Tidal flood-ebb asymmetry

[marcmestres]

Dear all,

I am having an issue with tides in my simulations that I don't understand since I am no tides expert. I am modelling a South American Pacific estuary using a ROMS-only COAWST implementation on a three-grid nested scheme, forced at the outer boundaries by the CMEMS Global product and at the surface by atmospheric data from ECMWF. The CMEMS forcing does not include tides, so I add them as an additional contribution to the sea level and currents using the TPX09 model, and creating a tidal forcing netcdf with the OTPS2FRC matlab files.
I define the SSH_TIDES and UV_TIDES CPP options, and undefine RAMP_TIDES, ADD_FSOBC and ADD_M2OBC. I don't use wetting/drying either, as my minimium depth is larger than the tidal amplitude.

The simulation runs fine for a test period of one month, but when I compare the ROMS output with sea level measurements from a tidal gauge inside the estuary. it shows a flood-ebb "asymmetry" that is not real. As shown in the attached figure, whereas the measurements clearly show a 6-hr flood-ebb alternation, the modelled flood tide lasts about 5 hours, whereas the ebbing phase lasts 7 hours. The tides outside of the estuary appear to be symmetric in this sense. I am wondering what could be the reason for this behaviour.

As I said, I am no expert in the use of tides, so any help would be appreciated.

Cheers

M

[jcwarner]

looks like model is taking longer to ebb, that could be due to friction too high on the landward side of this site.
you need to look at an animation and see how the flow looks.

[gli353]

It might have something to do with bathymetry.

You may want to have a look at http://www.coastalwiki.org/wiki/Tidal_a ... hodynamics (especially the section "Tidal wave deformation in the presence of friction and intertidal areas", which applies to friction-dominated estuaries with small tidal amplitude).

Now the flood lasts shorter and ebb lasts longer. This is the case when considering that the high-water crest, when propagating into an estuary, experiences less friction than does a low water trough (equation 7) there, so the celerity during flood is higher than sqrt(g*h), vice versa for the ebb, leading to shorter flood and longer ebb. A "storage area", say, tidal flats, may reverse this scenario, as shown by the equation 11 in the link: it may lead to weaker flood current and stronger ebb current, shortening the ebb duration (and lengthening the flood duration).

However, you mentioned that you have not activated wetting and drying, since the minimum depth has a value lower than the tidal amplitude. Have you checked this against the more up-to-date bathymetry data? We ran into a similar problem before, and it turned out that using a more accurate LiDAR dataset for bathymetry solved the problem (partially).

Gaoyang

[marcmestres]

Thank you, John and Gaoyang, I will look into the bottom friction characterization. The bathymetry within the estuary is interpolated from a set of rather high-resolution field data, so I don't think the problem will be due to that, but I will bear it in mind just in case.

M