En:using ww3:2013 fall waves school:waves 101
This page contains all the materials (or links to it) for the first day on the waves school.
- 1 Foreword
- 2 Course material
- 3 Tutorials: hands-on lab
- 3.1 Setting up the environment
- 3.2 Setting up a directory structure
- 3.3 Installing tools
- 3.4 Exercise 1: Reading and plotting spectral data
- 3.5 Exercice 2: Comparing modeled and measured spectral wave spectra
- 3.6 Exercice 3: Reading and plotting gridded data (regular and unstructured)
- 3.7 Exercise 4 : comparing modeled and measured seismic noise
- 3.8 Exercise 5 : comparing wave model and satellite altimeter data
- 4 Additional resources
Before we get started with this one week of waves, I need to make a few important points. First of all, in all processing or modelling activities, there are many choices that have to be made (what averaging, which value of this or that parameter ...). The danger of providing users with "ready to run" examples, as we do here, is that all these choices are forced onto the user in some kind of "black box", and the user may think of these are "default parameters". One of the goals of this week is to show you all these buttons and try to emphasize the important ones and their effects. Contrary to the wave model SWAN, WAVEWATCH III (hereinafter WWATCH or WW3) puts many of these parameters under your face in the *.inp files: there are no "default" parameters (except for namelists) nor "default" parameterization... This comes from the basic philosophy that you should not use a model if you do not understand what it is doing, and in WWATCH it forces you to go through some learning process: you need to take the driving lesson before you can actually start this car... and that can be frustrating, and understandingly has turned some people away.
Now, obviously, there is a limit to how far this logic can be pushed, and we have also tried to make it easier for us and for other users: there are many ways to go around the hurdles put by the large number of settings and switches ... We have made it easier to choose, but it will always be your choice. Most likely you will indeed start doing some processing and modelling by just copying over an example that runs. In preparing for these courses we asked again to ourselves: so why did I chose to do things this way instead of that way? I quite a few cases, the answer was ... "hum, I don't know, I looked at it some years ago and it was better like that" or "So-and-so told me to do it like this, why should I bother". Unfortunately we ran out of time to dig back on these problems, but we urge you to take the precious time to ask yourself, about each parameter setting or choice of method: why?
- Ifremer slides
These present general considerations on ocean waves: time and space scales, 1D and 2D spectra, typical spectral shapes and possible non-linear corrections. recorded on YouTube presented by F. Ardhuin
These discuss the different wave parameters that can be measured or estimated from numerical models. recorded on YouTube presented by F. Ardhuin
From buoy and satellites to models: some discussion of wave databases. recorded on YouTube presented by F. Ardhuin
- NCEP slides
- A short introduction to NCEP.
- Wave modelling at NCEP
- Hindcast database
- Data freely available from NCEP.
- and videos:
- YouTube video with NCEP presentation and wave modelling at NCEP (first 2 sets of slides) presented by H. Alves.
- YouTube video on 31 year Hindcast database .
- YouTube videp on free data from NCEP ... the end of that video carries into waves 201.
Tutorials: hands-on lab
Setting up the environment
We will be working on Linux-based PCs connected to the internet, with access to the caparmor.ifremer.fr cluster. PATH and other environment variables
Please keep a web browser open on the wiki. You should be able to log in on the wiki (using your Ifremer extranet login and password), so that you can also modify the contents of this and other pages.
Setting up a directory structure
Some of the data and tools used will be common to several tutorials and was put under the DATA or TOOLS folder in the "Ifremer Nautile thumbdrive" . Other data, specific to each lab exercise was gathered in TUTORIALS/TUTORIAL* or TUTORIALS/tutorial* .
You are thus advised to :
- Put the thumbdrive in a USB port
- open a terminal
- find out its name: ls /media -> will give you 'my_thumbdrive_name'
- cd ~
- cp -r /media/my_thumbdrive_name/TOOLS .
- cp -r /media/my_thumbdrive_name/DATA .
- cp -r /media/my_thumbdrive_name/TUTORIALS/* .
For those not present in Brest: you will find the list of necessary tools and data for each tutorial exercise. The tools can generally be found here and the data there, where each tutorial was packaged as a single tar file. For example: for the first exercise
the TWIST tool will be used for mesh visualization and editing. Now that you have copied over the TOOLS, please install and test the TWIST visualization package that uses the IDL run-time licence (if you have an IDL licence you could also run it after starting idl: in that case, following the IDL prompt, type @visu ).
cd TOOLS/IDL tar -xvf rt_visumain.tar.gz idl -rt=rt_visumain.sav
That last command should start a graphics interface. If you are not present in Brest, once you have installed IDL (see details here ) , the only files you need to download are : rt_visumain.tar.gz and rt_visumain.sav, which can be found in this ftp folder.
Exercise 1: Reading and plotting spectral data
You can now go to this page:
it also includes a link to YouTube where I have uploaded my screen capture. OK, it was the first time I did it, following Andres' advice, so hum, it could be better, but it certainly helps: YOU CAN PUT ME ON PAUSE! gee, I wish I could have done that to my professors, years ago.
The reason we start with FULL DIRECTIONAL SPECTRA is that, when you are running a wave model and getting funny behaviours on wave heights or anything else, you always have to remember than these wave heights come from a wave spectrum. If the wave heights look funny, it must be that the spectrum is VERY FUNNY. When you see it, you will see where the energy is coming from that makes it funny, and possibly you'll understand better what has gone wrong.
Following up now with Matlab, you can now go to this page:
First of all a reading script readWWNC_SPEC.m was missing in TOOLS/MATLAB on the ftp ... it has been added. Second, the wave directions in the matlab animation were not correct. I have thus updated the script example_read_spectrum.m This is now updated on the ftp
Exercice 2: Comparing modeled and measured spectral wave spectra
You can now go to this page:
Exercice 3: Reading and plotting gridded data (regular and unstructured)
For plotting with IDL, you can now go to this page:
For the matlab part, just follow the tutorial text (see link above)
Exercise 4 : comparing modeled and measured seismic noise
text for "exercise 4" . This exercise uses matlab only, but you can still use IDL to visualize the modelled noise sources.
Some more details on this page.
Exercise 5 : comparing wave model and satellite altimeter data
- book by L. Holthuijsen: Waves in Oceanic and Coastal Waters (Cambridge University Press)
- Ocean waves in Geosciences by F. Ardhuin and J.F. Filipot: french version and english version ... just begun translating and updating from the 2012 French version
- Rascle and Ardhuin, Ocean Modelling 2013
- Queffeulou et al. (2004) and updates: altimeter data calibration.
- Zieger et al. (JTECH 2009): more altimeter calibration
- Globwave reports: altimeter data quality
- Ardhuin et al., JGR 2011: understanding seismic noise sources
- Rascle and Ardhuin, Ocean Modelling 2013