Identification of backside events from the NOAA solar proton event list

Solar proton events, whose fluxes are larger than 10 particles cm^-2 sec^-1 ster^-1 for >10 MeV protons, have been observed since 1976. The NOAA solar proton event list from 1997 to 2006 shows that the most of the events are related to both flares and CMEs but several events (5/93) are related with only CMEs and some events (9/93) do not have enough flare information. In this study, we carefully identified the source location of these events. For this, we used Active Region maps, SOHO LASCO, MDI and GOES data. First, we examined the directions of CMEs related to the events. Second, we searched a major active region in the front solar disk for several days before the solar proton events occurred by taking into account two facts: (1) The location of the active region is consistent with the position angle of a given CME and (2) there were several flares in the active region or the active region is the largest among several candidates. As a result, we were able to determine active regions which are likely to produce proton events as well as their longitudes at the time of proton events by considering solar rotation rate. From this study, we found that the longitudes of 9 active regions are between 90°W and 150°W, the longitude of 3 active regions are between 100°E and 120°E, and the longitude of 2 active regions are located near the west limb. When the flare peak time is assumed to be the CME event time, we confirmed that the dependence of their rise times (proton peak time - flare peak time) on longitude are consistent with the previous empirical formula. Our results imply that misidentification of the NOAA proton event list should be caused by backside events and suggest a possibility that nearly all proton events are associated with flares.