Long‐term‐average, solar cycle, and seasonal response of magnetospheric energetic electrons to the solar wind speed

Among the interplanetary activity parameters the solar wind speed is the one best correlated with the energetic electron fluxes in the inner magnetosphere. We examine the radial and temporal characteristics of the 2–6 MeV electron response, approximating it in this paper with linear filters. The filter response is parameterized by the time delay (τ), measured from the time of solar wind impact, and the L shell (L). We examine solar cycle and seasonal effects using an 8‐year‐long database of Solar, Anomalous, and Magnetospheric Particle Explorer (SAMPEX)/ Proton Electron Telescope (PET) measurements at the radial range L = 1.1–10. The main peak P 1 of the long‐term‐average response is at (τ, L) = (2, 5.3) and has a simultaneous response over a wide range of radial distances, ΔL = 5. The duration of the response after the peak is inversely proportional to the L shell. The central part of the inner magnetosphere (L = 3.7–5.75) has a much more prolonged response (>10 days) than other parts. Prior to the main response, P 1 , a brief response, P 0 , of typically lower amplitude appears at (τ, L) = (0, 3), probably as a quasi‐adiabatic response to the compression of the magnetosphere by the solar wind pressure. Over the solar cycle the variation in solar wind input results in a systematic change of the position, amplitude, radial extent, and duration of the two peaks: during solar wind minimum the quasi‐adiabatic peak disappears, and the radial size of the responding region decreases; both are responses to low‐density, high‐speed streams. During solar minimum, the duration is at least 3 days (30%) longer than average, probably due to the sustained solar wind input. Systematic variations appear also as a function of season due to several magnetic and fluid effects. During equinoxes the coupling is stronger, and the duration is longer (by at least 2 days) compared to solstices. Between the two equinoxes the fall response has a significantly higher amplitude and longer duration than the spring equinox response. This is at least partly due to the higher GSE B y component during the observation time, which acts to increase the effective GSM B z component according to the Russell‐McPherron effect. The seasonal modulation of the response is consistent with the variation in the fluxes themselves [ Baker et al. , 1999]. The modulation is discussed in terms of the equinoctial and axial hypotheses [ Boller and Stolov , 1970; Russell and McPherron , 1973; Cliver et al. , 2000].