Feedback on my GDM model

Dear CROCO community, best regards.

I have been working on my model for the GDM. I can’t get the SSH and Loop Current (LC) simulated by my model to fit better based on satellite data. I have changed my parameters for horizontal viscosity and vertical mixing several times, but I don’t see noticeable changes. My model has no data assimilation, although I would expect to see a better representation of the LC. I have assumed a year of model spin-up before analyzing the results. Below are the parameters of my model:

image description

Below I share a few results and the satellite product, with which I am making the comparison.

CROCO is the figure on the left.

I understand that DUACS is The absolute dynamic topography and the scale of values is not the same. Therefore, I have additionally compared with GLORYS reanalysis. As the simulation runs more, the simulated LC has a different behavior than the satellite. In addition to the presence of several eddies that do not come out in the satellite.

Please let me know your opinions and recommendations to improve my model.

Thanks for your support!


As mesoscale ocean dynamics are highly nonlinear, the model cannot be expected to match the data of a given day without data assimilation. We can only compare statistics, i.e. the variance of the SSH over several years. An easy alternative to data assimilation is to relax T,S in CROCO to GLORYS data using the ROBUST_DIAG method.