Abstract
Urban ecosystems are vulnerable to extreme weather events due to environmental degradation induced by human activities. Massive deforestation and build-up of impervious areas in urban areas block infiltration of precipitation which now becomes surface runoff in urban watersheds. The rapid release of huge quantities of water after storms not only cause erosion of stream channels and floodplains but also dimmish a series of ecosystem functions (e.g., denitrification, plant nitrogen uptake, etc.) that serve as important roles to mitigate nutrient sources in uplands released to streams. Consequently, urban watersheds are suffering from many environmental issues locally, such as flash floods and excessive in-stream nutrient loads and sediment, and these upstream degradations propagate to downstream and coastal waterbodies and cause eutrophication and elevated levels of stream bed, threatening not only ecosystem health but also human lives there. To address these water-related issues, various efforts (e.g., low-impact development or LID) are introduced to manage water quality, restore the pre-urbanization flow regime, and mitigate risks from previous destructions in urban watersheds, receiving attentions of urban water managers these days. However, the effectiveness of these ecosystem restorations is heterogeneous among watersheds. Quantifying local outcomes and improvements could not been done without universally applied and reliable analytical frameworks.
Therefore, this dissertation focuses on building frameworks for analyzing and quantifying current states and upcoming improvements brought by ecosystem restorations in urban watersheds. Specifically, these frameworks allow urban managers to project nitrate reduction from stream restoration projects and associated socio-economic benefits for every 1,000-ft stream reaches in Baltimore metro areas (Chapter 2), quantify the possible changes in streamflow and upland ecological responses from each scenario of LID implementations in Scotts Level Branch, Baltimore (Chapter 3), and include human-induced nitrogen load from fertilization and septic wastewater to improve water chemistry analysis using our hydro-ecological model, RHESSys, in Baisman Run, Baltimore (Chapter 4).
Overall, the results of each chapter in this dissertation show a synthetic theme: Shifting ecohydrological conditions of urban watersheds back to pre-urbanization conditions is unattainable, no matter how massively ecological restoration practices are implemented. Human-induced land cover changes and nutrient inputs permanently alter hydrological flow regimes and nutrient cycles of urban ecosystems. Relying solely on green infrastructures would be insufficient to reverse current issues and, in some cases, may even exacerbate them. Coupling of grey and green infrastructures with regulating nutrient inputs to urban ecosystems is essential for future management practices and protection of local and downstream aquatic ecosystem health. The statistical and process-based models in this dissertation provide valuable and easy-to-use tools for decision makers to plan possible restoration scenarios spatially and evaluate corresponding responses of upland and streams in urban ecosystems systematically.
