Add Southern Hemisphere region

[cbegeman]

Add a region for the Southern Hemisphere

[cbegeman]

@xylar Can you take a look at this and let me know if you see any issues? I'm having issues using it with MPAS-Analysis but I think something might be off with my editable environment because when I use a modified pytest to check these features they seem fine.

[xylar]

@cbegeman, yes, it looks fine to me.

[cbegeman]

@xylar Would you be willing to give this a try in MPAS-Analysis? I was changing the regions option in timeSeriesOHCAnomaly but I don't know if I'm getting the syntax right on the region name that corresponds to Southern Hemisphere. It seemed to fail with or without an underscore.

[xylar]

@cbegeman, sure, I can give it a try tomorrow.

[xylar]

@cbegeman, I see the problem. timeSeriesOHCAnomaly uses output from the online Regional Stats analysis member. This is really unfortunate because those regions are hard-coded.

We could presumably write a post-processed version of OHC that could take a region from a mask but that doesn't yet exist. Or alternatively wouldn't OHC be trivial to calculate from mean temperature in a region (using timeSeriesOceanRegions), volume of the region, heat capacity, etc.? Or am I missing something?

[milenaveneziani]

@xylar: you are not missing anything. I always thought that OHC analysis tasks could be done in post-process mode (along with regional statistics, to be honest). In fact, I believe that the other OHC task, the climatological maps, is done using the layer-weighted averaged temperature.

[cbegeman]

Thanks @xylar and @milenaveneziani. It seems like post-processing OHC makes a lot of sense and could allow the AM to be removed in OMEGA. I'll implement it in MPAS-Analysis. This work was motivated by being able to compare our SORRM OHC with published values from CMIP6 models.