3D numerical modelling of a pulsed pumping process of a large DNAPL pool: In situ pilot-scale case study of hexachlorobutadiene in a keyed enclosure

A Two‐Stage Bayesian Data‐Driven Method to Improve Model Prediction

“…Numerical groundwater models can predict the spatial and temporal distribution characteristics of groundwater flow and contamination, and have been widely used in the scientific management and protection of groundwater resources, such as groundwater resource evaluation (Ascott et al., 2020; Berthold et al., 2004; Maddaus & Aaronson, 1972), coastal seawater intrusion problems (Koussis et al., 2015; Zang & Li, 2021), and groundwater pollution remediation (Giraud et al., 2018). Because the real earth system is very complex and some features are not directly observable, it is difficult for model users to conceptualize an accurate groundwater model with limited knowledge, which inevitably leads to random and systematic prediction errors (Demissie et al., 2009).…”