Initial sediment layers

[c.drinkorn]

Hi all,

I wonder if and how it would be possible to start a sedimentation simulation with initial sediment layers coming with e.g. thickness and densities, by running just the sediment module in ROMS. Or will I have to run a coupled SWAN simulation in order to do that?

[jcwarner]

The sediment part in roms can be run with just roms. The swan coupling provides wave information that can be used by the ocean model to:
- compute enhanced surface mixing due to tke wave breaking
- computed wave driven flow using vortex force approach
- compute enhanced bottom stresses due to combined action of waves+currents

so you can run the sed alone, but if you are studying a realistic coastal bay or shelf (for example) then you should work up to include waves.

[c.drinkorn]

The sediment part in roms can be run with just roms. The swan coupling provides wave information that can be used by the ocean model to:
- compute enhanced surface mixing due to tke wave breaking
- computed wave driven flow using vortex force approach
- compute enhanced bottom stresses due to combined action of waves+currents

so you can run the sed alone, but if you are studying a realistic coastal bay or shelf (for example) then you should work up to include waves.

Hi John,

thank you for your prompt reply!
I am rather studying a deep sea sediment body fed by a deep boundary current. That's why I did not consider SWAN coupling and tried to set up ROMS with the sediment module instead.

Looking into ana_sediment.F tells me what ROMS will expect as input if I turn the analytical option off. I could provide thickness and porosity but seemingly not density. Also, I do not have information of ripples geometry or in fact their geometry can simply be assumed to be zero. Age should start with zero, too, and then evolve with model time. So I could provide these sediment properties in the initial file, except for density. How could I have ROMS read densities from an input instead of using the ones from the in-file?

[c.drinkorn]

It seems to be possible by using a bottom boundary layer model... I will try something like that.
I'd be happy about any further hints.