[GSoC Project Proposal]: OCSMesh Parallelization

[AtiehAlipour-NOAA]

Project Description

Numerical meshes are the backbone of ocean models and mesh generation is often a laborious manual task. OCSMesh is a Python package that automates the generation of unstructured continuous (creek-to-ocean) meshes using the jigsaw-python library and triangles. The goal of this project is to increase OCSMesh’s efficiency by parallelizing the most time consuming steps, including DEM processing and jigsaw mesh generation. We expect the following main tasks:

• Pickling: The current version of OCSMesh is not pickle-able. So, some kind of Serialization or Memory sharing is needed.
• Refactoring: Changes in the package architecture will be needed to accommodate parallelization.
• Parallelization: Should be focused on HPC environments.

OCSMesh is an open source tool for preparation of unstructured mesh for coastal ocean model

• NOAA report: https://repository.library.noaa.gov/view/noaa/33879
• Repository: https://github.com/noaa-ocs-modeling/OCSmesh
• Documentation: https://noaa-ocs-modeling.github.io/OCSMesh/

Getting started with OCSMesh:

• https://github.com/SorooshMani-NOAA/OCSMeshTutorial/blob/main/UFS_Nov_2022/ocsmesh_basics.ipynb
• https://github.com/SorooshMani-NOAA/OCSMeshTutorial/blob/main/Tutorials_2023/golfo_san_jose.ipynb
• https://github.com/SorooshMani-NOAA/OCSMeshTutorial/blob/main/Tutorials_2023/Subsetting_Ian.ipynb (requires data not shared)

Expected Outcomes

A new version of OCSMesh that fully takes advantage of HPC resources and better fulfills the needs of the ocean modeling community. The faster mesh generation from this improvement will allow the automated creation of global meshes for ocean prediction (ocean temperature, salinity, currents, etc.) and efficient deployment of high resolution meshes during storm surge events

Skills Required

Python; Libraries: Xarray, Dask, MPI4Py, GIS libraries and data.

Additional Background/Issues

No response

Mentor(s)

Felicio Cassalho (@felicio93), Soroosh Mani (@SorooshMani-NOAA)

Expected Project Size

350 hours

Project Difficulty

Intermediate