Skin-temp vs SST vs soil temp

Good morning

I've been running some OpenIFS experiments at Tco95 L91. As usual OpenIFS reads soil temp at level 1 (STL1), sea ice fraction (CI) etc from the ICMCL file.

My question is: Should the soil temp at level 1 (STL1), skin temperature (SKT) and sea-surface temp (SST) be equal at all times over the ocean?

I have made a plot of monthly mean STL1 from the ICMCL file and SKT from the output. Both variables are very similar, but not identical.
On a map plot of SKT - STL1 it is clear that SKT - STL1 is around -0.1 K almost everywhere over the ocean.

Could this be the "cool-skin" parametrisation mentioned in the IFS documentation, or is this a known bias, or could there be an error somewhere?

Note: The ICMCL data was interpolated to regular grid using CDO (bi-linear) and the OpenIFS output on regular grid comes from XIOS (also bi-linear, I think...). So both STL1 and SKT are interpolated variables.

Cheers
Joakim