We model the 2012 May 17 solar energetic particle event by combining the AWSoM and iPATH codes. Using this combined approach, we investigate particle acceleration when the parent coronal mass ejection (CME) is still close to the Sun. We have obtained reasonable agreements between our simulation and observations made by multiple spacecraft. We follow the three-dimensional CME and the CME-driven shock from their initiation using the AWSoM code for a period of 3 hours when the shock is below ∼20 R s . Above 20 R s , iPATH2D is used to follow the CME-driven shock. The plasma properties from the AWSoM code are fed into the iPATH model, where particle acceleration at the shock front is modelled and the instantaneous energetic particle spectra are obtained. The subsequent transport of these energetic particles in the solar wind is followed using the iPATH model. We obtain both the instantaneous particle spectra and particle fluence at 1 au, and we then compare them with observations. To account for uncertainties of magnetic field connectivity from 1 au to the shock, as well as uncertainties of the shock profiles, our model’s results are obtained as an ensemble average where, instead of considering Earth as a single point location, we consider multiple locations within 10 degrees in longitude and latitude from Earth. The effect of this model uncertainty mimics that of the field line meandering, as suggested by Bian & Li, but is of different origin. We suggest that a trustworthy solar energetic particle forecast should be made in an ensemble average approach.

On 2017 September 10, a fast coronal mass ejection (CME) erupted from the active region (AR) 12673, leading to a ground level enhancement (GLE) event at Earth. Using the 2D improved Particle Acceleration and Transport in the Heliosphere (iPATH) model, we model the large solar energetic particle (SEP) event of 2017 September 10 observed at Earth, Mars and STEREO-A. Based on observational evidence, we assume that the CME-driven shock experienced a large lateral expansion shortly after the eruption, which is modeled by a double Gaussian velocity profile in this simulation. We apply the in-situ shock arrival times and the observed CME speeds at multiple spacecraft near Earth and Mars as constraints to adjust the input model parameters. The modeled time intensity profiles and fluence for energetic protons are then compared with observations. Reasonable agreements with observations at Mars and STEREO-A are found. The simulated results at Earth differ from observations of GOES-15. However, the simulated results at a heliocentric longitude 20° west to Earth fit reasonably well with the GOES observation. This can be explained if the pre-event solar wind magnetic field at Earth is not described by a nominal Parker field. Our results suggest that a large lateral expansion of the CME-driven shock and a distorted interplanetary magnetic field due to previous events can be important in understanding this GLE event.

Aims. We present the implementation of a coupling between EUropean Heliospheric FORcasting Information Asset (EUHFORIA) and improved Particle Acceleration and Transport in the Heliosphere (iPATH) models. In this work, we simulate the widespread solar energetic particle (SEP) event of 2020 November 29 and compare the simulated time-intensity profiles with measurements at Parker Solar Probe (PSP), the Solar Terrestrial Relations Observatory (STEREO)-A, SOlar and Heliospheric Observatory (SOHO), and Solar Orbiter (SolO). We focus on the influence of the history of shock acceleration on the varying SEP time-intensity profiles and investigate the underlying causes in the origin of this widespread SEP event.Methods. We simulated a magnetized coronal mass ejection (CME) propagating in the data-driven solar wind with the EUHFORIA code. The CME was initiated by using the linear force-free spheromak module of EUHFORIA. The shock parameters and a 3D shell structure were computed from EUHFORIA as inputs for the iPATH model. Within the iPATH model, the steady-state solution of particle distribution assuming diffuse shock acceleration is obtained at the shock front. The subsequent SEP transport is described by the focused transport equation using the backward stochastic differential equation method with perpendicular diffusion included. Results. We examined the temporal evolution of shock parameters and particle fluxes during this event and we find that adopting a realistic solar wind background can significantly impact the expansion of the shock and, consequently, the shock parameters. Timeintensity profiles with an energetic storm particle event at PSP are well reproduced from the simulations. In addition, the simulated and observed time-intensity profiles of protons show a similar two-phase enhancement at STA. These results illustrate that modeling a shock using a realistic solar wind is crucial in determining the characteristics of SEP events. The decay phase of the modeled time-intensity profiles at Earth is in good agreement with the observations, indicating the importance of perpendicular diffusion in widespread SEP events. Taking into account the possible large curved magnetic field line connecting to SolO, the modeled timeintensity profiles show a good agreement with the observation. We suggest that the broadly distorted magnetic field lines, which are due to a stream interaction region, may be a key factor in helping to improve our understanding of the observed SEPs at SolO for this event.