Impact of vertical coordinates on HYCOM solutions

[Alfredo Terrazas]

Hello HYCOM community,

I am currently investigating the impact of vertical coordinates on HYCOM solutions and I have a question regarding the observed differences.

I ran two experiments:

1. Using the vertical coordinates from GOMu0.04/expt_50.1

2. Using the vertical coordinates from GOMl0.04/expt_32.5

In both experiments, I kept all other settings identical. Specifically, the forcings, initial conditions, and boundary conditions were taken from expt_50.1, and I did not modify anything except the vertical coordinate configuration.

I am attaching a profile of speed from both experiments. In the plot, the red line corresponds to the profile with expt_50.1 coordinates, and the black line corresponds to the profile with expt_32.5 coordinates. Both solutions were interpolated to the same set of levels from NetCDF files, which are used to generate the figure.

The results show a significant divergence between the two experiments. My question is: Is such divergence expected solely due to changes in vertical coordinates, or could there be something in the configuration that is causing this behavior?

Thank you very much for your guidance.

Best regards,

Alfredo

[Alfredo Terrazas]

[Alan Wallcraft]

Since the two cases you used have different densities (sigma0 and sigma2), it is going to be hard to compare them.

See:

Xu, X., Chassignet, E. P., and A. J. Wallcraft, 2023, Impact of vertical resolution on representing baroclinic modes and water mass distribution in the North Atlantic PDF Document, Ocean Modelling, doi: 10.1016/j.ocemod.2023.102261 [pdf] PDF Document. 

Xiaobiao's standard setup is 32 layers, which is most similar to GOFS 3.0.  GOFS 3.1 adds 9 layers as the surface that are always in z-coordinates.

Alan.

[Alfredo Terrazas]

Thank you, Alan, for the explanation and the reference. I will check the Xu et al. (2023) paper.

Best regards.

Alfredo