Nature inspired probabilistic uncertainty analysis of a groundwater flow and radionuclide transport model for a uranium tailings pond area

Abstract

newline Uncertainty analysis is an integral part of model-based studies as it helps to acknowledge and quantify the inherent variability in the parameters used. It is mainly used to quantify the extent to which a particular input or result can genuinely represent reality. For environmental models, uncertainty mainly accounts for stochastic and subjective or epistemic uncertainty. When the input data is derived from insufficient observations and experiments, particularly in heterogeneous subsurface, challenges arise due to high spatial variation, leading to interpolating point values of crucial input parameters. It is essential to account for these spatial variations, especially when dealing with geologic and hydrogeologic heterogeneity within aquifer systems. Generally, probabilistic methods are used for uncertainty analysis. Researchers have explored approaches such as first-order and second-order reliability, importance sampling, subset simulation, and meta-modelling techniques. These methods aim to alleviate the computational demands of Montecarlo simulation and offer a quantifiable means to assess uncertainties within geological systems.