Open DamienIrving opened 6 months ago
Add potential evapotranspiration, potentially using the pyet package.
Use all the ensemble members from each model (i.e. r1i1p1f1
, r2i1p1f1
, etc) to calculate the adjustment factors (i.e. pool the data before calculating the quantile changes). If there's enough ensemble members, this will dramatically reduce the natural variability in the change signal and leave just the externally forced change signal. It will also mean there's much more data for each month, so the process will be less sensitive to outliers / have more data in the tails.
This issue keeps track of suggestions for future Quantile Delta Change datasets.