En:using ww3:2013 fall waves school

Example of unstructured grid and model results (color is wave height) for the French West Indies, case of 2013 Tropical storm Chantal. See http://www.previmer.org/previsions/vagues for real time forecasts and hindcast visualizations

aims and scope

The "waves school" will start on November 4 . It is intended for graduate students, post-docs, researchers, engineers or consultant that are dealing with ocean wave data of one form or another. The school will draw from the latest developments made possible by the IOWAGA, NOPP-waves, & Previmer, projects, and will benefit from the new (beta) release of the WAVEWATCH III (R) version 4 from NOAA/NCEP.

Local information

To prepare your stay in Brest during the ww3 training course, please find here all the information you need. For any question, send email to mickael.accensi@ifremer.fr


List of trainers

  • 01 Accensi, Mickael
  • 02 Alves Henrique
  • 03 Ardhuin, Fabrice
  • 04 Roland, Aron

List of registered trainees

  • 01 Anthony, Robert : confirmed
  • 02 Barbariol, Francesco: confirmed (last 3 days only)
  • 03 Benetazzo, Alvise: confirmed
  • 04 Benshila, Rachid: confirmed
  • 05 Chailan, Romain: confirmed
  • 06 Gallagher, Sarah: confirmed
  • 07 Gautier, Caroline: confirmed
  • 08 Garnier, Valerie: confirmed
  • 09 Gualtieri, Lucia: confirmed (2 days only)
  • 10 Leckler, Fabien: confirmed
  • 11 Lagodan, Sabique: confirmed
  • 12 Magne, Rudy: confirmed
  • 13 Mansal, Leon : canceled ... moved to march 2014
  • 14 Mediavilla, Dernis: confirmed
  • 15 Michaud, Eloise: confirmed
  • 16 Mora Escalante, Rodney: confirmed
  • 17 Pallares, Elena: confirmed
  • 18 Pianezze, Joris: confirmed
  • 19 Perez, Jorge: confirmed
  • 20 Pein, Johannes: confirmed
  • 21 Rawat, Arshad: to be confirmed
  • 22 Sepulveda, Andres: confirmed
  • 23 Stutzmann, Eleonore: confirmed (2 days only)
  • 24 Veras Guimaraes, Pedro: confirmed
  • 25 Camarana, Amaury

Waiting list for November 2013

  • 01 Yohia, Christophe

list for March 2014

  • 13 Mansal, Leon

+ Many other people

code and tools versions

Matlab or octave

Easy reading of NetCDF 4 requires a Matlab version R2010b or more recent. Otherwise you may transform NetCDF 4 in NetCDF 3 and use a Matlab version that is as old as 2008. Andres Sepulveda has started to adapt some of our basic matlab tools for *Octave* (an open source alternative to Matlab). You may contact him directly.


For visualization of model output (maps, time series ...) and mesh generation we will use a runtime version of IDL that will allow you to use our Wave Visualization and Analysis Tool . Please follow the indications on that page.


The WW3 code version to be used is alpha-4.12, with a tar-ball kindly prepared by Henrique Alves. A request for access to this code version will be sent to NCEP this week (October 7). Due to US gov. shutdown, a temporary ftp site for download may be set up at Ifremer. However please read the licensing agreement and [ ftp://ftp.ifremer.fr/ifremer/cersat/products/gridded/wavewatch3/pub/HINDCAST/publications/WW3_early_distribution.pdf | this clarification by NCEP ] .

Because it is an "alpha" version, there are still a few things that need to be ironed out ... thank you for your patience. On this page you will find a list of known issues ... The manual (which is contained in the package) is available [1] . The cover page says it corresponds to version 4.08 updated on August 25, 2011 ... This is not true! It does correspond to version 4.12 and was updated on July 2013, but it is still incomplete and your feedback will be very useful for completing it and making it easier to read. First time users of WW3 should we warned that it is written backwards, and you should start at chapter 5 if you want to install the model and work your way backwards towards chapter 1. If you just want to know what is in the model and what it does then you can start at chapter 1.

our to-do list


The program is designed in such a way that people that only use wave data and do not want to run a model can attend day 1 & 2 while the real brave geeks that actually want to install WAVEWATCH III and run the thing will be welcome to stay until late Friday.

  • Monday 4, 9-10 AM, Ifremer, Salon de l'ocean: Welcome, coffee, logistics, discussion of participants objectives (e-mail discussions before Monday 4 are welcome): what do you want to acheive / understand with wave data and models.
  • Monday 4, morning, starting at 10 AM, Salon de l'Océan (lectures)
basic principles of ocean waves: elevations, velocities pressure.... and spectra ( 1 hour: Fabrice Ardhuin)
different spectral wave parameters: what can be measured and what is usually measured ( 20': Fabrice Ardhuin)
short break (lunch ticket sales)
wave measurement databases: altimeter and SAR data, buoy data, seismic noise, IOWAGA & Previmer model databases ( 30': Fabrice Ardhuin)
the NCEP hindcast and forecast database ( 30': Henrique Alves)
examples of ocean engineering applications: long term statistics ... ( 30': Michel Olagnon)
Lunch break 1 PM - 2 PM
  • Monday 4, afternon, starting at 2 PM (hands-on practical training: computer rooms at IUEM )
Model and data file formats: ASCII & NetCDF, reading tools. (M. Accensi)
Exercise starting from spectra, visualizing and recomputing integrated parameters
using long wave data time series (??)
reading wave model output and altimeter data, performing simple error analysis on wave models
coffee break from 4 PM to 4:30 PM & Posters
comparing ocean wave model or buoy data and seismic noise records
closing at 6 PM.
  • Dinner somewhere ( downtown Brest, around 8 PM).
  • Tuesday 5: Physical parameterizations and numerics, 9 AM (lectures, Ifremer, Salon de l'Ocean)
the wave action equation and source terms, basic principles (H. Alves 40').
parameterizations in wave models and impacts on various quantities: from Hs to air-sea fluxes. Why results from

this WW3 different can be different from that or that version of WAM or SWAN. So why do we have different wave models? (F. Ardhuin 1h30 hours).

forcing fields: quality of wind fields, validation datasets
coffee break
wave model numerics: principles (A. Roland. 1 hour)
wave model numerics: the dark side of limiters, SWAN's "renormalization" .... (A. Roland, 50 minutes)
  • Tuesday 5: Physical parameterizations and numerics. afternoon (hands-on practical training)
analysis of different model hindcasts and different model resolutions. (spatial, possibly spectral too).
coffee break and posters from 4 PM to 4:30 PM.
evidence of discretization and numerical scheme effects on coastal wave model results
looking at directional spread
closing .. around 6 PM
  • Wednesday 6, morning (9 AM):
Installing and running WAVEWATCH III: code structure, pre-compilation, running the model: the basic steps, from ww3_grid to ww3_out* (H. Alves. 2 hours )
coffee break
introduction to grid types and multi-grid (F. Ardhuin 30 h).
specific aspects of unstructured grids (A. Roland & F. Ardhuin 1h)
lunch break (12:30 to 1:30 PM)
  • Wednesday 6, afternon (hands-on practical training)
setting up and running WW3 on single grids
coffee break and posters from 3:45 PM to 4:15 PM.
running a triangular mesh grid and using the results. From preprocessing to post-processing.
closing at 6 PM or before

  • Thursday 7, morning (9 AM, salon de l'océan)
grid generation for regular grids (H. Alves. 1 hour)
Running WW3 as a multi-grid system (M. Accensi. 1 hour)
Mesh generation with polymesh (A. Roland. 1 hour)
  • Thursday 7, afternoon: (hands-on practical training, computer rooms at IUEM )
one simple example with gridgen: a global + regional zoom
using polymesh
optimizing a mesh
  • Friday 8, morning, 9 AM (hands-on: computer room at IUEM):
to be defined (possibly something more interactive on the different projects of the participants):
Coffee break at IUEM (10h30 AM)
Finishing at 12h30



The material and tutorials will be accessible from here. You may already go to the page En:using_ww3_output

Additional resources