Author(s): Seth Nathaniel Linga

Linked Author(s):

Keywords: Sensitivity auditing epistemic uncertainties irrigation water withdrawal global water model assumption hunting

Abstract: Some Global Water Models (GWMs) incorporate modules specifically for estimating irrigation water usage, which help guide agricultural and water management strategies. However, the complexity of reality forces analysts to employ assumptions, simplifying the modelling chain or filling gaps in their knowledge. Uncertainties in the representation of the phenomenon of interest are part and parcel of any model-based exercise. Yet, traditional uncertainty and sensitivity analyses often overlook epistemic uncertainties stemming from modelling assumptions, leaving significant portions of the uncertainty space unexplored. We do not know the extent to which irrigation models are grounded on scientifically substantiated claims versus premises that are based more on practicality – pragmatic assumptions. This knowledge is important to define the robustness of irrigation models and their epistemic reach. Sensitivity auditing is an extension of sensitivity analysis where all assumptions in the model, both quantifiable and non-quantifiable, are examined. Using sensitivity auditing, we will expose potentially relevant assumptions embedded in the modelling of IWW by nine GWMs (CLM4.5, H08, LPJmL, MATSIRO, MPI-HM, PCR-GLOBWB, VIC, Water-GAP, WBMplus) that likely have non-negligible influence in the final estimation.

DOI:

Year: 2025