Gustavo Marques

Yes, this is using the Ferarri et al., 2010 stream function.

@StephenGriffies, thanks for the explanation. The snapshot above, taken from OM4_05, is already the depth integral (GMwork, CMOR equivalent is "tnkebto").

No, we have not. I propose we talk about this during the next MOM6 dev call.

Update: I've run OM4_05 with `KHTH_USE_FGNV_STREAMFUNCTION = False`. Still getting GMwork > 0, but now positive values occur mostly near the coastline (red values below, the maximum value is 7E-3)....

Update: Setting USE_STORED_SLOPES = False does not solve the problem.

I believe this issue still exists, although it would be good to confirm it using the latest main branch. My plate is pretty full right now, but I will try...

I run OM4_05 with the latest dev/gfdl for MOM6 and SIS2 and I still see positive values for GMwork. I definitely see improvements when compared to the first plot I...

Good question. I was about to say [this](https://github.com/NOAA-GFDL/MOM6/pull/1050/files#diff-f667ab3399aff360a13eefb0797f44b5489cca3f149cfcf41936579b20d87b95R1240-R1248) was the code change that made the difference. However, setting `USE_GM_WORK_BUG = True` results in very little difference, so I don't think...

@marshallward, shouldn't [this line](https://github.com/NOAA-GFDL/MOM6/blob/dev/gfdl/src/parameterizations/lateral/MOM_thickness_diffuse.F90#L1066) be changed to `Work_u(I,j) = Work_u(I,j) - G_scale`, just like you did in https://github.com/NOAA-GFDL/MOM6/pull/1050 ? Similarly for Work_v [here](https://github.com/NOAA-GFDL/MOM6/blob/dev/gfdl/src/parameterizations/lateral/MOM_thickness_diffuse.F90#L1331).

@StephenGriffies and @Hallberg-NOAA can you please clarify why the temperature of fprec (frozen precipitation) can be < 0C?