Bottom drag coefficient (BDC)

[tony1230]

This is a topic on how to give bottom drag coefficient(s).

In standard input file *.in, we have RDRG(RDRG2) and Zob used as bottom drag for momentum calculation. And they also important in tide's computing. As for a region, we firstly want to calculate the tides (usually M2/S2/K1/O1/N2/K2/P1/Q1) and keep adjusting the inputs, such as bottom drag coeff or sometimes the tidal harmonic constant (THC) if we got a irrational results of Co-amplitude and Co-phase. We can not start simulating the currents until we have self-confident in tides (The 'tides' i mean here are the elevation,amplitude and phase,i.e., astronomical tide). Now, here comes the question: how to give relatively accurate BDC(s) for each point of the grid

We can give a constant value over the domain if the calculating area is not so vast or give a specified value if we already have a clear distribution. There is one point must be bear in mind is that almost all variables are spatial variation in the ocean, not excepting BDC. The key point to my question would be much clear, i.e., how to do and based on what then we can calculate this variation for BDC.

As i can understand, water depth and sediment type (mainly the d50) may play the primary roles, but i dont understand how they take effects. I have never seen someone do this job (using depth and d50 to get BDCs). Is that a difficult or pointless job

Thanks for any point of view.

-Shou

[arango]

You can specify the bottom drag coefficient at every grid point! You can have those fields in the grid NetCDF. If you check Data/ROMS/CDL/grd_spherical.cdl, you will see that we can have:

Code: Select all

        double rdrag(eta_rho, xi_rho) ;
                rdrag:long_name = "linear bottom drag coefficient" ;
                rdrag:units = "meter second-1" ;
                rdrag:coordinates = "lon_rho lat_rho" ;
        double rdrag2(eta_rho, xi_rho) ;
                rdrag2:long_name = "quadratic bottom drag coefficient" ;
                rdrag2:coordinates = "lon_rho lat_rho" ;
        double ZoBot(eta_rho, xi_rho) ;
                ZoBot:long_name = "time invariant, bottom roughness length" ;
                ZoBot:units = "meter" ;
                ZoBot:coordinates = "lon_rho lat_rho" ;

This is used in conjunction with CPP option UV_DRAG_GRID. This has been discussed previously in this forum

You may also activate any of the Bottom Boundary Layer (BBL) models in ROMS, but that is another story...

[tony1230]

Yes, i know roms can read in BDCs at every grid point. Actually, we can do this for ca. two years by adding a variable in varinfo.dat and corresponding files, and give the values for BDC as a variable of grd.NC file. As in my applications, i have calibrated the BDCs for times, but still dissatisfy with something.

My question in the first post is how to give relatively accurate BDC(s) for each point of the grid. I want to know if we can compute the BDC distribution by using water depth and sediment types (d50).

Thanks,

-Shou

[arango]

As I mentioned above, you can use any of the BBL model formulations. However, you will need some sediment properties and wind-induced wave data (usually from wave models).