In this section
@misc{Lefauve:2025:10.5194/egusphere-egu25-18513,
author = {Lefauve, A and Zhu, L and Jiang, X and Kerswell, R and Linden, P},
doi = {10.5194/egusphere-egu25-18513},
title = {New insights into experimental stratified flows obtained through a physics-informed neural network},
type = {Other},
url = {http://dx.doi.org/10.5194/egusphere-egu25-18513},
year = {2025}
}
TY - GEN
AB - <jats:p>We develop a physics-informed neural network (PINN) to significantly augment state-of-the-art experimental data of stratified flows. A fully connected deep neural network is trained using experimental data in a salt-stratified inclined duct (SID) experiment. SID sustains a buoyancy-driven exchange flow for long time periods, much like an infinite gravity current. The data consist of time-resolved, three-component velocity fields and density fields measured simultaneously in three dimensions at Reynolds number= O(10^3) and at Prandtl or Schmidt number = 700 [1]. The PINN enforces incompressibility, the governing equations for momentum and buoyancy, and the boundary conditions at the duct walls. These physics-constrained, augmented data are output at an increased spatio-temporal resolution and demonstrate five key results: (i) the elimination of measurement noise; (ii) the correction of distortion caused by the scanning measurement technique; (iii) the identification of weak but dynamically important three-dimensional vortices of Holmboe waves; (iv) the revision of turbulent energy budgets and mixing efficiency; and (v) the prediction of the latent pressure field and its role in the observed asymmetric Holmboe wave dynamics. These results mark a significant step forward in furthering the reach of fluid mechanics experiments, especially in the context of stratified turbulence, where accurately computing three-dimensional gradients and resolving small scales remain enduring challenges.References[1] L. Zhu, X. Jiang, A. Lefauve, R. R. Kerswell, and P. F. Linden. New insights into experimentalstratified flows obtained through physics-informed neural networks. J. Fluid Mech., 981:R1, 2024.</jats:p>
AU - Lefauve,A
AU - Zhu,L
AU - Jiang,X
AU - Kerswell,R
AU - Linden,P
DO - 10.5194/egusphere-egu25-18513
PY - 2025///
TI - New insights into experimental stratified flows obtained through a physics-informed neural network
UR - http://dx.doi.org/10.5194/egusphere-egu25-18513
UR - https://doi.org/10.5194/egusphere-egu25-18513
ER -
Climate Science
Earth and Life Sciences
Energy and Low-Carbon Futures
Resources and Pollution
Economics and Finance
Health
View all publications and browse by year
Briefing papers and Briefing notes
Grantham Institute Outlooks
Evidence & submission papers
Infographics