High-Resolution Solar X-ray Spectroscopy from Archived Solar Maximum Mission Data

TL;DR

The paper reanalyzes archival high-resolution Solar Maximum Mission FCS channel 1 spectra (13.1–22.4 Å) from nonflaring active regions to extract physical parameters using Fe XVI satellites for temperature tagging and CHIANTI v11 for line intensities. It demonstrates Fe XVII lines persist down to $T < 3$ MK, reveals that the underintensity of Fe XVII lines 3C/3D is not due to resonant scattering but likely due to revised collisional excitation rates and Fe XVI satellite blending, and uses Fe XVI satellites to establish a temperature index $T_{1517}$ for spectral ordering. The study also discusses density sensitivity via the Fe XVII 3G/3H ratio, which remains in the low-density limit for these spectra but may reveal higher densities during impulsive flare stages; it identifies and interprets Fe XVI satellites near Fe XVII lines and examines two well-observed ARs to derive temperatures, emission measures, and thermal energies within the FCS field of view. The findings underscore the need for updated atomic data, inform the interpretation of archived solar spectra, and point toward instrument design improvements for future high-resolution crystal spectrometers, while laying groundwork for exploiting flare densities in future observations.

Abstract

Archived high-resolution X-ray spectra in the 13~Å to 22~Å range from the Flat Crystal Spectrometer (FCS), an instrument on the Solar Maximum Mission operating in the 1980s, are analyzed with reference to nonflaring active regions, and to the \ion{Fe}{17} line emission in light of laboratory and atomic data for nearby \ion{Fe}{16} satellites. The satellites allow temperature to be found for these relatively low-temperature spectra, at which more conventional temperature-dependent line ratios are unavailable. By this means, the spectra can be arranged by temperature, showing that the \ion{Fe}{17} lines are evident at temperatures of $<3$~MK. We confirm that the problem of the underintense Fe XVII 3C and 3D lines is not due to resonant scattering, and instead suggest that, for comparison with CHIANTI spectra, the problem may lie with a needed revision of collisional excitation rates. The line ratio 3G/3H is in theory density-dependent but for \ion{Fe}{17} the ratio is in the low-density limit. However, we suggest that spectra taken during the impulsive stage of flares might reveal a departure from this limit and so allow densities to be derived and hence properties of the flaring plasma. Suggestions for the design of future crystal spectrometers are made in the light of the fluorescence background in FCS spectra.

Figures (11)

• Figure 1: The 853 FCS channel 1 spectral scans, during flares and from nonflaring active regions, are illustrated here in stack form in order of a running spectrum number (indicated on the left-hand vertical axis) and plotted against wavelength (Å). These spectral scans cover the effective lifetime of FCS channel 1 (1980 February--1987 December). The colors indicate emission intensity, with high emission colored yellow and lower-level emission colored red or deep red. The strong lines most evident include those of Ne-like Fe (Fe17) lines (the most intense being at 15.015 Å, 15.262 Å, 16.777 Å, 17.050 Å, and 17.097 Å) and the H-like oxygen (O8) Ly-$\alpha$ line at 18.970 Å (the home position location for channel 1).
• Figure 2: Procedure for estimating background (pedestal) level for a single nonflaring active region FCS channel 1 spectrum (14.5-minute duration starting 1985 November 16 at 14:50:24 UT). Black histogram: observed spectrum plotted logarithmically (units of photon counts s$^{-1}$ 0.001Å$^{-1}$). Yellow curve: smoothed observed spectrum with a box-car average having a width 0.005 Å. Red dots: minima determined using an automated IDL routine, which are minima in the yellow curve subtracted by 0.2 in the logarithm. Orange curve: Third-order polynomial to fit the red dots. Lower blue curve: preliminary determination pf the pedestal by eye. Green curve: GOES 1--8 Å emission plotted logarithmically to illustrate the absence of large changes with time (and therefore wavelength) during the FCS scan. The vertical scale of this curve is arbitrary (as drawn it corresponds to GOES level A5.8). Upper blue curve: total flux in Ca19 BCS channel 1.
• Figure 3: The FCS average spectrum for 165 nonflaring active region spectra (in black) displayed in two sections (13.1--15.8 Å, 15.7--19.6 Å). A comparison CHIANTI v. 11 (with input coronal abundances of fel92b and ion fractions of bry06) with Fe16 satellites is shown as the red curve, calculated for temperature $T = 3.80$ MK and volume emission measure = $8.85 \times 10^{45}$ cm$^{-3}$. Principal lines are identified as in the last three columns of Table \ref{['tab:line_ids']} (Fe17 lines with the notation of par73). The ionization limit of H-like oxygen is also shown.
• Figure 4: Upper panel: Average nonflaring active region spectrum (linear scale) showing bands used for estimating temperatures. For the temperature ${T_{15{\rm sat}}}$, the ratio of Fe XVI satellites in the bands on the long-wavelength side of lines 3C and 3D (marked in gray) to Fe XVII line 3F (16.776 Å). For the temperature ${T_{17{\rm sat}}}$, the ratio of Fe XVI satellites in the bands on the long-wavelength side of Fe17 lines 3G and 3H (marked in gray) to Fe XVII line 3F. Lower panel: Same as upper panel but plotted on a logarithmic scale.
• Figure 5: Left: GOES temperature $T_{\rm G}$ plotted against the intensity ratio $F_{15{\rm sat}}/F_{3F}$ for both flare (in gray) and nonflaring active regions (blue points). The error bars represent statistical uncertainties. The red line is the best-fit $T^{-0.9}$ dependence. Middle: As for the left panel but with the intensity ratio $F_{17{\rm sat}}/F_{3G+3H}$ on the $x$-axis. The red line is the best-fit $T^{-1.5}$ dependence. Right: Plot of $T_{15{\rm sat}}$ against $T_{17{\rm sat}}$ showing the strong correlation (coefficient $r = 0.86$).