I'm runing a COAWST model, but i did some tests with only WRF.
I have some lakes which I transform into water body (I just put the whole lakemask to 0, and change these points in LU_INDEX to 16 (water body)), I checked the table parameters and looks nice.
After running for 2 days with the timestep of 10 seconds (actually model blow up with this time, I need 5 seconds but doesn't matter) and a Dx of 1km, I look at the SKIN TEMPERATURE (TSK), and doesn't look bad, but not what I expected.
In this gif (if you click you can see the animation) you can see the Berre lagoon, in France, in a piece of my domain, the bottom contour is TSK, the middle one is the water body in red, (the bottom part of this domain is the sea)
The top image is the variation coefficient of the temperature in time over a single line horizontal, corresponds with the arrow (Coef. Var), where the blue cross is the water body, is look that water is changing more over time than land.
The second plot is the evolution of the temperature over timesteps in a single point of land and a single point of water.
It also shows that water changes more than land.
And my final image is an autocorrelation.
Again I see that land points change the temperature more than water.
I'm not sure if is the correct behaviour, I guess that as we see in the first image, the thermal inertia of water is higher so it should have a lower variation than the land.
Does anyone know the reason? if is some problem of my set up, or maybe this is what should be and is fine.
I asked this in WRF forum, but I didn't get a reply. I would really appreciate your help
Jordi