Validating Nitrogen Removal Models with Field Bioretention Data

Abstract

Bioretention is a type of stormwater best management practice that can reduce stormwater runoff volume, and with appropriate design, simultaneously reduce nitrogen loads. To meet the water quality potential of bioretention systems, improved simulation models of the transport and transformations of nitrogen are needed. In this work, two versions of a three-nitrogen-pool (3P) model (3P-1 and 3P-m) are applied to simulate observed nitrogen transport through a field bioretention system. The 3P models provide predictions of net export of dissolved organic carbon, dissolved organic nitrogen, and inorganic nitrogen species. The 3P-m model has promising results: (1) relative to the Storm Water Management Model (SWMM), it reduced the prediction error of percent removal of total dissolved nitrogen by up to 10.6% and reduces the scaled root-mean square error of total dissolved nitrogen loads by up to 53.0%; (2) it successfully captured the correlation between impactful environmental factors and event mean concentrations of total dissolved nitrogen and nitrate nitrogen in underdrain effluent; (3) its predictions on denitrification showed good alignment with observed isotope results; and (4) it demonstrated higher stability in simulated time series among calibration results than 3P-1. The 3P-m is expected to assist the design of bioretention system and valve control strategies by providing more accurate predictions on nitrogen removal under the impacts of soil moisture content and temperature. Future research and collaboration opportunities are also proposed in this article.