Estimating cross-field particle transport at the outer midplane of TCV by tracking filaments with machine learning

Abstract Cross-field transport of particles in the boundary region magnetically confined fusion plasmas is dominated by turbulence. Blobs, intermittent turbulent structures with large amplitude and a filamentary shape appearing scrape-off layer (SOL), are known from theoretical experimental studies to be main contributor cross-field particle transport. The dynamics blobs differs depending on various plasma conditions, including triangularity ( δ ). In this work, we analyze dependence at outer midplane δ = + 0.38 , +0.15, −0.14, −0.26 Tokamak à Configuration Variable, using our novel machine learning (ML) blob-tracking approach applied gas puff imaging data. flux determined way same order as overall inferred KN1D, GBS, SOLPS-ITER simulations, suggesting that identified ML account for most SOL. Also, KN1D show decrease becomes more negative. blob-by-blob analysis result tracking reveals decreasing accompanied number fixed time, which tend have larger area lower radial speed. these connected sheath regime, velocity scaling consistent two-region model.