{"id":"arxiv/2604.23874","url":"https://arxiv.org/abs/2604.23874","tags":["physics.flu-dyn","cs.LG","math.DS","physics.comp-ph","physics.geo-ph"],"score":0,"title":"Deep Learning of Solver-Aware Turbulence Closures from Nudged LES Dynamics","domain":"physics.flu-dyn","source":"arxiv","authors":["Ashwin Suriyanarayanan","Dibyajyoti Chakraborty","Romit Maulik"],"reviews":0,"subject":"physics.flu-dyn","summary":"The differentiable physics paradigm may be leveraged as an a-posteriori approach for discovering turbulence closure models by embedding a neural network parameterization directly inside the solver and optimizing it given potentially sparse target data. This addresses a key limitation of a-priori learning where direct numerical simulation (DNS) data is used to approximate the subgrid stress with the assumption of a low-pass filter. Closures trained in this a-priori manner frequently lead to unstable deployments due to the mismatch between the assumed filter and the effect of numerical…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-04-26","reactions":{}}