Philip Chmielowiec

120km: (59,329 nodes, 87980 edges, 28571 faces) 480km: (3947 nodes, 5754 edges, 1791 faces) About a 15x increase in the number of each elements going from 480km to 120km

Results. The ~15x expect increase for linear scaling functions is present in most of these benchmarks. 

@anissa111 I've added you as a reviewer in case you were interested in taking a look, however don't feel obligated to actually give it a review!

Pinging @kafitzgerald if you want to take a look too!

> I noticed this also, possible to have a flag and only perform this operation if needed/specified? > > Like xarray, we should load the whole thing only when user...

> Will that 6 minute cost show up in some other operation that has to actually read the data? The 6 minutes is the total system time needed to execute...

For this dataset, and all SCRIP grids, the costly operation is when we find all the unique [lon, lat] pairs and construct the face nodes in the following code block,...

> That does seem like the only bit of code in the that does anything significant. I think in these cases we can recommend the user rewrite the mesh in...

> > > > Would decorating this method with numba play well? It might, especially if the overall time to convert to machine code is shorter than the execution time....

Just ran a quick check for a 500,000 face grid. I'll run some for larger grids too. With Numba: 892ms Without Numba: 17ms