Trainings

Dear participant, this page proposes a selection of training material that you might want to browse before the OP’24 symposium.

As an introduction to ocean prediction science you might want to browse the OceanPredict website and the ForeSea website. The ETOOFs guide, and the resources from 2017 International School “New Frontiers in Operational Oceanography” provide precious overviews of today’s operational oceanography.

In preparation of OP’24, we have selected below self-training material for the following two priority themes “data assimilation for ocean forecasting and reanalyses” and “artificial intelligence methods for operational oceanography”.

In a third part, you will also find general information delivered by the Copernicus Marine Service for the download and use of their products.

A one-hour “meet with experts” session, with live exchanges will take place during the symposium for each one of these three topics.

Meet with experts: Nov 20, 2024 - 12:45 pm - Room IX

Topic 1: Data assimilation for ocean forecasting and reanalyses

Material selected by the Data Assimilation task team (DA-TT) - OceanPredict

For those interested in delving into data assimilation for ocean forecasting, the following reading material may serve as a helpful starting point for expanding your understanding of the field:

This review (preprint) provides an overview of the DA methodologies used in operational oceanography and presents a list of data assimilation software used in many global and regional ocean forecasting systems.

This paper summarizes efforts on evaluation of ocean observing systems in ocean prediction systems.

This recent review article focuses on the impact of ocean observing systems in regional and coastal ocean prediction systems.

These articles are chosen as they provide a broad overview of the current state of data assimilation in ocean prediction systems.

Several of the open source data assimilation softwares provide hands-on tutorials. In this page we would like to highlight the Parallel Data Assimilation Framework (PDAF), the Ensemble and Assimilation Tool (EAT), and the Regional Ocean Modeling System 4-Dimensional Variational DA (ROMS 4D-Var), as experts with experience with these systems will be available to answer questions during the one-hour “meet with experts” at the OP24. Additionally, we would like to point attendees to the JCSDA SOCA package for coupled ocean-ice data assimilation package from the Joint Effort for Data Assimilation Integration (JEDAI)

The Parallel Data Assimilation Framework (PDAF) is a software environment for data assimilation. PDAF simplifies the implementation of the data assimilation system with existing numerical models. With this, users can implement a data assimilation system for their model of choice with less work and hence focus on applying data assimilation. A list of numerical models already linked with PDAF is available on the official website.

During the EGU General Assembly a short course called “Getting Started with Data Assimilation: Theory and Application” was organized, aimed at early career scientists and others who are new to data assimilation. The materials from the course includes slides on DA theory and basic concepts, examples of applications within geosciences, and a hands-on exercise. Complementary material from a similar course at the EGU 2019 is also available here.

The Ensemble and Assimilation Tool (EAT) is a 1D test bed for physical-biogeochemical data assimilation in natural waters. EAT builds on established open-source components for hydrodynamics (GOTM), biogeochemistry (FABM), and PDAF for data assimilation. EAT is described in a peer-reviewed article, and in the EAT GitHub Repository you’ll find instructions on how apply EAT for a provided example, as well as instructions for how to set up your own experiments with the EAT framework.

The Regional Ocean Modeling System (ROMS) provides tools to perform four-dimensional variational assimilation (4D-Var), as well as observation impact and sensitivity analyses. At the ROMS website you can find a tutorial covering the basics of 4D-Var and the application in ROMS.

All files and scripts necessary to run the DA tests are available for download from the ROMS GitHub repository.To run the experiments described in the tutorial, it is necessary to compile ROMS on your computer. The application used in the tutorials is, however, lightweight and can be run on a standard laptop PC.

Meet with experts: Nov 19, 2024 - 12:45 pm - Room IX

Topic 2: Artificial intelligence methods for operational oceanography

Material selected by Daria Botvynko et al. (IMT Brest Chaire ANR AI OCEANIX : Physics-Informed AI for Observation-driven Ocean AnalytiX | IMT Atlantique)

For those who are interested in learning more about artificial intelligence methods and how they apply to operational oceanography and ocean forecasting science, we have selected online courses providing a general introduction to artificial intelligence and neural networks, tutorials to get started with pytorch, and examples of applications to geosciences as experts with experience with these applications will be available to answer questions during the one-hour “meet with experts” at the OP24. Additional resources and links to jupyter notebook material are also provided.

Average duration: ~6 hours per course / ~15 minutes per video