Stormwater (SW) runoff in urban areas represents a major pathway for pollutant transfer to receiving waters. Best management practices (BMP) were introduced in the 1970s to help mitigate the negative effects of SW. In the 1990s, Stormwater management (SWM) ponds were established as a BMP to help increase the water quality of SW effluent. Many SWM ponds do not provide sufficient water quality treatment. Information on the internal processes influencing the reduction of total phosphorus (TP), soluble reactive phosphorus (SRP) and total suspended solid (TSS) concentrations in SWM ponds with different designs is lacking. Knowledge of the processes affecting TP, SRP and TSS retention can help improve the design of SWM ponds to enhance their treatment performance. The purpose of this thesis is to provide an assessment of the internal chemical processes that affect the trap efficiency (TE) and spatial and temporal variability of TP, SRP and TSS concentrations at two structurally different SWM ponds (Pond 45; conventional and Pond 33; hybrid extended detention) in Waterloo, ON. Water samples were collected at the inflow and outflow at the two SWM ponds during six storm events and 30 baseflow periods. A mass balance approach was used to quantify the TE of TP, SRP and TSS concentrations at each pond. Pond 33 had a TE of 24.3%, 26.7% and 66.8% for baseflow and stormflow samples of TP, SRP and TSS. Pond 45 performed much better with TE of 93.8%, 94.2% and 98% for TP, SRP and TSS concentrations. Pond 33 was a source of TP, SRP and TSS for 3, 4 and 2 storm events sampled during the field season, respectively. Pond 45 was a sink for all parameters on all storm events samples. The spatial and temporal variability of TP, SRP and TSS concentrations were examined to improve knowledge of external factors and internal processes that influence the TE of SWM ponds. The effects of storm magnitude, seasonality and vegetation growth and senescence on effluent water quality were investigated. Additionally, the role of sediment on P cycling in the ponds was evaluated by determining grain size distribution, porewater SRP concentrations, sediment geochemistry and mineralogy, and the sediment P buffering capacity. Vegetation senescence, anoxic conditions, porewater SRP concentrations, sediment characteristics and buffering capacity influenced the poor TE at Pond 33. Pond 45 had more favourable water column conditions, i.e. higher dissolved oxygen concentrations, therefore allowed greater amounts of P to adsorb onto sediment. Design and maintenance considerations are described to help improve the performance at Pond 33. Continual water quality monitoring of SW effluent will identify changes in quality and mitigation measures can be implemented to increase a SWM ponds performance.

Stormwater impoundments are one of many types of best management practices (BMP) designed and implemented to regulate water quantity and improve the quality of runoff from urban areas. Studies of water quality in urban impoundments have indicated that conventional designs are however, not very effective at removing solids and associated pollutants. Accordingly, many urban impoundments are being re-designed to improve downstream water quality. However, few studies have systematically monitored and quantified post-design water quality improvements of urban impoundments. This thesis examines changes in the water quality performance of an urban impoundment (Columbia Lake) in Waterloo, Ontario resulting from redesign of the lake for the pre-design period (2003 and 2004) and the post-design period (2006 and 2007). To achieve this goal, four years of water quality data collected at the inlet and outlet of Columbia Lake as part of the Laurel Creek Monitoring Program was measured. Water chemistry parameters included total phosphorus (TP), soluble reactive phosphorus (SRP), suspended solids (SS), dissolved oxygen (DO), pH and total dissolved solids (TDS). Inlet and outlet discharge (Q) were measured to determine the water retention time in the lake. Concentrations and loads of TP and SS for the post-design period (2006 and 2007) were compared to those for the pre-design period (2003 and 2004). During the pre-design period (2003 and 2004), inflow TP concentrations ranged from 18 to 372 µg L-1 with an average (mean ± standard error) of 56±7 µg L-1, while outflow TP concentrations ranged from 37 to 266 µg L-1 with an average of 116±6 µg L-1. Post-design TP concentrations ranged from 10 to 124 µg L-1 with an average of 53±5 µg L-1 and from 14 to 147 µg L-1 with an average of 44±3 µg L-1 at the inflow and outflow, respectively. Pre-design SS concentrations ranged from 1.8 to 168.5 mg L-1 with a mean of 19.0±3.2 mg L-1 and from 4.0 to 194.7 mg L-1 with a mean of 66.6±4.7 mg L-1 at the inflow and outflow, respectively. Post-design SS concentrations varied from < 0.1 to 25.8 mg L-1 with an average of 8.5±0.8 mg L-1 and from < 0.1 to 42.5 mg L-1 with an average of 14.5±0.8 mg L-1 at the inflow and outflow, respectively. Sedimentation/resuspension dominated the TP and SS transfer via Columbia Lake. Pre-design TP loads (log-transformed) strongly correlated with SS loads at the inflow and outflow (r = 0.661 and 0.777, p = 0.0001). These parameters were more strongly correlated during the post-design period (r = 0.794 and 0.915, r = 0.0001), which indicates that particulate P (PP) was a dominant fraction of TP and that the release of dissolved phosphorus (DP) from bottom sediments was considerably decreased following the redesign. No significant difference was observed between inflow and outflow SRP concentrations. Discharge strongly affected TP and SS loads at the inflow and outflow during the pre- and post-design periods (r > 0.79, p = 0.000 for all). After the redesign of Columbia Lake, the average net internal P loading rate decreased from 198% to 22% for TP. The primary factor influencing the observed decreased post-design TP and SS outputs was the removal of sediment from the lake. Bottom sediment removal and changes to the lake bathymetry reduced sediment resuspension and P desorption, which decreased the average net internal SS loading rate from 828% to 154%. The Columbia Lake Water Quality Model developed by Stantec Consulting Ltd. (2004) underestimated the post-design outflow TP and SS concentrations mainly because it did not include terms that account for factors such as bioturbation, wave induced resuspension and biological activity.

Stormwater (SW) runoff in urban areas represents a major pathway for pollutant transfer to receiving waters. Best management practices (BMP) were introduced in the 1970s to help mitigate the negative effects of SW. In the 1990s, Stormwater management (SWM) ponds were established as a BMP to help increase the water quality of SW effluent. Many SWM ponds do not provide sufficient water quality treatment. Information on the internal processes influencing the reduction of total phosphorus (TP), soluble reactive phosphorus (SRP) and total suspended solid (TSS) concentrations in SWM ponds with different designs is lacking. Knowledge of the processes affecting TP, SRP and TSS retention can help improve the design of SWM ponds to enhance their treatment performance. The purpose of this thesis is to provide an assessment of the internal chemical processes that affect the trap efficiency (TE) and spatial and temporal variability of TP, SRP and TSS concentrations at two structurally different SWM ponds (Pond 45; conventional and Pond 33; hybrid extended detention) in Waterloo, ON. Water samples were collected at the inflow and outflow at the two SWM ponds during six storm events and 30 baseflow periods. A mass balance approach was used to quantify the TE of TP, SRP and TSS concentrations at each pond. Pond 33 had a TE of 24.3%, 26.7% and 66.8% for baseflow and stormflow samples of TP, SRP and TSS. Pond 45 performed much better with TE of 93.8%, 94.2% and 98% for TP, SRP and TSS concentrations. Pond 33 was a source of TP, SRP and TSS for 3, 4 and 2 storm events sampled during the field season, respectively. Pond 45 was a sink for all parameters on all storm events samples. The spatial and temporal variability of TP, SRP and TSS concentrations were examined to improve knowledge of external factors and internal processes that influence the TE of SWM ponds. The effects of storm magnitude, seasonality and vegetation growth and senescence on effluent water quality were investigated. Additionally, the role of sediment on P cycling in the ponds was evaluated by determining grain size distribution, porewater SRP concentrations, sediment geochemistry and mineralogy, and the sediment P buffering capacity. Vegetation senescence, anoxic conditions, porewater SRP concentrations, sediment characteristics and buffering capacity influenced the poor TE at Pond 33. Pond 45 had more favourable water column conditions, i.e. higher dissolved oxygen concentrations, therefore allowed greater amounts of P to adsorb onto sediment. Design and maintenance considerations are described to help improve the performance at Pond 33. Continual water quality monitoring of SW effluent will identify changes in quality and mitigation measures can be implemented to increase a SWM ponds performance.

This thesis shows the effect social learning has on various stakeholders involved in a project aimed at tackling a stormwater challenge in the city of Uppsala in Sweden and if social learning is a useful tool to address such an issue. Due to the onset of climate change societies are having to deal with increasingly complex issues. Finding sustainable answers to these challenges is proving difficult so alternative methods such innovation competitions much like the one studied in this paper are becoming attractive alternatives to conventional climate change mitigations approaches. By using an active participation method, this thesis attempts to study whether or not social learning is taking place in the innovation competition and if it is having an impact on the innovation competition. The study found that social learning is in fact taking place during the meetings which were attended but whether or not the resulting knowledge created as a result of the social learning was being used to the advantage of the stakeholders was compounded by communication issues outside of the project meetings.

Stormwater runoff generated from urban areas can be a source of contamination and may negatively impact receiving waters. Best management practices, including the use of treatment wetlands, are recommended to minimize impacts and maintain the quality of water bodies receiving stormwater discharge. This study focuses on the viability of a natural wetland in the treatment of urban runoff. Kuhn Marsh is a natural urban wetland located in Dartmouth, NS. The wetland is approximately 2 ha in size and the primary inlet is a stormwater outfall servicing a 28 ha urban drainage area. Kuhn Marsh has been receiving stormwater generated from the urban drainage area for decades. A wetland drainage area of approximately 9 ha contributes to surface runoff downstream of the wetland inlet. Project objectives are defined as: (i) characterization of the hydrology and hydraulics of the wetland system, (ii) characterization of contaminant fluxes within the wetland system, and (iii) analysis of the treatment performance of Kuhn Marsh. Research strategies used to achieve project objectives include physical and hydrologic characterization of the wetland and contributing watersheds as well as surface and ground water quality analysis. Monitoring was conducted in the wetland during both baseflow and stormflow conditions from May 2011 through October 2012, with the exception of November 2011 to January 2012. Surface water samples were analyzed in the laboratory for TSS, TOC, TN, TP, turbidity, E.coli, and a suite of heavy metals including Fe, Pb, Cu, Cd and Zn. In-situ surface water monitoring included DO, temperature, conductivity and pH. Groundwater samples were analyzed for E.coli and microbial source tracking was performed on all well samples in addition to samples from the inlet and outlet of the wetland. Results from the well samples and the wetland outlet were inconclusive, however the wetland inlet showed human source bacteria indicating potential sewer cross connections within the stormwater system. It was determined that the wetland is an area of groundwater discharge, with groundwater accounting for an average of 50% of the volume discharging through the outlet control structure. Largely due to groundwater influence, Kuhn Marsh shows no peak flow dampening or volume reduction between inlet and outlet. Minimal hydraulic retention times, between 2 and 4 hours, were calculated during stormflow conditions, indicating potential short circuiting of flows through the wetland. Wetland treatment performance was analyzed on a concentration and mass reduction basis and on the number samples that exceeded parameter guidelines at the outlet of the wetland. Guideline exceedances were reported for the majority of samples taken and increases in concentration between inlet and outlet resulted in a larger number of samples exceeding guidelines at the outlet. Despite dilution from groundwater discharge, minimal to no concentration reduction was reported between the inlet and outlet of the wetland. Mass reduction did not occur between the inlet and outlet and Kuhn Marsh was found to be a source of all contaminants sampled. Results of this study show that Kuhn Marsh is no longer acting as a reservoir for stormwater contaminants and, based on the fact that the wetland has been receiving stormwater input on the order of decades, study results may be indicative of the long-term treatment capacity of a stormwater treatment wetland. In the future, comprehensive sampling of groundwater is recommended to determine if contaminants are entering the wetland via groundwater discharge, and if possible, surface water sampling should be conducted on a finer scale to better estimate mass fluxes and contaminant loading rates.

Impervious surfaces negatively affect urban hydrology by altering the depth, frequency and seasonal distribution of stormwater runoff. To assess the imperviousness of green roofs, a mathematical model was developed to simulate the stormwater response of a hypothetical green roof. The model is based on the physical processes that affect the green roof stormwater response and uses historic climate data. The results show that green roof imperviousness fluctuated according to climate conditions and precipitation sequence. Only 29% of the total precipitation received by the green roof resulted in runoff, however, the response varied substantially when evaluated at a daily interval. Runoff was eliminated during 82% of days with rain and a higher proportion of runoff disturbances were eliminated during the spring and summer compared to the fall. In comparison to an impervious surface, the green roof showed a reduction in the depth and frequency of runoff thereby improving urban hydrology.

Low impact development (LID), the practice of preserving and restoring natural water cycles in urban development, is considered the next step in stormwater management. Policy and governance play a strong role in the adoption of LID. There has been progress in implementing LID in the Greater Toronto Area but less progress in Southern Ontario in general. This research identifies barriers and opportunities to implementing LID in the context of policy and governance in Southern Ontario. The barriers, opportunities, and policy are identified using a focused literature review, then verified and further explored through key informant interviews. Data are synthesised to produce an Enhanced Governance Model (EGM) for implementing LID. The EGM is evaluated by key informants and further refined. Findings show that public education and provincial-level standards are fundamental to widespread adoption of LID. Five opportunities for jurisdictional integration in stormwater governance are identified and their implications discussed.

Current stormwater management practices (SMP) are not sufficient for maintaining predevelopment runoff volumes. Low impact development (LID) uses site scale SMP to reduce runoff. Bioretention cells, one practice within LID, are small planting beds designed to filter and infiltrate runoff using amended soil and vegetation. The bioretention cell can create a harsh soil moisture regime for plants that has not been adequately characterized. Bioretention cell construction, meteorological, and soil science data were built into the Happy Plant Model to determine how often bioretention plants were saturated and experienced water stress over a thirty year period. The model takes into account eight design factors: soil media depth and texture, gravel storage, ponding depth, drainage area, in situ soil infiltration rate, the landscape coefficient, and root zone depth. The Happy Plant model will aid future studies and landscape architecture practitioners with bioretention plant selection.

Research on green roofs has focused on grasses, sedums, and forbs. The aims of this thesis were to determine the potential of rooftop gardens (RTGs) in an urban setting to reduce local levels of CO2, remediate storm water runoff, and provide boutique vegetables for a restaurant. The garden roof footprint was 238 ft2, with 14% covered by vegetated boxes. The soil mixture used had 96% absorbency with 54.12 gallons of the 55 gallons of precipitation that fell within the rain catcher boxes absorbed. Total biomass production was 37.98 Kg of wet biomass and 5.04 Kg of dry biomass. The amount of CO2 removed equals 0.22 Kg ft-2. RTGs have a limited capacity to help sequester CO2, but retain precipitation in amounts similar to green roofs. The restaurant was provided with 4.7 Kg (wet weight) of produce (several varieties of tomatoes, peppers, and eggplant). These results support the utility of RTGs. / Bayer School of Natural and Environmental Sciences / Environmental Science and Management (ESM) / MS / Thesis

Thesis (M. S.)--Architecture, Georgia Institute of Technology, 2009. / Committee Chair: Dagenhart, Richard; Committee Member: Elliott, Michael; Committee Member: Green, David