Pollutant fate of groundwater LNAPLs under unknown pollution source information by stochastic methods

Groundwater organic pollution by LNAPLs is a global concern, and clarifying LNAPL pollutant fate is critical for remediation under unknown pollution source information. This study proposed an integrated framework combining pollution source identification, stochastic simulation and uncertainty analysis, which integrated data assimilation (ES) with Bayesian inference (DEMC_(B)) to develop the ES-DEMC_(B) algorithm for accurate inversion of LNAPL source information and acquisition of their posterior probability distributions. Based on the posterior distribution, Monte Carlo stochastic simulation was conducted to quantify uncertainty propagation, reveal the pollutant fate characteristics of groundwater LNAPLs, and analyze their correlation with source information. Validated at a petrochemical-contaminated site in Jilin City (benzene as the target LNAPL), the results showed that ES-DEMC_(B) realized efficient and accurate source identification, and stochastic simulation revealed clear LNAPL pollutant fate regularities: 98.2%-98.5% of benzene remained in the aquifer, 1.49%-1.64% were biodegraded, and only a tiny proportion flowed out through the aquifer boundary. This study clarifies the migration and transformation law of groundwater LNAPLs under unknown source information, and provides a reliable technical support for the design of targeted LNAPL pollution remediation strategies.