Authors: Brian T. Welsch (Univ. Wisc. - Green Bay), Yang Liu (Stanford University)

Energy released in solar flares and coronal mass ejections (CMEs) is believed to be stored in electric current systems flowing in the solar corona. Because coronal currents cannot be directly measured, their properties are often deduced from proxies, such as measures of photospheric twist derived from vorticity in the photospheric magnetic field. A tendency for some twist measures to decrease in the hours prior to large flares has been reported, which does not accord with a model in which (i) photospheric twist reliably indicates coronal currents’ strengths, and (ii) coronal currents intensify prior to large flares. Decreases in photospheric twist might arise from either (a) the passage of the apex of an Omega-shaped, twisted flux tube upward across the photosphere prior to a flare, or (b) flux cancellation, perhaps via photospheric magnetic reconnection. Or photospheric flows might simply untwist photospheric fields. All three models predict that the size of areas containing high-twist pixels would decrease prior to flares. The apex emergence and cancellation scenarios predict that changes in twist should occur primarily near polarity inversion lines (PILs) of the vertical magnetic field, the only loci where vertical flux can emerge or cancel. In contrast, untwisting could occur anywhere. Only the cancellation model predicts that magnetic flux should decrease within the areas where twist decreases. Here, we investigate changes in photospheric magnetic twist from several hours prior to just after several large (X-class) flares. We quantify: changes in the sizes of high-twist areas; where changes in twist occur relative to PILs; and changes in vertical flux within areas of significant twist changes. Notably, we use the field’s toroidal component, from a poloidal-toroidal decomposition (PTD) of the photospheric field, as one measure of twist. Where appropriate, we use SuperPosed Epoch (SPE) analysis to combine evolution from several events, to reveal patterns in pre- and post-event changes.