Documentation for a OceanParcels setup with POP model velocities for determining the origin of Atlantic AAIW

Content

• advection_routines: folder that allows for advection and basic analysis of particles in a repeated year framework

  • advection_execution.py: file includes the main advection routine full_execution, which allows for advection for a given time, merge_zarr allows for re-chunking and concatenation of output over multiple iterations.
  • displacement.py: file includes an optional setup for a displacement of particles that get stuck along the boundaries due to the Arakawa B-grid configuration of the POP model. The displacement kernel is a 3d interpretation of the displacement routine found in the OceanParcels Documentation
  • parcels_routines.py: file includes the core parts of the advection routine, following the OceaParcels syntax of fieldset, particle set and kernels. Further, it includes a start set for particles based on density (find_density), and a routine to remove stuck particles from advection (delete_stuck_particles).
  • post_process.py: file includes two classes that allow for creation and plotting of metadata from a dataset produced by advection_routines.merge_zarr. PlotableClass allows for plotting of ventilation position and transit time distributions, AnalyseTrajectory produces metadata, such as start and ventilation position for each trajectory, that are easier to handle than the whole dataset.

• vizualisation+results: folder to empathize results and general mechanisms of the particle advection

General Information

• Velocity fields are not included in this documentation, the setup still relies on precise (hard coded) filenames etc.
• The execution is performed with a config file, restarting the full_execution after 2 years of advection, due to limited computation time.

Development

• Upcoming uploads
  • config file and bash scripts for the execution
  • advanced analysis of output dataset
• extended soft coding
• run time saving adjustments in advection_routines/parcels_routines.delete_stuck_particles