Numerical simulations of shear-induced consecutive coronal mass ejections

Abstract

Context. It is widely accepted that photospheric shearing motions play an important role in triggering the initiation of coronal mass ejections (CMEs). Even so, there are events for which the source signatures are difficult to locate, while the CMEs can be clearly observed in coronagraph data. These events are therefore called ‘stealth’ CMEs. They are of particular interest to space weather forecasters, since eruptions are usually discarded from arrival predictions if they appear to be backsided, which means not presenting any clear low-coronal signatures on the visible solar disc. Such assumptions are not valid for stealth CMEs since they can originate from the front side of the Sun and be Earth-directed, but they remain undetected and can therefore trigger unpredicted geomagnetic storms. Aims. We numerically model and investigate the effects of shearing motion variations onto the resulting eruptions and we focus in particular on obtaining a stealth CME in the trailing current sheet of a previous ejection. Methods. We used the 2.5D magnetohydrodynamics package of the code MPI-AMRVAC to numerically simulate consecutive CMEs by imposing shearing motions onto the inner boundary, which represents, in our case, the low corona. The initial magnetic configuration consists of a triple arcade structure embedded into a bimodal solar wind, and the sheared polarity inversion line is found in the southern loop system. The mesh was continuously adapted through a refinement method that applies to current carrying structures, allowing us to easily track the CMEs in high resolution, without resolving the grid in the entire domain. We also compared the obtained eruptions with the observed directions of propagation, determined using a forward modelling reconstruction technique based on a graduated cylindrical shell geometry, of an initial multiple coronal mass ejection (MCME) event that occurred in September 2009. We further analysed the simulated ejections by tracking the centre of their flux ropes in latitude and their total speed. Radial Poynting flux computation was employed as well to follow the evolution of electromagnetic energy introduced into the system. Results. Changes within 1% in the shearing speed result in three different scenarios for the second CME, although the preceding eruption seems insusceptible to such small variations. Depending on the applied shearing speed, we thus obtain a failed eruption, a stealth, or a CME driven by the imposed shear, as the second ejection. The dynamics of all eruptions are compared with the observed directions of propagation of an MCME event and a good correlation is achieved. The Poynting flux analysis reveals the temporal variation of the important steps of eruptions. Conclusions. For the first time, a stealth CME is simulated in the aftermath of a first eruption, originating from an asymmetric streamer configuration, through changes in the applied shearing speed, indicating it is not necessary for a closed streamer to exist high in the corona for such an event to occur. We also emphasise the high sensitivity of the corona to small changes in motions at the photosphere, or in our simulations, at the low corona.