blows up in shallow water! wet_dry problem?

[FangFang]

I applied roms in a bay using tidal forcing.The minimux size of my grid is about 400*250m. Firstly i clampped the depth of all the grids to 5m, without definition of wet_dry option,using dt=120 and ndtfast=15,it runs well.Then i clampped the depth to 0.5m,and defined the wet_dry option,and set Dcrit=0.1m. With dt=120,ndtfast=15,it blowed up after it run for 10hour. With dt=60,ndefast=20,it blowed up at 35th hour. I found it is the ubar/vbar blowed up at wetting/drying edge. It seems that the wet/dry module cause the unstability. How to do with this problem? Should I set some special parameters when useing wet/dry module? Thanks for any advise.

[linzhenhua]

I guess this thread may be useful for your question.

viewtopic.php?f=19&t=1087

[FangFang]

I guess this thread may be useful for your question.

viewtopic.php?f=19&t=1087

thanks very much

[longmtm]

Please post your findings with the bug pointed out in the previous post, I'm also interested in seeing how well WET_DRY works

Thanks

Wen

[FangFang]

Please post your findings with the bug pointed out in the previous post, I'm also interested in seeing how well WET_DRY works

Thanks

Wen

It didn't help.I don't think it is the right solutions...

[arango]

We will kindly ask to be patient and stop using this option until we completely rework the wetting/drying algorithm. We already sketched a new strategy to take care of associated boundary conditions problems but all this needs to be tested. Otherwise, this forum will be full of complaints. This issue is near to the top of our priority list. Please be aware that John Warner and I have a heavy load of work plus our own projects. We cannot abandon our projects immediately to go and solve your problems.

[linzhenhua]

In such cases,I think it may be better for ROMS to issue an error and stop the running if wet_dry is defined,until it's could be used.

[arango]

Well, you also have the choice to turn off that option or figure out how to fix your application and stop complaining The wetting and drying works well for others and there is not reason for us to completely remove that option. This option is based on numerical discretization engineering and not based on first principles from the continuum equations. This means that it is application dependent and sometimes is very difficult to come up with generic solutions.

You perhaps needed to be reminded that ROMS is given to you free You need to read the open source license.