Problem in storm surge simulation

[Robina]

Hello,

We are trying to simulate a massive storm surge. For the bathymetry, we only consider the interested area which is far away from the storm center, and we do not take into account the entire bathymetry including the center of the storm in our simulation to reduce the simulation time. Afterwards, the result shows an increase in water up to 15 meters, and if we increase the AMR level to 4, we have an output up to 200 meters in some gauges, which shows the results are not acceptable. Is it possible that such a problem will arise if we do not consider the bathymetry of the center of the storm in our simulation?
If yes, may I ask why?
The last thing, what are your suggestions to us for getting the correct results?

[Kyle Mandli]

Bathymetry is needed everywhere but you can use very coarse bathymetry away from the coast/area of interest.  For instance you can use degree accurate bathymetry in the deep ocean.  The code will not run if the bathymetry does not cover (or nearly cover) the entire domain so you must be giving it something.

For the large surge heights my guess is that the simulation is failing/blowing up.  This can happen if the storm gets too close to boundaries so that would be the first thing I would check.  Otherwise, if you can provide some images of the simulation we can hopefully see what may be causing the problem.

Kyle

--
You received this message because you are subscribed to the Google Groups "claw-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to claw-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/claw-users/7a33bfaf-36a1-42a2-a6ea-8fd3f0cc40e9n%40googlegroups.com.

[Robina]

Hello Kyle,

Thank you!  Yes, this was the reason.   " This can happen if the storm gets too close to boundaries so that would be the first thing I would check."

 But why does the storm getting too close to boundaries cause a problem?

Regards,

Ibelat

[Kyle Mandli]

Hi Ibelat,

The boundary conditions handle outgoing waves fairly robustly but can sometimes have issues with incoming waves, which are not specified.  The best way to avoid this is to ensure that the forcing from the storm does not get too close to the boundary and therefore cause the incoming conditions to be forced or dependent on the storm rather than the just the fluid dynamics.  

If you are interested I have been working on a test example and some tools that looks at this problem a bit more explicitly.  You can find the code at https://github.com/mandli/boundary-tests.  The plan is to incorporate this into an example in the surge-examples repository (https://github.com/mandli/surge-examples/) along with some explanation and suggestions.

Kyle

To view this discussion on the web visit https://groups.google.com/d/msgid/claw-users/c6169f59-9274-4962-88dc-9432a203ba5cn%40googlegroups.com.

[Robina]

Thank you so much, Kyle.

Best,