When urbanization occurs, the removal of vegetation, compaction of soil and construction of impervious surfaces—roofs, asphalt, and concrete—and drainage infrastructure result in drastic changes to the natural hydrological cycle. Stormwater runoff occurs when rain does not infiltrate into soil. Instead it ponds at the surface and forms shallow channels of overland flow. The result is increased peak flows and pollutant loads, eroded streambanks, and decreased biodiversity in aquatic habitat. In urban areas, runoff is typically directed into catch basins and underground pipe systems to prevent flooding, however such systems are also failing to meet modern environmental goals. Green infrastructure is the widely evocative idea that development practices and stormwater management infrastructure can do better to mimic the natural hydrological conditions through distributed vegetation and source control measures that prevent runoff from being produced in the first place. This dissertation uses statistics and high-resolution, coupled surfacesubsurface hydrologic simulation (ParFlow.CLM) to examine three understudied aspects of green infrastructure planning. First, I examine how development characteristics affect the runoff response in urban catchments. I find that instead of focusing on site imperviousness, planners should aim to preserve the ecosystem functions of infiltration and evapotranspiration that are lost even with low density development. Second, I look at how the spatial configuration of green infrastructure at the neighborhood scale affects runoff generation. While spatial configuration of green infrastructure does result in statistically significant differences in performance, such differences are not likely to be detectable above noise levels present in empirical monitoring data. In this study, there was no evidence of reduced hydrological effectiveness for green infrastructure located at sag points in the topography. Lastly, using six years of empirical data from a voluntary residential green infrastructure program, I show how the spread of green infrastructure depends on the demographic and physical characteristics of neighborhoods as well as spatially-dependent social processes (such as the spread of information). This dissertation advances the science of green infrastructure planning at multiple scales and in multiple sectors to improve the practice of urban water resource management and sustainable development. / Doctor of Philosophy in City and Regional Planning

Master of Landscape Architecture / Department of Landscape Architecture/Regional and Community Planning / William P. Winslow III / Sports venues require large amounts of parking to facilitate the number of spectators attending an event. The parking, mostly surface, is underutilized when compared to traditional parking lots accommodating daily use. Large parking surfaces alter the natural hydrological cycle by generating large volumes of runoff. Over time, pollutants build up on a parking surface and are discharged into a stormwater drainage system during a rain event. The intent of the research is to investigate the use of Best Management Practices that ultimately reduce the pollutant loads created by stormwater runoff while creating amenities for spectators that could potentially generate revenue. The focus of the study will be on a 400 acre sports venue in Kansas City, Missouri, the Truman Sports Complex. To understand retrofitting stormwater management practices, one must understand how large parking lots are constructed and understand successful examples. Through the use of archival research, interviews and analysis of two parking lot case studies, Northgate Mall in Seattle, Washington and US Cellular Field in Chicago, Illinois, the research analyzed how retrofitting design solutions are utilized to reduce stormwater pollutant loads. Each case study documents a distinct type of retrofitting strategy; bioswales at Northgate Mall and permeable concrete pavers at US Cellular Field. Using the Design Point Method developed by the Center for Watershed Protection, the research analyzes each case study retrofit design solution - conceptually and post construction. The Design Point Method allowed the research to measure the success of retrofitting strategies and informed the research to as to how the strategies could be implemented at the Truman Sports Complex. The conclusion of the project is a retrofit design solution of a surface parking lot at the Truman Sports Complex. Using the Design Point Method as an analysis tool, the final study provides compelling evidence that retrofitting existing surface parking lots at sports venues using Best Management Practices provides a sustainable solution to reducing pollutant loads while creating the potential for enhancing the tailgate experience for the sports fan.

The potential real-time control (RTC) has to improve the performance of existing stormwater management systems is a topic of increasing interest as hydraulic and hydrologic modeling capabilities proliferate. The benefits of incorporating precipitation forecast data into a RTC algorithm to allow for prediction-based control of an urban watershed is explored using an EPA SWMM 5.1 watershed model. One reactive and two predictive RTC algorithms are simulated in various configurations across seven dry detention ponds located in the 162 hectare urbanized watershed. The hydraulic benefits they provide at the site and watershed outlet in regards to peak flow and the flow duration curve are compared to conventional, static control. The ponds retrofit with the novel predictive RTC algorithm had lower peak flows during 24-hour design storms more consistently than when retrofit with reactive RTC. The duration of erosive flows at the site level was decreased by the novel predictive RTC in most cases. Improvements at the watershed outlet depended on where RTC was applied as hydrograph compounding was observed during some RTC implementations. / Master of Science / The consequences of watershed urbanization on nearby waterways has become a more relevant concern as urbanization increases and climate change continues to develop. Conventional stormwater management practices are employed to control peak flows from urbanized drainage areas for certain design storm criteria. Real-time control (RTC) technology has the potential to enable existing stormwater facilities to improve their performance during storm events different from their design conditions. This study compares the performance of several reactive and predictive rule-based RTC algorithms simulated as retrofits on seven dry detention ponds in a 162 hectare urbanized watershed. The results indicate that RTC algorithms that use rainfall forecast data for predictive decision making have the most potential to reduce stream erosion when applied appropriately throughout the watershed.

Residential infill development and the associated increases in impervious cover and stormwater runoff have the potential to overwhelm aging infrastructure causing erosion, flooding, and ecological degradation of waterways. The lack of greenfield sites available for new development around urban centers coupled with a desire for maintaining a traditional neighborhood aesthetic drives residential infill and replacement of existing smaller homes by significantly larger structures, minimizing open space available for stormwater mitigation. An analysis of residential infill development trends in Fairfax County was performed to characterize the effects of intensified development. Development patterns were identified and assessed using digitized impervious cover data extracted from aerial imagery of the study area for six selected years from 2002 to 2015. The average annual increase in percent impervious area was computed to be 0.3%. Redevelopment was found to be mainly occurring in areas zoned as lower-density residential. The spatial distribution of redevelopment showed trends of clustered increases as opposed to isolated events. The hydrologic analysis performed revealed that from 2002 to 2015, there was an 8,930 m3 (7.2 acre-ft) increase in stormwater runoff volume, 38% of which is contributed to by runoff generated from development that is considered exempted from regulation in Fairfax County. Additionally, there was a 28 kg/year (62 lb/year) increase in total phosphorus loads attributable to the change in impervious cover due to residential development throughout the study area. The results of this study provide a quantitative basis for municipalities to amend policies regulating residential development and its associated stormwater management. / Master of Science / Residential infill and redevelopment is contributing to significant increases in impervious cover in suburban areas, resulting in increased stormwater runoff. This has the potential to overwhelm aging infrastructure causing erosion, flooding, and ecological degradation of waterways. The lack of greenfield sites available for new development around urban centers coupled with a desire for maintaining a traditional neighborhood aesthetic drives residential infill and replacement of existing smaller homes by significantly larger structures, minimizing open space available for stormwater mitigation. An analysis of residential infill development trends in a watershed within Fairfax County was performed to characterize the effects of intensified development. Development patterns were identified and assessed using digitized impervious cover data extracted from aerial imagery of the study area for six selected years from 2002 to 2015. The average annual increase in impervious area was evaluated against metrics such as lot size and location to determine historical trends and predict future rates of infill development. A hydrologic analysis was conducted to quantify the increase in stormwater runoff and associated quality of runoff impacts attributable to the change in impervious cover due to residential development throughout the study area. The results of this study provide a quantitative basis for municipalities to amend policies regulating residential development and its associated stormwater management.

Master of Science / Department of Biological & Agricultural Engineering / Stacy L. Hutchinson / To understand the region-specific effects of biota on function of bioretention cells, a lysimeter study was conducted at Kansas State University to determine how earthworms and native Kansas grasses impact runoff treatment and hydraulic function of a bioretention cell. This study also employed the Comprehensive Bioretention Cell (BRC) model to demonstrate how three seasons of growth could impact bioretention cell function. The model results of the first season of growth were then compared to field data. Results indicate that the interaction of plant roots and soil macrofauna over one growing season improved several aspects of bioretention cell function. The greatest increase in saturated hydraulic conductivity was in the treatment that included both plants and macrofauna. The presence of vegetation reduced ponding effects and increased water storage. Earthworm treatments had a lesser ability to store water. All treatments were effective in reducing the concentration of P in effluent. A large amount of N was released during all events from all treatments probably because of a high initial N content of the bioretention media. No treatment performed significantly better in improving water quality, indicating that macropore flow in the earthworm treatments did not induce a higher rate of pollutant transport.

Urbanization, which is ever increasing on a global scale, can negatively affect wildlife and habitat as well as biotic realms including air, soil, and water. One of the impacts of urbanization is urban runoff. A method commonly employed to mitigate runoff is the construction of stormwater management ponds (SWMPs). These ponds have created new habitat for urban wildlife, especially waterfowl such as Canada geese and mallard ducks. In some municipalities, increased numbers and densities of waterfowl near SWMPs are considered a nuisance due to their large numbers, considerable amount of fecal matter and because they may create health hazards. This research specifically examines the influence of species composition and height of vegetation around SWMPs on waterfowl use and whether changes in Escherichia coli counts are attributed to waterfowl use. Ponds were visited 140 times between April 2005 and June 2006. At each site, the number of Canada geese, mallard ducks, and other waterfowl species was recorded. Geese significantly preferred ponds with short vegetation (F = 53.45, p < 0.0001) and ducks exhibited no preference (F= 2.17, p = 0.347). The time (day) that observations were made was a factor indicating that there were temporal variations among seasons; geese were slightly more affected by time (F = 16.08, p < 0.0001) than ducks (F = 11.18, p < 0.001). This was not surprising given that waterfowl generally migrate locally across seasons. This result also supported the hypothesis that geese respond to changes in vegetation height. The influence of time coincides with the development of vegetation between spring and summer. Geese moved to the ponds with short vegetation as the growing season progressed, whereas they tended to avoid ponds with more naturalized vegetation due to the increased height. From a management perspective, municipalities can discourage nuisance geese at SWMPs through the naturalization of ponds and decreasing the level of maintenance via less mowing. There was no significant correlation between waterfowl use of ponds and E. coli counts. The Pearson’s correlation ranged from -0.152 to 0.990, associated p values ranged from 0.07 to 0.981. While further study is required, it appears that waterfowl do not offer any noticeable addition of E. coli to stormwater management ponds beyond what is already in the water column from other sources. Therefore, waterfowl may not present a health threat, at least within the pond itself.

Urbanization, which is ever increasing on a global scale, can negatively affect wildlife and habitat as well as biotic realms including air, soil, and water. One of the impacts of urbanization is urban runoff. A method commonly employed to mitigate runoff is the construction of stormwater management ponds (SWMPs). These ponds have created new habitat for urban wildlife, especially waterfowl such as Canada geese and mallard ducks. In some municipalities, increased numbers and densities of waterfowl near SWMPs are considered a nuisance due to their large numbers, considerable amount of fecal matter and because they may create health hazards. This research specifically examines the influence of species composition and height of vegetation around SWMPs on waterfowl use and whether changes in Escherichia coli counts are attributed to waterfowl use. Ponds were visited 140 times between April 2005 and June 2006. At each site, the number of Canada geese, mallard ducks, and other waterfowl species was recorded. Geese significantly preferred ponds with short vegetation (F = 53.45, p < 0.0001) and ducks exhibited no preference (F= 2.17, p = 0.347). The time (day) that observations were made was a factor indicating that there were temporal variations among seasons; geese were slightly more affected by time (F = 16.08, p < 0.0001) than ducks (F = 11.18, p < 0.001). This was not surprising given that waterfowl generally migrate locally across seasons. This result also supported the hypothesis that geese respond to changes in vegetation height. The influence of time coincides with the development of vegetation between spring and summer. Geese moved to the ponds with short vegetation as the growing season progressed, whereas they tended to avoid ponds with more naturalized vegetation due to the increased height. From a management perspective, municipalities can discourage nuisance geese at SWMPs through the naturalization of ponds and decreasing the level of maintenance via less mowing. There was no significant correlation between waterfowl use of ponds and E. coli counts. The Pearson’s correlation ranged from -0.152 to 0.990, associated p values ranged from 0.07 to 0.981. While further study is required, it appears that waterfowl do not offer any noticeable addition of E. coli to stormwater management ponds beyond what is already in the water column from other sources. Therefore, waterfowl may not present a health threat, at least within the pond itself.

Virginia's new stormwater regulations involve the use of the Runoff Reduction Method (RRM), which requires the product of the peak flow rate and runoff volume (Q*RV) from the one-year storm event in the post-development condition to be reduced to eighty percent of the pre-development Q*RV to protect against channel erosion. This study models different bioretention cell sizes in a developed watershed in Blacksburg, Virginia to determine the "performance" at both the sub-watershed and watershed levels. In addition, models of "optimal" bioretention cells sized to meet the RRM for each sub-watershed are evaluated. A direct relationship is determined between the size of the cell required to meet the RRM and the sub-watershed's Natural Resources Conservation Service (NRCS) curve number. However, the required size for some of the cells is much larger than those typically seen. With the RRM applied for all of the sub-watersheds, the resulting hydrograph at the watershed outlet has a lower peak than the pre-development condition. / Master of Science

This study uses the Natural Resources Conservation Service (NRCS) TR-55 method to quantify the increase in stormwater runoff volume from infill residential redevelopment, or mansionization, in a 34-acre residential subwatershed of Fairfax County, Virginia. Analysis of 10 redeveloped lots in the subwatershed showed an average increase in impervious cover from 8% to 28% after redevelopment, resulting in an average increase in runoff volume of 18% for the 10-year, 24-hour storm. From 1997 to 2009, the total impervious cover in the subwatershed increased from 18% to 25%, resulting in a calculated 6% increase in runoff volume. Low Impact Development (LID) techniques were modeled as retrofits in the subwatershed to mitigate the increase in runoff volume. Measures modeled include bioretention basins, infiltration trenches, amended soils, permeable pavement, and cisterns. Results indicate that placing bioretention basins or infiltration trenches on 0.5% of the subwatershed or amending 20% of the open space with soil composts would reduce the runoff volume back to the 1997 quantity for the 1-year, 24-hour storm. / Master of Science

While many studies have evaluated the hydrologic effects of bioretention at the site level, few have investigated the role bioretention plays when distributed throughout a watershed. This study aims to assess bioretention's effects on an urbanized watershed using two modeled scenarios: one where runoff from many land uses was routed through the practice, and another in which only runoff from large impervious areas was routed. Peak flows, volumes, and lag times from these models were compared to the watershed's current and predeveloped conditions. Both scenarios provided reductions in peak flows with respect to existing conditions for modeled storm events, sometimes to levels below the predeveloped condition. Neither case was able to reduce volumes to predevelopment levels; the option to treat impervious areas had a negligible effect on runoff volume. Both cases were able to extend lag times from the existing development condition. Based on these results, bioretention appears to have the capability to improve watershed hydrologic characteristics. Furthermore, only treating impervious areas could be a viable alternative when funds or space are limiting factors. / Master of Science