Evidence of a nonlinear transition from mitigation to suppression of the edge localized mode (ELM) by using resonant magnetic perturbations (RMPs) in the EAST tokamak is presented. This is the first demonstration of ELM suppression with RMPs in slowly rotating plasmas with dominant radio-frequency wave heating. Changes of edge magnetic topology after the transition are indicated by a gradual phase shift in the plasma response field from a linear magneto hydro dynamics modeling result to a vacuum one and a sudden increase of three-dimensional particle flux to the divertor. The transition threshold depends on the spectrum of RMPs and plasma rotation as well as perturbation amplitude. This means that edge topological changes resulting from nonlinear plasma response plays a key role in the suppression of ELM with RMPs. DOI: 10.1103/PhysRevLett.117.115001 Magnetic reconnection and the resultant topological change play an important role in plasma dynamics in both laboratory and space plasma physics research. The formation of an edge stochastic magnetic field caused by resonant magnetic perturbations (RMPs) is believed to be the reason for the suppression of periodic crash events near the plasma edge known as the edge localized mode (ELM) observed in the DIII-D tokamak [1]. The ELM causes transient heat loads to the plasma facing components and may degrade them on the next generation fusion device like ITER [2]. The reduction of free energy in the edge pressure gradient and current because of field stochasticity moves the plasma into a stable regime against the ELM [3]. This successful experiment motivated ELM control using RMPs in many other tokamaks [4][5][6][7]. However, the plasma response effect usually shields the external applied RMPs and may significantly reduce the magnetic field stochasticity [8][9][10][11], which makes this mechanism questionable. Different from topological change, the linear peelinglike magneto hydro dynamics (MHD) response has been found to play an important role in ELM control [12][13][14]. Nonlinear plasma response has been observed in the JET totamak [15]. The possible formation of a magnetic island near the plasma edge [16] with a toroidal Fourier mode number n ¼ 1 during ELM suppression by using n ¼ 2 RMP has been recently observed on DIII-D [17]. However, the key difference between ELM suppression and mitigation and the different roles of linear and nonlinear plasma response on ELM suppression are still not clear.In this Letter, we report the first observation of full ELM suppression using low n RMPs in slowly rotating plasmas with dominant radio-frequency (rf) wave heating, which is potentially important for the application of this method for a future fusion device. This is the first observation of full ELM suppression using RMPs in the medium plasma collisionality regime in EAST, and it expands beyond the previous observations of ELM suppression on DIII-D [1,3] and KSTAR [7]. It is found that not only the formation of a magnetic island near the edge [17] but also a critical leve...

Strong mitigation of edge-localized modes has been observed on Experimental Advanced Superconducting Tokamak, when lower hybrid waves (LHWs) are applied to H-mode plasmas with ion cyclotron resonant heating. This has been demonstrated to be due to the formation of helical current filaments flowing along field lines in the scrape-off layer induced by LHW. This leads to the splitting of the outer divertor strike points during LHWs similar to previous observations with resonant magnetic perturbations. The change in the magnetic topology has been qualitatively modeled by considering helical current filaments in a field-line-tracing code.

This paper reports on disruption prediction using a shallow machine learning method known as a random forest, trained on large databases containing only plasma parameters that are available in real-time on Alcator C-Mod, DIII-D, and EAST. The database for each tokamak contains parameters sampled ∼10 6 times throughout ∼10 4 discharges (disruptive and nondisruptive) over the last four years of operation. It is found that a number of parameters (e.g. P rad /P input , i , n/n G , B n=1 /B 0 ) exhibit changes in aggregate as a disruption is approached on one or more of these tokamaks. However, for each machine, the most useful parameters, as well as the details of their precursor behaviors, are markedly different. When the prediction problem is framed using a binary classification scheme to discriminate between time slices 'close to disruption' and 'far from disruption', it is found that the prediction algorithms differ substantially in performance among the three machines on a time slice-by-time slice basis, but have similar disruption detection rates (∼80%-90%) on a shot-by-shot basis after appropriate optimisation. This could have important implications for disruption prediction and avoidance on ITER, for which development of a training database of disruptions may be infeasible. The algorithm's output is interpretable using a method that identifies the most strongly contributing input signals, which may have implications for avoiding disruptive scenarios. To further support its real-time capability, successful applications in inter-shot and real-time environments on EAST and DIII-D are also discussed.