Anomalous cosmic rays
Interstellar atoms that drift into the heliosphere, lose an electron, and are accelerated in its outer reaches. Near solar minimum they dominate helium, nitrogen, oxygen and neon at a few MeV per nucleon.
What they are
Neutral atoms from interstellar space drift into the heliosphere unhindered by its magnetic field. Sunlight and the solar wind ionize them, the wind picks them up and carries them outward, and some are accelerated to between a few and a few tens of MeV per nucleon. They diffuse back in as "anomalous" cosmic rays (ACR): anomalous because they showed up as an unexplained bump at the low end of the cosmic-ray spectrum.
Two things set them apart. They are mostly singly charged, so for the same energy they are far harder to bend than fully stripped cosmic rays and reach much deeper into Earth's magnetic field. And they are very sensitive to the solar cycle: plentiful at solar minimum, almost gone at maximum.
How we model them
- Method
- Oxygen fitted to HELIOS, 1975 to 1977; other elements by fixed ratios; intensity as the neutron-monitor rate to the 25th power; radial gradients of 49, 35, 13 and 4 %/AU.
- Checked against
- The factor of about 500 between solar minimum and maximum in CREME96.
- Expect
- A factor 2 at solar minimum, 3 near maximum. The shape below 9 MeV/n is assumed.
- Oxygen at 1 AU. j(E) = 3.567×10⁻⁵ / (1 + (E / 6 MeV/n)^3.668) cm⁻² s⁻¹ sr⁻¹ (MeV/n)⁻¹, from 1 to 100 MeV/n. We fitted it to the HELIOS oxygen intensities at 9 to 13 and 13 to 28 MeV/n for 1975 to 1977 Marquardt et al. (2018), after subtracting our own galactic oxygen at the same modulation.
- Other elements. Fixed ratios to oxygen at equal energy per nucleon Reames (1999).
- Solar cycle. The intensity follows the neutron-monitor count rate to the 25th power Leske et al. (2005), with the rate tied to φ Usoskin et al. (2017). It is capped at the 1975 to 1977 level: the record cosmic-ray minimum of 2009 did not produce record anomalous cosmic rays Leske et al. (2013) Rankin et al. (2022).
- Distance. Gradients of 49 % per AU inside 1 AU, 35 from 1 to 5 AU, 13 from 5 to 15 AU and 4 beyond, stopped at 90 AU Rankin et al. (2022).
- Charge. Singly charged below 20 MeV/n and fully stripped above Tylka et al. (1997). This sets the rigidity used for geomagnetic shielding.
The same population, trapped near L = 2, is described under the radiation belts.
Assumptions and limits
- Below about 9 MeV/n the spectrum is flattened by assumption. The measurements we fitted do not reach lower.
- The ratios to oxygen ignore the fact that each species is modulated a little differently.
- Expect a factor 2 at solar minimum and a factor 3 near maximum, where the suppression is a factor of several hundred and poorly known.
Sources
| Used for | Source | Basis |
|---|---|---|
| ACR O spectrum at ~1 AU (fit to 9–13 and 13–28 MeV/n intensities, 1975–77) | Marquardt et al. (2018) Table 1 | primary measurement |
| ACR composition (He 5000, N 120, O 1000, Ne 70, Ar 4.2, Si 1.7, Mg 1.2, S 0.6; C < 10) | Reames (1999) Table 1 | primary |
| Solar-cycle scaling ∝ (NM rate)²⁵ | Leske et al. (2005); Usoskin et al. (2017) NM–φ relation φ = A/N + B: Usoskin et al. (2017) Eq. 4 | primary |
| Cross-checks of the scaling: ×~500 min→max; ×100+ suppression 2001–03 | Tylka et al. (1997); Leske et al. (2005) CREME96; Leske et al. (2005) | primary |
| Cap at the 1975–77 level for deeper minima | Leske et al. (2013); Rankin et al. (2022) Leske et al. (2013): 2009 ACRs did not exceed earlier qA>0 levels; Rankin et al. (2022): SC24/25 ≈ SC20/21 at 1 AU | primary |
| Radial gradients 49 / 35 / 13 / 4 %/AU | Rankin et al. (2022); Cummings et al. (1987) Rankin et al. (2022) Table 1 compiling Marquardt 2018, Webber 1977/1981, Cummings 1990/2009, Marsden 1999; Cummings et al. (1987) ICRC 20 for >21 AU | primary + compilation |
| Q = 1 below 20 MeV/n; trapped ACRs omitted | Tylka et al. (1997) CREME96 (SAMPEX charge-state results) | primary |