Coronal jets have been the subject of study for decades, particularly following their discovery in X-ray images in the 1990s . They consist of a brightening of small coronal loops at their base, and a jet of heated plasma launched upward along open magnetic field lines. A burst of magnetic reconnection between the short base loops and the surrounding field is the obvious candidate for explaining both features , and many jets fit this simple picture. However, a sub-set of jets have a more complex evolution; involving the ejection of cool, dense plasma with clear untwisting motions. A simple burst of reconnection cannot easily explain them . Something more complex is required.Relatively recent observations have revealed that this sub-set of jets usually involve the eruption of a so called “mini-filament”, a small-scale version of their larger cousins which erupt to produce Coronal Mass Ejections (CMEs) . A comparison of a mini-filament jet with a full-scale filament eruption is shown in Fig. 1. The similarities between the tiny and large-scale filament eruptions suggests the two phenomena may be related, i.e. that these jets are miniature CMEs.
Recently, we developed a numerical model for mini-filament jets and showed that they can indeed be thought of as scaled down CME eruptions . Thus, coronal jets can be thought of as miniature CMEs when they are generated by the eruption of a mini-filament. In this nugget, we discuss this model and how different realisations of the model can help explain the different jet dynamics in the run up to these tiny eruptions.
Scaling down the eruption process
Our model is essentially a scaled down version of the Breakout Model for large scale CMEs (see  and  for a full description). Figure 2 depicts schematically the different stages of the model. The initial magnetic field (Fig. 2a) consists of a 3D null point (zero point in the field) with a domed separatrix surface (separating open and closed field) and inner and outer spine lines. Starting with this field there are then 3 main stages:
- Filament channel formation: Some process forms sheared field lines (yellow) capable of supporting mini-filament material in the closed field. In our numerical model we use surface motions, but in principle this could be flux emergence or shearing and cancellation. The system remains in force balance since the magnetic tension of the cyan field lines (the strapping field) balances the outward magnetic pressure of the sheared field lines (the filament channel).
- Breakout: The increasing magnetic pressure of the growing filament channel expands the strapping cyan field lines. Eventually, this forms a current layer at the null above, which begins to reconnect the strapping field. This reduces the downward force on the filament and allows it to expand faster. Beyond a tipping point, the feedback between the reconnection at the null and the upward expansion of the filament channel becomes self-sustaining and the filament starts to rise. This is the so called Magnetic Breakout mechanism . As the filament channel stretches upwards at a steady pace, a second current layer forms beneath it. Additional reconnection in this layer turns the sheared field lines into a twisted flux rope.
- Eruptive jet: Finally, when the rising flux rope reaches the null current layer, it is reconnected on to open field lines. The transfer of twist on to open field drives non-linear Alfvén waves that create an untwisting jet.
This multi-step model explains how the mini-filaments erupt, how they produce jets and why the jets have untwisting motions. It also theoretically links these tiny eruptions with large-scale filament eruptions through the same eruption mechanism (for a detailed discussion see , ).
Explaining different observed behaviours
Are the different stages of the model observed though? The model places no constraints on where the mini-filament comes from and agrees well with the ultimate untwisting jet. The interesting thing observationally is whether reconnection outflows are observed before the main untwisting jet that match those we expect in the breakout phase. Some mini-filament jets do show this, particularly when the surrounding field is highly inclined . However, others in less inclined field regions do not. This suggests that the background field inclination is an important factor.
To better understand this, we used the Adaptively Refined Magnetohydrodynamics Solver (ARMS) code to run three simulations with different background field inclination angles: +22, 0 and -22 degrees. The angle is with respect to the vertical direction, Fig. 2a. Figure 3 shows the velocity magnitude during the jet eruption with vertical field. After a short burst a long, slow breakout phase occurs with weak vertical exhaust flows. For the other two cases we find faster, denser reconnection exhausts over a shorter period. The main reason is that in the non-vertical cases the angle of contact between the strapping (cyan) and overlying (red) field lines is increased, driving more intense reconnection at the null (for full details see ). Thus, for higher inclination angles what are likely to be more observable reconnection exhausts are produced, agreeing qualitatively with the observations.
Coronal jets are deceptively simple events that we are only now getting a clear understanding of. It is now apparent that in the cases where a mini-filament is involved they share similarities with large-scale filament eruptions. Such jets can therefore be thought of as mini-CMEs. We have introduced a model for jets driven by mini-filament eruptions that explains the link between these large-scale and small-scale events. Our work suggests that eruptions across vastly different scales in the corona can be understood within the same framework. The three realisations of the model we studied also help explain the differences between individual mini-filament jets. With further space missions on the horizon, particularly the improved views of the poles with Solar Orbiter, the future looks bright for further unpicking the secrets of these small but mighty events.
-  Shibata, K. et al. 1992, PASJ, 44, L173
-  Moore, R. L. et al. 2010, ApJ, 720, 757
-  Sterling, A. C. et al. 2015, Nature, 523, 437
-  Shen, Y. et al. 2012, ApJ, 745, 164
-  Wyper, P. F., Antiochos, S. K., & DeVore, C. R. 2017, Nature, 544, 452
-  Antiochos, S. K., DeVore, C. R., & Klimchuk, J. A. 1999, ApJ, 510, 485
-  Karpen, J. T., Antiochos, S. K., & DeVore, C. R. 2012, ApJ, 760, 81
-  Wyper, P. F., DeVore, C. R., & Antiochos, S. K., ApJ, 852, 98
-  Liu, C. et al., 2011, ApJL, 735, L18