IS it something wrong in vertical velocity(w)?

[hqz]

I find some dots in vertical velocity near eastern boundary(Fig.1). The configure of my model is as follow:
MY horizontal resolution is 5km,and the min and max depth is 5and 5000 meter,respectively. the vertical is 20 layer with theta_s=5,theta_b=0.4,hc=3.Model initialization is with woa13 temperature and salt for the month of January and no flow. The surface forcing is by monthly sustr,svstr swrad swflux,shflux derived from CFSR(1979-2009),and I also add the SST SSS, and dQdSST to model. The boundary condition is come from soda2.2.4(1958-2010),and i make a monthly mean. I run the model in cyclical climatology run for 20 year. My .h and .in file is show in attachment. Can Anyone help me?

Thanks in advance.

[kate]

What is the purpose of this simulation? I can't say I'm surprised to see noise in the w-field. In your case, the noise is in the deep part of your domain where the Kuroshio runs through it. Catching a corner of the Kuroshio like that might not be the best idea. Your grid will resolve a sharper current than the SODA boundary conditions will provide.

It has been a long time since I used monthly forcings. Maybe it's good enough for getting started with something, but you'll get more energetic forcing with higher temporal resolution, like 3-hourly. Again, what is your goal here?

[hqz]

What is the purpose of this simulation? I can't say I'm surprised to see noise in the w-field. In your case, the noise is in the deep part of your domain where the Kuroshio runs through it. Catching a corner of the Kuroshio like that might not be the best idea. Your grid will resolve a sharper current than the SODA boundary conditions will provide.

It has been a long time since I used monthly forcings. Maybe it's good enough for getting started with something, but you'll get more energetic forcing with higher temporal resolution, like 3-hourly. Again, what is your goal here?

My purpose of this simulation is make a spin up to get a initial condition for the real time simulation(i just want to study the upwelling around ZhouShan Island lacated in 122°E, 30°N), So I would like to know whether this will affect the region my research or the method to improve the result of w-field.

[kate]

To get an initial condition, I would not spin up that way. Instead, run a month with real forcing in hindcast mode, starting from say a HYCOM initial condition.

As for the w in the deep water, I never look at w. It is not one of the primary variables but is instead computed from horizontal divergences and convergences. It has always been noisy. If it bothers you, can you change the grid so that you don't include the Kuroshio at all?