1 University of Applied Sciences and Arts Northwestern Switzerland, Bahnhofstrasse 6, 5210 Windisch, Switzerland
Stepwise and permanent changes of magnetic fields can often be observed in active regions during flares1,2. These changes are important because they may allow us to probe the reconfiguration of coronal fields during flares and it has been speculated that they may be related to sunquakes. Past studies focused only on photospheric changes because chromospheric flare polarimetry is very rare. Here we show photospheric magnetic field changes derived from HMI and compare them with chromospheric magnetic field changes from DST/IBIS. We focus on the well-observed March 29, 2014 flare, for which we acquired polarimetric data in the chromospheric Ca II 8542 Å line3.
Figure 1 | The left panel shows the locations and sizes of the photospheric magnetic field changes (colored pixels). The black contours show umbrae, penumbrae and pores of the active region. The red contours show a weak sunquake, which was slightly offset from the magnetic field changes. The right panel shows the magnetic field evolution of 9 neighboring pixels from the small box in front of the red arrow in the left panel. The purple line is the arctan-fit and the retrieved value of the change is given in each panel.
We mapped the changes by fitting an arctan function to each pixel in a sequence of HMI 45 s near-real-time line-of-sight (LOS) magnetic field data (BLOS) with the solar rotation removed beforehand. For the chromospheric data, we used the weak-field approximation to derive BLOS and excluded pixels with low polarization or irregular profiles before the arctan-fit (Fig. 1). The advantage of HMI is the unlimited field of view (FOV, whole Sun), while IBIS only observed a smaller FOV (marked with white lines in Fig. 2).
Figure 2 | The locations and sizes of the magnetic field changes (colored pixels) for the photosphere (left column) and the chromosphere (right column). The backgrounds are G-band images with contours of RHESSI hard X-ray emission (top row), a Hinode Na I Stokes V image / HMI magnetogram (bottom left) and AIA 1600 Å (bottom right). Photospheric changes occur near the polarity inversion line, chromospheric changes near footpoints of loops.
By mapping the photospheric and chromospheric changes and overlying them on different backgrounds, we can investigate their relation with X-rays, white-light emission, coronal loops, and a filament eruption.
Chromospheric changes are stronger (< 640 Mx cm-2) and appear in larger areas than their photospheric counterparts (< 320 Mx cm-2). It is a little surprising that photospheric and chromospheric changes are near, but not co-spatial to hard X-ray emission, continuum emission, or a small sunquake, which indicates different mechanisms for these processes. Even more surprising is that the photospheric and chromospheric changes show differences in their timing, sign, and size, which means that a simple scenario of a large changing magnetic loop throughout the atmosphere is not a valid approximation.
Figure 3 | Possible scenario of untwisting loops to explain the observed line-of-sight magnetic field changes in the chromosphere at the footpoints of loops.
Past studies found contracting loops during and after flares, which were linked to a model of coronal implosions4. The model predicts that the magnetic energy in the corona should decrease after flares, which may appear as contracted loops and should lead to more horizontal magnetic fields, consistent with previous observations5. In our case, contracting loops would appear as chromospheric changes of opposite sign to those that we observed. One explanation could be that our loops increased in height, or that they untwisted in a certain direction, as depicted in Fig. 3. However, without high signal-to-noise vector magnetic field data, we unfortunately can only speculate. The next step would be to explore three-dimensional vector field changes with the newly available faster cadence HMI vector data6 and to co-observe more flares using chromospheric polarimetry to improve the statistics with respect to this first observation.
 Wang, H. 1992, SoPh, 140, 85
 Sudol, J. J., & Harvey, J. W. 2005, ApJ, 633, 647
 Kleint, L. 2017, ApJ, 834, 26
 Hudson, H. S. 2000, ApJL, 531, L75
 Petrie, G. J. D., & Sudol, J. J. 2010, ApJ, 724, 1218
 Sun, X., Hoeksema, J. T., Liu, Y., Kazachenko, M., & Chen. R. 2017, ApJ, in press (arXiv:1702.07338)