The relationship between magnetic gradient and magnetic shear in five super active regions producing great flares

We study the magnetic structure of five well-known active regions that produced great flares (X5 or larger). The six flares under investigation are the X12 flare on 1991 June 9 in AR 6659, the X5.7 flare on 2000 July 14 in AR 9077, the X5.6 flare on 2001 April 6 in AR 9415, the X5.3 flare on 2001 August 25 in AR 9591, the X17 flare on 2003 October 28 and the X10 flare on 2003 October 29, both in AR 10486. The last five events had corresponding LASCO observations and were all associated with Halo CMEs. We analyzed vector magnetograms from Big Bear Solar Observatory, Huairou Solar Observing Station, Marshall Space Flight Center and Mees Solar Observatory. In particular, we studied the magnetic gradient derived from line-of-sight magnetograms and magnetic shear derived from vector magnetograms, and found an apparent correlation between these two parameters at a level of about 90%. We found that the magnetic gradient could be a better proxy than the shear for predicting where a major flare might occur: all six flares occurred in neutral lines with maximum gradient. The mean gradient of the flaring neutral lines ranges from 0.14 to 0.50 G km ^-1, 2.3 to 8 times the average value for all the neutral lines in the active regions. If we use magnetic shear as the proxy, the flaring neutral line in at least one, possibly two, of the six events would be mis-identified.