Hi all,
I am currently coupling my ocean model with GOTM. While the model runs initially, it unfortunately crashes after a certain period. I am using a turbulence scheme based on the k-epsilon model, but I am unsure of the correct configuration for the gotmturb.nml file. GOTM is used as a module accessed by PSOM to estimate the mixing coefficients at each time step.
Based on the model output, turbulent mixing appears normal in most areas. However, there are a few water columns exhibiting unusually high turbulence that contribute to the overall model collapse. After running the model for 50 time steps, I observed anomalously high values of the eddy viscosity coefficient in the horizontal cross-section, as well as unusual flow velocity values. What steps can I take to address these anomalous turbulence results and ensure the stable operation of the 3D ocean model?
Thank you for your time and assistance.
dear haibo tang,
sorry for the late reply. the present turbulence library requires the turbulent diffusivities num and nuh from the previous time step. could you please provide them (as you do for tke and eps).
cheers, knut
--
You received this message because you are subscribed to the Google Groups "GOTM-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gotm-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gotm-users/b8e23575-1de6-4b00-a263-a46711becf4en%40googlegroups.com.