Surface temperature too high

dynamics-numerics-params
1

Hi everyone,
I am currently working on a configuration on the Gulf of Lion and I an stuck because my results show a surface temperature too high compare to buoy data or Mercator.
I initialize to model with MERCATOR data and use these as climatology and I use the bulk formulation with AROME dataset. Here is my cppdefs.h cppdefs.h

As you can see on the following pictures the model estimate a surface temperature that is too high compare to mercator reference.

The heat fluxes seems well calculated by the model. My guess is that the influence of the winds on the temperature is underestimated but maybe I have missed something on the configuration.

Thanks in advance

CROCO

image description
image description661×555 28.6 KB

MERCATOR

image description
image description818×687 102 KB

SHFLUX CROCO

image description
image description655×556 51.1 KB

2

I had a similar issue (and am about to trial the proposal I am suggesting to you myself) and I suspect that it is because optical characteristics (Jerlov water type) are hard-coded as type 1 in ~/OCEAN/lmd_swfrac.F, which is activated as a result of you activating GLS_MIXING in your cppdefs.

3

The Western Med is classified by Jerlov type IB, which corresponds to a lmd_Jwt=3 in lmd_swfrac.F, not lmd_Jwt=1.

I wonder would modifying your lmd_swfrac.F change the outcome? It would be great to hear if this works!

4

Did you get this to work?

5

Hi,
I manage to find that this was an issue in the interpolation of the initial and boundary conditions of my configuration.
I found that the wrong interpolation gave me temperature in shallow water that were almost constant on the vertical. The upwelling did not cool the water then.