Calculate property anomalies during Argo sampling

[oj-tooth]

Add capability to calculate property anomalies relative to monthly climatology during nearest-neighbour Argo sampling:

• Develop the capability to calculate property anomalies versus co-located monthly climatology profiles (i.e., T_anom = T_prof - T_clim).
• Add functionality to regrid the anomalies by adding interpolated anomalies back onto the monthly climatology field - used to capture the main structures (similar to OA).

[oj-tooth]

@b-barton @atb299

Propose that we make the output of the synthetic observation profile dataset include both the time-evolving field and climatology field to allow efficient computation of the anomaly without computing this for the entire domain. In the case of the SwapRegridder, we substitute the full-field into the climatology, whereas all other Regridders will perform interpolation (e.g., OI or IDW) on the anomalies.