Deep learning for Lagrangian drift simulation at the sea surface - Archive ouverte HAL Access content directly
Preprints, Working Papers, ... (Preprint) Year :

Deep learning for Lagrangian drift simulation at the sea surface

(1, 2, 3, 4) , (1, 2) , (3) , (4, 1) , (1, 2)
1
2
3
4

Abstract

We address Lagrangian drift simulation in geophysical dynamics and explore deep learning approaches to overcome known limitations of state-of-the-art model-based and Markovian approaches in terms of computational complexity and error propagation. We introduce a novel architecture, referred to as DriftNet, inspired from the Eulerian Fokker-Planck representation of Lagrangian dynamics. Numerical experiments for Lagrangian drift simulation at the sea surface demonstrates the relevance of DriftNet w.r.t. state-of-the-art schemes. Benefiting from the fully-convolutional nature of Drift-Net, we explore through a neural inversion how to diagnose modelderived velocities w.r.t. real drifter trajectories.
Fichier principal
Vignette du fichier
DriftNet2022.pdf (954.88 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03852489 , version 1 (16-11-2022)

Identifiers

Cite

Daria Botvynko, Carlos Granero-Belinchon, Simon Van Gennip, Abdesslam Benzinou, Ronan Fablet. Deep learning for Lagrangian drift simulation at the sea surface. 2022. ⟨hal-03852489⟩
0 View
0 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More