1. Solar energetic particle simulations in SEPServer How to deal with scale separation of thirteen orders of magnitude R. Vainio, A. Afanasiev, J. Pomoell University of Helsinki, FINLAND N. Agueda, B. Sanahuja, University of Barcelona, SPAIN M. Battarbee, E. Valtonen, University of Turku, FINLAND U. Ganse, P. Kilian, F. Spanier, University of Würzburg, GERMANY

2. Introduction Solar energetic particle (SEP) acceleration occurs in flares and coronal/interplanetary shocks driven by CMEs Electron (proton) spectra extend from suprathermal energies up to tens (thousands) of MeVs in large SEP events SEPServer develops simulation tools for studying SEP acceleration and transport as well as for studies of radiation from the coronal acceleration regions Extreme spatial scales involved: - electron inertial length ~ 1 cm - distance to the observer ~ 1 AU Scale separation = 10 13 !

3. Models used to tackle the problem Global dynamics Dynamics of the bulk plasma → MHD simulation Particle transport from the source to the observer → Monte Carlo simulation Ion acceleration Coronal Shock Acceleration → Monte Carlo simulation Electron acceleration PiC simulation Plasma emission PiC simulation

4. Electron transport modeling Electron transport from a coronal source to the observer treated using the focused transport equation: + advection and adiabatic cooling Solved using the Monte Carlo method: simulate a large number of individual particles and collect the statistics at the position of the observer

5. Service to the community Database of Green's functions (= response at 1 AU to impulsive injection at the Sun) simulated: 1600 cases for different transport parameters simulated and made available to the community. Green's functions allow semi-empirical modeling of coronal sources See poster by Agueda et al.

6. Solar eruptions and shocks An MHD model used to model shocks driven by CMEs Parameters of the shock location speed normal angle etc. extracted from simulations

7. CSA – Coronal shock acceleration Results from MHD fed to a Monte Carlo simulation that follows ambient ion populations interacting with the shock Particle acceleration results from scattering off Alfvén waves on both sides of the shock Alfvén wave growth approximately computed from the streaming of ions Turbulent trap caused by the wave growth bootstraps the acceleration Shock model: RH jump conditions + semi- empirical cross-shock potential Very sensitive to ambient ion population, especially for oblique shocks! But: not applicable to electrons (different wave modes).

8. Testing the streaming-induced wave growth for flare-accelerated protons (see poster by Afanasiev et al.) precipitating protons escaping protons new results previous results by Vainio & Kocharov (2001)

9. Modeling at electron scales: ACRONYM PiC-Code Fully relativistic, 2D/3D PiC Code Correctly models complete collisionless plasma microphysics Constrained to kinetic scales Kinetic Instabilities Wave Behaviour Particle Acceleration

10. Type II Radio Burst Emission Drift-accelerated electrons in CME foreshock excite electrostatic waves Nonlinear wave-wave interaction leads to fundamental and harmonic radio emission Reproducible with ACRONYM

11. Electron acceleration at shocks Realistic v th, v A, v S Resolved e - scales Few proton gyro radii Focused on microphysics Constrained by computation time

12. Discussion: Coupling models or combining results? The complete modeling of solar energetic particle events requires a treatment of scales varying from electron inertial length to global scales. Problem treated using three approaches in SEPServer: MHD Monte Carlo simulation: particles and Alfvén waves (for ions) PiC simulation Increased understanding in SEPServer obtained so far by linking model results rather than coupling the codes. But is coupling of codes necessary to model the complete dynamics? Possible approaches (within computational reach): MHD wave effects on bulk plasma dynamics Detailed structure of shocks accelerating ions Fluctuations at MHD and ion scales cascading to electron scales Interesting new problems for future projects!

13. Acknowledgements The research leading to these results has received funding from the European Union’s Seventh Framework Programme (FP7/2007-2013) under grant agreement n:o 262773 (SEPServer). Computational support was provided by the Centre de Serveis Científics i Acadèmics de Catalunya (CESCA), CSC – IT Centre for Science, and Jülich Supercomputer Centre. We also acknowledge support by the COST Action ES0803 “Developing space weather products and services in Europe”.