Advances in Space Res. vol. 38, Issue 7

pp. 1543-1546 2006

Temperature-sensitive Line Ratio Diagnostics based on Si Satellite-to-Resonance Line Ratios for 1s²-1snp Transitions

K. J. H. Phillips^a, J. Dubau^b, J. Sylwester^c, B. Sylwester^c, J. L. Culhane^d, G. A. Doschek^e, J. Lang^f

^aNational Research Council Senior Research Associate, NASA Goddard Space Flight Center, Greenbelt, MD 20771, U.S.A.

^bObservatoire de Paris, Section de Meudon, Place Jules Janssen, F-92195, Meudon, France

^cSpace Research Centre, Polish Academy of Sciences, 51-622, Kopernika 11, Wroc aw, Poland

^dMullard Space Science Laboratory, Holmbury St Mary, Dorking, Surrey RH5 6NT, United Kingdom

^eE. O. Hulburt Center for Space Research, Naval Research Laboratory, Washington, D.C. 20375, U.S.A.

^fSpace Science Dept., Rutherford Appleton Laboratory, Didcot, Oxon. OX11 0QX, United Kingdom

Abstract

Key words: Sun, X-ray, spectra

1  Introduction

Atomic parameters for the satellites have been calculated by one of us (J.D.) and are used here to derive theoretical intensity ratios of the satellites to the He-like ion lines. We use observations of these ratios during the course of long-duration X-ray flares observed by RESIK. We compare the observed and theoretical ratios using temperatures from GOES X-ray data, showing that the agreement is very close.

2  Observations and Analysis

We selected four long-duration flares observed by RESIK in 2002, used in a previous analysis (Phillips et al. 2003) of line intensities to obtain abundances of K, Ar, and S. The flares occurred on April 14/15, July 11, July 22/23, and July 26/27. Emission detected by channel 4 of RESIK covering the Si lines is diffracted by quartz crystal (diffracting plane ), 2d spacing 8.51 Ǻ and rocking curve (FWHM) 0.62 mǺ. The spectral resolution has been empirically determined to be 15.7 mǺ from line widths during periods of relative quiescent solar conditions. Fig. 1 shows nine spectra (data gathering intervals varying between 5 and 20 minutes) during the course of the several-hour decay of the July 26/27 flare, with times (UT) indicated. The Si  XIII lines at 5.681 Ǻ and 5.384 Ǻ are due to 1s²-1s3p and 1s²-1s4p transitions respectively, and are prominent at earlier times (as are He-like S lines at shorter wavelengths). At later times, the Si  XII satellite line features at 5.818 Ǻ and 5.565 Ǻ (called here A and B respectively) are relatively more intense; the 5.818 Ǻ feature is comparable in intensity to the 5.681 Ǻ line some nineteen hours after the flare maximum, near 21:22 UT on July 26. There is a marked decrease in temperature T, as indicated by the ratio of GOES X-ray channels, for later times in this flare, so the intensity ratio of the satellites to the He-like ion lines is clearly inversely related to T.

Figure 1: RESIK channel 4 spectra during the flare of 2002 July 26/27 (times, UT, indicated), illustrating the changing intensities of the Si  XII satellite features (A, 5.818 Ǻ and B, 5.565 Ǻ) relative to the Si  XIII (5.681 Ǻ) and (5.384 Ǻ) lines. These lines are indicated in the 10:03 UT plot. The fluorescence background has not been subtracted from these spectra.

RESIK spectra for times during all four flares were analyzed by finding best-fit Gaussian profiles to the line emission whenever this was feasible. A pre-flare spectrum was subtracted from each flare spectrum analyzed. For most spectra, the measured Si  XIII lines intensities have uncertainties (taking account of the pre-flare spectrum subtraction) of ~ 15%, while the Si  XII satellites have larger uncertainties for times near the peak of each flare when they are relatively much weaker. The ratios A/ and B/ were obtained for each time interval. For these same intervals, the GOES temperature T was obtained, again subtracting a pre-flare level taken at the same time as the RESIK pre-flare spectrum. The A/ ratios are plotted in Fig. 2 for the July 23 and July 26/27 flares, with error bars indicating estimated uncertainties.

Figure 2: Observed (points with error bars) values of Si XII  A/Si  XIII   plotted against T derived from the ratio of the two GOES channels over the same time intervals. Open circles: July 23 flare; filled circles: July 26/27 flare. The curve is the theoretical variation calculated using CHIANTI values for the Si  XIII line flux and data described in the text for the Si  XII satellites forming the line feature A.

3  Theoretical Line Intensities

The Si  XIII and lines are formed mainly by electron collisional excitation. With the spectral resolution of RESIK, there are two transitions involved, due to 1s^2 1S₀ - 1snp ¹P₁ and 1s^2 1S₀ - 1snp ³P₁, the former much more intense than the latter. The Si  XII satellites making up the A and B features are formed by dielectronic recombination. Gabriel (1972) gives details of the excitation rates, which are a function of T and an intensity factor F_sat which depends on radiative and autoionization rate coefficients from the doubly excited upper state. The CHIANTI atomic database and code gives line fluxes for the Si  XIII lines and for the Si  XII satellite feature A but not B. We therefore used calculations of wavelengths, values of F_sat and other atomic data for the satellites making up features A and B done by one of us (J.D.). Generally, satellites with transitions 1s² 2p  ²P_3/2 - 1s 2p  np ²D_5/2 and 1s² 2p  ²P_1/2 - 1s 2p np ²D_3/2 dominate A (n=3) and B (n=4); they correspond to the intense satellites j and k for the more familiar n=2 satellites (Gabriel (1972)). The theoretical ratios depend approximately on F_sat/T. That for all satellites making up feature A are plotted against T in Fig. 2, along with the observed points.

4  Results and Conclusions

There is a very close agreement between the observed points and theoretical curve plotted in Fig. 2 for the Si  XII A/Si  XIII line ratio. Most points agree to within the estimated uncertainties. There are fewer observed values of the Si  XII B/Si  XIII and the uncertainties are larger, but there is a general agreement with the theoretical curve. In this preliminary work, we are aware of the possible uncertainty in values of temperature derived from the two GOES channels using a standard routine in the SolarSoft IDL package. This package is based on ratios empirically derived by Thomas et al. (1985), though newer line and continuum data from CHIANTI will probably modify the dependence somewhat. In later work, we will use the ratio of the total flux in two RESIK channels as temperature-indicators which may be more satisfactory. Meanwhile, we may conclude that RESIK observations during long-duration flares of the Si  XIII 1s²-1s3p and 1s²-1s4p lines and nearby Si  XII satellites indicate their ratios to be a useful diagnostic of temperature.

RESIK observes corresponding satellite line features near He-like Ar (Ar  XVIII) lines (extreme end of channel 1 wavelength range) and He-like S (S  XV) lines (channels 2 and 3) during flares. Sylwester etal.(2004a) gives details. Depending on the flare temperature, these satellite features have intensities often exceeding the He-like ion lines themselves. With such lines, other temperature diagnostic information is therefore available. Recognizing the multi-thermal nature of flare plasmas, the ratios of satellites to He-like ion lines may give much more accurate differential emission measures than has hitherto been possible.

Satellites to the 1s²-1s3p and 1s²-1s4p lines in He-like Fe (Fe  XXV) have not, to the authors' knowledge, been observed in solar flare spectra with crystal spectrometer resolution, but they must have intensities that are stronger than Si, S, or Ar because of the strong Z-dependence of the satellite intensity factor F_sat. The entire complex of Fe XXV lines and satellites form a strong feature in the broad-band resolution of spectra obtained with the RHESSI solar flare mission, photon energies around 8 keV (wavelengths ~ 1.5 Ǻ). The ratio of the 8 keV feature to the stronger 6.7 keV line feature, also observed in RHESSI spectra, is being used to diagnoze the hottest parts of the flare plasma. Dennis et al. (2004) and Phillips (2004)) discuss this in more detail.

Acknowledgements

B.S, J.S. and M.S. acknowledge support from grants 2.P03D.002.22 and BPZ-KBN-054/P03/2001 of the Polish Committee for Scientific Research. K.J.H.P. acknowledges support from the U.S. National Research Council Research Associateship Program.

References

Dennis, B. R., Phillips, K. J. H., Sylwester, J., Sylwester, B., R. A. Schwartz, A. K. Tolbert, Thermal and Nonthermal Contributions to the Solar Flare X-ray Flux, Paper E2.3-0006-04, These Proceedings.

Gabriel, A. H., 1972, Dielectronic satellite spectra for highly-charged helium-like ion lines, M.N.R.A.S., 160, 99.

Phillips, K. J. H., 2004, The Solar Flare 3.8-10 keV X-ray Spectrum, Ap.J., 605, 921.

Phillips, K. J. H., Sylwester, J., Sylwester, B., & Landi, E. 2003, Solar Flare Abundances of Potassium, Argon, and Sulphur, Ap.J., 589, L113.

Sylwester, B., Sylwester, J., Siarkowski, M., Phillips, K. J. H., Culhane, J. L., Lang, J., Brown, C., & Kuznetsov, V. D., 2004, Identification of Lines in the Range 3.35 Å - 6.1 Å Observed in RESIK Spectra, Paper E1.9-0032-04, These Proceedings.

Sylwester, J., Culhane, J. L., Doschek, G. A., Oraevsky, V. N., Phillips, K. J. H., & Sylwester, B. 2002, RESIK observations of highly ionized argon and potassium X-ray emission lines in solar flares, Proc. Tenth European Solar Physics Meeting, Prague: ESA SP-506, 765.

Sylwester, J. et al. 2004, RESIK: A Bent Crystal X-ray Spectrometer for Studies of Solar Coronal Plasma Composition, Solar Phys. (submitted).

Thomas, R. J., Starr, R., & Crannell, C.-J., 1985, Expressions to determine temperatures and emission measures for solar X-ray events from GOES measurements, Solar Phys., 95, 323.

BACK