Validation of geometric proxies using negative and positive triangularity plasmas with varying elongation on TCV and multi-machine gyrokinetic simulations
Negative Triangularity (NT) tokamak plasmas can achieve H-mode-like confinement while avoiding Edge Localized Modes, making them an attractive reactor scenario. Predicting in which configuration this benefit occurs requires understanding how magnetic geometry controls turbulent transport. Using a set of geometric proxies based on the poloidally averaged magnetic drift velocity, its poloidal variation, and Finite Larmor Radius (FLR) effects, this work validates their predictive power against gyrokinetic simulations and experiments. First, the proxies are used to predict how the relative benefit of NT over Positive Triangularity (PT) evolves with elongation $κ$, forecasting a detrimental effect of NT below $κ\sim0.75$--$0.8$. This prediction is confirmed by linear and nonlinear GENE simulations in both ITG- and TEM-dominated regimes. Additionally, a dedicated experimental elongation scan on TCV shows that NT plasmas at $κ^{\rm LCFS}=1.5$ have $β_N$ and $H_{98y2}$ roughly 1.5 times larger than PT counterparts, while no difference is observed at $κ^{\rm LCFS}=0.95$. The analysis is then extended to realistic ITG-dominated scenarios from DIII-D, AUG, and JET, including up-down asymmetric equilibria where conventional triangularity-based descriptions break down. In all three devices, simulations of NT equilibria exhibit substantially reduced nonlinear heat fluxes relative to PT (factors of 1.8, 1.6, and 2.1, respectively). A systematic swap of geometric coefficients between NT and PT equilibria shows that magnetic drifts and FLR effects are the dominant stabilizing mechanisms across all machines, with parallel dynamics playing a secondary role. These results establish the geometric proxies as reliable, computationally inexpensive predictors of NT transport benefits across diverse shapes and turbulence regimes.