Spectroscopic analysis and RHD modeling of the first Ca II H and H-epsilon flare spectra from DKIST/ViSP

TL;DR

This study presents the first flare-time spectra of Ca II H and Hε from DKIST/ViSP during the decay phase of a GOES C6.7 flare and confronts them with 1D RADYN+RH forward modeling. By sampling electron-beam heating with the F-CHROMA grid and testing a conduction-driven scenario, the authors show that Hε widths are reasonably reproduced, while Ca II H red-wing broadening is significantly underestimated, suggesting missing physics or heating complexity. They implement a RADYN+RH bridge to synthesize spectra with NEQ hydrogen densities and perform intensity-calibrated, spatially-resolved comparisons, revealing that condensation density alone does not set Hε width and that line formation also depends on lower-chromospheric layers and formation height. The work highlights the need for expanded model grids, multi-line constraints, and potentially combined heating mechanisms to fully capture flare chromospheric broadening observed by DKIST, guiding future DKIST-era flare modeling efforts.

Abstract

We analyze decay phase observations of the GOES class C6.7 flare SOL2022-08-19T20:31 by the Visible Spectropolarimeter (ViSP) on the National Science Foundation's Daniel K. Inouye Solar Telescope (DKIST). The data include the first flare-time DKIST observations of the chromospheric Ca II H 396.8 nm and H-epsilon 397.0 nm spectral lines. These diagnostics have rarely been studied together during the modern era of high-resolution solar flare observations, and never at the spectral and spatial resolution of the DKIST. We directly compare DKIST spectra to state-of-the-art RADYN+RH simulations, including one heated by a nonthermal electron beam and one by in-situ thermal conduction. While certain salient properties of the spectra such as the width of H-epsilon are reproduced, the models severely underestimate the width of Ca II H in the red wing and fail to reproduce the exact relative intensity of Ca II H to H-epsilon. The models exhibit a range of condensation electron densities spanning over an order of magnitude. Unlike the modeled lower-order Balmer-series lines, we find that the width of H-epsilon is not solely related to the condensation properties; the widths and intensities are also sensitive to the deeper flare layers. We outline possible avenues towards improvement of flare models, such as a comprehensive evaluation of flare heating mechanisms in the context of both impulsive and decay phase high-resolution data.

Figures (8)

• Figure 1: Overview of GOES/XRS data during flare observations. We show the light curve in GOES/XRS 1-8 Å SXR (black) with shaded regions indicating the periods when DKIST was observing. DKIST was observing exclusively during the periods indicated by colored shading (blue, red, yellow). The red shaded region indicates the period when DKIST was observing at flare-time. DKIST was undergoing instrument calibrations between the blue pre-flare period and the flare observations. The data in the final, yellow-shaded window are used for intensity calibration in this work. The two flares observed by GOES during this period occurred in a different active region, AR13081.
• Figure 2: SDO/AIA images from 2022 August 19 for three time-steps before (20:30 UT, panels (b) and (e)), during (20:42 UT, panels (a), (c), (f)), and after (21:10 UT, panels (d) and (g)) DKIST flare observations. Panel (a), (b) through (d), and (e) through (g) show the 304 Å, 1600 Å, and 131 Å images respectively. Panels (a), (c), and (f) correspond to the beginning of flare observations with DKIST. The approximate FoV of VBI (black) and the ViSP raster scan (red) are shown in panels (a), (c), and (f). The ViSP FoV intersects with the southern flare ribbon as seen in 1600 Å images. The coronal loops corresponding to the ribbons seen in the 1600 Å images are visible in the 131 Å images in panels (e)-(g).
• Figure 3: VBI H$\alpha$ observations of the GOES C6.7 class flare in NOAA active region 13078 on 2022 August 19. The bright flare ribbon and dark filament structures, used for co-alignment with ViSP spectra, are clearly visible, particularly in the first image frame (upper left). There are significant variations in image quality. The VBI-red field of view is 69" x 69". Helioprojective latitude and longitude values refer to those determined from co-alignment with SDO (Appendix \ref{['sec:app2']}).
• Figure 4: Result of co-aligning DKIST/ViSP @series Ca II Ca II Ca II H 396.8 nm integrated intensity maps with DKIST/VBI TiO and H$\alpha$ images from the first ViSP scan and first VBI image and determining proper spatial coordinates by comparison to SDO, as discussed in Appendix \ref{['sec:app2']}. (a) The result of transforming VBI image and ViSP integrated intensity map into the SDO basis, with the ViSP @series Ca II Ca II Ca II H integrated intensity from the first scan at 20:42:07 UT overlaid on a VBI TiO filter image. (b) The same as panel (a), but with the @series Ca II Ca II Ca II H integrated intensity overlaid onto the VBI-red H$\alpha$ filter image at 20:42:07 UT. We indicate the location of the ribbon leading and trailing edges (defined in Section \ref{['sec:CaIIvary']}). In both panels we indicate the direction of the ViSP scan, with the four slit positions corresponding to panels (a)-(d) in Figure \ref{['fig:flare_summary_19_august']} respectively. The raster step is 1.99" and the width of the slit is enlarged for clarity. Black dots in (a) correspond to the positions of maximum intensity, corresponding to the earliest intensity profiles in Figure \ref{['fig:flare_summary_19_august']}.
• Figure 5: Intensity-calibrated ViSP spectra. In panels (a)-(d) we show @series Ca II Ca II Ca II H 396.8 nm and H$\epsilon$ 397.0 nm emission line profiles for individual slit positions per scan. Spectral profiles are taken at the position of maximum intensity in the brightest flare ribbon. In panel (e) we combine all spectra from panels (a)-(d) in the observations between 20:42:07 UT and 20:46:33 UT (red shaded region in Figure \ref{['fig:goes_summary']}(a)). For the first scan at 20:42:07UT, the spectra correspond to the positions marked by the black dots in Figure \ref{['context']}(a).