RHESSI Line and Continuum Observations of Super-hot Flare Plasma

We use RHESSI high-resolution imaging and spectroscopy observations from ~5 to 100 keV to characterize the hot thermal plasma during the 2002 July 23 X4.8 flare. These measurements of the steeply falling thermal X-ray continuum are well fit throughout the flare by two distinct isothermal components: a super-hot (T > 30 MK) component that peaks at ~44 MK and a lower-altitude hot (T < 25 MK) component whose temperature and emission measure closely track those derived from GOES measurements. The two components appear to be spatially distinct, and their evolution suggests that the super-hot plasma originates in the corona, while the GOES plasma results from chromospheric evaporation. Throughout the flare, the measured fluxes and ratio of the Fe and Fe-Ni excitation line complexes at ~6.7 and ~8 keV show a close dependence on the super-hot continuum temperature. During the pre-impulsive phase, when the coronal thermal and non-thermal continua overlap both spectrally and spatially, we use this relationship to obtain limits on the thermal and non-thermal emission.

💡 Research Summary

This paper presents a comprehensive analysis of the 2002 July 23 X4.8 solar flare using the Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI). By exploiting RHESSI’s high‑resolution imaging spectroscopy (∼1 keV energy resolution, ∼2 arcsec spatial resolution) over the 5–100 keV range, the authors characterize the thermal plasma throughout the flare’s pre‑impulsive, impulsive, and decay phases.

The spectral fitting strategy began with a model consisting of a single isothermal component plus a non‑thermal power‑law. Residuals below ∼15 keV revealed an additional, cooler continuum, prompting the adoption of a two‑isothermal model: (1) a “super‑hot” component with temperatures exceeding 30 MK (peaking at ≈44 MK) and (2) a “hot” component with temperatures ≤ 25 MK that tracks the GOES‑derived temperature and emission measure. The non‑thermal component was modeled as a power‑law (or double power‑law when required) with a low‑energy cutoff constrained primarily as an upper limit because the thermal emission dominates at low energies. The isothermal spectra were generated with the CHIANTI v5.2 atomic database, allowing the extraction of emission measures (Q = nₑ² V) and, together with imaging‑derived source volumes, the estimation of electron densities and thermal energies.

Key diagnostics come from the Fe XXV (≈6.7 keV) and Fe‑Ni (≈8 keV) line complexes. By fitting these lines as Gaussian profiles and subtracting the underlying continuum, the authors measured line fluxes and the Fe/Fe‑Ni flux ratio. Both the absolute line fluxes and their ratio correlate tightly with the temperature of the super‑hot continuum. At lower temperatures (≲ 25 MK) the ratio’s temperature dependence steepens, providing a sensitive thermometer that remains valid even during the pre‑impulsive phase when thermal and non‑thermal emissions overlap spatially and spectrally. Using this empirical relationship, the authors place constraints on the super‑hot temperature (≈29–37 MK) and on the low‑energy cutoff of the non‑thermal electrons (≤ 20–27 keV) during the early phase.

Imaging analysis was performed on RHESSI CLEAN maps in the 6.2–8.5 keV band. The 50 % intensity contour defined the source size; after correcting for the point‑spread function, the authors derived an ellipsoidal volume (V = 4πab²/3) with ∼23 % uncertainty. By assigning half of this volume to each thermal component, they computed electron densities of ≈2 × 10¹¹ cm⁻³ for the super‑hot plasma and ≈1 × 10¹¹ cm⁻³ for the hot plasma. The centroids of the 6–7 keV, 9–12 keV, and 17–18 keV images shift linearly with the fractional contribution of the super‑hot component, indicating that the two thermal sources are separated by about 1.5–2 arcsec, with the super‑hot source located higher in the corona. Throughout the flare the separation grows modestly, and the super‑hot source sometimes appears elongated along the direction away from the footpoints.

Energetically, at the time of the super‑hot temperature peak (≈00:28:30 UT) the super‑hot plasma contains ≈10³¹ erg of thermal energy, while the hot component holds a comparable amount. Their energy densities are ∼4.8 × 10³ erg cm⁻³ (super‑hot) and ∼5.9 × 10³ erg cm⁻³ (hot). To magnetically confine these plasmas, coronal magnetic field strengths of ≳ 350 G (super‑hot) and ≳ 380 G (hot) are required. A survey of 37 large flares shows that all X‑class flares that reach super‑hot temperatures demand fields > 220 G, whereas cooler M‑class flares can be confined with fields as low as 60 G, underscoring the importance of strong coronal fields for achieving > 30 MK temperatures.

During the pre‑impulsive phase a hot (≈25 MK) coronal plasma co‑exists with a non‑thermal HXR source, yet footpoint HXR emission is weak or absent, suggesting that chromospheric evaporation is not the primary source of the super‑hot plasma. The authors therefore favor a scenario in which the super‑hot component is generated directly in the reconnection region: Joule heating, followed by Fermi acceleration and betatron heating as reconnected field lines contract, raises the plasma temperature. The hot component, by contrast, appears later, tracks the GOES measurements, and is consistent with chromospheric evaporation driven by the precipitating non‑thermal electrons.

In summary, the paper demonstrates that RHESSI’s combined imaging and spectroscopy can resolve two distinct thermal populations in a major solar flare. The super‑hot coronal plasma, with temperatures up to ∼44 MK, is spatially separate from the cooler GOES‑like plasma and is likely produced by direct heating in the reconnection region rather than by evaporation. This dual‑temperature structure, its temporal evolution, and its magnetic confinement requirements provide valuable constraints for flare energy‑release models and highlight the critical role of strong coronal magnetic fields in producing super‑hot flare plasma.