Empirically modeled global distribution of magnetospheric chorus amplitude using an artificial neural network

摘要: [1] Accurate knowledge of the global distribution magnetospheric chorus waves is essential for radiation belt modeling because it provides a direct link to understanding losses and acceleration processes. In this paper, we report on newly developed models amplitudes based in situ measurements interplanetary magnetic field (IMF) solar wind parameters as well geomagnetic indices using an artificial neural network technique. We find that speed IMF BZ are most influential affect evolution chorus. The variations outer (L ≥ 7) inner (5 ≤ L < regions, respectively, correlated with BZ. addition, parameter-based model generally results slightly higher correlation between measured modeled than any other including AE, Kp, Dst. shows amplified near prenoon sector during geomagnetically disturbed conditions. With increasing southward location peak amplitude moves from midnight sector, which due enhanced electron injection midnight. also present comparison diffusive transport simulations electrons interacting two models, index-based models. outside plasmapause can dynamic even inside plasmasphere populations act seed population particles plasmapause. One weakness our trained early phase cycle 24 where very few strong storms occurred. Therefore, might not be valid reproducing activity under extremely conditions, should updated future once such conditions become available.