Assessing Uncertainty of a Biofilter Micropollutant Transport Model MPiRe

MPiRe (Micro-Pollutants In RaingardEns) model was developed to predict both flows and removal of micropollutants by stormwater biofilters. It is a conceptual 1D model that includes sorption/desorption, biodegradation and volatilization processes. This paper presents an uncertainty evaluation of MPiRe using the GLUE methodology with atrazine as a representative pollutant. The uncertainty analysis shows that the soil-water partitioning coefficient (normalized to organic carbon content) is the most sensitive model parameter, while there is some correlation between sorption parameters and high uncertainty in the degradation rate estimation. It is hypothesized that the correlation between sorption parameters can be diminished by choosing two different combinations of calibration parameters (e.g. variations of their mutual products), and this hypothesis will be further tested. The practical implication of this analysis is that particular care should be given to measurements of initial outflow concentrations of events (to decrease the uncertainty in the degradation rate estimation). Additionally, if it is necessary to prioritize between monitoring procedures, the most attention should be given to sorption kinetics.