The radiation belts
Protons and electrons trapped in Earth's magnetic field. They set the dose in most Earth orbits, and they are why the South Atlantic matters to a spacecraft at 400 km.
What they are
A charged particle in Earth's field spirals around a field line, bounces between the two hemispheres, and drifts around the planet: protons westward, electrons eastward. Particles on the right orbits stay for months to years. They form two belts.
- The inner belt, from about 1.2 to 2.5 Earth radii, holds protons up to hundreds of MeV.
- The outer belt, centered near 4 to 5 Earth radii, holds electrons of up to several MeV and changes from day to day with solar activity.
Earth's field is not centered on the planet. It is weakest over the South Atlantic, so there the inner belt dips lowest: the South Atlantic Anomaly. A spacecraft in low orbit collects nearly all of its trapped-proton dose in the few passes a day that cross it.
Positions in the belts are given by L, roughly the distance in Earth radii at which a field line crosses the magnetic equator, and by B/B₀, the field strength relative to its minimum on that line.
How we model them
- Method
- AP8 and AE8 through IRBEM on their native field epochs, with the South Atlantic Anomaly drifted 0.3° per year; low-orbit protons above 15 MeV doubled.
- Checked against
- The flight data compiled by Armstrong & Colborn (2000), ±25 %.
- Expect
- ±25 % above 15 MeV below 2000 km. A factor 2 to 3 below 15 MeV and in medium Earth orbit.
- The models. NASA's AP8 for protons, 0.1 to 400 MeV Sawyer & Vette (1976), and AE8 for electrons, 0.04 to 7 MeV Vette (1991), in their solar-minimum and solar-maximum versions. Through IRBEM the differential electron flux starts at 0.05 MeV. They are evaluated through the IRBEM library IRBEM library in SpacePy SpacePy 0.7.
- The field epoch. The models were built with the magnetic field of the 1960s. Used with today's field they predict an unrealistic rise in low-orbit flux Heynderickx et al. (1996). We evaluate them on their own epochs, as the ECSS standard prescribes ECSS-E-ST-10-04C, and move the map 0.3° per year westward to follow the drift of the anomaly Armstrong & Colborn (2000).
- A flight-data correction. Below 2000 km, protons above 15 MeV are doubled. Armstrong & Colborn compared AP8 with flux, dose and activation data from five missions and found the measurements about twice the model, to within 25 % Armstrong & Colborn (2000).
- In the browser. IRBEM cannot run in a browser, so orbit averages for a grid of orbits are computed ahead of time. See data.
Trapped anomalous cosmic rays
- Method
- One hundred times the interplanetary anomalous oxygen in a shell at L = 2 ± 0.12; nitrogen and neon at half their interplanetary ratios.
- Checked against
- SAMPEX, 1992 to 1998.
- Expect
- A factor 3. The dependence on altitude and field strength is assumed.
Anomalous cosmic rays that lose their remaining electrons in the upper atmosphere become trapped in a narrow shell near L = 2 Cummings et al. (1993) Selesnick et al. (1995). There the intensity is about a hundred times the interplanetary value Leske et al. (1999) and follows the same solar cycle Selesnick et al. (2000).
Assumptions and limits
- AP8 and AE8 are static averages from the 1960s and 1970s. They do not know about storms, and the outer belt in particular can differ from them by an order of magnitude on a given day.
- Below 15 MeV in low orbit, AP8 is uncorrected and unvalidated Heynderickx et al. (1999).
- AP8 orbit averages are generally taken to be good to about a factor 2 CREME-MC help, "AP8 Accuracy".
- Trapped protons in low orbit are strongly directional. We treat them as isotropic.
- The 3D explorer places belt tracers with a tilted, offset dipole, not the full field. That makes the anomaly too wide in low orbit; the live readouts are rescaled so that their daily mean matches the orbit average. See visual encoding.
- The Python engine accepts spectra from the newer AE9/AP9 models AFRL IRENE (AE9/AP9/SPM) in place of AP8/AE8.
Sources
| Used for | Source | Basis |
|---|---|---|
| Trapped protons | Sawyer & Vette (1976) NSSDC/WDC-A-R&S 76-06 | primary model |
| Trapped electrons | Vette (1991) NSSDC/WDC-A-R&S 91-24 | primary model |
| Model evaluation, IGRF, McIlwain L | IRBEM library; SpacePy 0.7 IRBEM library via SpacePy 0.7 | code |
| Native field epochs (AP8MIN 1964, AP8MAX 1970, AE8 ~JC60) | ECSS-E-ST-10-04C; Heynderickx et al. (1996) ECSS §9.2.1.1b, Table 9-1; Heynderickx et al. (1996): modern field ⇒ unrealistic secular flux increase at LEO | standard + primary |
| LEO trapped-proton correction ×2 above 15 MeV below 2000 km (±25 %); SAA drift 0.3°/yr west | Armstrong & Colborn (2000) SAIC-TN-99020 (NTRS 20000021571): APEX, CRRES, NOAA, LDEF, DMSP flux/dose/activation data | primary (NASA contractor report) |
| Context: AP8MIN at <600 km overestimates low-E, underestimates >30 MeV | Heynderickx et al. (1999) PSB97 model | primary |
| Context: AP8 orbit averages reliable to ~×2; LDEF low by ×2 | CREME-MC help, "AP8 Accuracy" CREME-MC help "AP8 Accuracy" (Vanderbilt) | review |
| AP9/AE9 replacement path (user-supplied spectra) | AFRL IRENE (AE9/AP9/SPM) vdl.afrl.af.mil/programs/ae9ap9; Windows binary on registration, source on request | model |
| Dose-depth | Seltzer (1994); IRBEM library NIST IR 5477 (via IRBEM) | primary |
| Used for | Source | Basis |
|---|---|---|
| Trapped N, O, Ne >15 MeV/n near L = 2, of interplanetary ACR origin; N/O, Ne/O, C/O below interplanetary | Cummings et al. (1993); Selesnick et al. (1995) | primary |
| Trapped ≈ 100 × interplanetary intensity; both ×5 from 1992 to 1997; ~1e-5–1e-4 cm⁻² s⁻¹ sr⁻¹ (MeV/n)⁻¹ | Leske et al. (1999) SAMPEX/MAST, ACE/SIS | primary |
| Solar-cycle dependence follows the interplanetary source | Selesnick et al. (2000) | primary |
| Not included: 1–4 MeV/n trapped ACR at L ~ 2.3 | Mazur et al. (2000) | primary |