Eff ects of Soil Hydraulic and Transport Parameter Uncertainty on Predic ons of Solute Transport in Large Lysimeters

Advanced numerical simula on models can poten ally help improve guidelines for irriga on and salinity management. Many simula on model parameters have considerable uncertainty, and ideally that uncertainty should be refl ected in model predic ons and recommenda ons. In this work, we inves gate solute transport predica on intervals that can be generated by propaga ng model parameter uncertainty using Monte Carlo techniques. Flow and transport is simulated with a standard numerical model, while soil parameters and their uncertainty are es mated with pedotransfer func ons. Generalized global sensi vity coeffi cients are computed to determine the parameters having the greatest impact on transport predic on and uncertainty. Simula ons are compared with Br transport measured under unsaturated condi ons in large lysimeters packed with clayey soil materials. In a 48 cm tall, homogeneous soil profi le, model predic on intervals provided a reasonably good descrip on of a single, rela vely “noisy” breakthrough curve. In replicated 180 cm tall, layered soil profi les, model structural errors limited the accuracy of the predic on intervals under one irriga on water treatment, whereas under another treatment the predic ons tracked the me course of the data reasonably well but tended to overes mate solute concentra ons. The width of the predic on intervals tended to be small rela ve to the range of transport variability that existed across replicated lysimeters, par cularly at shallow depths. Addi onal work aimed at opera onal fi eld tes ng of model predic on uncertainty is needed if advanced water management models are to reach their full poten al.