Interpolator mask

[s-good]

Description

The interpolator currently does not use the mask variable so it returns interpolated values when it shouldn't. In the case of nearest neighbour time interpolation, this results in the model value from the last time point always being returned.

Issue(s) addressed

Resolves #95

Dependencies

None.

Impact

Only affects JOPA runs where the model background has multiple time steps. The configuration files have recently been updated to all use linear interpolation so there will be no effect from this change.

Checklist

• I have updated the unit tests to cover the change
• New functions are documented briefly via Doxygen comments in the code
• I have linted my code using cpplint
• I have run the unit tests
• I have run mo-bundle to check integration with the rest of JEDI and run the unit tests under all environments

[s-good]

I've tested the code on MetOp SST data using the sith suite and it appears to work. However, I'm not 100% confident I understand the way the mask variable is supposed to be used. @twsearle, please could you give an initial review and if you think it looks OK I will work on updating the unit test reference data.

[twsearle]

This looks correct to me - its slightly different from how others implement it in that we ensure the default value is always the missing data indicator, however I still think this lines up with the behaviour of the generic atlas interpolator here:
https://github.com/JCSDA-internal/oops/blob/c531e3fde8a313349f7a27f3da00f8719858ad00/src/oops/generic/AtlasInterpolator.cc#L53

[twsearle]

@odlomax Can you confirm this is the correct approach to implementing the oops interpolator "time mask" variable?