SEP Measurements

The SEP models in CREME are based on satellite measurements of particles throughout the range of energies and elements relevant for SEU/SEE studies.

These measurements come primarily from GOES (for protons) and the University of Chicago's Cosmic Ray Telescope on IMP-8 (for heavy ions). The latter instrument is at present unique in that it measures SEP fluxes at energies sufficiently high to penetrate the minimum shielding typical in satellite systems.

The CREME SEP models do not rely in any significant way on theoretical or semi-empirical assumptions about the fluxes, spectra shapes, or relative elemental abundances.

Specifically, the CREME SEP models are based on measurements of the extended episode of solar particle emission which occurred on 19-27 October 1989. This episode is the most severe SEP event observed since at least August 1972. In particular:

• The October 1989 episode qualifies as a "99% confidence-level worst case" in the JPL flare model's three highest proton energies, >10, >30, and >60 MeV. [The proton fluence of the August 1972 event, which has often been used in earlier SEP models, is at the 99% level for 10 and 30 MeV, but not for 60 MeV. See Majewski et al., IEEE Trans. N.S. 42, 2043 (1995); and Tylka et al., IEEE Trans. N.S. 44, 2150 (1997).]
• The October 1989 episode produced the largest heavy ion fluences observed by the Chicago/IMP-8 instrument since its launch in 1973. The episode formally corresponds to a 99% confidence level worst-case event, when compared to the entire Chicago/IMP-8 solar particle database. [See Tylka et al., IEEE Trans. N.S. 44, 2140 (1997).]
• The October 1989 episode produced the highest SEU rates ever observed in the TDRS-1 Attitude Control System (ACS) RAM since its launch in 1983 [See Tylka et al. IEEE Trans. N.S. 43, 2758 (1996) and references therein.]

The SEP fluxes, nuclear transport, and SEU rate calculation techniques used in CREME96 have been validated by comparison with on-orbit SEU data:

• "Single Event Upsets Caused by Solar Energetic Heavy Ions" by Allan J. Tylka, William F. Dietrich, Paul R. Boberg, Edward C. Smith, and James H. Adams, Jr., IEEE Transactions on Nuclear Science, 43, No. 6, 2758-2766 (December 1996).

The SEP measurements used in the CREME models extend to ~800 MeV/nuc (somewhat less for CNO nuclei, where the SEP spectra intersect the galactic cosmic-ray background at ~500 MeV/nuc.) The model should therefore accurately describe these "worst case" SEP environment except under very thick shielding (> 100 g/cm2 = 14 inches Al equivalent) . The functional forms used to fit the experimental SEP data are purely empirical; they were chosen simply because they gave good fits to the data. The SEP models use these functional forms to make extrapolations beyond ~800 MeV/nuc, but these extrapolations (which are irrelevant for SEE applications) should not be taken seriously.