Is there a balance between SSH and currents?

[LeaF_7]

Hi,

I am running a simulation in the Yellow Sea and Bohai sea. The model has an open boundary at the south end. The grid has been rotated so we have this open boundary at the east.

model domain and depth

The horizontal resolution is 1km with 40 vertical layers. And we applied tidal forcing at the open boundary. Atmospheric forcing files are from ERA-interim. The quadratic bottom drag coefficient was set to be 0.0015 in our model. We have 2 ADCP measuring the current profile near the coast(sitting at the bottom and look up).

ADCP location

ADCP stations.png (68.15 KiB) Viewed 3603 times

The current data looks fine when comparing the model and the observation at ADCP1 location(10m and 20m means the distance above the ADCP). But the sea surface elevation is being underestimated by the model.

velocity

SSH

ssh_adcp1.png (145.95 KiB) Viewed 3603 times

Some post in the forum have discussed this, that maybe we should use a spatially varied bottom drag coefficient. If I decrease the rdrg2, the SSH will become better, but the current in the model will be much larger than our observations.

I wonder if anybody familiar with this, could this be caused by the error in topography? The model depth near the observation is about 3~4 metres shallower than our measurement. Or maybe there are some other parameters that affect the SSH or the current independently?

Sorry for the messy information, I hope I am clear on the questions here. The cluster I am using have some issues now so I am not able to download the log file.

Any suggestion is appreciated!
Thanks,
Fan

[rsoutelino]

I think bathymetry is one thing to look at. Inaccurate bathy can cause discrepancies in tidal elevations. Note that under or over estimation of real depth in remote areas can also affect the local sea level due to effects in the tidal wave propagation. Might worth sanity-checking or accuracy-checking the tidal constituents data that is forcing the model at the boundaries as well.

[LeaF_7]

Thanks for your rely there!

I will try to have a look at the ssh at the open boundary.

This has been a problem for me for a long time.

One of my supervisor suggested that the ADCP may have some uncertainty in compass so it might be bad to compare the u and v velocity, another way is to compare the magnitude and directions. So I made some plots as follows.

I guess the bottom drag (here I use quadratic drag) majorly forms the tidal current amplitude vertical profile here, so away from the bottom (like 10m and 20m above the bottom in my figure above) will be less effected from the bottom drag.

But at the same time, adding bottom drag will reduce the SSH in the model, which is my major confusion here, the simulated zeta is too small compare to my observations. Just as in my original posts.

Maybe I just use the wrong way to calculate the SSH from the ADCP pressure data?

Any advice is appreciated.
Thanks in advance.

Fan