Data-Derived Operational Boundaries of RMP ELM Suppression in ASDEX Upgrade and DIII-D

Linear discriminant analysis (LDA) is performed on datasets of discharges from the ASDEX Upgrade and DIII-D tokamaks, revealing unexpected decision boundaries in elongation (κ) and internal inductance for RMP ELM suppression access. Suppression of ELMs by application of RMP fields has been demonstrated in many tokamaks, however, the access criteria are not fully understood. In this work, LDA enables decision boundaries for determining a discharge’s classification as ELMy or ELM-suppressed to be derived for each device and compared to known experimental threshold conditions. In particular, a clear decision boundary between κ and lower triangularity (δ_l) is found, which indicates suppression loss at high κ when at elevated δ_l. This is consistent with observations from ASDEX Upgrade. Furthermore, including discharges with different heating mixes blurs the boundary between ELMy and ELM-suppressed phases, indicating access criteria strongly related to heating power. Finally, consistency of the decision boundaries between devices is also explored. This work provides a further understanding of the parameter space required for ELMs to be suppressed by RMPs and can inform real-time ELM control schemes.