Evaluating Watershed-Scale Hydrologic Impacts of Green Stormwater Infrastructure Across Three U.S. Urban Climates

Nayeon Kwak; Virginia Smith; Kelly Good

Urban stormwater management increasingly relies on green stormwater infrastructure (GSI) to mitigate runoff and restore hydrologic function in developed watersheds. However, quantifying the influence of GSI at the watershed scale remains difficult, as most existing studies have focused on small subdivisions or isolated installations. This ongoing project seeks to characterize how the hydrologic response of urban watersheds to GSI implementation varies with factors such as climate, geography, and seasonality, and evaluate the potential of gridded precipitation products to enhance the replicability and scalability of urban hydrology assessments. This work develops a transferable, data-driven approach to evaluate the hydrologic impacts of GSI across three U.S. urban watersheds—in Philadelphia, Atlanta, and Austin—representing diverse climates and physiographic settings where GSI has been extensively implemented. To improve the accuracy and reproducibility of rainfall inputs, the study compares gridded precipitation datasets with point gage observations and assess how data source influences hydrologic interpretation. Precipitation data are paired with USGS streamflow records to examine hydrograph features and temporal changes in watershed response over the past decade, during which substantial GSI implementation has occurred in each watershed. Both upstream and downstream subwatersheds are analyzed to explore spatial variability in flow response and potential GSI influence. Findings are expected to provide an approach for consistent, city-scale evaluation of GSI performance and to improve understanding of how distributed green infrastructure collectively shapes urban watershed behavior.

Assessing Active and Passive Sampling Techniques for Emerging Contaminants Using Nested Watershed Monitoring

Henry Kibuye; Tamie Veith; Tyler Groh; Heather Preisendanz

Even at trace levels, emerging contaminants (ECs) in surface water resources pose potential adverse ecological and human health impacts. The assessment of monitoring approaches and sampling methods used in tracking spatial-temporal patterns of ECs is key to designing efficient monitoring networks. In this study, a nested watershed monitoring approach was implemented to quantify 24 ECs over the 2023-2024 growing seasons at five sampling sites within the Halfmoon Creek watershed of the Chesapeake Bay basin. Using grab sampling and polar organic chemical integrative samplers (POCIS), both active and passive stream samples were collected every two weeks to (1) compare the effectiveness of the methods in documenting spatial-temporal patterns of ECs, and (2) examine the relationship between estimated POCIS time-weighted average (TWA) and grab sample concentrations. POCIS sampling detected ECs at equal or higher frequencies than grab sampling did, while grab sampling showed higher seasonal variability in concentrations. Atrazine, simazine, clothianidin, and caffeine were the most frequently detected ECs, present in at least 68% of both grab and POCIS samples. Nested watershed monitoring enabled the identification of hotspot sites, and multivariate analysis indicated that POCIS captured greater differences in EC profiles between sites than grab samples. Neutral ECs at environmental pH showed greater POCIS sorption and had higher estimated POCIS TWA than grab samples concentrations. These results highlight the importance of multiple-site monitoring and utilizing both grab and passive sampling for comprehensive water quality assessment. Additionally, the protonation state of target compounds at environmental pH should be considered when selecting passive samplers.

Hydroanalytics for Watershed-Scale Nutrient Dynamics: Linking Field Monitoring, GIS, and Machine Learning in Central Florida

Xiaofan Xu

Watersheds in Central Florida are characterized by low topographic relief, intensive land management, and strong surface water-groundwater connectivity, creating unique challenges for nitrogen (N) and phosphorus (P) assessment and management. Conventional watershed modeling and statistical approaches often struggle to represent these complexities, particularly under variable hydrologic conditions. This study presents a hydroanalytics framework that integrates field monitoring data, GIS, and machine learning to analyze watershed-scale N and P dynamics in Central Florida.

Long-term hydrologic and water quality monitoring data from multiple managed watersheds were integrated with GIS-derived spatial attributes, including land use, soil properties, drainage infrastructure, and hydrologic connectivity metrics. Machine learning models were developed to predict total nitrogen (TN) and total phosphorus (TP) concentrations and loads. A comparative analysis evaluated nutrient responses across watersheds with contrasting land use composition and management intensity and compared machine learning performance against conventional regression-based models.

Results show that the hydroanalytics framework improved predictive accuracy for both TN and TP. Compared to regression approaches, machine learning models reduced prediction error by approximately 22–35% for TN and 18–31% for TP and achieved coefficients of determination (R²) exceeding 0.70 for TN and 0.65 for TP in several watersheds. Comparative analysis revealed distinct nutrient response behaviors, with agriculturally influenced watersheds exhibiting higher flow-dependent TN transport, while urbanized watersheds showed more event-driven TP responses. Feature importance and sensitivity analyses identified hydrologic connectivity, land use intensity, and antecedent moisture conditions as dominant controls on nutrient variability.

The proposed hydroanalytics approach demonstrates a transferable and interpretable framework for evaluating N and P dynamics in low-relief, hydrologically connected watersheds and supports applications in nutrient mitigation planning, best management practice evaluation, and adaptive watershed management in subtropical regions.