Modeling Streamflow and Nitrogen Dynamics in an Urban Watershed: A Mechanistic Approach

Abstract

This thesis develops and tests a process based model of streamflow and nitrogen concentration in the urban stream network of the Bassett Creek watershed. The model treats the watershed as a control volume. This allows for a simplified but physically grounded representation of nitrogen inputs, transformations, and losses. Streamflow is used as the primary control variable. Hydrologic variability is modeled as a key factor shaping nitrogen export over time. This model uses a forward Euler method for two differential equations of streamflow and nitrogen dynamics. These equations are implemented manually in MATLAB to emphasize clarity and sensitivity to parameter changes. The model is calibrated using long term data on rainfall, streamflow, and nitrogen concentrations. Sensitivity analyses are used to identify which parameters most strongly influence outcomes. Results show that the streamflow model captures general runoff behavior with moderate accuracy. The nitrogen model reveals useful qualitative trends but struggles with predictive power due to limited data and the complexity of underlying processes. By grounding the model in first principles and real world observations, this thesis offers an accessible alternative to black box models. It is particularly well suited for environments with limited data and exploratory research in urban watershed systems.