MSMakoto Sasaki

Makoto Sasaki

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Makoto Sasaki has a total of 420 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Wolkowicz-Styan Upper Bound on the Hessian Eigenspectrum for Cross-Entropy Loss in Nonlinear Smooth Neural Networks

    by Yuto Omae, Kazuki Sakai, Yohei Kakimoto, Makoto Sasaki, Yusuke Sakai, Hirotaka Takahashi

  • Coordinate-invariant singular value decomposition for modal analysis of turbulence in magnetized plasmas

    by T. Kodahara, M. Sasaki, K. Yano, A. Okuno, H. Seto, Y. Kawachi, T. Kobayashi, F. Kin

    We propose a mode decomposition method based on singular value decomposition (SVD) that enables the extraction of orthogonal basis functions in multi-dimensional general curvilinear coordinate systems, applied to plasma turbulence simulation data. By including the Jacobian weighting, which accounts for the local volume element in curvilinear coordinates, the proposed method allows the mode decomposition to satisfy both the mathematical orthogonality of the SVD bases and the physical space orthogonality. Applying the method to turbulence data in cylindrical plasmas, it is demonstrated that identical bases can be extracted to those obtained by the conventional SVD with Cartesian coordinate transformation. Furthermore, in the analysis of energy interactions that require accurate evaluation of higher-order derivatives, the proposed method effectively reduces numerical noise compared with the conventional SVD. Applying the coordinate-invariant SVD to toroidal plasma turbulence, we further demonstrate that it preserves modal orthogonality even in non-orthogonal field-aligned coordinates. This method enables coordinate-invariant quantitative analysis of interactions among turbulent structures at lower spatial resolution, compared to standard SVD.

  • Open Access

    Rapid damping of magnetic turbulence induced by collisional merging process of field-reversed configurations

    by Daichi Kobayashi, Taichi Seki, Tsutomu Takahashi, Hinata Suda, Ann Sekine, Makoto Sasaki, Sean Dettrick, Dan Barnes, Tomohiko Asai

    The damping process of magnetic turbulence induced by the collisional merging of field-reversed configurations (FRCs) was investigated using the FAT-CM device at Nihon University. Two FRC-like plasmoids collided at super-Alfvénic and supersonic speeds. During the collision phase, intense magnetic turbulence developed and temporarily disrupted the FRC-like reversed magnetic field structure. Nevertheless, the plasmoids relaxed into an FRC-like state within several tens of microseconds. High-frequency turbulence components were rapidly damped, particularly under kinetic conditions, suggesting the possible involvement of kinetic effects in the damping process. Additionally, energy conversion from kinetic to magnetic and subsequently to thermal energy was observed. These results suggest that kinetic effects may play an important role in both the rapid damping of magnetic turbulence and the associated plasma heating during the collisional merging process.

  • Information propagation in predator-prey dynamics of turbulent plasma

    by Tomohiro Tanogami, Makoto Sasaki, Tatsuya Kobayashi

    Magnetically confined fusion plasmas exhibit predator-prey-like cyclic oscillations through the self-regulating interaction between drift-wave turbulence and zonal flow. To elucidate the detailed mechanism and causality underlying this phenomenon, we construct a simple stochastic predator-prey model that incorporates intrinsic fluctuations and analyze its statistical properties from an information-theoretic perspective. We first show that the model exhibits persistent fluctuating cyclic oscillations called quasicycles due to amplification of intrinsic noise. This result suggests the possibility that the previously observed periodic oscillations in a toroidal plasma are not limit cycles, but quasicycles, and that such quasicycles may be widely observed under various conditions. For this model, we further prove that information of the zonal flow is propagated to turbulence. This result suggests that turbulence behavior may be predictable to a certain extent based on zonal flow characteristics.

  • Open Access

    Bifurcation of drift wave turbulence trapping by zonal flows near density limit

    by Makoto Sasaki, Go Yatomi, Yohei Onuki, Tatsuya Kobayashi, Yuichi Kawachi, Fumiyoshi Kin, Kimitaka Itoh

    We investigate drift wave - zonal flow interactions near the density limit and demonstrate transitions of turbulence trapping states. Nonlinear simulations of the modified Hasegawa-Wakatani model reveal two robust types of turbulence energy localization: valley-trapping near zonal flow troughs at small adiabaticity (hydrodynamic regime) and hill-trapping near crests at large adiabaticity (adiabatic regime). Around the transition, turbulence exhibits bistability, dynamically switching between valley- and hill-trapping, and also shows indications of multi-stable behavior depending on the adiabaticity. The Lagrangian auto-correlation time of turbulence changes markedly across the bifurcation between the trapping states, indicating that turbulent vortices become long-lived in the adiabatic limit and short-lived in the hydrodynamic limit. Eigenmode analysis reproduces these patterns and shows that the selection is governed by the relative magnitude of the drift wave phase velocity and the zonal flow amplitude. These findings provide a simple selection rule for turbulence energy localization and offer new insight into drift wave - zonal flow dynamics in regimes relevant to the density limit.


  • Open Access

    Uncertainties in Physics-informed Inverse Problems: The Hidden Risk in Scientific AI

    by Yoh-Ichi Mototake, Makoto Sasaki

    Physics-informed machine learning (PIML) integrates partial differential equations (PDEs) into machine learning models to solve inverse problems, such as estimating coefficient functions (e.g., the Hamiltonian function) that characterize physical systems. This framework enables data-driven understanding and prediction of complex physical phenomena. While coefficient functions in PIML are typically estimated on the basis of predictive performance, physics as a discipline does not rely solely on prediction accuracy to evaluate models. For example, Kepler's heliocentric model was favored owing to small discrepancies in planetary motion, despite its similar predictive accuracy to the geocentric model. This highlights the inherent uncertainties in data-driven model inference and the scientific importance of selecting physically meaningful solutions. In this paper, we propose a framework to quantify and analyze such uncertainties in the estimation of coefficient functions in PIML. We apply our framework to reduced model of magnetohydrodynamics and our framework shows that there are uncertainties, and unique identification is possible with geometric constraints. Finally, we confirm that we can estimate the reduced model uniquely by incorporating these constraints.

  • Zero-dimensional modeling of drift wave turbulence by using Bayesian regression

    by M Sasaki, Y Mototake, T Kobayashi, F Kin, G Yatomi, Y Kawachi

    We present a Bayesian model selection framework to identify reduced dynamical equations describing the evolution of volume-integrated turbulence energy in magnetically confined plasmas. Using data from two-dimensional Hasegawa–Wakatani simulations with varying adiabatic parameters, we consider a candidate model with polynomial nonlinearities up to ninth order and a coupling term to zonal flow energy. All combinations of these terms are evaluated based on Bayesian model evidence. The optimal model includes only linear growth, quadratic nonlinear saturation, and zonal flow suppression. While subcritical turbulence was not observed in the present dataset, the selection framework successfully excluded redundant higher-order terms, confirming its robustness. Crucially, this approach enables data-driven identification of nonlinear saturation and turbulence-zonal flow coupling, which are often introduced heuristically in conventional models. These results suggest that the proposed framework can serve as a useful tool for constructing interpretable reduced models of plasma turbulence and for examining key mechanisms involved in nonlinear energy regulation.


  • Spatiotemporal dynamics of turbulent fluctuations in the presence of toroidal zonal flow in reversed field pinch plasma

    by Yuichi Kawachi, Takashi Nishizawa, Makoto Sasaki, Tatsuya Kobayashi, Abdulgader Almagri, Noah Hurst

    In this study, we investigate the spatiotemporal dynamics of fluctuations in the presence of a toroidally directed zonal flow using two pairs of radial capacitive arrays with finer spatial resolution than the observed toroidal zonal flow. Conditional sampling is employed to evaluate frequency- and time-resolved behavior with respect to the toroidal zonal flow phase. We identify three distinct types of fluctuations, each exhibiting different spatiotemporal dynamics synchronized with the toroidal zonal flow. These modes are modulated in space and time in accordance with the phase of the toroidal zonal flow. The mechanism underlying the formation of spatiotemporal structures is discussed in terms of long-range correlations and wave-kinetic characteristics

  • Open Access

    Investigation of low-frequency fluctuation modulating turbulence during L–H transition in Heliotron J

    by F. Kin, Y. Kondo, M. Sasaki, A. Miyashita, R. Matsutani, K. Ozawa, T. Kodahara, S. Kobayashi, T. Minami, S. Kado, et al.

    In this study, we observed a distinct low-frequency (LF) fluctuation in the n/m = 4/9 configuration under electron cyclotron heating (ECH) and neutral beam injection heating—conditions favorable for the L–H transition in Heliotron J [Sano et al., Nucl. Fusion 45, 1557 (2005)]. This LF mode, with a frequency of approximately 14 kHz, was detected during the L–H transition using Doppler reflectometry (DR), electron cyclotron emission (ECE), magnetic probes (MP), Dα emission, and edge divertor probes (DP). It modulates edge turbulence amplitude and extends radially, with toroidal and poloidal mode numbers of n/m = 1/3. It is found that the LF mode detected by ECE is correlated with the perturbation of the Dα emission, suggesting the mode contributes to the cross-field edge transport. At the L–H transition, while the LF mode persists, modulation of edge turbulence is suppressed, as evidenced by the reduced amplitudes in Dα emission and DP signals. These findings reinforce the importance of turbulence in the edge transport modulation induced by the LF mode.


  • Open Access

    Estimation of Electrostatic Potential Fluctuations in Kelvin-Helmholtz Turbulence in Linear Plasmas using Multi-Scale CNN

    by Shuta Hoshino, Makoto Sasaki, Ryohtaroh T. Ishikawa, Motoki Nakata

    We demonstrate the estimation of electrostatic potential fluctuations in dynamically varying Kelvin-Helmholtz turbulence using multi-scale convolutional neural network. The turbulence field is obtained from simulations based on a reduced fluid model in cylindrical magnetized plasmas. The target turbulence shows limit-cycle oscillations, and coherent and spiral structures are generated and annihilated repeatedly. High accuracy of the prediction is realized for the electrostatic potential field, and the estimation of the particle flux calculated from the predicted potential agrees with the answer with 98.4% accuracy. Behavior of the prediction accuracy is also discussed by changing the hyper parameters, such as the number of filters and the size of the training data.