“…Some results of the Geant4 simulation were clearly deficient and we have supplemented them with experimental data in the highest E c bin: for elastic scattering of protons on all targets the Geant4 simulation underestimated the cross sections for E p ≤ 15 MeV and these have been replaced with results obtained from the EXFOR database [http://www-nds.iaea.org/exfor/exfor00.htm]; the reaction p + p → 2 H+ π + was not represented in the simulation and was added for 300 < E p < 3000 MeV [ Ramaty and Lingenfelter , 1969] assuming an isotropic secondary distribution, which is valid near the reaction threshold [ Drochner et al , 1998], and the reaction p + 4 He → 2 H+ 3 He, also not represented, was added for 40 < E p < 400 MeV [ Selesnick and Mewaldt , 1996]. In the cases of secondary 6 Li, 7 Li, and 7 Be the Geant4 results led to total cross sections that were significantly lower than experimental data [ Moskalenko and Mashnik , 2003], but they were used without supplementation because differential cross section data were not available; it is likely that the missing secondaries would be at low enough energy (in the LAB frame) so as not to contribute to the radiation belt model.…”