Speaker
Description
We investigate turbulent fluctuations observed around low-to-high confinement (L-H) and high-to-low confinement (H-L) transitions and the effects of externally applied edge-localised resonant magnetic perturbations (ERMPs) in KSTAR. Beam Emission Spectrometry (BES), Electron Cyclotron Emission Imaging (ECEI/ECE) and Mirnov Coils (MC) enable analysis of time-series fluctuations of electron density and temperature ($n_e$ and $T_e$) and the time-derivative of the poloidal magnetic field ($dB_{\theta}/{dt}$) respectively. Time-dependent probability density functions (PDFs) of fluctuations are approximated by binning data in 2 ms windows centred every 1 ms. These time-series PDFs are analysed using the information-geometric measures information rate $\Gamma$ and length ($\mathcal{L} = \int{\Gamma}dt$) alongside the conventional kurtosis $\kappa$ and variance $\sigma^2$. The information rate effectively captures the effect of strong transients such as ELMs and neutral beam blips that abruptly perturb the time-series PDFs away from equilibrium. Across multiple plasma densities ($n_e = 1.5 - 3.57 \times 10^{19} m^{-3}$) we find $\mathcal{L}$ indicates that L-H transitions strengthen the radial correlation of $n_e$ and $T_e$ fluctuations in the edge region. For the highest density $3.57 \times 10^{19}m^{-3}$, the information lengths of $T_e$ and $n_e$ have similar radial profiles, implying strong correlation between the two. However, at a slightly lower density $3.2 \times 10^{19}m^{-3}$, the behaviour differs significantly, associated with the presence of a strong $n=2$ coherent mode. This highlights the importance of hidden variables determining turbulence characteristics and reorganisation across the L-H transition. Typically, the highest fractional change in $\mathcal{L}$ across transitions is near the pedestal shoulder. Additionally, we observe the L-H transition leads to a transient reduction of $\sigma^2$, while the H-L transition produces an increase. ERMPs induce greater change in information length of $T_e$ and $n_e$ fluctuations at any fixed radial location when applied in the L-mode than H-mode, suggesting a stronger effect of ERMPs on the L-mode. In summary, information-geometric quantities can capture rapid events in non-equilibrium plasma evolution, offer insight into their radial correlation and provide a complementary diagnostic that tracks statistical reorganization, not just amplitude.