Description
Integrated modeling provides a platform on which the predictive capabilities of current models can be systematically verified and validated across devices, scenarios, and regimes [1]. This contribution presents the scientific program and the first results of the TSVV-H project, focused on coordinating integrated modeling efforts to improve reliability for scenario development and performance prediction. The work is based on multi-machine studies using the High Fidelity Pulse Simulator (HFPS) [2] and the IMAS data framework as the primary integrated modeling platform. At the same time, TSVV-H aims to act as a forum for European integrated modeling efforts, through benchmark activities with other integrated modeling frameworks such as ETS [3] and ASTRA [4]. Initial activities focus on establishing and extending the integrated modeling infrastructure and validation frame-work. This includes the consolidation of HFPS, the coupling and verification of reduced surrogate models, and integration of external workflows. Fast and accurate surrogates, such as FACIT, TGLFNN and SOLPSNN, can be used to greatly accelerate the work and can be verified against higher fidelity models thanks to the modularity of the workflows.
External, IMAS-compatible workflows can be coupled to enhance the modeling capabilities, for example by including MHD activity with EQSTABIL and fast particle transport with ATEP [5]. Early validation and benchmarking efforts are carried out on reference scenarios from JET, AUG, WEST, TCV, and ITER, targeting steady-state and time-dependent plasmas, including high-beta regimes, impurity transport, and radiative transients. In parallel, automated workflows are being developed to enable large-scale, multi-machine validation using standardized IMAS data and synthetic diagnostics.