Why the current modeled is wrong?

[qianfangcui]

I used the climate data to spin up the model in the East China Sea, but the current especially the Kuroshio can't be modeled well,who can tell me how to solve this problem?

[kate]

Honestly, there's more than one way to get a current like the Kuroshio wrong. It took people a lot of trial and error to get something close to right for the Gulf Stream, a similar current in the Atlantic. Specifically, a climatology of T and S is going to give you something strange because it's a blending of many states of the current instead of one snapshot of the current. You might do better using boundary/initial conditions from a larger scale model that does data assimilation.

Other failure modes have to do with horizontal resolution and with location of your domain boundaries.