Cut investigated domain

[reneschubert]

Name of your institution / research group

GEOMAR / Ocean Dynamics

Information request

In our case, we are cutting the SWOT-like simulated data for a specific region (here a nested domain in the Agulhas).
The interpolated swaths contain however the full SWOT-swath, each 16 Mb, and are filled with NaNs in the largest parts of the swath. Would it be helpful to cut the investigated region from the swath, i.e. the region with data?

[reneschubert]

I think, to speed up the procedure, a restriction of the interpolation to the investigated model domain would be great

[contreras-m]

Yes, it would be great the larger the domain as possible (just avoid sponge layers)

[reneschubert]

Thank you. My question is rather, if we should hand-in the full simulated SWOT swath (which includes NaNs in the parts that are not in our model domain, which is the largest part of the SWOT swath) or if we should cut the NaN-part of the SWOT-swath to save memory.

[contreras-m]

Thanks for the explanation,

Yes, you should include the NaN, since the idea is that the interpolation outputs have the same SWOT files format ( although this implies using more memory). Having the SWOT file format will make it easier for us to do the analysis.

I hope this answers your question.