For validation of tides at the domain of the North Sea, I decomposed SSH and U, V- velocity components from the results by using the pytides python tool.
Then I compared tidal amplitudes from the results at the points in the center of the domain with the amplitudes from the in-situ data, provided by the Flanders Institute (VLIZ), for three major modes (M2,S2 and N2).
And they happened to be twice bigger in ROMS, but the phases were OK.
![](https://pp.userapi.com/c841025/v841025391/5f8d/LpyEwpA4dRQ.jpg)
The I compared the amplitudes spatially with the original TPXO Forcing file, and everywhere at the domain they are also twice bigger (!), except the place around the tidal node, because its location slightly differs from the original.
![](https://pp.userapi.com/c841025/v841025391/5fa0/j1dXpLJP3H8.jpg)
Then we reconstructed tidal ellipses out of u and v velocity components from ROMS results, and the major and minor axes and the inclination angle from the original forcing file, and compared them for the points at the center of the domain. And there is the same ratio of 2!
![](https://pp.userapi.com/c841025/v841025391/5faa/QLvA9-JoGkg.jpg)
Why does it happen?
In the attachements, there are the header file and the .in file.