Features

Main Features

OceanTracker is primarily designed for coastal environments which is reflected in the supported models and implemented physics.

• We support a large set of hydrodynamical models, both with structured (i.e. rectangular grid) and unstructured (triangular or other multi-nodal cells) grid types. Many common models are already supported out of-the-box. If yours isn’t, feel free to reach out. We are happy to implement new ones. The current list includes SCHISM, FVCOM, ROMS, NEMO and GLORYS, Delft3D, and Copernicus i.e. CMEMS-catalog. For more details see supported hydrodynamical models.

• To enable a large set of research questions we support the following particle behaviors and physical processes: dispersion, settling and re-suspension from bottom based on critical shear velocities, tidal-stranding, varying buoyancies, particles splitting and culling. Users can also easily add their own ones to the computational pipeline.

• We are using the model ourselves, and spend quite some time on making its application efficient. This is done primarily in two ways: by reducing user effort to run and analyze the model and by making the model computationally fast to run.

  • We try to keep the time low that a user has to faff with the model configuration through a modular configuration structure. For large runs, output data management can become the bottleneck. We offer on-the-fly particle statistics to avoid writing and reading large trajectory files to disk. Supported hydrodynamical models are auto-detected and mapped accordingly and particle-tracking output can be quickly plotted with the included plotting routines (see [placeholder] link to the how tos).

  • Computational efficiency was the main focus since starting out as existing models weren’t fast enough for problems we tried to tackle. We believe we are currently the fastest model available for unstructured grids (Vennell et al., (2025)). This was done through plenty of code optimization (e.g. BC-walk), just-in-time compilation using numba, and built-in parallelization via threading.

For more detailed descriptions about individual features see section below:

• Readers of hydrodynamical model data
• Vertical grid types and support
  • Description of the vertical grid types supported in oceantracker
  • Overview table
• Supported hydrodynamical models
• Computational speed
• Computational pipeline
  • Building a Computational Pipeline
  • Computational Steps
  • Mechanisms Enabling Computational Pipeline Adaptability
• Data Structures
  • Field Role
  • Particle Property Role
• Manager Roles
  • Fields Group Manager Role
  • Particle Group Manager Role
• Reader Role
  • Interpolator Role
• Particle Release Groups Role
• Velocity Modifiers Role
• Trajectory Modifiers Role
• On-the-fly Particle Statistics Roles
• Integrated Models
• Integration methods
• Interpolation
• Particle properties
• Trajectory modifiers
• On-the-fly particle statistics
  • Overview of types of particle statistics
  • Formal descriptions
• Model restart and continue feature
  • Model restart, i.e. ‘hotstart’
  • Continue the simulation
  • Preliminary results