A Statistical Study to Determine the Origin of Long-duration Gamma-Ray Flares

Two scenarios have been proposed to account for sustained. >= 30 MeV gamma-ray emission in solar flares: (1) prolonged particle acceleration/trapping involving large-scale magnetic loops at the flare site, and (2) precipitation of high-energy (> 300 MeV) protons accelerated at coronal/interplanetary shock waves. To determine which of these scenarios is more likely, we examine the associated soft X-ray flares, coronal mass ejections (CMEs), and solar energetic proton events for (a) the long-duration gamma-ray flares (LDGRFs) observed by the Large Area Telescope on Fermi, and (b) delayed and/or spatially extended high-energy gamma-ray flares observed by the Gamma-ray Spectrometer on the Solar Maximum Mission, the Gamma-1 telescope on the Gamma satellite, and the Energetic Gamma-Ray Experiment Telescope on the Compton Gamma-Ray Observatory. For the Fermi data set of 11 LDGRFs with > 100 MeV emission lasting for >=similar to 2 hr, we search for associations and reverse associations between LDGRFs, X-ray flares, CMEs, and SEPs, i.e., beginning with the gamma-ray flares and also, in turn, with X-class soft X-ray flares, fast (>= 1500 km s(-1)) and wide CMEs, and intense (peak flux >= 2.67 x 10(-3) protons cm(-2) s(-1) sr(-1), with peak to background ratio > 1.38) > 300 MeV SEPs at 1 au. While LDGRFs tend to be associated with bright X-class flares, we find that only one-third of the X-class flares during the time of Fermi monitoring coincide with an LDGRF. However, nearly all fast, wide CMEs are associated with an LDGRF. These preliminary association analyses favor the proton precipitation scenario, although there is a prominent counter-example of a potentially magnetically well-connected solar eruption with > 100 MeV emission for similar to 10 hr for which the near-Earth > 300 MeV proton intensity did not rise above background.