Neoclassical Diffusion of Radiation‐Belt Electrons Across Very Low L‐Shells

Abstract

In the presence of drift‐shell splitting intrinsic to the International Geomagnetic Reference Field magnetic field model, pitch angle scattering from Coulomb collisions experienced by radiation‐belt electrons in the upper atmosphere and ionosphere produces extra radial diffusion, a form of neoclassical diffusion. The strength of the neoclassical radial diffusion at L < 1.2 exceeds that expected there from radial‐diffusion mechanisms traditionally considered and decreases with increasing L‐shell. In this work we construct a numerical model for this coupled (radial and pitch angle) collisional diffusion process and apply it to simulate raw count‐rate data observed aboard the Gemini spacecraft for several years after the 1962 Starfish nuclear detonation. The data show apparent lifetimes 10–100 times as long as would have been expected from collisional pitch angle diffusion and Coulomb drag alone. Our model reproduces apparent lifetimes for >0.5‐MeV electrons in the region 1.14 < L < 1.26 to within a factor of 2 (comparable to the uncertainty quoted for the observations). We conclude that neoclassical radial diffusion (resulting from drift‐shell splitting intrinsic to International Geomagnetic Reference Field's azimuthal asymmetries) mitigates the decay expected from collisional pitch angle diffusion and inelastic energy loss alone and thus contributes importantly to the long apparent lifetimes observed at these low L‐shells.

On Energetic Electron Dynamics During Geomagnetic Quiet Times in Earth's Inner Radiation Belt due to Atmospheric Collisional Loss and CRAND as a Source

“…The <D αα > is several orders of magnitude higher near the South Atlantic Anomaly (SAA) (~longitude = 300°) and at smaller equatorial PAs (PA < 65°), since the altitude of electron's mirror point in these regions is much lower (as illustrated in Li et al, ). These <D αα > results are comparable to previous calculations based on different ionospheric plasma models (e.g., Selesnick, , that used the global core plasma model [Gallagher et al, ] and Cunningham et al, , that used the IRI 2007 model [Bilitza & Reinisch, ]). The bounce‐averaged energy loss rates of 304‐keV electrons at L = 1.25 calculated based on equation are shown in Figures g–i.…”

“…As in the work of Cunningham et al (), we use the NRLMSISE‐00 model (Picone et al, ) for the number densities of neutral H, He, N, N 2 , O, O 2 , and Ar as a function of altitude, latitude, and the F 10.7 index, and we sample the model output at each point along the magnetic field line at a given location, date, and time using a fixed value of F 10.7 = 70 and Ap = 4. For charged species, we use the H + , He + , O + , NO + , and O 2 + output of the IRI 2012 (Bilitza et al, ).…”

“…The other boundary condition is at α eq = 90° for each ϕ . The term represents the energy loss effect and is treated as a 1‐D advection term (Cunningham et al, ). The energy losses of quasi‐trapped and trapped electrons are solved separately in the drift‐collision‐source model using the energy spectrum of quasi‐trapped and trapped electrons from DEMETER observations (Figures j–l) updated every 30 days, which is a sufficiently high frequency to capture the changes visible in Figures j–l.…”

“…Here we will use the term “collisions” to refer to the combined effects of PA scattering due to Coulomb collisions and Coulomb drag. The lifetime of ~500‐keV electrons at L < 1.2 (e.g., Cunningham et al, ; Imhof et al, ) and of 304‐keV electrons at L < 1.4 (e.g., Selesnick, ) calculated from satellite observations is much longer than theoretical estimates from atmospheric collisions, indicating that additional sources exist. Inward radial diffusion of electrons from higher L has been proposed as an additional source (e.g., Selesnick, et al, , suggested inward diffusion as a source of electrons up to L = 1.3), although it was noted quite earlier that the radial diffusion coefficient (inferred from satellite observations) in this scenario shows a decreasing trend as L values increase (at L = 1.15–1.21 from Newkirk & Walt, ), while radial diffusion calculated directly from theory is generally stronger at larger L (Fälthammar, ).…”

On the Use of Different Magnetic Field Models for Simulating the Dynamics of the Outer Radiation Belt Electrons During the October 1990 Storm

“…They specifically concluded that the model's underestimation when compared to data at high values was due to local acceleration from wave-particle interactions, which was not addressed by their model. In the same perspective, the VERB-3D (e.g., Kim et al, 2011;Shprits et al, 2009;Subbotin et al, 2010) code and the DREAM-3D code (e.g., Cunningham et al, 2018;Tu et al, 2013) are fed with PSD data derived from a variety of magnetic fields (preprocessing step), but they still account for a dipole field when the coordinate transformations are required through the resolution of the Fokker-Planck equation. Nevertheless, the absence of local acceleration in the 1D-RFP model is a strong limitation for storm times.…”