White-light emission and photospheric magnetic field changes in flares

From RHESSI Wiki

Revision as of 08:09, 21 September 2020

Introduction

The enhancement of the continuum in the emission spectrum of a solar flare likely accounts for a substantial part of the total energy released. The emission in the optical window is usually called white-light (WL) emission (Ref. [1]), and often detected by broad-band sensors including the human eye, as in the original Carrington flare of 1859. But the exact mechanism that produces the WL emission still remains unclear. In addition, it is still puzzling why in some flare we do not detect WL emission, even in some powerful X-class flares, while some weaker flares show it clearly. Because of the large energies involved in WL emission, investigating these open questions gives us direct insight into the energy dissipation in flares. Almost 30 years ago, the first evidence of permanent and rapid changes of the photospheric magnetic field (hereafter ΔB_LOS during flares was observed (Ref. [2]). These changes are irreversible, in the sense that they persist for time scales much longer than the flare itself. Nowadays, finite ΔB_LOS is a common phenomenon observed during major flares (Ref. [3]), and even during sometimes during minor ones as well (Ref. [4]) The chromospheric field may also vary in a similar manner (Ref. [5]) The mechanism that produces a ΔB_LOS is still unclear, but the evidence that the disturbance of the field is observed at lower layers might give insight into the energy transport from the corona to the photosphere.

Co-spatiality between the WL and ΔB_LOS has been reported in small flare samples. Here we study the WL emission for a large sample of 75 flares and analyse the relationship between the WL emission and ΔB_LOS statistically, as summarized in Figure 1.

Figure 1: Number of flares that showed WL emission and ΔB_LOS as a function of the GOES class.

Results

During 44 of 75 events (59\%), WL enhancements were observed, while 59 (79\%) had detectable ΔB_LOS. The WL kernels tend to be elongated and have the strongest emission in the middle, becoming diffuse towards the edges. These findings on the fine structure of the kernels of the WL emission agree with previous reports. In one small C5.6 flare (SOL2011-05-27T16:43), we observed clear WL emission although we did not detect ΔB_LOS. The opposite case, ΔB_LOS without WL emission, was also observed in a major X-class flare (SOL2012-07-12T16:50); see Figure 2. However, this X1.4 flare had a long duration, about 2 hours and a relatively extended impulsive phase. This suggests a relationship between impulsiveness and WL emission (see also Ref. [6]).

Larger and more energetic flares have been found to display stronger WL emission, which also covers larger areas. The WL enhancement and WL area are found to be related to the GOES class of the flare following a power-law distribution (see Figure 2). This relationship does not depend on the location of the flare after 4correcting the areas for foreshortening.

Figure 2: Left, correlation of WL intensity and GOES flux; right, area of the WL emission as a function of the GOES X-ray flux (black circles). Straight lines are the best-fit of a power-laws. See Ref. [4] for the details.

The co-spatial relationship between WL and the ΔB_LOS has been statistically verified. Figure 3 shows a sub-sample of the flares and the locations of the WL emission (orange), ΔB_LOS (blue), and those locations where both WL emission and ΔB_LOS are co-spatial (red). The ΔB_LOS and the WL emission areas are found to be related and they often overlap. The size of their areas are comparable, but the overlap varies between the events and it is not a 1:1 relationship. In addition, we did not find a difference in ΔB_LOS areas or flux between those flares and non-WLF. Consequently, we suggest that both the ΔB_LOS and WL emission are coupled in some way, but that their origin may be different.

Figure 3: Flares in our sample that showed WL emission and ΔB_LOS. The background image is the line-of-sight magnetic of the same region clipped at +-800 G. The coloured pixels show the location of ΔB_LOS (blue), WL emission (orange) and the overlapping areas (red).

Conclusion

The total WL enhancement, the area of the WL emission, and ΔB_LOS are all correlated with the strength of the flare (the GOES class). These results show that the required energy to perturb the magnetic field is larger for larger ΔB_LOS. We did not detect any fundamental distinction between those flares where we detected WL emission, flares with associated ΔB_LOS, and flares where neither of these two phenomena were seen. We found that the WL emission and the ΔB_LOS are closely linked, but note differences that suggest complicated relationships.

References

[1] "The importance of solar white-light flares"

[2] "Evolution of vector magnetic fields and the August 27 1990 X-3 flare"

[3] "Longitudinal Magnetic Field Changes Accompanying Solar Flares"

[4] "A Statistical Study of Photospheric Magnetic Field Changes During 75 Solar Flares"

[5] "First Detection of Chromospheric Magnetic Field Changes during an X1-Flare"

[6] "Characteristics that Produce White-light Enhancements in Solar Flares Observed by Hinode/SOT"