Forecast of solar proton flux profiles for well-connected events

Abstract

We have developed a forecast model of solar proton flux profiles (> 10 MeV channel) for well-connected events. Among 136 solar proton events (SPEs) from 1986 to 2006, we select 49 well-connected ones that are all associated with single X-ray flares stronger than M1 class and start to increase within 4 h after their X-ray peak times. These events show rapid increments in proton flux. By comparing several empirical functions, we select a modified Weibull curve function to approximate a SPE flux profile. The parameters (peak flux, rise time, and decay time) of this function are determined by the relationship between X-ray flare parameters (peak flux, impulsive time, and emission measure) and SPE parameters. For 49 well-connected SPEs, the linear correlation coefficient between the predicted and the observed proton peak fluxes is 0.65 with the RMS error of 0.55 log10(pfu). In addition, we determine another forecast model based on flare and coronal mass ejection (CME) parameters using 22 SPEs. The used CME parameters are linear speed and angular width. As a result, we find that the linear correlation coefficient between the predicted and the observed proton peak fluxes is 0.83 with the RMS error of 0.35 log10(pfu). From the relationship between error of model and CME acceleration, we find that CME acceleration is an important factor for predicting proton flux profiles.