Implement /compute/WATMAN

[tjmisko]

Need to resolve issue with CWUEFF: How to handle changes in water use efficiency over time?

Best method is probably a moving average method. Talk with professor about this for dthe dynamic data. Also will apply to deforestation and carbon capture

[maxstrr]

need to apply moving average/rolling window for change in water use efficiency intermediate; also need to remove outliers

[tjmisko]

WATMAN compute should not store any observations in clean API data until we decide how to compute for the dynamic indicators.

• [ ] Remove the insertion line in the compute route until we decide how to compute
• [ ] Present analysis on how to compute water management for dynamic data
  • [ ] Should we include water use efficiency as a policy indicator?
    • [ ] What factors determine water use efficiency?
  • [ ] Previously we were using change over time to assess policy.
    • [ ] Try computing the change over a rolling window: look at the difference today from a baseline average computed over the 10-5 years prior.
      • e.g. 2010, the value would be the precentage change in water use efficiency from the average in the period from 2000 to 2005.
      • Do this for as many years as we can, and report on the data coverage
  • [ ] What would be the advantages and disadvantages of looking at just water use efficiency?