It is generally recognized that there are at least two distinct venues for particle acceleration at the Sun: at CME-driven shocks, through first-order Fermi and shock-drift acceleration; and at flares, in association with reconnection, probably through resonant wave-particle interactions. Compositional signatures, such as elemental ratios (especially Fe/O), ionic charge states, isotopic abundances (especially 3He/4He), and the relative abundances of ultraheavy ions (with atomic numbers Z>30), distinguish these two sources of interplanetary particles at energies of a few MeV/nucleon. But as we move to energies of tens of MeV/nucleon and above, the very large SEP events associated with CME-driven shocks often show the compositional signatures of flare particles. This ostensible blurring in the distinction between the two SEP mechanisms has been one of the most persistent and controversial puzzles of the Cycle 23 observations. At least two hypotheses have been advanced to address these observations: (1) a direct flare component at high energies, in which the flare alone produces the higher-energy particles, while the CME-driven shock dominates particle production at lower energies; and (2) an indirect flare component, in which flare activity provides seed particles that are then accelerated to their observed energies by the CME-driven shock. In this paper, I will present the case for the second hypothesis as the likely explanation for most, if not all, of this ‘mixing’ of composition signatures. In particular, I will show how high-energy SEP spectral and compositional variability can be understood as natural consequences of evolution in the shock-normal angle as the CME-driven shock moves outward from the Sun, coupled with a compound seed population comprising at least flare ions and solar-wind suprathermals. I will also show why the direct-flare hypothesis provides a less satisfactory explanation for the observations.

It is widely accepted that in the smallest solar energetic particle (SEP) events the ions observed in the interplanetary medium are accelerated by flare processes. These ion increases, that have unusual relative abundances, are accompanied by electrons. The electrons are responsible for type III radio emissions that trace the open field lines emerging from the flare region. Since very energetic SEP events exhibit rapid onsets near the time of an associated flare it is highly likely that these first particles are accelerated in this flare. Just like in small SEP events, type III bursts and electrons are observed but the type III bursts last longer and the flares are larger. However in energetic events it is clear that interplanetary shock acceleration also occurs. The relative contribution from this additional source depends on a number of factors including the location of the observer and the structure of the interplanetary medium. A direct flare contribution in large events is often rejected on the grounds of a supposed lack of magnetic connection to the flare region and the delays observed between flare emissions and the inferred release of particles at the Sun. However observations do not support the supposed small connection region (flare events have been seen when the connection footpoint was more than 70 degrees from the flare) or the assumption of scatter-free transport required by the release time estimates

The experiments SEPICA onboard ACE and STOF onboard SOHO provide information on suprathermal heavy ion ionic charge states in solar energetic particle (SEP) events over the extended energy range of ~ 0.02 to 0.70 MeV/n. For events correlated with interplanetary shocks the ionic charge states in this energy range are mostly compatible with solar wind (SW) charge states or differ from SW charge states, e.g. for Fe, by only 1 to 2 charge units. In SEP events of short duration, low intensity and enrichment in heavy ions and ³He (usually called 'impulsive events'), we find significantly higher ionic charge states of ~10 -12 (Mg), ~11-14 (Si) und ~14-20 (Fe), with a remarkable increase of the mean ionic charge in this energy range by several charge states ( ~ 2-5 for Fe). This large increase of the mean ionic charge with energy at energies < 1MeV/n can be best explained by additional ionization of the ions by charge stripping in a sufficiently dense environment, during or after the acceleration. We compare the measured mean ionic charge of Mg, Si, and Fe with model calculations for the mean ionic charge as a function of particle energy, including the effect of energy dependent impact ionization. We show how these ionic charge measurements, in combination with model calculations and information on acceleration time scales, can be used to infer the altitude of the acceleration region, low in the corona.

We investigate 191 fast forward transient shocks observed by the ACE spacecraft from 1 February 1998 to 28 October 2003 and classify the energetic particle response associated with the passage of these shocks. For most shocks (~150) magnetic field and plasma data were available, and we were able to compute shock parameters. We present the frequency distributions of the angle between the upstream magnetic field and the normal to the shock, the Alfvenic Mach number, the shock normal speed in the frame of reference of the ambient plasma, and the plasma density and magnetic field ratios. We observe a trend for more quasi-perpendicular shocks. The Alfvenic Mach numbers show values less than about 7, the shock normal speeds less than about 350 km/s, and the density and magnetic field compression ratios values below 4.5. A few exceptions above these values were also observed. We compare our frequency distributions with those observed in the 1980s by ISEE-3, and the Helios 1 and 2 spacecraft. There is a trend for faster and stronger shocks to have stronger effects on the energetic particle intensities. However, the parameters of the shock do not determine unequivocally the characteristics of the energetic particle event observed at the passage of the shock.

Solar energetic particles in "gradual" events are accelerated efficiently by the process of diffusive shock acceleration at a CME-driven shock, a process in which the hydromagnetic waves excited by the energetic protons play a crucial role. New calculations of the upstream wave intensity are presented for the case in which ions are injected with a low energy at the shock, and in which a seed population of energetic ions with a prescribed distribution function is advected into the shock. In both cases the predicted wave intensity and ion distribution functions, their dependence on the angle between the shock normal and the upstream ambient magnetic field, the fractionation of ions upstream of the shock, the composition of the escaping ions, and the form of the high-energy cutoff are all presented. Selected comparisons with recent observations will also be featured

Results are presented from a fully time-dependent nonlinear model of the shock acceleration and transport of solar energetic particles coupled self-consistently to coronal Alfven waves. The model is based on the focused particle transport equation coupled to the wave kinetic equation. These equations express the conservation of particles and wave action in their respective phase spaces. For typical plasma parameters at a few solar radii, and starting with low wave intensities in the corona (mean free paths ~ 1 AU), preliminary results show that 20 – 40 keV suprathermal protons, at 1% of the ambient solar wind density, are accelerated to ~ 70 MeV in ~ 10 minutes by a traveling parallel shock of Alfven Mach number 3.3. The highest energy reached in the preliminary simulation is limited by the size of the simulation box and the spatial grid size. The simulation demonstrates the critical roles of particle focusing and wave excitation by streaming protons in enabling shock acceleration to bootstrap rapidly. It is also essential to have full pitch-angle dependence in the resonant wave-particle interaction, for only then can waves excited by small-pitch-angle low-energy particles scatter large-pitch-angle high-energy particles back to the shock, spreading and speeding particle acceleration from low to high energy. The acceleration time scale is expected to shorten for (a) shocks at lower coronal heights with larger magnetic field and higher suprathermal particle density , (b) larger ambient Alfven wave intensity, and (c) quasi-perpendicular shock. I will report on additional results from this continuing study as a contribution to the current debate on competing acceleration mechanisms in gradual SEP events.

Solar energetic particles (SEP) are accelerated in flares and at shocks in front of coronal mass ejections. The largest data bases are obtained by satellites in Earth orbit or at the Lagrange point, such as IMP, GOES or SOHO. These observations have shaped our understanding of SEP events. Observations at other radial distances, however, add significantly to our understanding for several reasons: for instance, in multi-spacecraft observations of individual events from different locations it becomes possible to disentangle the information about propagation and acceleration folded into the signal SEP. Shocks evolve and eventually interact as the propagate outwards: again, observations from different radial distances are helpful to understand the development of the shock´s acceleration efficiency during its propagation. Additional effects leading to variations in SEPs with radial distance are the asymmetric nature of the guiding interplanetary magnetic field combined with the rather radial propagation of the interplanetary shock and the interaction of shocks and particles with corotating interaction regions at larger radial distances.
In this talk, relevant observations from spacecraft at different radial distances (Helios, Voyager, Pioneer, partly also Ulysses, and for comparison earth-bound s/c) will be reported and compared to numerical models of acceleration and propagation to confirm our present understanding of particle acceleration and propagation and, more importantly, to identify crucial open questions in this field.

One of the outstanding puzzles of particle acceleration at interplanetary shocks is the nature of the mechanism at quasi-perpendicular shocks. At quasi-parallel shocks, it is well-known that the spatial gradient associated with energetic particles streaming away from the shock excites upstream waves which can then self-consistently scatter particles back and forth across the shock. First-order Fermi acceleration or diffusive shock acceleration is therefore closely tied to the self-consistent generation of the scattering wave field, but wave excitation is quenched with increasing shock obliquity. The enduring problem therefore is what provides for the scattering of particles at a quasi-perpendicular shock. The problem is integrally related to the nature of perpendicular particle transport. We use a recently developed model, Non-linear Guiding Center Theory, of the perpendicular component of the diffusion coefficient to investigate diffusive shock acceleration at quasi-perpendicular shocks. We investigate the time-scales for particle acceleration, the injection problem, and the accelerated time-dependent spectrum. The theoretical results are compared to observations of particle acceleration at highly oblique interplanetary shocks that exhibit very little wave activity. We also discuss recent SEP results which have been interpreted as demonstrating prompt acceleration at a perpendicular shock.

Numerical simulations of the evolution of interplanetary shocks exhibit the growth of distortions of the shock front and non-uniformity of energetic particles from place to place on the shock surface. In this work we use multi-spacecraft (ACE, Wind, SOHO, and Genesis) observations near Earth to search for such effects. About half the shocks studied are inconsistent with planar structures or with spherical structures with a radius of 1 AU. There are also differences in the details of the energetic particle fluxes seen at ACE and Wind.

The evolution of differential energy spectra and distribution functions of H+, He++ and He+ across shock boundaries are examined to gain insight into the acceleration of solar wind and pickup ions by various types of shocks. Data from the SWICS and HISCALE instruments on Ulysses are combined in order to span the wide energy range from 0.6 keV to almost 5 MeV. A number of different types of shock crossings are studied including forward and reverse shocks, CME shocks and the Jovian bow shock. Upstream of and far removed from shocks, ion velocity distributions are found to have hard suprathermal tails extending to high energies. Downstream distributions have higher particle densities in the tails, especially at energies corresponding to a few times the solar wind speed. Beyond about ten times the solar wind speed the density increase from upstream to downstream is relatively small. The presence of preexisting suprathermal tails provides particles with sufficiently high speeds for them to be readily injected into a shock acceleration mechanism. Using these observations made from 1 to 5.4 AU we will speculate on the acceleration of suprathermal ions in the distant heliosphere that is presently being probed by the Voyager 1 spacecraft.

Inner source pickup ions most likely originate from the interaction of the solar wind with dust in interplanetary space. Several model have been proposed that have these ions generated through origination from or interaction with dust. In regard to the emerging pickup ion distribution the models can be subdivided in two categories: either the newly generated ions start with the velocity of the dust, or after penetration of the dust solar wind ions start with a distribution between solar wind and very low speeds. In both cases a velocity distribution emerges, which is genuinely suprathermal, but peaks below and close to the solar wind speed. In the first case this is solely due to adiabatic cooling from close to the sun to the observer, in the latter case it is partly due to energy loss in the grains. Based on the observation that interstellar He⁺ pickup ions are preferentially injected and accelerated it has been suggested that also inner source ions may contribute significantly to energetic particles in interplanetary space. However, inner source ions have not been found in the CIR population that is accelerated mostly outside 1 AU and observed in Earth’s orbit. Since inner source ions are generated close to the sun one might expect that they could contribute to particles accelerated at interplanetary traveling shocks. We will present results from the search for such ions in shocks that appear to accelerate substantial amounts of He⁺. The results are discussed in the light of models of inner source pickup ion distributions and in terms of their implications for the injection of ions into the acceleration process.