Magnetically confined fusion devices, such as the stellarator TJ-II, are highly complex infrastructures where plasma evolution is monitored through advanced diagnostic systems. Failures in these diagnostics can result in missing or unreliable data, compromising operational stability and experiment reliability. This work proposes a data-driven nowcasting approach to estimate missing plasma density signals using alternative sensor inputs, specifically validated for Electron Cyclotron Resonance Heating (ECRH) regimes. Our methodology uses a Two-Stage Learning Framework that integrates a Variational Autoencoder for latent feature extraction with Recurrent Neural Network blocks and a Multi-Head Attention mechanism. This architecture is designed to capture long-range temporal dependencies by broadcasting a single latent representation across multiple time steps, thus merging local and global temporal features. Extensive experiments on 201 ECRH-heated discharges from the TJ-II dataset demonstrate that deeper recurrent architectures, particularly those employing Long Short-Term Memory units, significantly outperform alternative models in nowcasting plasma density. The best-performing model achieves superior accuracy with a Weighted Mean Absolute Percentage Error (WMAPE) of 0.0328, complemented by low mean squared error and high coefficients of determination. This proof-of-concept study demonstrates the technical feasibility of deep learning-based nowcasting for diagnostic substitution in well-controlled ECRH operational conditions, establishing a methodological foundation for future extension to broader stellarator regimes.

Magnetic confinement fusion experiments operate with extensive diagnostic coverage and high per-shot costs. Although machine operation is guided primarily by expert knowledge, theoretical understanding and decades of experimental experience, many aspects of scenario development still require exploring actuator trajectories whose consequences are uncertain until a discharge is executed. In this context, it is appealing to develop fast surrogate models capable of forecasting selected diagnostic signals, enabling in silico evaluation of candidate scenarios before committing machine time. Such capabilities would contribute to the long-term vision of digital twin infrastructures, which are beginning to gain traction in fusion but are not yet standard tools. As a proof of principle, we forecast the time evolution of the ACTON275 1 carbon 4+ (CV) line-emission signal in the TJ-II stellarator. The diagnostic consists of a 275 mm focal-length spectrometer tuned to lambda = 227.1 nm; its signal correlates with line-averaged electron density, impurity sourcing and global radiation, making it a sensitive proxy for fuelling efficiency and impurity dynamics. Photons are converted into a voltage trace via a photomultiplier and amplifier stage and stored in the TJ-II database. We compare a classical baseline support vector regression (SVR) with three neural architectures: echo state networks (ESNs), long short-term memory networks (LSTMs) and an autoencoder-equipped LSTM (AE-LSTM), for multivariate time series forecasting of ACTON275 under electron cyclotron resonance heating (ECRH) conditions. Model selection is formulated as a multi-objective search favouring high average R-2 on unseen shots, low dispersion of R-2 and compact architectures compatible with real-time deployment. The best models reproduce the build-up and decay of CV emission with useful accuracy across a substantial fraction of discharges, suggesting that data-driven surrogates can form a building block towards future digital twin components for stellarator operation.

This study reports on a set of experiments designed to clarify the impact of the rotational transform on confinement quality at the TJ-II stellarator. For this purpose, the net plasma current is controlled using external coils, resulting in the modification of the rotational transform profile. Significant and systematic variations of the edge electron density gradients (up to $50\,\%{-}60\,\%$ ) and the plasma energy content ( $20\,\%{-}30\,\%$ ) are achieved. The explanation of this behaviour relies on the placement of low-order rational surfaces in relation to the edge gradient region, which affect local turbulence fluctuation levels, facilitating the formation of zonal flows and concomitant transport barriers. This hypothesis is confirmed experimentally on the basis of a broad array of diagnostic measurements. Calculations based on a resistive magnetohydrodynamic turbulence model provide qualitative support for this hypothesis, clarifying the impact on confinement of specific rational surfaces and highlighting the complex nature of magnetically confined fusion plasmas.

The maximum achievable density in a tokamak is restricted by the so-called density limit. In this contribution, we observe that a MARFE precedes the final chain of events leading to disruption in ramp up L-mode discharges, resulting in strong power dissipation and plasma cooling, which likely makes the plasma unstable to MHD modes. However, this phase is preceded by an initial cooling governed by enhanced radial transport due to higher fluctuations and a more favorable cross-phase for transport, which likely impacts the MARFE formation. The increase in radial turbulent transport near the separatrix is primarily observed following the detachment of the outer divertor. This occurs when an X-point radiator forms and the edge collisionality rises, indicating that the increased collisionality influences the turbulence characteristics in the plasma boundary. The causal connection between the events that precede the disruption is discussed.

After a long device enhancement phase, scientific operation resumed in 2022. The main new device components are the water cooling of all plasma facing components and the new water-cooled high heat flux divertor units. Water cooling allowed for the first long-pulse operation campaign. A maximum discharge length of 8 min was achieved with a total heating energy of 1.3 GJ. Safe divertor operation was demonstrated in attached and detached mode. Stable detachment is readily achieved in some magnetic configurations but requires impurity seeding in configurations with small magnetic pitch angle within the edge islands. Progress was made in the characterization of transport mechanisms across edge magnetic islands: Measurement of the potential distribution and flow pattern reveals that the islands are associated with a strong poloidal drift, which leads to rapid convection of energy and particles from the last closed flux surface into the scrape-off layer. Using the upgraded plasma heating systems, advanced heating scenarios were developed, which provide improved energy confinement comparable to the scenario, in which the record triple product for stellarators was achieved in the previous operation campaign. However, a magnetic configuration-dependent critical heating power limit of the electron cyclotron resonance heating was observed. Exceeding the respective power limit leads to a degradation of the confinement.

The development of reduced models provide efficient methods that can be used to perform short term experimental data analysis or narrow down the parametric range of more sophisticated numerical approaches. Reduced models are derived by simplifying the physics description with the goal of retaining only the essential ingredients required to reproduce the phenomena under study. This is the role of the gyro-fluid code FAR3d, dedicated to analyze the linear and nonlinear stability of Alfvén Eigenmodes (AE), Energetic Particle Modes (EPM) and magnetic-hydrodynamic modes as pressure gradient driven mode (PGDM) and current driven modes (CDM) in nuclear fusion devices. Such analysis is valuable for improving the plasma heating efficiency and confinement; this can enhance the overall device performance. The present review is dedicated to a description of the most important contributions of the FAR3d code in the field of energetic particles (EP) and AE/EPM stability. FAR3d is used to model and characterize the AE/EPM activity measured in fusion devices as LHD, JET, DIII-D, EAST, TJ-II and Heliotron J. In addition, the computational efficiency of FAR3d facilitates performing massive parametric studies leading to the identification of optimization trends with respect to the AE/EPM stability. This can aid in identifying operational regimes where AE/EPM activity is avoided or minimized. This technique is applied to the analysis of optimized configurations with respect to the thermal plasma parameters, magnetic field configuration, external actuators and the effect of multiple EP populations. In addition, the AE/EPM saturation phase is analyzed, taking into account both steady-state phases and bursting activity observed in LHD and DIII-D devices. The nonlinear calculations provide: the induced EP transport, the generation of zonal structures as well as the energy transfer towards the thermal plasma and between different toroidal/helical families. Finally, FAR3d is used to forecast the AE/EPM stability in operational scenarios of future devices as ITER, CFETR, JT60SA and CFQS as well as possible approaches to optimization with respect to variations in the most important plasma parameters.

A pellet-induced enhanced confinement (PiEC) phase, with general characteristics similar to those reported for the stellarator W7-X, is observed after single pellet injection (>10 19 H atoms) into the neutral beam injection heated phase of plasmas in the mid-sized heliac-type stellarator TJ-II. In addition to a step-like increase in density, plasma diamagnetic energy content rises significantly with respect to that of reference discharges, energy confinement time is similarly enhanced when compared to International Stellarator Scaling law predictions (Yamada et al 2005 Nucl. Fusion 45 1684) renormalized for TJ-II, and the triple product, n e · T i · τ E , exhibits a clear bifurcation towards an improved confinement branch when compared to the branch product predicted by the same law. In this work, multiple pellets are injected in series into NBI-heated plasmas in the TJ-II and post-injection plasma performance is reported and discussed. For instance, a charge-exchange recombination spectroscopy diagnostic reveals significantly increased core ion temperatures after pellet injection compared to temperatures achieved in comparable reference plasmas, this pointing to increased ion energy content and improved ion energy confinement during a PiEC phase. It is also found that enhanced performance is independent of whether co- or counter-NBI heating beam is employed. Finally, record stored diamagnetic energy content and plasma beta values are achieved when the largest available pellets are employed. The results indicate that pellet injections extend the operational regime well beyond limits previously achieved in TJ-II without pellets.

We provide an overview of activities carried out at the TJ-II stellarator for improving our understanding of- and developing plasma physics models for particle density profiles in stellarators. Namely, we report on recent progress in turbulent particle transport simulation, validation of pellet deposition models, density profile shaping for performance control and new experimental techniques for edge turbulence and plasma-neutral interaction.

In this work, we report on the results obtained by measuring several turbulent quantities well inside the plasma edge by means of a Langmuir probe during dynamical rotational transform scans in the TJ-II stellarator, while applying a radial electric field to the edge plasma using a biasing probe. By calculating the intermittence parameter from floating potential measurements, we are able to identify a major low order rational surface and hence relate the probe measurements to the local value of the rotational transform. Based on the former, we are able to show that the poloidal plasma velocity (and hence radial electric field) has a significant radial structure that is clearly related to the rotational transform profile and in particular the lowest order rational surfaces in the range studied. The poloidal velocity is also affected by the edge biasing. The particle flux gamma was also found to exhibit a radial pattern, as did the flow shear suppression term omega(ExB), but the relation of the former to the low-order rational surfaces was less clear. We surmise that this lack of direct correspondence is due to an unknown term in the turbulence evolution equation: the instability growth rate, gamma. We make use of a reduced Magnetohydrodynamic turbulence model to interpret the results. Overall, a picture is obtained in which the plasma self-organizes towards a state with a clear radial pattern of the radial electric field, in line with expectations from some numerical studies describing the spontaneous formation of an "E x B staircase', consisting of alternating layers with fast and slow radial transport. In this state, the radial profiles of various quantities (density, temperature, pressure) will not be smooth.

Enhanced confinement is observed in neutral beam injector (NBI)-heated hydrogen discharges made in the stellarator TJ-II after the injection of a single cryogenic fuel pellet into the plasma core. In addition to the expected increase in electron density, ne, in the core after pellet injection (PI), the plasma diamagnetic energy content is seen to rise, with respect to similar discharges without PI, by up to 40%. Furthermore, the energy confinement time, τEdiag, as determined using a diamagnetic loop, is enhanced when compared to predictions obtained using the International Stellarator Scaling law [H. Yamada et al., Nucl. Fusion 45, 1684 (2005)] and the triple product, ne · Ti · τEdiag, exhibits a clear bifurcation point toward an improved confinement branch as compared to the branch product predicted by this scaling law. In general, once such a pellet-induced enhanced confinement (PiEC) phase has been established, it is characterized by steepened radial density gradients, by more negative plasma potential in the core, more negative radial electric fields, Er, across a broad plasma region, as well as by reductions in density and plasma potential fluctuations in the density gradient region. In addition, experimental observations show increased peaking of core radiation losses, this pointing to edge/core plasma decoupling. In parallel, neoclassical simulations of reference and PiEC plasmas predict increased particle and energy confinement times during a PiEC phase together with a more negative Er profile. Qualitative rather than quantitative agreement with experimental parameters is found, indicating that turbulence seems to play a significant role here. In summary, single cryogenic pellet injection facilitates the achievement of an enhanced operational regime that was previously not observed in NBI-heated discharges of the TJ-II.

An improvement in confinement associated with the injection of pellets has been observed in TJ-II and other devices. By means of a simple model, we study here the modification of turbulent transport by the injection of pellets in TJ-II and how this modification affects the particle confinement time. The analysis of the results shows the relationship of the confinement improvement with the evolution of the shear flow due to turbulence, especially near low order rational surfaces. This work is based on pellet injections into the neutral beam injection heated phase of TJ-II plasmas. For this, single, double, or triple hydrogen pellet injection are considered and the post-injection evolution of selected plasma parameters are analyzed and compared.

Understanding turbulence in magnetically confined fusion plasmas is an important issue on the long road to fusion energy as a power source for humanity. Fusion plasmas are extreme, strongly driven systems, far from thermodynamic equilibrium, in which turbulence plays a major role. This book reviews some of the observations that reveal the complex nature of plasma turbulence in detail and explains their implications. The change in paradigm needed to understand the observations is then discussed and mathematical models are highlighted that might address the dynamics better than traditional models. Along the way, a range of advanced analysis tools needed to characterize turbulence in this complex regime are presented and illustrated, including non-linear analyses, causality, and intermittence. The book aims to stimulate their broader adoption for the analysis of data from both experiments and numerical turbulence simulations. Part of IOP Series in Plasma Physics. Key features • Reviews observations that reveal the complex nature of turbulence in fusion plasmas • Provides advanced data analysis techniques to characterize complex turbulence • Includes non-linear spectral analyses, causality, intermittence • Covers self-organization, long-range correlations, long-time memory, models for effective transport • Includes spontaneous structure formation from turbulence such as zonal flows, MHD modes

We present recent highlights from the most recent operation phases of Wendelstein 7-X, the most advanced stellarator in the world. Stable detachment with good particle exhaust, low impurity content, and energy confinement times exceeding 100 ms, have been maintained for tens of seconds. Pellet fueling allows for plasma phases with reduced ion-temperature-gradient turbulence, and during such phases, the overall confinement is so good (energy confinement times often exceeding 200 ms) that the attained density and temperature profiles would not have been possible in less optimized devices, since they would have had neoclassical transport losses exceeding the heating applied in W7-X. This provides proof that the reduction of neoclassical transport through magnetic field optimization is successful. W7-X plasmas generally show good impurity screening and high plasma purity, but there is evidence of longer impurity confinement times during turbulence-suppressed phases.

TJ-II stellarator results on modelling and validation of plasma flow asymmetries due to on-surface potential variations, plasma fuelling physics, Alfvén eigenmodes (AEs) control and stability, the interplay between turbulence and neoclassical (NC) mechanisms and liquid metals are reported. Regarding the validation of the neoclassically predicted potential asymmetries, its impact on the radial electric field along the flux surface has been successfully validated against Doppler reflectometry measurements. Research on the physics and modelling of plasma core fuelling with pellets and tracer encapsulated solid pellet injection has shown that, although post-injection particle radial redistributions can be understood qualitatively from NC mechanisms, turbulence and fluctuations are strongly affected during the ablation process. Advanced analysis tools based on transfer entropy have shown that radial electric fields do not only affect the radial turbulence correlation length but are also capable of reducing the propagation of turbulence from the edge into the scrape-off layer. Direct experimental observation of long range correlated structures show that zonal flow structures are ubiquitous in the whole plasma cross-section in the TJ-II stellarator. Alfvénic activity control strategies using ECRH and ECCD as well as the relation between zonal structures and AEs are reported. Finally, the behaviour of liquid metals exposed to hot and cold plasmas in a capillary porous system container was investigated.

This work explores the impact of an imposed radial electric field on the intermittence parameter in magnetically confined plasmas. The intermittence is sensitive to both the magnetic configuration (dominant helical modes or low order rational surfaces) and to poloidal flows or radial electric fields. This behaviour was verified both in numerical turbulence calculations using a resistive magnetohydrodynamic model, and using Langmuir probe data obtained in experiments at the TJ-II stellarator. It is shown that the intermittence parameter can be used to detect when the local plasma rotation velocity, with respect to the laboratory frame of reference, is minimum.

In this article, we report the latest results for the impacts of the isotope mass and the plasma heating scenario on the spatiotemporal characteristics of the low-frequency zonal flow (ZF)-like electrostatic oscillation found in the plasma edge of the TJ-II stellarator. The radial profile of the toroidal cross-coherence in low-frequency floating potential fluctuations (1 < f < 25 kHz), known as long-range correlations (LRCs), was explored along the plasma boundary region by a dual Langmuir probe system. The study was performed using electron cyclotron resonance heating (ECRH)- and neutral beam injection (NBI)-heated plasmas with two different isotope concentrations: pure hydrogen plasmas and plasmas dominated by deuterium ( ≈ 70 % D ). The highest toroidal coherence levels are detected for floating potential oscillations at less than 10 kHz in hydrogen and deuterium plasmas, in both the ECRH and NBI heating schemes. The radial extent of the LRC profile is clearly larger in the ECRH than in the NBI-heated plasma scenarios. While the maximum amplitude of the LRC does not show a significant dependence on isotope mass, the relation between the LRCs’ radial sizes in plasmas and different isotope masses increases with the isotope mass in NBI-heated scenarios. These results provide the first experimental observation of ZFs with a larger radial size in deuterium than in hydrogen plasmas.

In this work, we study spontaneous electron to ion root transitions in TJ-II using Langmuir probes. By scanning the probe position on a shot to shot basis, we reconstruct a spatiotemporal map of the evolution of important turbulent quantities in the plasma edge region. We pay particular attention to the evolution of the cross phase between transport-relevant variables, showing the spatiotemporal evolution of this quantity for the first time, revealing the outward propagation of the changes associated with the transition. We also compute the intermittence parameter, which allows us to conclude that the turbulence, although its amplitude increases, condenses in a reduced number of dominant modes and becomes less bursty. The causal relationship between variables is studied using the transfer entropy, clarifying the interactions between the main variables and offering a rather complete picture of the complex evolution of the plasma across the confinement transition.

This work estimates the degree of turbulent intermittence of the plasma potential measured by a heavy ion beam probe in the core plasma region of the TJ-II stellarator. It is shown that the intermittence varies in a significant way with the plasma state (ion or electron root). In addition, radial minima of the intermittence are found to be associated with the location of topological structures of the flow associated with some important low-order rational surfaces. The local pressure gradient was also estimated, and a clear correlation was found between the steepening of the pressure gradient and the deepening of the minima of the intermittence, suggesting that the minima are associated with pressure gradient driven modes. By estimating the rotation velocity of the plasma from the measured plasma potential, it was possible to make a rough reconstruction of the two-dimensional radial–poloidal map of intermittence, thus clarifying the topological structure of the intermittence. The experimental results were put into context by comparing with simulations performed using a resistive magneto-hydrodynamic turbulence model.