8
$\begingroup$

I am looking for a few options of particle tracking for ocean modeling to try to do a comparison of different codes.

$\endgroup$
3
  • 1
    $\begingroup$ Are you interested in coming up with your own code or are you just talking about current available tools? $\endgroup$ May 6, 2014 at 17:43
  • 1
    $\begingroup$ I have several codes myself (both online and offline) that I want to be able to compare with other codes. $\endgroup$
    – arkaia
    May 6, 2014 at 17:51
  • 1
    $\begingroup$ Same here. I never used any external package or online tracking in models, but I implemented my own tracking because it was relatively quick and simple compared to learning to use a new tool. My guess is that the difference between tools will be mostly in how time integration and spatial interpolation is handled. $\endgroup$ May 6, 2014 at 19:39

2 Answers 2

12
$\begingroup$

The main difference between online and offline trajectory calculation is that the online is implemented as part of the model, and the trajectories are part of the model output. Offline trajectory calculation is a separate piece of software that takes Eulerian model output fields as input and calculates the trajectories.

The main advantage of online particle tracking is that the displacement of a particle or tracer can be evaluated every time step, no matter how small it is. As a result, you can obtain a true particle trajectory based on what model produces. The disadvantage of the online tracking is that the user needs to know the particle release location and time ahead of running the model. The models listed below (at least) support online Lagrangian trajectory calculation:

On the other side, offline particle tracking can only use the Eulerian fields as often as they are output. Thus, there will be an error in the calculated trajectory depending on the frequency of the output. On the other side, the main advantage of offline calculation is that the user can experiment and play around with deployment locations, times, area of coverage etc., because trajectory calculation typically takes much less computational effort than the actual model. There are several options here:

  • Connectivity Modeling System (CMS) takes gridded NetCDF files as input and produces Lagrangian trajectory output. Can handle fields from various Arakawa-staggered grids, and may produce probabilistic trajectories, diffusivity specified by user (optional). Compatible with at least HYCOM and ROMS output.

  • Ariane also takes gridded NetCDF files and is compativle with at least HYCOM, ROMS and NEMO models.

  • TRACMASS seems to be compatible with many circulation models. It also has a Python interface.

All models and tools listed in this answer are freely available for download under different open source licenses.

$\endgroup$
2
  • $\begingroup$ Pretty sure that MIKE21 / MIKE3 allows for online tracking, although it may not call it that. $\endgroup$ May 5, 2014 at 7:54
  • $\begingroup$ I did not know about CMS. Pretty cool. Thanks $\endgroup$
    – arkaia
    May 7, 2014 at 15:50
6
$\begingroup$

We are in the final stages of a new book on numerical particle tracking for the coastal ocean. Originally as part of that book and now separate from it, we have developed an evolving list of routines for particle tracking. It can be access at http://www.nefsc.noaa.gov/drifter/particles.html.

I don't mean to detract from the answer by @IRO-bot. I just wanted to complement his results.

Edits and extensions can be emailed to James Manning @NOAA

We provide a list of algorithms that is not meant to be complete. We separate the algorithms into two categories: 1) code run offline using available model results and 2) web-based particle tracking that does not require any execution by the user (point and click). Many of the web-based routines are now being shared and applied in multiple regions of the globe.

Example Offline Particle Tracking Packages

  • LTRANS: The Larval TRANSport Lagrangian model is an offline particle-tracking model that uses stored predictions of a 3D hydrodynamic model, specifically the Regional Ocean Modeling System (ROMS). Although LTRANS was built to simulate larvae, it has been adapted to simulate passive particles and other planktonic organisms. LTRANS is written in Fortran 90 and is designed to track the trajectories of particles in three dimensions using a 4th order Runge-Kutta scheme for advection and random displacement to account for vertical turbulence. Contact E. North (U. Maryland), [email protected], http://northweb.hpl.umces.edu/LTRANS.htm
  • Ariane: Ariane is an offline computational tool (Fortran 90/95) to calculate 3D streamlines from the output velocity field of a finite difference circulation model such as OPA-NEMO, ROMS, or Symphonie. Contact B. Blanke from IFREMER), [email protected], http://stockage.univ-brest.fr/~grima/Ariane/
  • LarvalMap: Is a web service build on a python library that provides a common data model and methods for particle tracking. It was originally developed as a 3D web-based larval fish and invertebrate transport model connected to ROMS circulation fields for the northeast Pacific. It can be used with any Data Access Protocol (DAP) accessible circulation dataset formatted using the network Common Data Format for Climate and Forecast (NetCDF CF). Contact: [email protected]. https://github.com/asascience-open/paegan-transport
  • PartTrack: PartTracker is a particle tracking research tool currently being developed in MatlabTM which is designed to be used with both unstructured and structured model grids and has previously been applied to outputs from SELFE (Semi -implicit Eulerian-Lagrangian Finite- Element model) for the purposes of assessing wastewater and larval dispersal. PartTracker applies the time adaptive composite 4th-5th order Runge-Kutta (RK45) numerical scheme to ensure required minimum standards of accuracy are achieved for individual particle paths. The application of the RK45 approach ensures consistency in particle path accuracy across hydrodynamically complex domains and a range of user defined time steps with the time adaptive aspect of the approach reducing computational overheads in hydrodynamically simple flows by allowing for long time steps. Knight BR, Zyngfogel R, Forrest B 2009. PartTracker – a fate analysis tool for marine particles. In Proceedings of Australasian Coasts and Ports Conference 2009. Dawe I, ed. New Zealand Coastal Society, Wellington. pp 8. Contact Ben Knight (Carthron Institute, New Zealand.
  • WebDrogue: Beta-release Java graphical user interface used to obtain drift predictions from user-specified points in the model domain and can be installed under MS-Windows and Linux. The drift trajectories are computed using circulation derived from the tides, the seasonal mean circulation, wind-driven circulation, and a surface-wind drift (Hannah, et al.2004) Contact: Jason Chaffey at [email protected] http://www.bio.gc.ca/science/research-recherche/ocean/webdrogue/index-eng.php
  • Two-Way PTM: In order to determine the source of particle arriving at one location, particle tracks can be run backwards to determine a probable source, then forward to determine the robustness of the results given a prescribed randomness. This is described in Isobe, et al, 2009, for example, where it is applied to the East China Sea. Contact [email protected]
  • ICTHYOP: Developed to study how physical factors (e.g., currents, water temperature) and biological factors (e.g., egg buoyancy, larva growth) affect the dynamics of fish eggs and larvae. It gathers physicians and biologists together around a common tool, easily available and user-friendly. Contact: [email protected] http://www.ichthyop.org
  • OILTRANS: "an off-line particle-transport model coupled to the most up to date operational met-ocean model forecasts. Formulations for the dominant oil fate processes of spreading, advection, diffusion, evaporation, emulsi- fication and dispersion have been encoded, providing the model with the ability to accurately predict the horizontal movement of surface oil slick, the vertical entrainment of oil into the water column and the mass balance of spilled oil." Contact: [email protected] See Berry, A., T. Dabrowski, and K. Lyons. 2012. The oil spill model OILTRANS and its application to the Celtic Sea. Marine Pollution Bulletin. 64 pp. 2489-2501. Some documentation reported here.
  • Sediment Transport: A community models for coastal sediment transport are under development. Some notes on this project are provided here where there is a list of existing models. Contact [email protected].

    Example Web-Based Particle Tracking Routines

  • SCCOOS tracks the Monteray Bay flows off California (http://ourocean.jpl.nasa.gov/SCB/scbmangen.jsp)
  • NCSTATE: tracks flow in large Northwest Atlantic ROMS/HYCOM grid Contact: Ruoying He http://omgsrv1.meas.ncsu.edu:8080/ocean-circulation-useast2/trajectory.jsp)
  • ALAVAI: Visualization tool for a surface ocean drift forecast. Environmental Modelling and Software. Otero et al., 2013. Contact: Pablo Otero (https://github.com/PabloOtero/Alavai)
  • Connie Tool: An “online tool for marine scientists and managers to investigate patterns of marine spatial connectivity. It has been applied in areas such as larval dispersion and recruitment studies, and the development of scenarios and risk assessments for contaminant dispersion.†Contact: [email protected] (http://www.csiro.au/connie2/)
$\endgroup$
3
  • $\begingroup$ Thanks, that is quite a comprehensive list. It would be good if you can expand your answer with the list of tools my answer did not mention. $\endgroup$ May 7, 2014 at 15:35
  • $\begingroup$ @IRO-bot: Let me know if this works. I know you have already mentioned Ariane in your answer, but I decided to keep the list from the website. $\endgroup$
    – arkaia
    May 7, 2014 at 15:48
  • $\begingroup$ Yes, looks great. $\endgroup$ May 7, 2014 at 16:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.