A drift wave based transport model is used to self-consistently predict the time evolution of temperature and density profiles in JET H mode tokamak discharges. It is found that the same theoretically derived gyro-Bohm transport model previously used to simulate systematic scans of L mode discharges is equally successful in modelling JET ELMy H mode plasmas, implying that core transport is not intrinsically different from L mode confinement. The only difference between the L mode and H mode simulations results from the boundary conditions (i.e. density and temperature pedestals), which are taken from experimental data in both cases. Here, standardized experimental data from 16 JET H mode discharges in the ITER Profile Database are used, including dimensionless parameter scans in relative gyro-radius ρ*, collisionality ν and plasma β. Imperfections in dimensionless similarity for three pairs of scans in relative gyro-radius cause a purely gyro-Bohm transport model to exhibit worse than gyro-Bohm confinement. For the β scan, the model indicates a somewhat stronger β dependence than that observed with a thermal energy confinement scaling of Bτ ∝ β-0.7. More than half of the predicted β scaling is found to result from finite-β effects in the model. The model demonstrates a collisionality scaling of Bτ ∝ ν*-0.3 with some unfavourable dependence arising from neoclassical transport in the plasma core region. The overall goodness of fit obtained when comparing the global and local predictions results in a root mean square error averaging less than 13% for the total stored energy and averaging less than 9% for the density and temperature profiles relative to the maximum experimental values. When E × B shear effects are added to the model, the resulting change in the mean RMS error for the temperature profiles is less than 4%.
Predictive simulations of JET ELMy H-mode beta and collisionality scaling experiments