One view of major Solar Energetic Particle (SEP) events is that these (proton-dominated) fluxes are accelerated in heliospheric shock sources created by Interplanetary Coronal Mass Ejections (ICMEs), and then travel mainly along interplanetary magnetic field lines connecting the shock(s) to the observer(s). This places a particular emphasis on the role of the heliospheric conditions during the event, requiring a realistic description of the latter to interpret and/or model SEP events. The well-known ENLIL heliospheric simulation with cone model generated ICME shocks is used together with the SEPMOD particle event modeling scheme to demonstrate the value of applying these concepts at multiple inner heliosphere sites.

We report a pair of oppositely directed sudden impulses (SI), in the geomagnetic field (ΔX), at ground stations, called SI+ – SI− pairs, that occurred between 1835 UT and 2300 UT on 23 April 1998. The SI+ – SI− pair, was well correlated with corresponding variations in the solar wind density, while solar wind velocity and the southward component of the interplanetary magnetic field (Bz) did not show any correspondence. This event had no source on the visible solar disk but was associated with a rear-side fast partial halo coronal mass ejection (CME) and an optically occulted M1.4 class solar flare behind the west limb. This event was unique in that one could clearly identify variations in ΔX at ground stations with solar wind parameters.

A scenario is presented whereby CMEs and interplanetary shocks are consequences of a large scale rearrangement of the coronal magnetic field induced by the disconnection of field lines from the solar surface due to the emergence of flux with opposite polarity. In this scenario the CME is the mass released from the previously closed structure and the interplanetary shock is formed by the injection of faster solar wind from an extended or newly created coronal hole which results from the opening of the field lines. Here CMEs and interplanetary shocks are associated events, but not cause-effect related. Observational and computational evidence supporting this view is provided.