Help~Got too high T and too low S in the bottom layer

[leon]

Hi all,

I'm using ROMS-SWAN-Sediment coupled model to simulate river sediment transport on continental shelf. However, at the river mouth, the T goes too high, about 300, and the salinity goes to low, about -300, after 7 months running. The river sediment concentration is really high, about 50 kg/m^3, and the wrong TS only happened on the last vertical layers. So I wonder, whether the high sediment concentration would cause the wrong T,S caculation since I defined SED_DENS? Any suggestions on this problem? Thank you so much.

Best,

leon

[jcwarner]

if you do a simulation without defining sed_dens do you get the same problem?

[leon]

Hi John,

Thank you for the reply. I just tried to not use sed_dens, but I got the same problem. The TS are totally wrong. Do you have any suggestion? Thanks

[jcwarner]

many people have had problems at river mouths. here are some things you should think about: are you using mpdata? how many cells wide is the river? what is the max value of velocity compared to courant number? how did you distribute vertical river layers? is the flow always coming out of hte river? etc etc. there are many posts on this issue. since it is not specific to sediment, you should focus on the T and S first and then put sed back in.

[leon]

Thank you for the advice. The weird thing is that when I just run the circulatio part without sediment submodel, everything is fine. And the false TS only occur at the bottom layer when running the sediment model. Anyway, thank you for your suggestion.

[jcwarner]

maybe there is an issue with the sediment being specified in your river input file. if you turn off sed_dens, then there should be no effect of sed on the currents. what if you make the sed=0 from the river?