Joe Hamman

> Or do you mean some sort of "re-normalizing"? Say, if the weights for a destination point are [0.25, 0.25, nan, nan], you would like the weights to be converted...

> If configured properly, the mask on the source grid will lead to an appropriate normalization for weight factors with masked entries excluded in the final weight matrix. The destination...

@JiaweiZhuang - thanks for doing some extra leg work. I agree we need to hear from the ESMF crew how best to proceed. In the end, I'll be happy if...

Thanks both of you for keeping this conversation going. I've been busy working on other things but recently ran into how CDO treats this case. For conservative area remapping, there...

Thanks @JiaweiZhuang - this is exactly what I was going for.

@JiaweiZhuang - circling back here. What do you think the next steps should be to resolve this issue. Your example above is compelling to the point I would like to...

Something that may be of interest on the xESMF roadmap is to move the repo to another namespace. [xarray-contrib](https://github.com/xarray-contrib) comes to mind as an obvious option. This may help increase...

> My major concern is how can xarray represent unstructured grids (as discussed in docs)? My thought here is that we can follow existing conventions for this. There are existing...

> If xarray doesn't give additional advantage for analyzing unstructured mesh, it doesn't seem to worth the effort to use it as the frontend for ESMPy regridding. I'm guessing my...

@bekozi - thanks for revising this issue. > This is out of the current scope of my package. I'm curious what it would take to scope this issue such that...