Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.tesis.lebedev.ru/en/docs/1373.pdf Äàòà èçìåíåíèÿ: Mon May 26 00:00:00 2008 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:10:00 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: earth's atmosphere

ISSN 1063-7737, Astronomy Letters, 2008, Vol. 34, No. 1, pp. 33­51. c Pleiades Publishing, Inc., 2008. Original Russian Text c S.V. Shestov, S.A. Bozhenkov, I.A. Zhitnik, S.V. Kuzin, A.M. Urnov, I.L. Beigman, F.F. Goryayev, I.Yu. Tolstikhina, 2008, published in Pis'ma v Astronomicheski Zhurnal, 2008, Vol. 34, No. 1, pp. 38­57. i

Solar EUV Spectra Obtained during the SPIRIT Experiment Onboard the ° CORONAS-F Satellite: A Catalog of Lines in the Range 176­207 A
S. V. Shestov1, 2 * , S. A. Bozhenkov3 , I. A. Zhitnik1 , S. V. Kuzin1 , A. M. Urnov1, 2 ** , I. L. Beigman1, 2 , F. F. Goryayev1 , and I. Yu. Tolstikhina
1 2

1

Lebedev Physical Institute, Russian Academy of Sciences, Leninskii pr. 53, Moscow, 119991 Russia Moscow Institute of Physics and Technology, Institutskii per. 9, Dolgoprudnyi, Moscow oblast, 141700 Russia 3 Institute of Plasma Physics, Juelich Research Center, Germany
Received April 28, 2007

° Abstract--We present a catalog of 65 spectral lines in the range 176­207 A recorded by the RES spectroheliograph in active regions and flares during the SPIRIT experiment onboard the CORONASF satellite. We have identified 51 lines. The relative intensities of lines recorded during the M6.5 (GOES) flare of September 16, 2001, are given. The data processing technique is discussed. PACS numbers : 95.55.Fw; 95.75.Fg; 96.60.P-; 96.60.Tf; 95.85.Mt; 95.30.Ky DOI: 10.1134/S1063773708010052 Key words: CORONAS-F, SPIRIT, EUV spectrum, catalog of lines in solar flares and active regions.

the spectra of active regions and flares recorded in the ° 176­207 A channel of the RES spectroheliograph. This paper is a continuation of the paper by ° The spectral range 180­210 A is of great interBeigman et al. (2005), which is devoted to processing est in spectroscopically diagnosing the solar coronal and identifying the extreme ultraviolet (EUV) spectra obtained with the RES spectroheliograph of the plasma. This range contains many spectral lines of SPIRIT instrumentation onboard the CORONAS-F multiply charged ions formed in a wide temperature satellite (see Zhitnik et al. 2005). The RES spectro- range, 8 â 105 -3 â 107 K. Since there is a number of heliograph includes two channels to obtain monochro- spectral lines of iron ions at various ionization stages ° matic EUV (176­207 and 280­330 A) solar im- (from Fe VIII to Fe XXIV), the emitting plasma can ages and X-ray channels to image the Sun in the be diagnosed by abundance-insensitive methods. The ° monochromatic MgXII 8.42 A line. A detailed de- sensitivity of Fe XII­Fe XV line ratios to density alplasma density to be determined in scription of the RES channels and their basic charac- lows the emitting 12 the range 108 -10 cm-3 (Brickhouse et al. 1995). teristics are given in Kuzin et al. (1997) and Zhitnik et al. (2006). The overall objectives of the SPIRIT The Sun was observed in the spectral range 186­ experiments and the first results of observations are 201 A by the RES-C spectroheliograph onboard the ° presented in Zhitnik et al. (2002). A brief overview CORONAS-I satellite (see Zhitnik et al. 1998). The of the observational data and preliminary results of RES-C data were used both to determine the plasma modeling active plasma structures in the solar corona density in active regions and to identify spectral lines based on EUV spectra and X-ray images are provided in this range and to refine atomic parameters (Zhitnik in Zhitnik et al. (2006) and Urnov et al. (2007). The et al. 1999). The solar observations in the spectral specifications of the RES EUV channels, the data range 176­207 A were continued by the RES spec° processing technique, and a full catalog of spectral troheliograph of the SPIRIT instrumentation. ° lines in the range 280­330 A compiled from the Here (see Table 1), we present a catalog of spectral observations of the X3.4 flare of December 28, 2001, ° lines in the range 176­207 A compiled from the specand the compact active region NOAA 9765 are given trum for the M5.6 (GOES) flare of September 16, in Beigman et al. (2005). In this paper, we analyze 2001, at 03 : 39-03 : 53-04 : 18 UT. The flare spec* trum was recorded on September 16, 2001, at 03 : E-mail: sshestov@dgap.mipt.ru ** E-mail: urnov@sci.lebedev.ru 59 UT with an exposure time of 37 s. We also used the
33

INTRODUCTION


34

SHESTOV et al.

° Table 1. Parameters of the observed lines in the range 176­207 A N 1 2 3 4 5 6 7 ° , A 176.09 176.34 176.71 176.93 177.24 178.02 179.24 I ,arb.units 1070 ± 128 956 ± 132 822 ± 128 2625 ± 162 6900 ± 234 3799 ± 178 1409 ± 141 2365 ± 157 15381 ± 1500 823 ± 85 170 ± 51 3225 ± 112 ... 230 ± 54 324 ± 51 130 ± 66 2386 ± 120 2742 ± 107 434 ± 55 1726 ± 81 112 ± 40 1044 ± 81 4780 ± 130 ,pixels 2.68 ± 0.24 2.21 ± 0.24 2.14 ± 0.25 2.23 ± 0.09 2.62 ± 0.06 2.83 ± 0.09 3.06 ± 0.24 3.13 ± 0.15 ... 2.20 ± 0.16 2.24 ± 0.50 3.00 ± 0.07 ... 2.41 ± 0.41 1.93 ± 0.21 2.04 ± 0.91 3.02 ± 0.12 3.14 ± 0.08 1.77 ± 0.16 2.49 ± 0.08 1.77 ± 0.49 2.43 ± 0.14 2.82 ± 0.05 Ion Ni XV ... Ni XV Ca XV + b Fe X Fe XI + b Ni XV SIX 8 9 179.75 180.42 Fe XI Fe XI Fe X 10 11 12 13 14 15 181.11 181.87 182.17 182.3 182.71 183.41 Fe XI Ca XV Fe XI Fe X ... Ar XIV Ca XIV 16 17 18 19 20 183.81 184.03 184.52 184.77 185.20 ... ... Fe X Fe XI Fe VIII Ni XVI 21 22 186.23 186.60 Fe XII Ca XIV Fe VIII 23 186.87 Fe XII Fe XII SXI 24 25 187.38 187.92 109 ± 67 746 ± 134 8198 ± 245


° C , A .11 ... .74 .93 .24 .06 .27 .28 .76 .41 .41 .14 .90 .17 .31 ... .45 .46 ... ... .54 .80 .21 .23 .24 .61 .60 .88 .85 .84 ... .92 .97 .23 .30 .19

I ... ... ... .95 .24 .05 .27 ... .75 .41

II ... ... ... ... .24 ... .28 ... ... .38

.14 .91 .17 ... ... .45 ... ... ... .54 .80 .21 ... .60

.13 ... .17 ... ... ... ... ... ... .53 .80 .22 ... .62

.87

.87

2.71 ± 1.36 4.26 ± 0.66 3.42 ± 0.08

... Fe XXI Ar XIV Fe XI Fe XI Fe XII

... .95

... .96

26

188.25

.23

.21

ASTRONOMY LETTERS

Vol. 34 No. 1

2008


SOLAR EUV SPECTRA Table 1. (Contd.) N 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 ° , A 188.50 188.66 188.78 189.06 189.39 189.69 190.00 I ,arb.units 911 ± 230 313 ± 170 474 ± 167 1059 ± 139 102 ± 42 193 ± 47 848 ± 33 585 ± 96 190.35 191.00 378 ± 59 238 ± 77 809 ± 110


35

,pixels 2.92 ± 0.70 1.69 ± 0.63 2.46 ± 0.81 4.39 ± 0.48 1.97 ± 0.61 1.95 ± 0.35 3.62 ± 0.16 2.42 ± 0.28 2.85 ± 0.76 2.55 ± 0.30 ... 2.89 ± 0.06 2.05 ± 0.34

Ion ?Fe XII SXI Ar XI ?Fe XI+ b ... Fe XI Fe X Fe XII SXI Ar XI Fe XII SXI Fe XXIV Fe XII ... Ca XVII Fe XI Fe XII Fe XI Ca XIV ?Ni XVI ?Ar XIV ... Fe XII Fe XII Fe XIII Fe XII ... Fe XIII ... Fe XI S VIII Fe XII

° C , A .45 .68 .81 .13 ... .72 .04 .07 .36 .95 .05 .27 .03 .39 ... .82 .83 .52 .52 .87 .02 .40 ... .12 .18 .54 .64 ... .43 ... .55 .55 .58 ... .02 ...

I ... ... ... .14 ... ... .04

II .49 .66 .82 .19 ... .72 .05 ... ... ... .26 ... .39 ... .88 .81

.37 .04

191.25 192.03 192.39 192.61 192.82

.27 .03 .40 ... .86

11396 ± 1000 5659 ± 164 661 ± 123 4851 ± 186 1722 ± 51


2.94 ± 0.06 ... 3.08 ± 0.96 2.63 ± 1.14 2.25 ± 0.32 ... ... ... ... ... ... ...

193.54

11488 ± 1500 1344 ± 466 533 ± 428 647 ± 114 478 ± 150 17628 ± 1500 8176 ± 610 445 ± 200 2197 ± 250 947 ± 100 4426 ± 400

.51

.51

193.83 193.99 194.35 194.66 195.19

.87 .05 .41 ... .13

.87 .03 .40 ... .12

196.62

.52 .63 .01 .44 ... .55

.52 .64 ... ... ... .56

197.07 197.42 197.83 198.53

52 53 54

199.24 200.01 200.41

1601 ± 150 11864 ± 1000 565 ± 40

... ... ...

... Fe XIII ...

... .02 .37

... .02 ...

ASTRONOMY LETTERS

Vol. 34 No. 1 2008


36 Table 1. (Contd.) N 55 56 ° , A 200.67 201.15 I ,arb.units 569 ± 40 21866 ± 2000

SHESTOV et al.

,pixels ... ...

Ion ... Fe XIII Fe XII

° C , A ... .13 .14 .74 .76 .71 .04 .71 .61 .63 .16 .83 .80 .26 .65 .95 ...

I ... .13

II ... .12

57

201.72

2518 ± 250

...

Fe XII Fe XII Ar XIII

.74

.76

58 59

202.08 202.63

40323 ± 4000 1600 ± 200

... ...

Fe XIII ?Fe XI S VIII Fe XI

.05 ... ... ... .18 .82

.04 ... ... ... .16 .82

60 61

203.19 203.84

8065 ± 500 52795 ± 5000

... ...

Fe XIII Fe XIII Fe XIII

62 63 64 65

204.26 204.63 204.95 206.32


6052 ± 1000 2180 ± 200 7782 ± 800 12380 ± 1000

... ... ... ...

Fe XIII Fe XVII Fe XIII + b ...

.25 .65 .94 .25

.26 .67 .95 .26

° Note. Notation: N is the line number; is the measured wavelength; I is the line intensity; is the line width (1 pixel is 0.025 A); ion is the identified ion; C is the fractional part of the wavelength of the identified line from the CHIANTI database (Landi et al. 2006); I is the fractional part of the line wavelength from the spectral catalog of flares by Dere (1978); II is the fractional part of the line wavelength from the spectral catalog of active regions by Brosius et al. (1998); marks the lines that are clearly seen only in the flare spectrum; +b denotes the line intensity attributable to an unidentified blend; denotes the intensities of blended lines estimated from the branching ratios; ? means that questions remain in the line identification. The lines given under the same number are unresolvable. For further explanations, see the text.

spectra of active regions (NOAA 9650, 9653, 9655, 9659, 9660, 9665, 9695, 9697, 9698, 9715, 9716, 9725) recorded on October 14­17, November 22, and December 5, 2001, with various exposure times. The spectra of active regions were used to check the line identifications and the spectral sensitivity calibration of the spectroheliograph (see Table 2). DATA PROCESSING AND CALIBRATION One-Dimensional Spectrum Construction The spectroheliogram is a sequence of monochromatic images of the full solar disk compressed in

the dispersion direction. The primary data processing consists in subtracting the background on the spectroheliogram, rotating the image, and constructing the source's one-dimensional spectrum. The primary spectroheliogram processing technique for the ° 280­330 A channel as an example was described by Beigman et al. (2005). Figure 1a shows the spec° troheliogram for the 176­207 A channel recorded on September 16, 2001, at 03 : 59 UT. Figure 1b shows the same spectroheliogram after the background subtraction; the arrow indicates the dispersion direction. The one-dimensional spectrum is a spectroheliASTRONOMY LETTERS Vol. 34 No. 1 2008


SOLAR EUV SPECTRA

37

ogram section in the dispersion direction and a convolution of the source's emission spectrum with the spatial distribution of its intensity along the dispersion axis. Thus, information about the source's spatial intensity distribution along the dispersion axis is generally required to reconstruct and interpret the one-dimensional spectrum. However, the spatial distribution is unimportant if the source's size in this direction is smaller than the width of the point spread function that corresponds to the source's 1 size of R . Solar flares and most active regions 5 meet this requirement. Figure 2a shows the spectroheliogram recorded on September 16, 2001, at 03 : 59 UT; the dispersion direction is horizontal and the wavelength increases to the right. The dashed line represents the one-dimensional spectrum that corresponds to the M5.6 (GOES) flare. The most intense lines are labeled. Figure 2b shows a magnified portion ° of the spectroheliogram near = 193 A, where we see relatively "cold" Fe XII lines, which are intense on the entire solar disk, and "hot" flare regions identified as

Ca XVII and Fe XXIV lines. In this paper, the onedimensional spectrum was averaged over five pixels in a direction perpendicular to the dispersion direction. In this direction, the width of the point spread function is 2 pixels; 1 pixel corresponds to 5 . 4. The subsequent spectroheliogram processing procedure consists in the wavelength calibration of the one-dimensional spectrum obtained (based on reference lines), the determination of its parameters-- the intensity (given the spectral sensitivity of the instrument) and width, and the identification of spectral lines. Based on a geometrical model of the spectroheliograph, Beigman et al. (2005) showed that the wavelength could be expressed by a second-degree polynomial of the line position (in pixels) in the onedimensional spectrum. The polynomial coefficients can be determined using reference lines for the spectral region under consideration. In this channel, the following reference lines were used to determine the wavelengths:

° 177.24 A (FeX 3s2 3p ° 184.54 A (FeX 3s2 3p

5 5 3 3 2 2

2 2 4 4 3 3

P3/2 - 3s2 3p4 (3 P )3d P3/2 - 3s2 3p4 (1 D)3d S S
3/2 3/2

2

P3/2 ), S
1/2

2 4 4

),

° 192.39 A (FeXII 3s2 3p ° 193.52 A (FeXII 3s2 3p

- 3s2 3p2 (3 P )3d - 3s2 3p2 (3 P )3d
3 3

P1/2 ), P3/2 ),

° 200.02 A (FeXIII 3s2 3p ° 204.95 A (FeXIII 3s2 3p

P1 - 3s2 3p3d P2 - 3s2 3p3d

D2 ), D1 ).

The accuracy of determining the wavelengths by this ° method is 0.04 A. To test the technique, Fig. 3 compares the experimental and theoretical wavelengths. The spectral sensitivity of the instrument is determined mainly by the wavelength dependence of the multilayer mirror reflectance. The mirror reflectance was measured before the flight of the instrument; the normalized reflectance is shown in Fig. 4. To check the calibration accuracy of the spectral sensitivity, we used line intensity ratios insensitive to emitting plasma parameters. The results are presented in Table 2. Most of the experimentally measured ratios agree with the theoretical values with an accuracy of 30%, which corresponds to the accuracy of measuring the mirror reflectance.
ASTRONOMY LETTERS Vol. 34 No. 1 2008

Line Parameter Determination
To determine the positions and parameters of spectral lines, we fitted the spectrum in the chosen narrow wavelength range. This method proved to be efficient in analyzing strongly blended spectra (Brooks et al. 1999; Thomas and Neupert 1994; Lang et al. 1990; Brosius et al. 1998, 2000; Beigman et al. 2005). The emission intensity distribution in the chosen wavelength region 1 ­2 can generally be written as
N

i() = P + Q +
i=1

Ai

(1)

â exp -

1 2

- i

i 0

2

,

= [1 ,2 ],


38

SHESTOV et al.

° Table 2. SPIRIT line intensity ratios in the range 176­207 A Lines SXI 188.68/191.27 , % Fe X 177.24/184.54 , % Fe X 184.54/190.04b , % Fe XI 178.06b/182.17 , % Fe XI 192.83b/188.23 , % Fe XI 189.13/189.72 , % Fe XII 192.39/193.51 , % Fe XII 192.39/195.12 , % Fe XII 193.51/195.12 , % Fe XIII 197.43/204.95b , % Fe XIII 197.43/201.13b , % Fe XIII 201.13b/204.95b , % Fe XIII 200.02/204.26 , % Fe XIII 200.02/203.16 , % Fe XIII 203.82/200.02 , % 1 0.39 56 2.52 5 1.91 7 1.18 6 0.80 4 ... ... 0.42 5 0.32 9 0.78 10 0.18 17 0.10 20 1.79 11 1.83 15 1.44 18 4.94 28 2 ... ... 2.76 9 1.49 7 0.93 20 0.36 8 0.55 105 0.52 13 0.36 14 0.69 15 0.19 16 0.10 20 1.90 11 1.82 15 1.21 21 5.35 26 3 0.39 49 2.90 9 ... ... 0.69 9 0.28 4 1.00 32 0.50 14 0.33 15 0.65 14 0.16 18 0.08 25 1.91 11 1.61 17 1.76 15 5.54 25 4 ... ... 3.15 9 ... ... 1.00 13 0.38 8 ... ... 0.48 6 0.24 8 0.50 14 0.19 16 0.10 20 1.86 11 1.65 17 1.44 18 6.57 21 5 0.59 32 2.60 6 1.42 8 0.95 14 0.36 3 1.43 50 0.54 15 0.38 13 0.71 14 0.19 16 0.09 23 2.31 9 2.00 14 1.31 20 5.39 26 6 ... ... 2.55 6 1.56 8 0.85 7 0.37 3 1.26 37 0.59 15 0.40 15 0.68 15 0.17 18 0.09 23 1.77 11 1.85 15 1.38 20 4.61 30 5.40 4.55


Mean 0.46

Theory 0.57


Ratio 0.81

2.64

1.93

1.37

1.56

3.21

0.49

0.89

0.27

3.30

0.34

0.21

1.62

1.04

1.30


0.8

0.45

0.42

1.07

0.29

0.26



1.12

0.65

0.61



1.07

0.18

0.65

0.28

0.09

0.24

0.38

1.92

3.35


0.57

1.79

1.56

1.15

1.42

1.68



0.85

1.19

Note. The data are given for the flare (1) and the active regions (2­6). Mean--the value obtained by averaging 2­6 values; theory-- the theoretical value based on CHIANTI data (Landi et al. 2006); --based on data by Brickhouse et al. (1995); --based on data by Landi (2002); ratio -- the mean-to-theory ratio; b marks a blended line; is the measurement error.

ASTRONOMY LETTERS

Vol. 34 No. 1

2008


SOLAR EUV SPECTRA

39

60.20° ± 0.02°

()

(b)

Fig. 1. (a) Original spectroheliogram recorded on September 16, 2001, at 03 : 59 UT. (b) The same spectroheliogram after the background subtraction; the arrow indicates the dispersion direction on the spectroheliogram.

()

Fe XI 180.4 å

Fe XII 186.9 å

Fe XI 188.3 å

Fe XIII 202.1 å

Fe XIII 203.8 å

(b)

Fe XXIV 192.0 å Ca XVII 192.8 å

Fe XII 192.4, 193.5, 195.1å

Fig. 2. (a) Spectroheliogram recorded on September 16, 2001, at 03 : 59 UT; the dispersion axis is horizontal. The dashed line represents the one-dimensional spectrum that corresponds to the M5.6 (GOES) flare. Some of the spectral lines are labeled. ° (b) A magnified portion of the spectroheliogram near = 193 A.

where P and Q describe the local background, N is the number of spectral lines, Ai , i , and i are 0 the peak intensity, wavelength, and width of spectral line i, respectively. The line shape is assumed to be Gaussian; the total line intensity can be calculated using the formula (2) Ii = 2Ai i . In fitting, the parameters P , Q, Ai , i , and i are 0 chosen using the 2 test (see, e.g., William et al. 2001; Brandt 1998). The error distribution is assumed to be normal and the standard deviation for each
ASTRONOMY LETTERS Vol. 34 No. 1 2008

point is given by the Poisson formula: = i(). The method allows the error in the parameters to be estimated and yields an estimate for the quality of the fit. In most cases, the error in the line parameters is determined precisely by the fitting procedure, despite the error introduced at the background subtraction step and the detector noise. However, there was a defocusing in the RES channel under consideration, causing a broadening of the lines and and a distortion of their shapes near the long-wavelength edge of the range (see Fig. 5a). Because of asymmetry in the line profiles,


40
0.10

SHESTOV et al.
1.0 0.8 Reflectance 180 185 190 195 Wavelength, å 200 205
l1

0.05 exper ­ theor, å

0.6 0.4 0.2 0

0

­ 0.05

­ 0.10 175

165 170 175 180 185 190 195 200 205 Wavelength, å

Fig. 3. Difference between the experimental and theoretical wavelengths for the flare of September 16, 2001.

Fig. 4. Normalized multilayer mirror reflectance versus wavelength.

the Gaussian fitting method in our case is applicable ° for wavelengths < 196 A, where the asymmetry ° in the profile is minor. For > 196 A, we used a different technique to determine the line parameters: the intensities were summed over the entire line profile: Ij =
l0

(3) The match between the experimental and theoretical (CHIANTI) line ratios. We checked the identification of the flare and active region spectra based on density-insensitive line ratios. Table 2 gives the experimental line intensity ratios along with the CHIANTI theoretical ones. The catalog of spectral lines compiled from the spectrum of the M5.6 (GOES) flare recorded on September 16, 2001, at 03 : 59 UT is given in Table 1. The table contains the measured wavelengths of the spectral lines, their parameters (intensities and widths), and suggested identifications; based on it, we make a comparison with the wavelengths from the CHIANTI catalog and from the spectral catalogs of flares (Dere 1978) and active regions (Brosius et al. 1998). The RES flare spectrum is shown in Figs. 6­ 8; the observed lines are marked by vertical bars. Below, we provide explanations in the identification of individual lines. The following notation is used: "+"-- the line is observed; "-"--the line is not observed; "b"--the line is blended; "l"--the line that does not ° fall within the recorded wavelength range 176­207 A; "u"--there remain questions in the line identification discussed in the text. S VIII (logT = 5.9). 2s2 2p5 2 P - 2s2p6 2 S . ° Jlow - Jup , A Note 1/2 - 1/2 202.61 +b 3/2 - 1/2 198.55 +b
ASTRONOMY LETTERS Vol. 34 No. 1 2008

i()d.

(3)

The "effective" peak intensity of such a line was calculated from the formula Aj = Ij / 2j , where j is the "reduced" line width dependent on wavelength j (see Fig. 5b). Below in the texts and the tables, we use the total line intensities I obtained from Eqs. (2) and (3), while the peak intensities A and A are used to estimate the contribution from weak or blended lines. LINE IDENTIFICATION The line identification method used here corresponds to that applied by Beigman et al. (2005). The line identification was based on the following criteria: (1) The coincidence of wavelengths with previously recorded ones. The CHIANTI package, version 5.2 (Landi et al. 2006), was used as the wavelength standard. We also made a comparison with the spectral catalog of flares (see Dere 1978) and active regions (Brosius et al. 1998). (2) The presence of a multiplet structure in the experimental spectrum.


SOLAR EUV SPECTRA

41

1.0 Relative intensity 0.8 0.6 0.4 0.2 0 ­12

4.5 () Line width, pixel 4.0 3.5 3.0 2.5 2.0 1.5 12 173.03 182.87 192.99 Wavelength, å 203.37 (b)

1

2

4 3 8

­8

­4 0 4 Coordinate, pixel

° Fig. 5. (a) Change of the line profile with increasing wavelength in the range 176­207 A: 1, 177.24 ° (b) Increase in the line width toward the long-wavelength edge of the range 176­207 4, 203.82 A. experimental data; the solid line represents the linear fit = -7.73(±1.4) + 0.055(± 0.008).

° ° ° A; 2, 192.39 A; 3, 201.12 A; ° The squares indicate the A.

2000 Intensity, arb. units 1500 500 1000 500 0 176 177 178 179 0 180 181 181 182 Wavelengths, å

183

184

185

186

Fig. 6. Spectrum of the September 16, 2001 flare. The vertical bars mark the lines included in Table 1.

° The S VIII 202.61 A line is blended with the ° (3s2 3p4 1 D2 - 3s2 3p3 (2 D)3d 1 P1 ) Fe XI 202.71 A ° line. The S VIII 198.55 A line is blended with the ° (3s2 3p4 1 D2 - 3s2 3p3 (2 D)3d 3 P1 ) Fe XI 198.55 A 2P ° and Fe XII 198.58 A (3s2 3p3 1/2 - 2 3p2 (1 D )3d 2 P 3s 3/2 ) lines. The S VIII line intensity ratio I (198.55)/I (202.61) calculated from the branching ratio using the CHIANTI package is insensitive to density and temperature and is 2.2. SIX (logT = 6.0). 2s2 2p ° 179.28 A.
41

SXI (logT = 6.3). 2s2 2p

23

P - 2s2p

33

S.

° Jlow - Jup , A Note 0-1 1-1 2-1 186.84 +b 188.68 191.27 + +

D2 - 2s2p

51

P1 =

The experimental intensity ratio I (188.68)/ I (191.27) is close to the theoretical one (see Table 2). ° ° Consequently, the S XI 188.68 A and S XI 191.27 A lines are unblended. The intensity of the blended ° S XI 186.84 A line was estimated from the intensity of ° the S XI 191.27 A line and the branching ratio (Landi et al. 2006) ISXI (186.84) = 0.19ISXI (191.27) = 0.19(810 ± 111) = 156 ± 21. (4)

° The SIX 179.28 A line is blended with the Ni XV 2 3p2 3 P - 3s2 3p3d 3 D ) line. ° 179.27 A(3s 2 3
ASTRONOMY LETTERS Vol. 34 No. 1 2008


42

SHESTOV et al.

1000 Intensity, arb. units 1000 500 500

0 186

187

188

189

190

0 191 192 191 Wavelengths, å

193

194

195

196

Fig. 7. Spectrum of the September 16, 2001 flare (the continuation of Fig. 6). The vertical bars mark the lines included in Table 1.

600 500 400 300 200 100 0 196

ntensity, arb. units

1500 1000 500 0 201 201 202 203 204 205 206 Wavelengths, å

197

198

199

200

Fig. 8. Spectrum of the September 16, 2001 flare (the end of Fig. 6). The vertical bars mark the lines included in Table 1.

° The experimental intensity of the 186.87 A line is 4780 ± 130. Thus, the contribution from the ° S XI 186.84 A line is 3% and is at the level of ° the error in the 186.87 A line intensity, while the 2 3p3 2 D 2 23 2 ° Fe XII 186.88 A(3s 5/2 - 3s 3p ( P )3d F7/2 ) and 3s 3p
2

Ar XI (log T = 6.3). 2s2 2p

43

P - 2s2p

53

P.

° Jlow - Jup , A Note 1-0 0-1 1-1 2-1 1-2 2-2 187.09 - - - +

190.98 +b 189.58

Fe XII
2 (3 2

186.85

° A

(3s 3p

2

3

2

D3/2 -

P )3d F5/2 ) lines make a major contribution.

184.52 +b 194.10 188.81

2s 2p

2

21

D2 - 2s2p

21

° P1 = 190.36 A.

° The spectrum exhibits the 190.35 A line with an intensity of 378 ± 59. The theoretical S XI line ratio I (190.36)/I (191.27) is sensitive to density and reaches values >0.3 at log ne > 11 cm-3 . Thus, the ° ° 190.35 A line can be identified as S XI 190.36 A.

° The Ar XI 190.98 A line is blended with the ° (3s2 3p3 2 P3/2 - 3s2 3p2 (3 P ) â Fe XII 191.05 A ° 3d 2 D5/2 ) line. The blending of Ar190.98 A is con° firmed by the absence of the Ar XI 189.58 A line in the experimental spectrum whose intensity should ° account for 0.79 of the Ar XI 190.98 A intensity (based on CHIANTI data; Landi et al. 2006).
ASTRONOMY LETTERS Vol. 34 No. 1 2008


SOLAR EUV SPECTRA

43

° The Ar XI 184.52 A line is blended with the ° Fe X 184.54 A (3s2 3p5 2 P3/2 - 3s2 3p4 (1 D)3d 2 S1/2 ) line. Estimating the contribution from the ° Ar XI 184.52 A line based on the intensity of ° by assuming that the latter is unAr XI 190.98 A blended yields I
Ar XI

(184.52) = 1.45IAr XI (190.98) 1.45(238 ± 77) = 345 ± 112.

(5)

° The intensity of the 184.54 A line on the spectroheliogram is Itotal (184.54) = 2737 ± 107. Consequently, ° the contribution from Ar XI 184.52 A does not exceed 13%. Ar XIII (log T = 6.5). 2s2 2p2 3 P - 2s2p3 3 P . ° Jlow - Jup , A Note 1-0 0-1 1-1 2-1 1-2 2-2 205.95 - - -l - -l

° The Ar XIV 183.45 A line is blended with the ° Ca XIV 183.46 A(2s2 2p3 4 S3/2 - 2s2p4 4 P1/2 ) line. ° The Ar XIV 191.40 A line is not observed, although, judging by the branching ratio (Landi et al. 2006), it should be approximately twice as strong as ° the Ar XIV 183.45 A line. Consequently, the observed ° ° 183.45 A line is determined by the Ca XIV 183.46 A emission. ° The Ar XIV 187.97 A line is blended with the ° 2s2 2p2 1 D2 - 2s2p3 3 D1 ) line. The Fe XXI 187.92 A( ° Ar XIV 180.29 A line is not observed. Let us estimate ° its intensity based on the intensity of Ar XIV 187.97 A by assuming that there is no blending and on the branching ratio (Landi et al. 2006) by taking into account the calibration factor, (6) IAr XIV (180.29) = 0.63 â 0.18 â IAr XIV (187.97) 0.11(746 ± 134) = 82 ± 15, AAr 1 I 2 2
XIV

201.71 +b 205.80 211.01 205.29 210.47

(180.29)

(7)

Ar XIV

(180.29) = 16 ± 3.

The peak intensity is at the noise level; besides, it ° should be noted that the Ar XIV 187.97 A line is blended. 2s2 2p 2 P - 2s2p2 2 S . ° Jlow - Jup , A Note 1/2 - 1/2 194.40 + 3/2 - 1/2 203.35 +b ° The Ar XIV 194.40 A line is observed. The in° tensity of Ar XIV 203.35 A is estimated from the ° 194.40 line to be 80; however, the 203.35 A line is not observed due to the presence of strong close Fe XII and Fe XIII lines. Ca XIV (log T = 6.6). 2s2 2p3 4 S - 2s2p4 4 P . ° Jlow - Jup , A Note 3/2 - 1/2 183.46 +b 3/2 - 3/2 186.61 +b 3/2 - 5/2 193.87 +

° The Ar XIII 201.71 A line is blended with the ° strong Fe XII 201.74 A (3s2 3p3 2 P1/2 - 2 3p2 (1 D )3d 2P ° and Fe XII 201.76 A 3s 1/2 ) 2 3p3 2 P 2 3p2 (1 D )3d 2 S (3s 3/2 - 3s 1/2 ) lines. Since the ° line is not observed in the experiAr XIII 205.80 A mental spectrum, an upper limit for the intensity of ° Ar XIII 201.72 A can be estimated from the branching ° ratio. The intensity of Ar XIII 205.80 A should be a ° factor of 1.33 higher than that of Ar XIII 201.71 A (CHIANTI), which, given the calibration factor, corresponds to the same intensity, i.e., IAr XIII (205.80) ° IAr XIII (201.71). The intensity of 205.80 A cannot exceed 200 and, thus, the contribution from ° ° Ar XIII 201.71 A to the experimental 201.72 A line does not exceed 10%. Ar XIV (log T = 6.5). 2s2 2p 2 P - 2s2p2 2 P . ° Jlow - Jup , A Note 1/2 - 1/2 183.45 +b 3/2 - 1/2 191.40 -u 1/2 - 3/2 180.29 - 3/2 - 3/2 187.97 +b
ASTRONOMY LETTERS Vol. 34 No. 1 2008

° The Ca XIV 183.46 A line is blended with the ° (2s2 2p 2 P1/2 - 2s2p2 2 P1/2 ) line. Ar XIV 183.45 A However, as follows from the discussion of the Ar XIV identification, the line is determined mainly by the ° Ca XIV 183.46 A emission. ° The Ca XIV 186.61 A line is blended with the ° Fe VIII 186.60 A (3p6 3d 2 D3/2 - 3p5 3d2 (3 F ) 2 F5/2 )