Modeling the effect of the submerged zone on nitrogen removal efficiency of a bioretention system under dry-wet alterations

Abstract

Bioretention systems are effective stormwater treatment systems. Many studies have been conducted on the nitrogen regulation effects of bioretention systems. However, few studies had investigated the variation of nitrogen concentration in the submerged zone during dry periods and evaluated the effect of the submerged zone on nitrogen removal efficiency under dry-wet alterations. Furthermore, existing models cannot fully consider the impact of the submerged zone on nitrogen removal. This paper presents the validation, prediction uncertainty, parameter sensitivity analysis, and scenario simulations with different submerged zone depths of a proposed model. We conducted 13 consecutive rainfall events for model validation under dry-wet alternations in a mesocosm bioretention system with a submerged zone. The collected data mainly includes outflow rates, outflow nitrogen concentrations (NH4-N, NO3-N, and TN) in wet periods, and submerged zone nitrogen concentrations (NH4-N, NO3-N, and TN) in dry periods. The results show that (1) the Nash efficiency coefficients in the wet and dry periods were all above 0.5 during calibration and validation phases; (2) the prediction uncertainty of outflow/submerged zone nitrogen concentration (NH4-N, NO3-N, and TN) can be significantly reduced by using NSE values of the wet and dry periods as likelihood objective functions than only using NSE values of the wet periods; (3) the sensitivity analysis shows that outflow nitrogen concentrations (NH4-N, NO3-N, or TN) are sensitive to most of nitrogen module parameters in the submerged zone; (4) in the long-term dry-wet alterations simulation, the increase of submerged zone depth can improve the annual average nitrogen removal rate, while it does not always improve nitrogen removal rates in single rainfall event. If a rainfall event with a short antecedent dry period (<1.5 d) and the inflow nitrogen concentration in the current event is less than that in the previous event, the increase of the submerged zone depth will even reduce the nitrogen removal rate.