CMIP6 ts versus tos for Pontential Intensity calculation

As per the title, I am interested in computing the potential intensity (PI) and wondering what the difference between surface temperature (ts) over the ocean and sea surface temperature (tos) in CMIP6 is. I ask this because I want to compute PI over the whole global grid, and the tos data have, for many models, a different grid, and for some, I suspect that regridding would introduce a considerable amount of bias. I want to know if I am introducing a considerable amount of bias by proxying tos by ts.

To be more precise, I am using monthly data.

I searched online for the relationship between the two but to no avail.

Thank you very much