Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.tesis.lebedev.ru/en/docs/1376.pdf Äàòà èçìåíåíèÿ: Mon May 26 00:00:00 2008 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:10:05 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: annular

Astronomy Letters, Vol. 31, No. 1, 2005, pp. 37­56. Translated from Pis'ma v Astronomicheski Zhurnal, Vol. 31, No. 1, 2005, pp. 39­58. i Original Russian Text Copyright c 2005 by Zhitnik, Kuzin, Urnov, Beigman, Bozhenkov, Tolstikhina.

Extreme Vacuum Ultraviolet Solar Spectra Obtained during the SPIRIT Experiment aboard CORONAS-F: ° A Catalog of Lines in the Range 280­330 A
I. A. Zhitnik1 , S. V. Kuzin1 , A. M. Urnov1, 2 * , I. L. Beigman1, 2 , S. A. Bozhenkov1, 2 , and I. Yu. Tolstikhina
1

1

Lebedev Institute of Physics, Russian Academy of Sciences, Leninski i pr. 53, Moscow, 119991 Russia 2 MoscowPhysicotechnical Institute, Dolgoprudnyi, Moscow oblast, 141700 Russia
Received June 15, 2004

° Abstract--We present a catalog of 100 lines in the wavelength range 280­330 A detected by the RES-C spectroheliograph in solar active regions and flares during the SPIRIT experiment aboard the CORONAS-F orbital station. We identified 54 lines. The line intensities recorded during the X3.4 (GOES) solar flare of December 28, 2001, are given. The data reduction procedure is discussed. c 2005 Pleiades Publishing, Inc. Keywords: CORONAS-F, catalog of lines in solar flares and active regions.

INTRODUCTION Extreme vacuum ultraviolet (XUV) solar spectra are of great interest and are widelyused to solve many problems in astrophysics and the spectroscopy of ° multiplycharged ions. The spectral range 180­330 A contains intense ionic lines of almost all of the abundant (in astrophysical plasma) elements excited over a wide temperature range, 105 ­107 K. In contrast to X-ray spectra, the XUV spectra of a hot plasma are not affected by nonthermal (beam) electrons due to the relatively low line excitation thresholds and are effectivelyused to studythe differential emission measure (DEM), the abundances of various elements, etc. The sensitivityof the relative intensities of XUV lines to the electron temperature and density allows the plasma structures in the solar atmosphere to be diagnosed (Jordan 1974, 1979; Doschek 1991; Zhitnik et al. 1999). This is necessary for modeling its structure and solving fundamental problems: the heating of the solar corona, the origin of the solar wind, etc. Apart from important information about the energy accumulation, transfer, and release, high-resolution spectra for the coronal plasma of active regions and flares also make it possible to experimentallyrefine the wavelengths and to identify the lines that correspond to both optically allowed and forbidden transitions in multiply charged ions. Elucidating the physical conditions for the generation of solar XUV lines and compiling the catalogs of these lines are inextricably
*

E-mail: urnov@sci.lebedev.ru

linked with one another and represent two aspects of the same problem aimed at studying their excitation mechanism under typical conditions in the solar atmosphere. By now, a series of catalogs have been compiled using spectra both from the full solar disk (Malinovsky and Heroux 1973; Behring et al. 1972, 1976) and from separate areas of the quiet corona (Brooks et al. 1999) and active regions under flareless conditions (Thomas and Neupert 1994; Brosius et al. 1998, 2000). The published spectroscopic data on solar flares are much more scarce. Observational spectra are difficult to obtain for solar flares, because these events are relativelyrare and because the size of the source is small. Regular observations with a widefield instrument are required to obtain them. The main source of information about flare lines is the catalog compiled from the measurements performed with the S 80A spectroheliograph aboard the Skylab station (Dere 1978). This instrument recorded the spectral images of the Sun with high spatial (2 ) and spectral ° (0.1 A) resolutions. However, the large size of the ° Sun, which was 25 A on the wavelength scale, led to an overlap of the spectral images and made it difficult to restore the spectra. Regular observations of the full solar disk in monochromatic XUV spectral lines were performed with the RES-C spectroheliograph aboard CORONAS-I in 1994 (Sobel'man et al. 1996). A catalog of coronal plasma lines in the wavelength ° range 177­207 A was compiled from these observations (Zhitnik et al. 1998), and a procedure for

1063-7737/05/3101-0037$26.00 c 2005 Pleiades Publishing, Inc.


38

ZHITNIK et al.

3

analysis of the wavelength measurement accuracy, and the line identification procedure based on analysis of the relative line intensities of the multiplet structure, are described. DESCRIPTION OF THE INSTRUMENT

1

2 4

Fig. 1. Optical scheme of the spectroheliograph: (1) entrance filter, (2) diffraction grating, (3) multilayer X-ray mirror, and (4)detector.

determining the physical plasma parameters using these data was presented (Zhitnik et al. 1999). An outgrowth of this instrument is the XUV spectroheliograph for the wavelength ranges 180­210 ° and 280­330 A that was part of the SPIRIT instrumentation aboard CORONAS-F (Oraevskii and Sobel'man 2002). This station was placed in a nearEarth polar orbit with a height of 500 km on July 31, 2001. Apart from the spectroheliograph, the SPIRIT instrumentation includes XUV telescopes ° for the wavelength range (171, 175, 195, 285, 304 A) and a spectroheliometer to record the solar images ° in the Mg XII 8.42 A line (Zhitnik et al. 2002). Regular SPIRIT observations began in August 2001 during the maximum of the solar cycle 23. More than 300 000 soft X-ray and XUV images and spectroheliograms of the Sun, including more than 1000 XUV spectroheliograms, have been obtained to date. In this paper, we analyze the spectra of solar flares and other active structures in the solar atmosphere taken with the RES-C spectroheliograph as part of the SPIRIT experiment. This instrument can simultaneously obtain about 150 images of the entire Sun in individual monochromatic lines in ° two spectral ranges: 180­210 and 280­330 A. We present a complete catalog of lines in the range 280­ ° 330 A that was compiled from the observations of the active region NOAA 9765 and the X3.4 solar flare of December 28, 2001 (20:02­21:32) (GOES). The spectrum of the flare was taken at 21:21:44 on December 28, 2001; the spectra of the active region were taken at the following times: 14:38:11 on December 29, 2001; 16:16:43 on December 29, 2001; and 04:52:44 on December 30, 2001. The exposure time for all of the frames used was 37 s. The lines recorded only during the flare are marked separately in the catalog. The relative line intensities and widths are given for the X3.4 flare of December 28, 2001. The spectroheliogram reduction procedures, including the wavelength and spectral sensitivity calibration, the

The spectroheliograph was a slitless spectrometer with a flat objective reflecting diffraction grating at the entrance (Fig. 1). The grating was set at a small grazing angle, 1 ­2 , which led to the "spectral compression" of the angular aperture of the Sun in the diffracted emission of an individual spectral line: the angular size of the image in the dispersion direction decreased proportionally to the ratio of the sines of the grazing and diffraction angles. The emission diffracted on the grating was focused by a normalincidence multilayer mirror on the sensitive area of an imaging detector. Characteristics of the main optical elements used in the XUV RES-C spectroheliograph are given in Table 1. The calibration procedure for the main optical elements and the entire spectroheliograph was described in detail byKuzin et al. (1997). DATA REDUCTION AND CALIBRATION The spectroheliogram was a sequence of monochromatic images of the full Sun compressed in the dispersion direction. The initial spectroheliogram is illustrated by Fig. 2a. The data reduction procedure consisted in subtracting the background, constructing the one-dimensional spectrum, determining the observed line parameters, and calibrating the wavelengths and the spectral sensitivity of the instrument. The background intensity at observed peak line intensities from 20 to 16 000 counts was 50­300 counts and exhibited a significant gradient (below, counts are designated as arbitraryunits). The background subtraction algorithm was refined on a large number of frames; the error of the algorithm was 5 counts. The spectroheliograms before and after the background subtraction are compared in Fig. 2.

Constructing the One-dimensional Spectrum
The spectrum is a one-dimensional section of the spectroheliogram in the dispersion direction. The dispersion direction of the instrument does not coincide . . with the CCD rows. The slope is = 2763 ± 002, which corresponds to tan = 0.5234 ± 0.0004. The measurement error of the tangent allows the relative position of the region under study on the solar disk at the opposite ends of the spectroheliogram to be determined with an accuracy higher than 7 . The position of a point on the dispersion axis is determined
ASTRONOMY LETTERS Vol. 31 No. 1 2005


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA Table 1. Elements of the spectroheliograph Range ° 176­207 A Diffraction grating Line frequency, lines mm Grazing angle, deg
-1

39

° 280­330 A

3600 1.2 Multilayer mirror 1.4

Mirror shape Focal length, mm Coating Peak reflectance, %

Sphere 830

Paraboloid 570 Mo­Si

30 Entrance filter

23

Material Transmission in working range, % 12

Al 0.15 ± 0.05 µmin thickness 11

Detector. Image tube (IT) with open solar-blind MCP + electrically cooled CCD array IT photocathode IT amplification Number of CCD pixels Input pixel size, µm ADC width, bit Readout time, s MgF2 up to 10
5

1024 â 1152 20.8 â 20.8 14 2

by both the wavelength and the geometrical position on the Sun. However, the geometrical position is 1 unimportant if the size of the emitting region R 5 (a flare or an active region). The size of the region in the image is smaller than the point spread function of the instrument due to the compression; i.e., the distance along the dispersion direction corresponds to the wavelength. The spectroheliogram compression in the dispersion direction causes the spectrum to be averaged over the extent of the region along the dispersion axis. For the spectra used here, the contribution of nearby regions is negligible, since for our analysis we selected regions close to or above the limb at the upper and lower edges of the spectroheliograms. All of the spectra used were averaged in a
ASTRONOMY LETTERS Vol. 31 No. 1 2005

direction perpendicular to the dispersion axis over five pixels; one pixel corresponds to about 7 on the solar disk, and the point spread function of the instrument is two pixels in width. As an example, Fig. 3 shows the position of the December 28, 2001 flare on the solar disk, the height at which the spectrum is viewed, and the spectrum averaging region.

Determining the Line Parameters
To determine the line positions and intensities, we used a method of fitting the emission intensity in the chosen narrow part of the spectrum by taking into account the expected number of lines and the local background. The latter may be represented as a constant or a linear function after the removal of the main


40

ZHITNIK et al.

(a)

(b)

Fig. 2. Comparison of the spectroheliograms before and after applying the background subtraction procedure: (a) initial spectroheliogram; (b) reduced spectroheliogram.

()

284.2

292.0

303.8 (b)

320.6

335.4

Fig. 3. (a) Position o to the flare images in highlighted region of indicate the region of

f the December 28, 2001 flare on the solar disk. The bright pointlike areas at the lower limb correspond individual lines. The rectangle highlights the region that is shown below on an enlarged scale. (b) The the spectrum: the middle straight line indicates the center of the flare; the upper and lower straight lines the spectrum averaging perpendicular to the dispersion axis, and the bright regions are the flare images.

gradients. The method proved an exceptionallyuseful tool for analyzing strongly blended spectra (Brooks et al. 1999; Thomas and Neupert 1994; Lang et al. 1990; Brosius et al. 1998, 2000). The line shape was assumed to be Gaussian. The parameters included the peak line intensities, widths, and positions. The lines were fitted by the 2 method (see, e.g., Press et al. 2001; Brandt 2003). The error distribution was assumed to be normal; the standard deviation for each I point was given by the Poisson formula: i = Ii , where Ii is the intensity of point i. This method also allows the errors in the parameters to be estimated

and yields an estimate for the quality of the fit. An example of a fit is shown in Fig. 4. In most cases, the error in the peak intensity is determined precisely by the fitting procedure, despite the error introduced at the background subtraction step and the presence of detector noise. For some of the lines, the error can be slightly larger than that determined from the confidence interval, because the profile must be determined in narrow segments where the behavior of the background cannot always be predicted unambiguously. An analysis of the dependence of the line width on its position in the spectrum (Fig. 5) shows that the width in the interval under
ASTRONOMY LETTERS Vol. 31 No. 1 2005


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA

41

300 3 Intensity, arb. units 200 Width, pixels

2

100 1 0 400 800 Coordinate (pixels) 1200

0 590

600 610 Coordinate (pixels)

620

Fig. 5. Line width versus position in the spectrum.

Fig. 4. An example of fitting the emission profile by minimizing the 2 value, 2 = 0.19.The squares indicate the experimental data. The dashed curves represent three separate lines, and the solid curve represents the combined fit.

291.98 (Ni XVIII 3s
2

2

S

315.02 (Mg VIII 2s 2p 319.84 (Si VIII 2s2 2p3

1/2 - 2P 3/2 4S 3/2

3p

2

P3/2 ),
2 4

- 2s2p2 - 2s2p4
2 1

P3/2 ), P5/2 ),

327.03 (Fe XV 3s3p 3 P2 - 3p The resulting polynomial is

D2 ).

consideration is almost constant and is determined by the point spread function and the Doppler broadening. The Doppler broadening is given by
Dop

= 262.571 + 0.043049P +1.600 â 10-6 P 2 , (3) where P is the coordinate in pixels, and is the wave° length in A. The calibration accuracy was checked by comparing the measured wavelengths with the wavelengths from the CHIANTI database for all of the reliably identified lines. We see from Fig. 6 that ° the difference does not exceed ±0.040 A almost for all of the lines except the line corresponding to the Fe XIII 3s2 3p2 3 P1 - 3s3p3 3 P1 transition (the CHI° ANTI wavelength is 312.11 A). The measured wave° length turns out to be closer to 312.2 A, in agreement ° given by Thomas with the wavelength of 312.17 A and Neupert (1994), within the adopted error limits ° of 40 mA.

=

c

2kT M

1/2

.

(1)

For an iron ion at temperature T = 107 Kand = ° ° 300 A, the Doppler width is Dop 0.05 A. The mean ° line width in the experimental spectrum is 0.1 A. Thus, the width is determined mainly by the point spread function. The total line flux is (2) I = 2A, where A is the peak line intensity. Since the line widths are constant, we compare below the peak intensities rather than the total fluxes.

Determining the Wavelengths
Based on reference lines, we calibrated the wavelength scale of the instrument using a quadratic polynomial. The sufficiency of the quadratic polynomial was checked in a geometrical model of the instrument. As the reference lines, we chose the lines that satisfied the following criteria: a sufficient intensity, observability in previous experiments, a high accuracy of determining the position, and coverage of the entire range bya complete set of lines. We selected the following reference lines: 288.17 (Ni XVI 3s2 3p 2 P1/2 - 3s3p2 2 D3/2 ), 288.42 (S XII 2s2 2p
2

Calibrating the Spectral Sensitivityof the Instrument The wavelength dependence of the multilayer mirror reflectance makes a major contribution to the spectral sensitivity. When calibrating the instrument, we measured the angular dependence of the mirror reflectance at a fixed wavelength. The angular dependence was recalculated to the wavelength dependence using the Bragg condition 2d cos = m,
where d is the period of the multilayer structure, m is the order of interference, and is the angle of incidence. The recalculation result is shown in Fig. 7. To check the calibration, we used the intensity ratios of the lines with a common upper transition

P1/2 - 2s2p

2

2

D3/2 ),

ASTRONOMY LETTERS

Vol. 31 No. 1 2005


42 Table 2. Checking the calibration of the instrument Lines Fe XIII 312.11/321.40 Fe XIII 311.55/320.81 Mg VIII 317.03/313.74 Si VIII 314.36/319.84 Si VIII 316.22/319.84 SXI 285.59/281.40 SXII 288.42/299.78 Fe XV 327.03/312.57b Fe XV 307.75/321.77 Fe XV 292.28/304.89b Ni XVIII 291.98/320.57

ZHITNIK et al.

Without calibration 3.09 ± 0.11 0.23 ± 0.02 0.63 ± 0.03 0.36 ± 0.01 0.84 ± 0.02 0.73 ± 0.08 7.27 ± 0.45 0.75 ± 0.03 1.04 ± 0.05 0.26 ± 0.01 2.62 ± 0.16

With calibration 2.42 ± 0.09 0.18 ± 0.02 0.67 ± 0.03 0.31 ± 0.01 0.75 ± 0.02 0.61 ± 0.07 8.61 ± 0.53 1.22 ± 0.05 0.79 ± 0.04 0.28 ± 0.01 2.24 ± 0.12

Theory 1.94 0.13 0.71 0.35 0.66 0.49 9.26 1.81 0.69 0.35 2.00

Relative 1.25 ± 0.05 1.38 ± 0.15 0.94 ± 0.04 0.89 ± 0.03 1.14 ± 0.03 1.24 ± 0.14 0.93 ± 0.06 0.67 ± 0.03 1.14 ± 0.06 0.80 ± 0.03 1.12 ± 0.06

Note. Lines--the intensity ratio under consideration; without calibration--the intensity ratio without the calibration factors; with calibration--the intensityratio with the calibration factors; theory--the theoretical intensityratio; relative--the experimental intensity ratio with the calibration factor (the column "With calibration") divided bythe theoretical value (the column "Theory"); b--the line is blended.

level or other ratios that do not depend on the emitting plasma parameters. The results are summarized in Table 2. With the exception of the ratios that include blended lines, the experimental values agree with the theoretical values within 3 , which corresponds to a calibration accuracyof about 20%.

RESULTS AND DISCUSSION The identification was based on the following criteria: (1) The coincidence of wavelengths with previously measured ones. The CHIANTI database was used as the source of wavelengths.
1.0 Mirror reflectance

0.08 Difference, å

0

0.6

­ 0.08 280 300 320 Wavelength, å

0.2 260 320 290 Wavelength, å

Fig. 6. Difference between the experimental and theoretical wavelengths for the December 28, 2001, flare.

Fig. 7. Relative reflectance of the multilayer mirror versus wavelength. ASTRONOMY LETTERS Vol. 31 No. 1 2005


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA

43

Intensity, arb. units

600 400

800

400 200 0 280 282 0 284 284 Wavelengths, å

286

288

Fig. 8. Spectrum of the December 28, 2001, flare. The vertical bars mark the lines included in Table 3.

1800 Intensity, arb. units 800 1200 600 0 290 292 400

0 295 294 Wavelengths, å

297

299

Fig. 9. Spectrum of the December 28, 2001, flare (the continuation of Fig. 8).

(2) The presence of multiplet lines in the experimental spectrum. (3) The match between the experimental and theoretical intensity ratios of the lines with a common upper level. In several cases where criteria 2 and 3 could not be used, we compared the line intensities in various areas on the Sun (flares, active regions, the quiet Sun) and the intensities of the lines belonging to the same isoelectronic sequence. In addition, we imposed the requirement on each line that it exceeded in amplitude a 3 noise level (10 counts). The spectrum is shown in Figs. 8­14; the observed lines are marked by vertical bars. The identification results are presented in Table. 3. This table contains the following information: the line numbers; the measured wavelengths, the peak line intensities with errors and relative calibration of the instrument, the line widths with errors, the identified ions, the fractional parts of the CHIANTI wavelengths of the identified lines, the fractional parts of the line wavelengths from the spectral catalog of a
ASTRONOMY LETTERS Vol. 31 No. 1 2005

flare (Dere 1978), the fractional parts of the line wavelengths from the spectral catalog of an active region by Thomas and Neupert (1994), and the fractional parts of the line wavelengths from the spectral catalog of an active region by Brosius et al. (2000); the lines observed only in the flare spectra are marked by "." Below, we comment on the line identification with an indication of the transitions considered. The ions are arranged in order of increasing atomic number of the element and charge. For each ion, we gives the temperature of the maximum ion abundance (CHIANTI) and the terms between which the transitions are observed. The multiplet structure is presented in the form of tables that list the total angular momenta of the transitions and the corresponding CHIANTI wavelengths. When comparing the experimental line intensity ratios that are insensitive to the plasma parameters, we used the spectra of the December 28, 2001, flare and the limb active regions: at 14:38:11 on December 29, 2001; at 16:16:43 on December 29, 2001; and at 04:52:44 on December 30, 2001. In several cases, we compare the intensities of the line under discussion in the December 28, 2001, flare and in the


44

ZHITNIK et al.

Intensity, arb. units

800

800

400

400

0 300 302

0 304 305 Wavelengths, å

307

309

Fig. 10. Spectrum of the December 28, 2001, flare (the continuation of Fig. 9).

Intensity, arb. units

800

800

400

400

0 310 312

0 315 314 Wavelengths, å

317

319

Fig. 11. Spectrum of the December 28, 2001, flare(thecontinuation ofFig. 10).

Intensity, arb. units

800

400

400

200

0 320 322

0 325 324 Wavelengths, å

327

329

Fig. 12. Spectrum of the December 28, 2001, flare(thecontinuation ofFig. 11).

active region observed at 14:38:11 on December 29, 2001. Below, the following notation is used: "+"-- the line is observed; "-"--the line is not observed; "b"--the line is blended; "l"--the line does not fall ° within the recorded wavelength range 279­330 A;

"u"--there are questions in the line identification that are discussed in the text. He II (log T = 4.9). 1s 2S - 2p 2P .
Jlow - Jup 1/2 - 1/2 1/2 - 3/2 ° , A 303.79 303.78
Vol. 31 No. 1

Note +b +b
2005

ASTRONOMY LETTERS


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA

45

200 Intensity, arb. units
Intensity, arb. units 12 000 8000 4000 0

100

0 330 332 Wavelengths, å 334

302

303 304 Wavelengths, å

305

Fig. 13. Spectrum of the December 28, 2001, flare (the continuation of Fig. 12).

Fig. 14. Spectrum of the December 28, 2001, flare in the region of the resonance HeII line.

The doublet is not resolved. Since the image of the solar disk is very intense in the helium line, only the peak intensity was determined for it. CIV (log T = 5.1). 1s2 2p 2P - 1s2 4s 2S .
Jlow - Jup 1/2 - 1/2 3/2 - 1/2 ° , A 296.86 296.91 Note +b +b

Mg VIII (log T = 5.9). 2s2 2p 2 P - 2s2p
Jlow - Jup 1/2 - 1/2 3/2 - 1/2 1/2 - 3/2 3/2 - 3/2 ° , A 313.74 317.03 311.77 315.02

22

P.

Note + + +b +

The doublet is not resolved, which leads to a large line width. Mg VII (log T = 5.8). 2s2 2p2 1D2 - 2s2p3 1D2 . ° The 319.02 A line is blended with the Ni XV 2 3 P - 3s3p3 3 D line. 3s 3p 2 3 ° 2s2 2p2 1 D2 - 2s2p3 1 P1 . The 280.74 A line.
2

° 2s2 2p2 1 S0 - 2s2p3 1 P1 . The 320.52 A line is 2S 2P blended with the Ni XVIII 3s 1/2 - 3p 1/2 line. The upper level of the 2s2 2p2 1 S0 - 2s2p3 1 P1 transition coincides with the upper level of the 2s2 2p2 1 D2 - ° 2s2p3 1 P1 (280.74 A) transition. The intensity of the ° Mg VII 320.52 A line was estimated from the inten° sity of the Mg VII 280.74 A line and the branching ratio (CHIANTI): I
Mg VII

All multiplet components are observed. The ° Mg VIII 311.77 A line is blended with the Ni XV ° 3s2 3p2 3 P1 - 3s3p3 3 D2 311.76 A line. The blending is confirmed bythe difference between the experimental and theoretical intensityratios I (311.77)/I (315.02) by a factor of 1.6 (Table 4). The intensity of the blending line was estimated from the intensity of ° the Mg VIII 315.02 A line and the branching ratio (CHIANTI): INiXV (311.77) = Itotal (311.77) - 0.19IMgVIII (315.02) = 87 ± 6. T he experimental intensity ratio I (317.03)/I (313.74) is equal to its theoretical value within 1.33 (Table 4). Al VII (log T = 5.8). 2s2 2p3 2 D - 2s2p4 2 D.
Jlow - Jup 3/2 - 3/2 5/2 - 3/2 3/2 - 5/2 5/2 - 5/2 ° , A 309.02 309.09 309.06 309.13 Note +bu +bu +bu +bu

(320.52) = 0.20IMg

VII

(280.74) = 38 ± 3.

The total line intensity is Itotal (320.52) = 1828 ± 261. The contribution of the Mg VII 2s2 2p2 1 S0 - 2s2p3 1 P1 transition to the total intensity is about 2% and is less than the measurement error of the total ° intensity. Therefore, the 320.52 A line was identified 2S 2P as Ni XVIII 3s 1/2 - 3p 1/2 .
ASTRONOMY LETTERS Vol. 31 No. 1 2005

The lines of this multiplet could have contributed ° to the observed unidentified 309.14 A line.


46 Table 3. Parameters of the observed lines N 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 ° , A 279.27 279.76 280.19 280.74 281.14 281.43 282.45 283.17 284.20 285.55 285.83 286.43 287.27 287.86 288.16 288.42 288.96 289.17 289.74 290.32 290.73 291.03 291.61 291.98 292.31 292.51 292.82 293.17 293.75 293.94 294.46 294.78 295.07 295.32 296.11 296.99 297.39 297.78


ZHITNIK et al.

A,arb.units 50 ± 12 365 ± 17 576 ± 19 188 259 121 42 34 13 696 118 179 30 52 72 321 546 99 214 116 110 194 298 290 4605 357 651 311 117 54 17 47 137 199 136 586 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 13 18 10 8 15 140 11 12 9 9 8 29 18 17 19 9 8 10 20 13 433 55 18 14 10 9 8 8 10 11 9 13

,pixels 0.96 ± 0.26 1.76 ± 0.07 2.22 ± 0.06 1 1 2 2 1 .74 .60 .95 .68 .33 ... 2.33 2.08 1.91 .55 .11 .97 .83 .92 .95 .08 .60 .92 .10 .07 ... .04 .14 .42 .13 .56 .36 .84 .08 .73 .16 .62 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.11 0.11 0.38 0.70 0.70 ... 0.35 0.21 0.64 0 0 0 0 0 0 0 .30 .45 .23 .16 .47 .23 .21

Ion ... Cr XXII Fe XVII Al IX Mg VII ... SXI Al IX ... Fe XV SXI SXI Fe XVII AlIX Fe XII ... Ni XVI SXII ... Fe XIV Cr XIV Mn XIII ... Si IX Fe XII SXI Ni XVIII Fe XV Fe XXII Si IX Cr XXI ... ... ... ... ... ... Si IX Si IX CIV CIV ... ...

° C , A ... 0.74 0.14 0.14 0.74 0.09 0.40 0.42 ... 0.16 0.59 0.82 0.42 0.38 0.26 ... 0.17 0.42 ... 0.15 0.75 0.72 ... 0.69 0.05 0.58 0.98 0.28 0.49 0.81 0.11 ... ... ... ... ... ... 0.11 0.21 0.86 0.95 ... ...

I 0.21 0.70 0.14 0.74 ... 0.42 ... ... 0.17 0.57 0.83 ... 0.23 0.23 0.14 0.40 ... 0.15 0.72 ... 0.71 0.00 ... 0.98 0.26 0.48 0.79 0.15 0.78 ... 0.47 0.72 0.01 ... 0.12 ... ... 0.34 ...
Vol. 31

II ... ... ... 0.75 ... 0.44 0.43 0.15 0.16 0.58 0.83 ... ... ... 0.16 0.40 ... 0.17 ... ... 0.69 0.01 ... 0.99 0.31 ... 0.80 ... ... ... ... ... ... ... 0.14 ... ... ... ...
No. 1 2005

III ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­





1 2 1 2 1 1 2 2 2 2 2 1 2 2 2 1 1 1 2 1 2 2



0.30 0.49 0.18 0.17 ... 0.18 0.19 0.26 0.27 0.48 1.12 0.47 0.26 0.18 0.24 0.05

46 ± 5 257 ± 12 52 ± 7

4.39 ± 0.80 1.91 ± 0.10 2.38 ± 0.79

ASTRONOMY LETTERS


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA Table 3. (Contd.) N 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 ° , A 298.10 298.27 298.46 299.28 299.55 299.79 300.00 300.32 300.89 301.23 302.2 303.4 303.8 304.93 305.83 306.32 307.00 307.79 308.34 308.58 309.14 309.29 311.56 311.78


47

A,arb.units 75 ± 62 218 ± 133 176 ± 44 60 ± 10 300 ± 13 71 ± 44 90 ± 10 121 ± 11 22 ± 8 62 ± 10 3212 ± 301 1614 ± 303 16 468 ± 505 972 ± 19 60 ± 9 448 ± 15 24 ± 11 227 ± 14 181 ± 15 250 ± 12 194 ± 30 395 ± 38 95 ± 28 121 ± 3 87 ± 6


,pixels 2.04 ± 1.95 1.75 ± 0.76 2.54 ± 0.62 2.35 ± 0.78 2.07 ± 0.25 1.48 ± 0.80 2.47 ± 1.45 2.23 ± 0.52 1.99 ± 0.90 1.55 ± 0.29 ... ± ... ... ± ... ... ± ... 2.18 ± 0.43 2.01 ± 0.39 1.93 ± 0.06 1.64 ± 0.90 1.87 ± 0.14 1.81 ± 0.22 2.05 ± 0.21 4.07 ± 0.38 1.87 ± 0.15 1.60 ± 0.40 2.16 ± 0.39 2.91 ± 0.16

Ion Ni XV ... ... ... SXII SXII ... ... ... ... Ca XVIII Si XI He II Fe XV ... ... ... Fe XV ... Fe XI ... ... Fe XIII Mg VIII Ni XV Fe XIII Fe XV Co XVII Fe XIII Mg VIII Si VIII Mg VIII Si VIII Mg VIII ... Fe XV

° C , A 0.15 ... ... ... 0.54 0.78 ... ... ... ... 0.19 0.33 0.78 0.89 ... ... ... 0.75 ... 0.55 ... 0.24 0.55 0.77 0.76 0.11 0.56 0.54 0.87 0.74 0.36 0.02 0.22 0.03 ... 0.60

I ... ... ... ... 0.50 ... ... ... ... ... 0.21 0.33 0.79 0.80 ... 0.29 ... 0.76 ... 0.54 ... ... ... 0.74

II 0.12 0.20 ... ... 0.53 ... ... 0.25 ... ... 0.17 0.32 0.78 0.87 0.84 ... ... ... ... 0.58 0.21 ... 0.56 0.78

III ­ ­ ­ ­ 0.55 ... ... ... ... ... 0.32 0.32 0.79 0.87 0.79 ... 0.97 0.76 ... 0.55 0.23 ... 0.58 0.77

312.20 312.58

402 ± 14 315 ± 32 319 ± 37 107 ± 8 288 ± 13 110 ± 11 636 ± 18 359 ± 12 170 ± 10 30 ± 15 63 ± 16


0.18 0.57 ... ... 0.74 0.31 0.02 0.19 0.02 ... 0.61

0.17 0.55 ... 0.87 0.74 0.36 0.02 0.22 0.01 ... ...

0.17 0.57 ... 0.90 0.73 0.36 0.02 0.21 0.02 0.48 0.64

1.84 ± 0.08 2.68 ± 0.42 2.16 ± 0.10 1.77 ± 0.22 1.98 ± 0.05 2.18 ± 0.08 2.45 ± 0.17 1.89 ± 1.55 1.94 ± 0.67

312.89 313.76 314.36 315.03 316.19 317.04 317.38 317.61

ASTRONOMY LETTERS

Vol. 31 No. 1 2005


48 Table 3. (Contd.) N 73 74 ° , A 318.13 319.04 A,arb. units 331 ± 14 288 ± 13 453 ± 14 75 ± 12 1828 ± 261 585 ± 258 ... ± ... 274 ± 13 24 ± 11 57 ± 10


ZHITNIK et al.

,pixels 1.99 ± 0.08 1.96 ± 0.15 2.48 ± 0.10 1.71 ± 0.61 ... ± ... 2.59 ± 0.45 ... ± ... 2.10 ± 0.14 1.20 ± 0.62 1.84 ± 0.37 2.29 ± 0.10 1.41 ± 0.46 2.85 ± 0.23 2.72 ± 0.32 1.86 ± 0.59 2.28 ± 1.03 3.40 ± 0.49 2.33 ± 0.36 1.84 ± 0.23 1.94 ± 0.61 2.58 ± 0.10 2.21 ± 0.77 3.26 ± 5.00 2.11 ± 3.96 1.94 ± 2.53 3.55 ± 0.86 1.61 ± 0.84 2.78 ± 1.25

Ion Fe XIII Ni XV Mg VII

° C , A 0.13 0.06 0.02 0.84 ... 0.57 0.81 0.40 0.77 ... ... ... ... 0.98 ... ... ... ... ... 0.03 ... 0.27 ... ... ... ... ... ... ...

I 0.12 0.03

II 0.12 0.02

III 0.13 0.03

75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100

319.83 320.20 320.57 320.78 321.50 321.81 322.30 322.68 323.59 323.86 324.91 325.40 325.68 325.91 326.35 326.78 327.03 327.90 328.27 328.74 329.43 329.70 329.92 330.22 330.75 331.07


Si VIII ... Ni XVIII Fe XIII Fe XIII Fe XV ... ... ... ... Fe XV ... ... ... ... ... Fe XV ... Cr XIII ... ... ... ... ... ... ...

0.84 ... 0.55 0.80 0.47 0.78 ... 0.72 0.57 0.85 0.97 0.40 ... ... 0.36 0.80 0.03 0.86 0.26 ... ... ... 0.92 0.17 ... 0.03

0.84 ... 0.56 0.80 0.46 0.78 ... 0.70 ... ... ... 0.42 ... ... 0.35 ... 0.03 ... 0.26 ... ... ... ... ... ... 0.07

0.84 0.25 0.56 0.81 0.48 0.79 ... 0.73 ... ... 0.98 0.39 0.78 ... 0.32 ... 0.03 0.85 0.26 0.75 ... ... 0.92 0.22 ... 0.99

557 ± 20 62 ± 14 133 ± 8 279 ± 14 156 ± 37 62 ± 20 129 ± 12 346 ± 22 572 ± 58 55 ± 12 571 ± 20 40 ± 10 31 ± 29 60 ± 69 59 ± 102 158 ± 18 48 ± 17 73 ± 18

Note. N is the line number; is the experimental wavelength; A is the peak line intensity with its error; is the line width with its error; Ion denotes the identified ion; C is the fractional part of the wavelength of the identified line from the CHIANTI database; I is the fractional part of the line wavelength from the spectral catalog of a flare by Dere (1978); II is the fractional part of the line wavelength from the spectral catalog of an active region by Thomas and Neupert (1994); III is the fractional part of the line wavelength from the spectral catalog of an active region byBrosius et al. (2000); denotes the lines that are clearlyobservable onlyin the flare spectra; is the intensities of the blended lines estimated from the branching ratios (for a further explanation, see the text).

ASTRONOMY LETTERS

Vol. 31

No. 1

2005


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA Table 4. Checking the Mg VIII identification using SPIRIT data Ratio 1 2 3 4 Mean Theory 0.32 0.01 0.67 0.03 0.71 0.19

49
22

Al IX (log T = 6.0). 2s2 2p 2 P - 2s2p
Jlow - Jup 1/2 - 1/2 3/2 - 1/2 1/2 - 3/2 3/2 - 3/2 ° , A 282.42 286.38 280.14 284.03

P.

Note + +b +b -

311.77/315.02 0.32 0.33 0.33 0.29 0.02 0.02 0.02 0.02

317.03/313.74 0.68 0.63 0.68 0.66 0.10 0.07 0.04 0.05

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 1­4 values; theory--the theoretical ratio based on CHIANTI data.

Table 5. Checking the Al IX identification using SPIRIT data Ratio 1 2 3 4 Mean Theory

° The Al IX 286.38 A line is blended with the 5 3p 3 D - 2p5 3d 3 F 286.42 A line. The ° Fe XVII 2p 1 2 experimental line intensity ratio I (286.38)/I (282.42) is equal to its theoretical value within 1 (Table 5). Therefore, the contribution of the Fe XVII ° 2p5 3p 3 D1 - 2p5 3d 3 F2 286.42 A line to the total in° tensity must be negligible. The Al IX 280.14 A line is 5 3p 3 P - blended with the Fe XVII 2p 2 ° ° 2p5 3d 3 D3 280.14 A line. The Al IX 284.03 A line is ° blended with the Fe XV 3s2 1 S0 - 3s3p 1 P1 284.16 A line. Si VIII (log T = 5.9). 2s2 2p3 4 S - 2s2p4 4 P .
Jlow - Jup ° , A 314.36 316.22 319.84 Note + +u +

286.38/282.42 0.71 0.88 0.81 1.57 0.25 0.25 0.21 0.49

0.86 0.13

0.81

3/2 - 1/2 3/2 - 3/2 3/2 - 5/2

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 1­4 values; theory--the theoretical ratio based on CHIANTI data.

Table 6. Checking the Si VIII identification using SPIRIT data Ratio 1 2 3 4 Mean Theory 0.75 0.02 0.31 0.01 0.35 0.66

316.22/319.84 0.78 0.84 0.74 0.69 0.04 0.04 0.03 0.03

All multiplet component are observed. The intensity ratio of the multiplet lines depends weakly on the density (Brosius et al. 2000). The experimental ratio I (316.22)/I (319.84) differs from its theoretical value by 4.5 (Table 6); this could result from the ° blending of the 316.22 A line with an unknown line ° (e.g., 316.1 A) that is not observed separately in the published spectrum, but was observed in experiments (Thomas and Neupert 1994; Brosius et al. 2000). The experimental ratio I (314.36)/I (319.84) differs from its theoretical value by 4 (Table 6). Si IX (log T = 6.1). 2s2 2p
Jlow - Jup 1-0 0-1 1-1 2-1 1-2 2-2
23

P - 2s2p

33

P.

° , A 292.81 290.69 292.86 296.21 292.76 296.11

Note +b + +b +b +b +b

314.36/319.84 0.28 0.32 0.31 0.31 0.03 0.02 0.02 0.03

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 1­4 values; theory--the theoretical ratio based on the data of Brosius et al. (2000). ASTRONOMY LETTERS Vol. 31 No. 1 2005


50

ZHITNIK et al.

Table 7. Checking the S XI identification with SPIRIT data Ratio 1 2 3 4 Mean Theory 0.61 0.07 0.49

285.59/281.40 1.00 0.54 0.58 0.74 0.12 0.12 0.11 0.13

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 2­4 values; theory--the theoretical ratio based on CHIANTI data.

Table 8. Checking the S XII identification with SPIRIT data Ratio 1 2 3 4 Mean Theory 8.61 0.53 9.26

° We see that the intensity of the S XI 291.81 A line is much lower than that of the close Ni XVIII ° 3s 2 S1/2 - 3p 2 P3/2 291.98 A line, I (291.98) = 4605 ± ° 433. Therefore, the S XI 291.81 A line is not observed ° line is wide, which in the spectrum. The 281.40 A may be due to the contribution from a close line, for example, Ni XXIII 2s2 2p2 3 P1 - 2s2 2p2 1 S0 at ° 281.56 A in the CHIANTI list. The intensity ratio I (285.59)/I (281.40) differs from its theoretical value by 1.71 (Table 7). The ratio for the flare differs from the ratios for the active regions, which may be due ° to the blending of the S XI 285.59 A line in the flare ° with an unknown line. The S XI 291.57 A and S XI ° 291.58 A lines are not resolved. The intensities of ° ° S XI 291.57 A and S XI 291.58 A were determined ° from the intensity of the S XI 281.40 A line and the branching ratio (CHIANTI): I
SXI

288.42/299.78 - 8.34 8.83 8.70 - 0.88 0.89 1.04

(291.57) = 0.014ISXI (281.40) = 1.7 ± 0.1, I
total

° We see that the intensity of the S XI 291.57 A 2 3 P - 2s2p3 3 D line is much lower than the 2s 2p 2 1 total intensity of the lines and the measurement error ° of the latter. Therefore, the 291.58 A line was identi2 2p2 3 P - 2s2p3 3 D 291.58 A. ° fied as 2s 2 3 All multiplet components are observed. The 292.86, S XII (log T = 6.4). 2s2 2p 2 P - 2s2p2 2 D. ° lines as well as the 296.11 and 292.81, and 292.76 A ° Jlow - Jup , A Note ° 296.21 A lines are not resolved. 1/2 - 3/2 288.42 + Si XI (log T = 6.2). 2s2 1 S0 - 2s2p 1 P1 . 3/2 - 3/2 299.78 + ° The 303.33 A line. Since this line is close to the ° 3/2 - 5/2 299.54 + resonance He II 303.8 A line, only the peak intensity is given for it in Table 3. All multiplet components are observed. The exSXI (log T = 6.3). 2s2 2p2 3 P - 2s2p3 3 D. perimental ratio I (288.42)/I (299.78) differs from its theoretical value by 1.23 (Table 8). Table 8 gives ° Jlow - Jup , A Note no ratio for the published flare spectrum, because ° 0-1 281.40 + the 300.00 A line strengthens significantly (Fig. 15), causing the measurement error in the parameters of 1-1 285.59 +u ° the 299.78 A line to increase greatly. 2-1 291.57 +b Ca XVIII (log T = 7.0). 1s2 2s 2 S - 1s2 2p 2 P .
2

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 2­4 values; theory--the theoretical ratio based on CHIANTI data.

(291.58) = 290 ± 13.

1-2

285.82

+

2-2 2-3

291.81 291.58

- +b

Jlow - Jup 1/2 - 1/2 1/2 - 3/2

° , A 344.76 302.19

Note -l +

All of the multiplet components, except for S XI ° 291.81 A, are observed. The intensity of the S XI ° 291.81 A line was determined from the intensity of ° the S XI 285.82 A line and the branching ratio (CHIANTI): ISXI (291.81) = 0.14ISXI (285.82) = 25 ± 2.

° The Ca XVIII 302.19 A line is observed. This line in the images is observed only in hot regions: active regions, flares, and above-limb structures (Fig. 16). The contribution of the Fe XV 3s3p 3 P0 - ° 3p2 3 P1 302.34 A line to the total intensity does not exceed 10% and does not exceed the error in the
ASTRONOMY LETTERS Vol. 31 No. 1 2005


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA
Flare Active region

51

400 Intensity, arb. units

200

0 299.5 300.5 Wavelength, å 301.5

Fig. 16. The pointlike region to the left of the bright central line corresponds to the flare image in the Ca XVIII ° 302.19 A line.

° Fig. 15. Comparison of the S XII 299.78 A line in the active region and in the flare. The vertical bars mark the observed lines for which the fractional part of the wavelength (to the second decimal place) is given.

° intensity of the 302.19 A line; an estimate of the contribution is given when discussing the Fe XV identification. Both Ca XVIII doublet lines were observed byDere (1978); their intensity ratio is about 2, which confirms the identification. Cr XIII (log T = 6.2). 3s2 1 S0 - 3s3p 1 P1 . The ° 328.27 A line. Cr XIV (log T = 6.4). 2p6 3p 2 P - 2p6 3d 2 D.
Jlow - Jup 1/2 - 3/2 3/2 - 3/2 3/2 - 5/2 ° , A 289.75 301.83 300.30 Note +b - -u

observed in the spectrum. However, as Keenan et al. (2003) showed, it follows from the SERTS mea° surements that the observed 300.32 A line does not belong to CrXIV; therefore, the identification of this line is open to question. Cr XXI (log T = 7.0). 2s2 1 S0 - 2s2p 3 P1 . The ° 293.11 A line. A comparison of the line intensity in the active region and in the flare (Fig. 17) argues for the suggested identification. Cr XXII (log T = 7.1). 1s2 2s 2 S - 1s2 2p 2 P .
Jlow - Jup 1/2 - 1/2 1/2 - 3/2 ° , A 279.74 222.98 Note + -l

° Only the Cr XIV 289.75 A line, blended with the 2 3p 2 P 2 2P ° Mn XIII 3s 3/2 - 3s3p 1/2 289.72 A line, is ° observed. The Cr XIV 301.83 A line is not observed. The maximum possible (in view of the blending) in° tensity of Cr XIV 301.83 A was estimated from the ° intensityof the Cr XIV 289.75 A line and the branching ratio (NIST): I (301.83) = 0.18ICrXIV (289.75) 21 ± 2. ° We see that the Cr XIV 301.83 A line is much weaker than the close Ca XVIII 1s2 2s 2 S1/2 - ° 1s2 2p 2 P3/2 302.19 A line. Therefore, it is not ob° served in the spectrum. The Cr XIV 300.30 A line ° could correspond to the unidentified 300.32 A line
ASTRONOMY LETTERS Vol. 31 No. 1 2005

° The 279.74 A line is observed. The intensity ratio ° of the Ca XVIII 1s2 2s 2 S1/2 - 1s2 2p 2 P3/2 302.19 A ° and Cr XXII 279.74 A lines must roughly correspond to an abundance ratio of 105.85­6.30 0.35 (Allen 1973), because both lines belong to the same isoelectronic sequence. Since the 1s2 2s 2 S1/2 - 1s2 2p 2 P1/2 transition is observed for chromium, we obtain 2ICr XXII (279.74) = 0.22 ± 0.02. ICa XVIII (302.19) The theoretical ratio exceeds the measured ratio by a factor of about 1.6. Probable causes of the difference is inaccuracy in the calibration of the spectral sensitivity of the instrument (given when describing the Fe XV identification), inaccuracy in the abundances, and the fact that the equality must be only approximate. A comparison of the line intensities in the active region and in the flare (Fig. 18) argues for the suggested identification. Dere (1978) observed both doublet lines.

CaXVIII

.28 .55 .78 .00 .32

.89

.23


52
Flare Active region

ZHITNIK et al.
Flare Active region

2000 Intensity, arb. units

Intensity, arb. units

400

.27

.76

.19 .45

.74

.14

.43

1000

.61

.31

.51

.82

.17

200

0 291.5 293.0 292.0 292.5 Wavelength, å 293.5

0 279.5 280.5 Wavelength, å 281.5

° Fig. 17. Comparison of the Fe XXII 292.49 A and Cr XXI ° lines in the active region and in the flare. 293.11 A The vertical bars mark the observed lines for which the fractional part of the wavelength (to the second decimal place) is given.

° Fig. 18. Comparison of the Cr XXII 279.76 A line in the active region and in the flare. The vertical bars mark the observed lines for which the fractional part of the wavelength (to the second decimal place) is given.

Mn XIII (log T = 6.2). 3s2 3p 2 P - 3s3p
Jlow - Jup 1/2 - 1/2 3/2 - 1/2 1/2 - 3/2 3/2 - 3/2 ° , A 277.43 289.72 272.09 283.91

22

P.

Note -l +b - -

The large errors of these ratios stem from the fact that ° the Fe XII 287.26 A line has a low intensity. Fe XIII (log T = 6.2). 3s2 3p2 3 P - 3s3p3 3 P .
Jlow - Jup 1-0 0-1 1-1 2-1 1-2 2-2 ° , A 312.87 303.30 312.11 321.40 311.55 320.81 Note + - +u + + +

° The Mn XIII 289.72 A line is blended with the 6 3p 2 P 6 3d 2 D ° Cr XIV 2p 1/2 - 2p 3/2 289.75 A line. Mn XIV (log T = 6.3). 3s S0 - 3s3p P1 . The ° 304.85 A line. Blended with the Fe XV 3s3p 3 P2 - 2 3 P 304.89 A line. As we showed when describing ° 3p 2 ° the Fe XV identification, the Mn XIV 304.85 A line contributes no more than 20%. FeXI (log T = 6.1). 3s2 3p4 1 D2 - 3s3p5 1 P1 .The ° 308.55 A line. Fe XII (log T = 6.2). 3s2 3p3 2 D - 3s3p4 2 P .
21 1

Jlow - Jup 3/2 - 1/2 3/2 - 3/2 5/2 - 3/2

° , A 283.45 287.26 291.05

Note - + +

All of the multiplet components, except for the ° ° 303.30 A line blended with the SiXI 303.33 A line, are observed. The experimental intensity ratio I (312.11)/I (321.4) differs from its theoretical value by 5.33 (Table 10). A probable cause of the dif° ference is the blending of the 312.11 A line with an unknown line. The experimental intensity ratio I (320.809)/I (311.553) differs from its theoretical value by 2.5 (Table 10). ° 3s2 3p2 1 D2 - 3s3p3 1 D2 . The 318.13 A line. Fe XIV (log T = 6.3). 3s2 3p 2 P - 3s3p ° the Fe XIV 289.15 A line is observed.
Jlow - Jup 1/2 - 1/2 3/2 - 1/2 ° , A 274.20 289.15
22

S .Only

° The Fe XII 283.45 A line is not observed. This line ° may be obscured by the strong Fe XV 284.16 A line. The experimental intensityratio I (287.26)/I (291.05) differs from its theoretical value by 3.58 (Table 9).

Note -l +

ASTRONOMY LETTERS

Vol. 31

No. 1

2005


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA

53

FeXV (log T ° 284.16 A line. Si very intense, only for this line. 3s3p 3 P - 3p2
Jlow - Jup 1-2 2-2

= 6.4). 3s2 1 S0 - 3s3p 1 P1 . The nce the image of the solar disk is the peak intensity was determined
1

D.
° , A 312.57 327.03 Note +b +

(Table 11). This could be attributable to a calibration ° error or optical vignetting near 284.16 A. It is also possible that the plasma is not optically thin in this line. 3s3p 3 P - 3p2 3 P .
Jlow - Jup 1-0 0-1 1-1 2-1 1-2 2-2 ° , A 317.60 302.34 307.75 321.77 292.28 304.89 Note + - + + + +b

All multiplet components are observed. The Fe XV ° 312.57 A lineisblended with theCo XVII 3s 2 S1/2 - ° 3p 2 P3/2 312.54 A line, as confirmed bythe difference between the experimental and theoretical intensity ratios I (327.03)/I (312.57) by more than 10 (Table 11). The difference between the experimental and theoretical ratios I (327.03)/I (312.57) for the flare spectrum is larger than that for the spectra of the active regions (Table 11), because the blending line gives a larger contribution (the decimal logarithm of the temperature of the maximum abundance for Co XVII 6.7 against 6.4 for Fe XV (CHIANTI)). The ° intensity of the Fe XV 312.57 A line was estimated ° line and the branchfrom that of the FeX V 327.03 A ing ratio (CHIANTI): I
Fe XV

(312.57) = 0.55IFe

XV

(327.03) = 315 ± 32.

The ratio I (327.03)/I (284.16) depends weakly on the temperature (less than 10%) and the density (no more than 20%), being about 0.01 (Brickhouse et al. 1995). The close match between the theoretical and experimental values in previous experiments (Table 12) confirms the identification of the Fe XV ° 327.03 A line. In the SPIRIT experiment, the ratio differed from its theoretical value by 7.67 , although it remained constant for the whole series of spectra
Table 9. Checking the Fe XII identification with SPIRIT data Ratio 1 2 3 4 Mean Theory 6.15 0.65 8.48

All of the multiplet components, except for the ° ° 302.34 A line, are observed. The Fe XV 302.34 A line must be a factor of 1.4 more intense than the ° Fe XV 307.75 A line; i.e., its intensity must be about ° 318 ± 20, while the intensity of the 302.3 A line (the experimental wavelength, identified as the Ca XVIII ° 1s2 2s 2 S1/2 - 1s2 2p 2 P3/2 302.19 A line) is 3212 ± 301. Thus, this iron line can contribute about 10%. The experimental intensityratio I (307.75)/I (321.77) differs from its theoretical value by 2.5 (Table 11) and confirmed the identification. The Fe XV 3s3p 3 P2 - ° 3p2 3 P2 304.89 A line is blended with the Mn XIV 2 1 S - 3s3p 1 P 304.85 A line. The Fe XV 292.28 A ° ° 3s 0 1 line in the spectra of the active regions is clearly observed in the wing of the Ni XVIII 3s 2 S1/2 - ° 3p 2 P3/2 291.98 A line, and the line position, intensity, and width can be determined accurately. In the flare spectrum, an intense Fe XXII 2s2 2p 2 P3/2 -
Table 10. Checking the Fe XIII identification with SPIRIT data Ratio 312.11/321.40 1 - - 2 3 4 Mean Theory 2.42 0.09 0.18 0.02 0.13 1.94

2.51 2.64 2.15 0.19 0.17 0.16

291.05/287.26 5.68 7.68 10.03 5.47 1.01 1.69 2.76 1.07



311.55/320.81 0.16 0.17 0.22 0.16 0.09 0.03 0.04 0.03
Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 1­4 values; theory--the theoretical ratio based on CHIANTI data. ASTRONOMY LETTERS Vol. 31 No. 1 2005

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 1­4 values; theory--the theoretical ratio based on CHIANTI data.


54

ZHITNIK et al.

Table 11. Checking the Fe XV identification using SPIRIT data Ratio 327.03/312.57 327.03/284.16 307.75/321.77 292.28/304.89 1 0.89 0.15 0.041 0.008 0.83 0.06 ­ ­ 2 1.37 0.08 0.034 0.004 0.74 0.08 0.34 0.03 3 1.15 0.07 0.030 0.004 0.75 0.11 0.26 0.02 4 1.29 0.08 0.033 0.004 0.81 0.11 0.27 0.02 mean 1.22 0.05 0.033 0.003 0.79 0.04 0.28 0.01 0.35 0.69 0.01 Theory 1.81

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--the active region observed at 16:16:43 on December 29, 2001; 4--the active region observed at 04:52:44 on December 30, 2001; mean--the value obtained byaveraging 1­4values; theory--the theoretical ratio based on CHIANTI data.

° 2s2p2 4 P3/2 292.49 A line appears near the Fe XV ° 292.28 A line; as a result, the parameters of Fe XV ° can be determined only with a large error. 292.28 A Therefore, Table 11 gives the experimental ratios I (292.28)/I (304.89) only for the active regions. The experimental intensityratio I (292.28)/I (304.89) differs from its theoretical value by 7 due to the ° contribution of Mn XIV 3s2 1 S0 - 3s3p 1 P1 304.85 A noted above, which can estimated from this difference ° as 20%. The Fe XV 3s3p 3 P1 - 3p2 3 P0 317.60 A line is observed, but a significant error in its intensity, attributable to the presence of the uniden° tified 317.42 A line and the Mg VIII 2s2 2p 2 P3/2 - ° 2s2p2 2 P1/2 317.03 A line near it, cannot be ruled out. 3s3p 1 P1 - 3p
21

° 2p5 3p 3 D - 2p5 3d 3 F . The Fe XVII 286.42 A 2 2p 2 P line is blended with Al IX 2s 3/2 - 2 2P ° the Fe XVII line probably 2s2p 1/2 286.38 A; makes a minor contribution to the total intensity, as we see from the discussion of the Al IX identification. ° The Fe XVII 284.17 A line is obscured by the Fe XV 2 1 S - 3s3p 1 P 284.16 A line. ° 3s 0 1
Jlow - Jup 1-2 2-2 3-2 2-3 3-3 3-4 ° , A 286.42 211.48 217.47 269.42 279.21 284.17 Note +b -l -l -l - -
24

° S0 . The 324.98 A line.

FeXVII (log T = 6.9). 2p5 3p 3 P - 2p5 3d 3 D.
Jlow - Jup 0-1 1-1 2-1 1-2 2-2 2-3 ° , A 285.44 244.46 ... 275.54 287.12 280.14 Note - -l ... -l - +b

Fe XXII (log T = 7.1). 2s2 2p 2 P - 2s2p
Jlow - Jup 1/2 - 1/2 3/2 - 1/2 1/2 - 3/2 3/2 - 3/2 3/2 - 5/2 ° , A 247.20 349.32 217.32 292.49 253.17

P.

Note -l -l -l + -l

° Only the Fe XVII 280.14 A line is observed; it is blended with the Al IX 2s2 2p 2 P1/2 - ° 2s2p2 2 P3/2 280.14 A line.

° The Fe XXII 292.49 A line is observed. A comparison of the line intensities in the active region
ASTRONOMY LETTERS Vol. 31 No. 1 2005


EXTREME VACUUM ULTRAVIOLET SOLAR SPECTRA ° Table 12. Checking the FeXV 327.03 A identification in previous works Ratio 327.03/284.16 1 0.012 0.002 2 0.090 0.002 3 0.010 0.002 4 0.009 0.002 5 0.006 0.002

55

Table 13. Checking the Ni XVIII identification using SPIRIT data Ratio 1 2 3 4 Mean Theory 2.24 0.12 2.00

291.98/320.57 2.51 2.09 1.98 2.29 0.25 0.20 0.37 0.22

Note: 1--Active region (Thomas and Neupert et al. 1994), 2-- active region (Brosius et al. 1996), 3--quiet Sun (Brosius et al. 1996), 4--active region (Brosius et al. 1996), 5--above-limb spectrum (Brosius et al. 1996).

Note: 1--The flare of December 28, 2001; 2--the active region observed at 14:38:11 on December 29, 2001; 3--theactiveregion observed at 16:16:43 on December 29, 2001; 4--theactiveregion observed at 04:52:44 on December 30, 2001; mean--the value obtained by averaging 1­4 values; theory--the theoretical ratio based on the data byNeupert and Kastner (1983).

and in the flare confirms the suggested identification (Fig. 17): this line is clearly observed only in the flare spectrum. Co XVII (log T = 6.7). 3s 2 S - 3p 2 P .
Jlow - Jup 1/2 - 1/2 1/2 - 3/2 ° , A 339.50 312.54 Note -l +

° The Ni XV 324.65 A line is a factor of 95 weaker ° than the Ni XV 298.15 A line (NIST); as we see, the line intensity INi XV (324.65) is verylow: I
Ni XV

(324.65) = 0.011INi

XV

(298.15) = 0.79 ± 0.65.

° The Co XVII 312.54 A line is blended with Fe XV 3 P - 3p2 1 D ° 3s3p 1 2 312.57 A. The intensity of ° Co XVII 312.54 A was estimated from the equation I
Co XVII

° The Ni XV 311.76 A line is blended with the ° line and contributes up to 40% Mg VIII 311.77 A to the total intensity; the equation yields 87 ± 6. The ° Ni XV 324.35 A line must be weaker than the Ni XV ° 311.76 A line bya factor of 95.23 (NIST): I
Ni XV

(312.54) = Itotal (312.57) - IFe XV (312.57) = 319 ± 37.

(324.35) = 0.011INi

XV

(311.76) = 0.96 ± 0.07.

° The second Co XVII 339.5 A double line was observed in the spectra byDere (1978) and Thomas and Neupert (1994). Ni XV (log T = 6.4). 3s2 3p
Jlow - Jup 0-1 1-1 2-1 1-2 2-2 2-3
23

P - 3s3p

33

D.

° , A 298.15 312.03 324.65 311.76 324.35 319.06

Note + - - + - +

° Thus, the Ni XV 324.35 A line is very weak and is not observed in the spectrum. Note that the ° 324.35 and 324.65 A lines were not observed in previous experiments either, probably because of their extremely low intensity (Behring et al. 1976; Dere 1978; Thomas and Neupert 1994; Brosius et al. 2000). Recently, Brooks et al. (1999) reported the observation of the extremely weak unidentified ° 324.32 and 324.64 A lines in the spectrum of the quiet Sun. It maywell be that they correspond to the nickel lines, but the published intensities were measured with errors of 100 and 71%, respectively. Thus, the spectrum presented here exhibits three lines, Ni XV ° 298.15, 311.76, and 319.06 A. Ni XVI (log T = 6.4). 3s2 3p 2 P - 3s3p
Jlow - Jup 1/2 - 3/2 3/2 - 3/2 3/2 - 5/2 ° , A 288.17 313.23 309.18
22

D.

° The Ni XV 319.06 A line is blended with Mg VII 2 2p2 1 D - 2s2p3 1 D 319.02 A. The Ni XV 312.03 A ° ° 2s line must be a factor of 5.76 weaker than the Ni XV ° 298.15 A line (NIST), as the intensity of this line is low and much smaller (3%) than that of the nearby ° strong Fe XIII 312.11 A line: I
Ni XV

Note + - +bu

(312.03) = 0.17INi I
Fe XIII

XV

(298.15) = 13 ± 11,

° The intensity of the Ni XVI 313.23 A line can ° be estimated from that of the Ni XVI 288.17 A line, because theyshare a common level: INi
XVI

(312.11) = 402 ± 14.
Vol. 31 No. 1 2005

(313.23) = 0.03INi

XVI

(288.17) = 10 ± 1.

ASTRONOMY LETTERS


56

ZHITNIK et al. 4. S. Brandt, Statistical and Computational Methods in Data Analysis (North-Holland, Amsterdam, 1976; Mir, "OOO Izdatel'stvo AST," Moscow, 2003). 5. N. S. Brickhouse, J. C. Raymond, and B. W. Smith, Astrophys. J., Suppl. Ser. 97, 551 (1995). 6. D. H. Brooks, G. A. Fishbacher, A. Fludra, et al., Astron. Astrophys. 347, 277 (1999). 7. J. W. Brosius, J. M. Davila, and R. J. Thomas, Astrophys. J., Suppl. Ser. 106, 143 (1996). 8. J. W. Brosius, J. M. Davila, and R. J. Thomas, Astrophys. J., Suppl. Ser. 119, 255 (1998). 9. J. W. Brosius, R. J. Thomas, and J. M. Davila, Astrophys. J. 543, 1016 (2000). 10. CHIANTI, http://www.solar.nrl.navy.mil/chianti.html. 11. K. P. Dere, Astrophys. J. 221, 1062 (1978). 12. G. A. Doschek, Extreme Ultraviolet Astronomy, Ed. byR. F. Malina and S. Bowyer (Pergamon Press, 1991). 13. C. Jordan, Astron. Astrophys. 34, 69 (1974). 14. C. Jordan, Progress in Atomic Spectroscopy,Ed. by W. Hanle and H. Kleinpoppen (Plenum, New York, 1979), Part B. 15. F. P. Keenan, A. C. Katsiyannis, J. W. Brosius, et al., Mon. Not. R. Astron. Soc. 342, 513 (2003). 16. S. V. Kuzin, I. A. Zhitnik, A. A. Pertsov, et al., J. XRaySci. Technol. 7, 233 (1997). 17. J. Lang, H. E. Mason, and R. W. McWhirter, Solar Phys. 129, 31 (1990). 18. M. Malinovsky and L. Heroux, Astrophys. J. 181, 1009 (1973). 19. A. V. Mitrofanov and S. Yu. Zuev, Spectral Properties of Thin-Film X-ray Filters Based of Track Membranes Surface (2002). 20. W. M. Neupert and S. O. Kastner, Astron. Astrophys. 128, 181 (1983). 21. NIST, http://www.physics.nist.gov/cgibin/AtData/ lines_form. 22. V. N. Oraevskii and I. I. Sobel'man, Pis'ma Astron. Zh. 28, 457 (2002) [Astron. Lett. 28, 401 (2002)]. 23. H. W. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flanery, Numerical Recipes in Fortran 77. The Art of Scientific Computing (Univ. Cambridge, 2001). 24. I. I. Sobel'man, I. A. Zhitnik, A. P. Ignat'ev, et al., Pis'ma Astron. Zh. 22, 604 (1996) [Astron. Lett. 22, 539 (1996)]. 25. R. J. Thomas and W. M. Neupert, Astrophys. J., Suppl. Ser. 91, 461 (1994). 26. I. A. Zhitnik, O. I. Bougaenko, V. A. Slemzin, et al., ESA SP-506 2, 915 (2002). 27. I. A. Zhitnik, S. V. Kuzin, V. N. Oraevskii , et al., Pis'ma Astron. Zh. 24, 943 (1998) [Astron. Lett. 24, 819 (1998)]. 28. I. A. Zhitnik, S. V. Kuzin, A. M. Urnov, et al., Mon. Not. R. Astron. Soc. 308, 228 (1999).

° The intensity of the Ni XVI 313.23 A line is at the ° line may be blended, as noise level. The 309.18 A suggested bythe large line width. Ni XVIII (log T = 6.7). 3s 2 S - 3p 2 P .
Jlow - Jup 1/2 - 1/2 1/2 - 3/2 ° , A 320.57 291.98 Note + +

All multiplet components are observed. The experimental intensity ratio I (291.98)/I (320.57) differs from its theoretical value by 2 (Table 13). CONCLUSIONS We have presented a catalog of 100 lines in the ° wavelength range 280­330 A detected bythe RES-C spectroheliograph in the SPIRIT experiment aboard CORONAS-F in flares and active regions. We used the spectra for the X3.4 flare of December 28, 2001, and the active region NOAA 9765. Based on the spectral image reduction, analysis of the relative line intensities, and CHIANTI data, we calibrated the wavelengths and intensities of the instrument. The ° accuracy of determining the wavelengths is 40 mA, ° the spectral resolution is 0.1 A, and the calibration accuracyof the relative intensities is 20%.Among the 100 presented lines, 15 are observed only in flares. We identified 54 lines and noted clear blends. For the December 28, 2001, flare, we gave the relative line intensities. ACKNOWLEDGMENTS We wish to thank I.I. Sobel'man and O.I. Bugaenko for their help. This work was supported in part by the NATO (grant PST.CLG.979372), the Russian Foundation for Basic Research (project nos. 0302-16053, 02-02-16613, 02-02-17272) and the following programs of the Russian Academy of Sciences: Nonstationary Processes in Astronomy, Solar Wind (N16), and Optical Spectroscopy and Frequency Standards. REFERENCES
1. C. W. Allen, Astrophysical Quantities, 3rd ed. (Athlone Press, London, 1973; Mir, Moscow, 1977). 2. W. E. Behring, L. Cohen, and U. Feldman, Astrophys. J. 175, 493 (1972). 3. W. E. Behring, L. Cohen, U. Feldman, and G. A. Dosheck, Astrophys. J. 203, 521 (1976).

Translated by V. Astakhov

ASTRONOMY LETTERS

Vol. 31

No. 1

2005