Date of Award

Spring 4-15-2026

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Civil, Construction, and Environmental Engineering

First Advisor

Anthony Parolari

Second Advisor

Isabelle Horvath

Third Advisor

Walter McDonald

Abstract

Urbanization of previously natural land cover has spurred the innovation and implementation of green stormwater infrastructure (GSI) to combat increased surface runoff due to impermeable areas and a changing climate. Bioretention has been a popular choice in design due to its small footprint and proven effectiveness in retaining stormwater runoff and limiting discharges to receiving surface waters. However, their event-to-event performance variability is difficult to explain due to a number of confounding and interacting influences that distort cross-site analyses including inter-event drying and antecedent soil moisture conditions, geometric size and media properties, and event magnitude. This study applies and evaluates a physically-based retention model, then reformulates the contributing variables into a set of parameters without dimensions integrating event size and system design. Dimensionless variables eliminate scale dependence, enabling direct comparison and generalization across bioretention systems with differing physical characteristics. Frequency matching was also evaluated as a method to reduce the noise in retention data as result of inter- and intra-event characteristics and its use in design storm modeling. Frequency matching proved to be useful in removing stochastic noise from the analysis of bioretention systems and emphasized the applicability to distribution-based bioretention design. The dimensionless approach reduced the description of retention performance to a smaller set of governing parameters linking event size to system properties. Models using normalized rainfall and inflow volume outperformed their dimensional counterparts for conventionally-under drained systems, exhibiting lower residual standard error and AIC, and enabling improved general performance evaluation across sites. The emergence of the dimensionless characteristic surface length relative to the available storage depth ratio further suggests that lateral flow processes influence retention, challenging the common assumption that spatially uniform flow is an adequate description of bioretention behavior. Additionally, the native soil properties were identified as a primary control on system performance yet are often unmeasured or unreported in practice. These findings highlight the need for full system understanding and the power of incorporating dimensionless process and event descriptors into the evaluation of bioretention systems.

COinS