Data assimilation for PFOA transport in vadose zone using EnKF with modified HYDRUS-1D
摘要
Per-and polyfluoroalkyl substances (PFASs) in subsurface media become a significant environmental concern due to their persistence, toxicity and ubiquitous occurrence. And the transport mechanisms of PFAS are not yet fully understood. Recent studies reveal that air-water interface adsorption (AWIA) contributes 50–75% of total PFAS retention and surpasses traditional solid-phase adsorption. However, key aspects of this process, particularly the AWIA coefficient Kaw, remain poorly constrained. Therefore, an original AWIA module was added to HYDRUS-1D source code to simulate perfluorooctanoic acid (PFOA) transport in vadose zones. The model was validated against experimental data from sand column studies. Furthermore, the data assimilation (DA) method via ensemble Kalman filter (EnKF) was firstly coupled with the PFOA transport model via assimilating breakthrough curves (BTCs) observations. Study results demonstrated Kaw and longitude dispersivity αL were the first and second primary parameters for the simulated BTCs. EnKF method was efficient to improve the simulated BTCs of PFOA and inverse the Kaw and αL jointly. DA results were better with joint inversion of both parameters than that with single parameter inversion. Moreover, the closer initial guessed means for Kaw and αL to the ideal value, the better DA results. Kaw and αL were inversed and optimized via EnKF even with good simulations without DA. The study results provide a scientific basis for handling the environmental problems caused by the abuse of PFOA in the past years.