Comparison of Turbulence and Transport Characteristics between Global Flux-driven and Gradient-driven Gyrofluid Simulations in Tokamak Plasmas

We perform global nonlinear simulations of ion temperature gradient (ITG) turbulence in a concentric circular tokamak geometry with the gyrofluid code \textsc{GF2-BOUT++} and compare the dynamics in flux-driven and gradient-driven global simulations. The flux-driven simulations exhibit extended radial profiles and longer correlation lengths of the turbulent ion heat flux ($Q_i(r,t)$), whereas the gradient-driven simulations display characterstics of local transport process. Both approaches present broad $1/f$-type power spectra for $Q_i$, indicative of self-organized criticality (SOC)-like transport avalanches. The probability distribution functions of the heat fluxes $Q_i$ in the flux driven simulations are more non-Gaussian with enhanced skewness and kurtosis at lower heating power. Despite the similar levels of the total ion heat fluxes, the two approaches yield markedly different normalized ITG length ($R_0/L_{T_i}$), which we attribute to the differences in low-frequency $\bm{E}\times\bm{B}$ flow shear. Strong and persistent $\bm{E}\times\bm{B}$ shear in the gradient-driven simulations sustains the same heat flux at higher $R_0/L_{T_i}$ than in the flux driven ones, whereas weaker shear in flux-driven simulations permits stiff profiles with the critical normalized gradient $\sim 6$, similar to full-$f$ gyrokinetic results.