We should have a test to ensure that we don't produce any NaN values anywhere there is valid data. Use the Scott RPS data as the benchmark. If there are NaNs in that data outside the domain (e.g. oceans), then check that we have valid data everywhere in that domain. If the Scott data is zero outside the domain, then check that we have zero NaNs period.
We should have a test to ensure that we don't produce any NaN values anywhere there is valid data. Use the Scott RPS data as the benchmark. If there are NaNs in that data outside the domain (e.g. oceans), then check that we have valid data everywhere in that domain. If the Scott data is zero outside the domain, then check that we have zero NaNs period.