AP8 Accuracy

Numerous authors have assessed the present-day accuracy of the AP8MIN/MAX trapped proton models by comparing model predictions with on-orbit dose and SEU-rate data. Such comparisons have supported the conclusion that AP8 results, at least for orbit-averaged quantities, are generally reliable to within a factor of two. A partial list of such comparative studies is as follows:

Comparisons with Dose Measurements:

1. Armstrong et al. compared AP8 predictions with thermoluminescent detector (TLD) and activation dosimetry measurements from the LDEF satellite (28.5^o, 478 km mean altitude, April 1984-January 1990). The radiation exposure on LDEF came mainly at solar minimum, and these dosages were extracted from well-shielded dosimeters, in which the trapped electron contribution was negligible. These comparisons indicated that the AP8 trapped proton fluxes in LDEF's low-altitude, low-inclination orbit were systematically low by a factor of two. [For further information on these LDEF studies, contact Dr. Tony W. Armstrong, tony.w.armstrong@cpmx.saic.com; 931-468-2603]
2. Comparisons between AP8MIN calculations and APEXRAD, a new empirical low-altitude dose model based on direct dose measurements from the APEX satellite, also show similar factor-of-two discrepancies.
3. Similar comparisons have been carried out for dose measurements aboard Space Shuttle flights. These dose measurements were made on numerous flights, at two well-shielded locations on the Shuttle mid-deck, with minimum shielding ~1.25 g/cm² and median shielding >5 g/cm², so that trapped electron dose was negligible. At ~300-600 km altitude:
   • In 28.5^o Shuttle orbits, AP8 sometimes overpredicts (up to a factor of ~2) and sometimes underpredicts (up to a factor of 2), depending on shielding, altitude, and time during the solar cycle.
   • In mid-inclination Shuttle orbits (52-57^o), the AP8 models tended to overpredict the proton dose by factors up to ~2-3.
   • [For further information on these Shuttle comparisons, contact Dr. Tony W. Armstrong, tony.w.armstrong@cpmx.saic.com; 931-468-2603.]

Comparison with SEU Measurements:

1. Petersen [IEEE Trans. Nucl. Sci. 44, 2174-2187 (1997); Figure 7] recently reviewed reports on trapped-proton induced SEU rates in low-Earth orbits. In this review, Petersen accepted the experimenters' conclusions, without any detailed examination of their SEU data or the methodologies by which they made their AP8 calculations. Petersen found 22 such reports in the literature which, on average, indicated that the AP8 models tended to underpredict the observed SEU rates by a factor of ~0.7. However, the distribution of reported results was quite broad, and included cases where observations and predictions disagreed by more than an order of magnitude.
2. Petersen [1997; ibid.] also compared AP8MAX predictions with SEU-rates observed in the CRRES Micro-Electronics Package (MEP) prior to the March 1991 disturbance. CRRES flew in a highly-elliptical near-equatorial orbit, which carried it through the heart of the proton belts. Petersen's comparisons suggested that the general agreement was quite good, with an average predicted-to-observed ratio of ~1. However, Petersen's calculations neglected significant details of the CRRES orbital trajectory, such as the precession in the orbit's argument of perigee. For a more recent comparison, which takes into account the details of the CRRES orbit, see CRRES/MEP results.