BC3 Webinar: Artificial Intelligence applied to ocean waves forecast: A deep learning approach to predict significant wave height.

“Artificial Intelligence applied to ocean waves forecast: A deep learning approach to predict significant wave height.“

Post.Researcher Felipe Minuzzi

Department of Pure and Applied Mathematics
Federal University of Rio Grande do Sul (UFRGS), Brazil

Abstract:

Better predictions of ocean waves conditions are essential for a handful of industries that define its operations based on the outcome of these analysis. Almost all engineering applications that happens in the ocean, from transportation to renewable energies, going across offshore platforms and alerts of catastrophic events, not to mention geosciences research benefits from an accurate description of sea state. We present a framework using long short term memory (LSTM) as a deep learning technique to forecast significant ocean wave heights, based on the ERA5 database available through Copernicus Climate Data Store (CDS) implemented by ECMWF (European Center for Medium Range Forecast). The predictions are made for different locations in the Brazilian coast, ranging both shallow and deep waters. Experiments are conducted using only the historical series, and the influence of other variables as inputs for training is investigated. The results shows that a data-driven methodology can be used as a surrogate to the computational expensive physical models, and achieve accuracies that are near 95% in the best cases.

About the Lecturer

Felipe Minuzzi is an Applied Mathematician that currently holds a position as Postdoctoral Researcher at the Federal University of Rio Grande do Sul (Brazil), where his research is focused on applying artificial intelligence algorithms in physical oceanography. Prior to that, he finished his PhD at the same University, working in applications of fluid mechanics, with an exchange period at the Karlsruhe Institute of Technology, Germany.