Superhot (> 30 MK) flare observations with STIX: Joint spectral fitting

TL;DR

The paper addresses the challenge of diagnosing hot and superhot solar flare plasma when attenuators suppress low-energy X-rays. It introduces joint spectral fitting of STIX BKG and imaging detector spectra using SUNKIT-SPEX, modeling two thermal components, a nonthermal component, and albedo, with a calibration-binding parameter to harmonize the detectors. Analyzing 32 large flares, the study shows that two thermal components better describe GOES X-class spectra, with a superhot component ($T_{SH} \ge 30\,\mathrm{MK}$) often dominating flux above $15\,\mathrm{keV}$ at the temperature peak, and a typical superhot EM fraction of 5–10% relative to the hot EM. The work also finds a strong GOES–$T_{SH}$ correlation ($r \approx 0.77$) and quantifies the superhot component’s limited contribution to GOES, illustrating a clear multithermal flare structure and demonstrating that joint STIX fitting is a robust, practical diagnostic tool with implications for energy budgeting and future HXR instrumentation.

Abstract

Spectroscopic analysis of large flares (>X1) in the hard X-ray (HXR) range offers unique insights into the hottest (> 30 MK) flare plasma, the so-called superhot thermal component. To manage the high count rates in large flares, an attenuator is typically placed in front of the HXR detectors. However, this significantly limits the spectral diagnostic capabilities at lower energies, and consequently, it restricts the analysis of the lower temperatures in flares. The Spectrometer/Telescope for Imaging X-rays (STIX) on board the Solar Orbiter mission was designed to observe solar flares in hard X-rays. The imaging detectors use an attenuator during periods of high flux level. In contrast, the background (BKG) detector of STIX is never covered by the attenuator and is therefore dedicated to measure the unattenuated flux using differently sized apertures placed in front of the detector. We aim to demonstrate that joint spectral fitting using different detector configurations of STIX allows us to reliably diagnose both the hot and the superhot components in large flares. We jointly fit the HXR spectra of the STIX BKG detector and the STIX imaging detectors using SUNKIT-SPEX software package to determine the spectral parameters of both the hot and superhot thermal components in solar flares. Using joint fitting on 32 STIX flares, we corroborate that for GOES X-class flares, the HXR spectrum is better represented by two thermal components instead of an isothermal component. At the temperature peak time, the superhot HXR flux above 15 keV is typically stronger than the hot HXR flux. The GOES long-wavelength channel is dominated by the hot component with a superhot contribution up to 10%. This paper demonstrates that joint spectral fitting of the same detector type with different attenuation schemes is a simple and powerful method to monitor multithermal flare plasma.

Figures (6)

• Figure 1: Spectrum of the estimate X5-class flare SOL20230103 recorded by the imaging detectors of STIX. The spectrum is integrated over a time range of 10 s between 06:26:50-06:27:00 UTC (Earth time). The measured spectrum is the same in both panels, only the fitting model is different: For the top panel, a photon model including an isothermal model (red), a nonthermal model (green), and an albedo component (grey) is fitted. The total fit is given in black. In the bottom panel, we used the same components for the photon model together with a second thermal component (orange). Below each spectrum, the residuals (data minus model divided by the uncertainty on the data points) and the $\chi^2$ are given. The green bar indicates the energy range used for fitting.
• Figure 2: Spectrum of the estimate X5-class flare SOL20230103 measured by the BKG detector (left plots) and the imaging detectors (right plots, same as Fig. \ref{['Fig: Single Fit Imaging']}) of STIX. The spectrum is integrated over a time range of 10 s between 06:26:50-06:27:00 UTC (Earth time). The measured spectrum is the same in the top and bottom panels, only the fitting model is different: For the top panel, a photon model including an isothermal model (red), a nonthermal model (green), and an albedo component (grey) is jointly fitted to the two spectra. The total fit is given in black. In the bottom panel, we used the same components for the photon model together with a second thermal component (orange). Below all spectra, the residuals (data minus model divided by the uncertainty on the data points) and the $\chi^2$ are given. The green bar indicates the energy ranges used for fitting.
• Figure 3: Results of the statistical analysis using 32 STIX flares. In the left plot, the temperature of the thermal emission as a function of the GOES flare class is shown. The x-axis is the GOES peak flux estimated on the low-energy counts from STIX, and the y-axis gives the temperature of the thermal emission in MK. The blue and red dots represent the spectral fitting results for the superhot and the hot thermal component from 32 large STIX flares observed over the past three years (2022-2024). The grey crosses are the results from the statistical analysis in Caspi_2014 with RHESSI data. The right plot shows the temperature as a function of the emission measure of the superhot component.
• Figure 4: Panel (a) shows the histogram plot of the ratio in percentage between the superhot EM and the hot EM using the 32 STIX flares. Panel (b) gives the ratio in percentage between the estimated GOES flux of the superhot component and the GOES peak flux (= GOES flare class). Panel (c) shows the GOES flux as a function of the EM. The different colors correspond to different temperatures, from blue to orange, as the temperature increases.
• Figure 5: MCMC analysis of the fit shown in Fig. \ref{['Fig: Joint Fit']} (b). The diagonal plots show the histogram plots of the individual parameters fitted in the photon model. The other plots show the correlations between two different parameters.