Skip to content

Latest commit

 

History

History
26 lines (19 loc) · 1.09 KB

File metadata and controls

26 lines (19 loc) · 1.09 KB

Title

TurbidityMapping

Short description of the project

Create a pipeline for performing turbidity mapping using a pre-trained 4dvarnet model over Wadden Sea region which contains Dutch Wadden Sea and German Wadden Sea.

Expected benefit of the project

Create a service for turbidity Mapping and analysing evaluation metrics and visualiations. This inference allows us to benchmark the model agaist a classical method called Dineof with the ML based 4dvarnet.

Datasets

  • Dataset is provided by CMEMS services, it is also uploaded on EDITO cloud
  • A pretrained weight is uploaded on EDITO cloud as well

Codebase

Description of the software packages that will be used (with links to github repos): https://github.com/nguyenthuynga/4dvarnet-starter

Compute resources

4DVarNet model training and inference needs heavy computation, and since the current GPU resources limit our inference to a month of data instead of the whole year at one. We will require H100 GPU for performing training/inference in the future.

People involved

List of people involved: Fabio, Minh, Nga, Shashank