On the estimation of solar energetic particle injection timing from onset times near earth

We examine the accuracy of a common technique for estimating the start time of solar energetic particle injection based on a linear fit to the observed onset time versus 1/(particle velocity). This is based on a concept that the first arriving particles move directly along the magnetic field with no scattering. We check this by performing numerical simulations of the transport of solar protons between 2 and 2000 MeV from the Sun to the Earth, for several assumptions regarding interplanetary scattering and the duration of particle injection, and by analyzing the results using the inverse velocity fit. We find that, in most cases, the onset times align close to a straight line as a function of inverse velocity. Despite this, the estimated injection time can be in error by several minutes. Also, the estimated path length can deviate greatly from the actual path length along the interplanetary magnetic field. The major difference between the estimated and actual path lengths implies that the first arriving particles cannot be viewed as moving directly along the interplanetary magnetic field.