Juan-Carlos Martínez1, Hugh Hudson1,2, Pascal Saint-Hilaire1, Sébastien Couvidat3
1. Space Sciences Laboratory, UC Berkeley, Berkeley, CA 94720, USA
2. School of Physics and Astronomy, University of Glasgow, Glasgow G12 8QQ, UK
3. W. W. Hansen Experimental Physics Laboratory, Stanford University, Stanford, CA 94305, USA
Observations
Two X-class flares on the 2013 May 13 were observed with HMI, revealing at least two distinct kinds of sources (chromospheric and coronal). During the SOL2013-05-13T02:17 and SOL2013-05-13T16:00 events, the HMI front camera (the standard HMI intensity) shows several features above the limb: short-lived jet-like early sources associated with the impulsive phase of the flare (Figure 1: red box in left frame), long-lived loop-like later sources (Figure 1: red box in right frame), and an enhanced glow along the limb from the chromosphere above the flaring active region. Although both events show these distinct features we will focus on the SOL2013-0513T16:00 event as it is the better observed case.
Figure 1 | HMI intensity images at three times during SOL2013-05-13T16:01, and six-point spectra averaged over the boxed regions. Left, an early low-altitude jet-like source; middle, a northern footpoint (here with a different scaling); right, the late-phase loop-like source. The image panels have a simple contrast enhancement (factor of 80) in the corona, beginning a few pixels above the limb itself (black line). The mean spectra refer to selected pixel groupings and have the simultaneously measured background levels subtracted. The averaged spectra derived from the left circular polarization (LCP) data is shown by the red circles and the red line shows the best Gaussian fit to the data; similarly blue shows the right circular polarization (RCP). In each spectrum the black solid and dashed curves show the reference spectrum from its background region for each polarization, consisting of the non-flaring parts of the image annulus that encompass the source pixels.
Figure 2 shows the coronal sources as height-versus-time plots and time series, along with GOES and RHESSI X-ray data. The late sources move steadily outward for some tens of minutes and eventually vanish outside the truncated radius of the HMI images; the short-lived brightenings during the impulsive phase remain at lower altitudes. The sources are generally fainter than the background brightness due to stray light from the solar disk. In this work we subtract a preflare image, which minimizes the effects of scattered light from the solar atmosphere and from within the instrument itself. We find the background light level to be quite stable and to have a lower time-series fluctuation level than the on-disk images. The image brightness in the annulus is about 0.5%–2% of the disk-center image brightness, but the rms fluctuation in a difference image, which limits the sensitivity, is typically only about 0.02%.
Figure 2 | Time series plots for the two flares studied here. The upper panels show the GOES long-wavelength channel (gray fill), the RHESSI 30 – 100 keV flux (black fill, with some gaps), and two histograms with HMI summed intensities in the early source region (red) and in a field of view containing all of the observed phenomena (purple), both early and gradual. The light curves show excess fluxes, relative to the start time, normalized to their maxima. The lower panels show height vs. time information for HMI; note the spike at about 16:00 UT in the lower-right panel, which appears in only a few 45 s data frames. The limb reference position is at elongation 949′′.06.
RHESSI Observations
Figure 3 shows the impulsive-phase development of SOL2013-05-13T16:01 as seen by the AIA in the EUV, and by RHESSI in soft and hard X-rays. The footpoint hard X-ray sources are well-matched with the white-light emission patches (Figure 4 upper left frame). The AIA EUV images (six panels as labeled) show foreground/background sources that appear to be loops in the active region resulting from early stages of the flare development. The jet-like source originates near the northern footpoint.
Figure 3 | Gradual-phase sources, with image times WL: 16:25:22.7 (WL), 16:25:28.1 (1600A), and 16:25:21.5 (193Å). The HMI 193Å image shows an absorption feature just at the lower edge of the soft X-ray source, which itself must be optically thin.
The RHESSI soft X-ray images reveal a source at a comparable height, but displaced laterally from the jet-like HMI source. The AIA 193Å pass band also shows a feature at this location, confirming that it has a hot component via the Fe xxiv contribution to this pass band. This hot source is not observed by HMI or the more sensitive AIA at the 1600Å pass band. This situation is made more interesting by the fact that the HMI jet-like source does not have a hot counterpart observed by RHESSI or AIA hot channels (RHESSI in general cannot detect sources with temperatures below 8 MK). This suggests that the HMI emission is not the bremsstrahlung tail from a hot source, but free – bound emission from plasma at temperatures too low to emit X-ray radiation detectable by RHESSI. This possibility of low-temperature ejecta into closed-field structures in the impulsive phase of a flare does not fit naturally into our conventional understanding and hence may represent something new about flare development. The RHESSI hard X-ray contours, on the other hand, reflect the well-known footpoint sources and hint at a loop-top source as well.
The HMI loop-like sources, observed during the gradual phase of the events, appear to be easier to interpret. We see a close relationship with the loops seen in the AIA 1600Å bands, but not directly in the EUV bands, where most of the emission is from larger-scale structures (Figure 3). This component and the coronal soft X-ray sources seen by RHESSI are concentrated along the apex of an arcade of loops, whose axis roughly lies along the line of sight but with the northern footpoints noticeably displaced to the west, as judged from the HMI images. They also suggest a cusp structure that reflects the open field lines created by the associated CME (e.g., Ref[1]).
Discussion
These observations reveal a new and important application of HMI to flare studies. Even without optimization for coronal observation, HMI difference images reveal a variety of effects in the low corona, and provide good signal-to-noise ratio. Among other things, these observations will help to characterize the “mass cycle” by which flares inject plasma into coronal magnetic field structures.
References
[1] Hiei, E., Nakagomi, Y., & Takuma, H. 1992, PASJ, 44, 55