Innocenti, 1982]{landi82} Landi Degl'Innocenti, E., 1982, Solar Physics, {\bf 79}, 29.
762: 
763: \bibitem[Landolt-B\"ornstein, 1982]{lb82} Landolt-B\"ornstein, Volume 2, Astronomy and Astrophysics, subvolume b, Stars and Star clusters, 1982, Springer-Verlag, (98-100).
764: \bibitem[Leroy et al, 1983]{leroy83} Leroy, J.L., Bommier, V., Sahal-Br\'echot, S, 1983, Solar Physics, {\bf 83}, 135.
765: \bibitem[Lin et al, 1998]{lin98} Lin, H., Penn, M.J., Kuhn, J.R., 1998, ApJ, {\bf 493}, 978.
766: \bibitem[L\'opez Ariste et al, 2005]{ariste05} L\'opez Ariste, A., Casini, R., 2005, A\&A, {\bf 436}, 325.
767: \bibitem[Merenda et al, 2005]{merenda05} Merenda, L., Trujillo Bueno, J., Landi Degl'Innocenti, E., Collados, M., 2005, ApJ, {\bf 642}, 544.
768: 
769: \bibitem[Pozhalova, 1988]{pozhal98} Pozhalova, zh. A., 1988, Sov. Astron., 65(5).
770: \bibitem[Ramelli et al, 2006a]{ramelli06a} Ramelli, R., Bianda, M., Merenda, L., Trujillo Bueno, J., 2006a, in ``Solar Polarization Workshop 4'', eds. R. Casini and B.W. Lites, ASP Conf. Series Vol. {\bf 358}, 448.
771: \bibitem[Ramelli et al, 2006b]{ramelli06b} Ramelli, R., Bianda, M., Trujillo Bueno, J., Merenda, L., Stenflo, J.O., 2006b, in ``Solar Polarization Workshop 4'', eds. R. Casini and B.W. Lites, ASP Conf. Series Vol. {\bf 358}, 471. 
772: \bibitem[R\"uedi et al, 1995]{ruedi95} R\"uedi, I., Solanki, S.K., Livingston, W., 1995, A\&A, {\bf 293}, 252.
773: 
774: \bibitem[Rybicki \& Hummer, 1991]{rybicki91} Rybicki, G.B., Hummer, D.G., 1991, A\&A, {\bf 245}, 171.
775: \bibitem[Rybicki \& Hummer, 1992]{rybicki92} Rybicki, G.B., Hummer, D.G., 1992, A\&A, {\bf 262}, 209.
776: \bibitem[Tobiska, 1991]{tobiska91} Tobiska, W. Kent, 1991, Journal of Atmospheric and Terrestrial Physics, {\bf 53}, 1005.
777: \bibitem[S\'anchez-Andrade Nu\~no et al, 2007]{bruno07} S\'anchez-Andrade Nu\~no, B., Centeno, R., Puschmann, K.G., Trujillo Bueno, J., Blanco Rodr\'\i guez, J., Kneer, F., 2007, A\&A Letters, in press.
778: 
779: \bibitem[Seaton, 1962]{seaton62} Seaton, M.J., 1962, Proc. Phys. Soc. London, {\bf 79}, 1105.
780: \bibitem[Socas-Navarro et al, 2005]{sn05} Socas-Navarro, H., Elmore, D., 2005, ApJ, {\bf 619}, 195.
781: \bibitem[Socas-Navarro et al, 1997]{sn97} Socas-Navarro, H., Trujillo Bueno, J., 1997, ApJ, {\bf 490}, 383.
782: \bibitem[Trujillo Bueno \& Asensio Ramos, 2007]{jtb07} Trujillo Bueno, J., Asensio Ramos, A., 2007, ApJ, {\bf 655}, 642.
783: \bibitem[Trujillo Bueno et al, 2002]{jtb02} Trujillo Bueno, J., Landi Degl'Innocenti, E., Merenda, L., Collados, M., Manso Sainz, R., 2002, Nature, {\bf 415}, 403.
784: \bibitem[Trujillo Bueno et al, 2005]{tb05} Trujillo Bueno, J., Merenda, L., Centeno, R., Collados, M., Landi Degl'Innocenti, E., 2005, ApJ, {\bf 619}, 191.
785: 
786: \bibitem[Van Tegemorter, 1962]{regemorter62} Van Regemorter, H., 1962, ApJ, {\bf 136}, 906.
787: \bibitem[Vernazza, Avrett \& L\"oser, 1981]{val81} Vernazza, J.E., Avrett, E.H., L\"oser, R., 1981, ApJS, {\bf 45}, 635.
788: 
789: \end{thebibliography}
790: 
791: 
792: 
793: \end{document}
794: