open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am trying to propagate (stationary) extreme wave spectra inside a domain with TOMAWAC. The extreme waves seem to propagate fine, but I have some concerns.

1. Is it relevant to use TOMAWAC for this purpose ? (is TOMAWAC able to propagate extreme conditions ?)

2. Extreme waves tend to break at shallow water depths (I used Battjes & Janssen's model for bathymetric breaking).

This is a problem since at some of the points where the wave broke (0m in the model), a standard extreme value analysis would give extreme waves that are > 0m. On the other hand, I don't think removing the bathymetric breaking would be a good idea. Do you know if there would be an alternate parametrization of wave breaking I could try ?

I know I can modify GAMMA2 or use Miche's criterion for instance, but there are a lot of options and I would like to pick your brain, if you have suggestions...

Thanks a lot.

Hello Raphael,

I discussed your point with Michel Benoit. We understand that there are nodes of your spatial grid where the Hmo is equal to zero. This seems to be due to depth induced breaking, and is likely related to the fact the time step used for this sink term is too high at these places.

In order to better analyse your case, could you :

1. tell us what are the value of the following parameter in the steering file (or, better if possible, send me the steering file) :
TIME STEP/ PAS DE TEMPS
NUMBER OF ITERATIONS FOR THE SOURCE TERMS / NOMBRE DE SOUS-ITERATIONS POUR LES TERMES SOURCES
NUMBER OF BREAKING TIME STEPS/ NOMBRE DE SOUS-PAS DE TEMPS POUR LE DEFERLEMENT

2. send images showing, in the area where Hm0 = 0, the mesh and the bottom elevation contours with indication of spatial scale in X and Y axis, together with a plot of significant wave height over the same area.

With this info, we should be in postion to give recommendation for updating/testing parameters related to modelling wave breaking.

Regards,

Hello,

Here are my .cas file and some snapshots (the mesh, the bathymetry and 19y-Hm0) with a zoom on the area where most of the problematic breaking occurs.

The coordinates are in Lambert93, so X and Y are in meters.

As you will see I did not change the parameters NOMBRE DE SOUS-PAS DE TEMPS POUR LE DEFERLEMENT nor NOMBRE DE SOUS-ITERATIONS POUR LES TERMES SOURCES so these are set to their default values.

Thanks for your help,

RD

Hello,

Michel Benoit and I have looked at your result and we would advise you to make some test.

Firstly we would take a smaller time step. As your period is 10s, in 300s (your time step) the waves travel 2.4 km. You'd better use a time step of 60s.

Secondly you could use a NUMBER OF BREAKING TIME STEPS of 5.

This is finally not so important but the value of your FREQUENCE DE PIC INITIALE looks more like a period than a frequency.

Finally to judge your result we would need a profile of the hM0 and the heigth of water along the 4 cardinal directions to check if hm0 is null when the heigth is null. On your results we can only see that hm0 is small where heigth is small.

Regards,