ABSTRACT Title of Dissertation: HIGH-VOLUME RAINFALL IMPACTS AND ADAPTATION IN THE U.S. MID- ATLANTIC UNDER CLIMATE CHANGE AND URBANIZATION Ibraheem Muhammad Pasha Khan, Doctor of Philosophy, 2018 Dissertation directed by: Professor Klaus Hubacek Department of Geographical Sciences Water over-abundance has negative effects on proper functioning of ecosystem services. The increase in heavy precipitation events and hence stormwater quantity, due to climate change and urbanization, is a major flooding concern. These events also affect ecosystem processes leading to soil erosion and sedimentation. This dissertation draws from different disciplines and involves quantification of hydrological extremes, assessment of stormwater management resilience and analysis of impacts on ecosystem services under anticipated future changes in the U.S. Mid- Atlantic region. In Chapter 2 of the dissertation, use of precipitation capture depth and findings of likely increase in heavy precipitation events is relevant to flooding concerns at small watershed scales (~3 km2) and are valuable planning-level information for municipal stormwater management. Estimates developed in this dissertation of changes in water volume and resultant on-site infrastructure costs can help stakeholders and managers in planning for flood mitigation and protection of ecosystem services. In addition, the use of capture depth percentiles such as d85, d90, d95, and d99, have the potential to serve as meaningful hydrologic indicators for stormwater management planning. In Chapter 3, the finding of likely higher erosion rates and sediment yield in the future is a point of concern and relevant for effective land use planning. The approach to estimate representative calibration values for sediment delivery ratio model, at small scale (~3 km2) urban watersheds, is valuable for ungauged sites replacing average or theoretical calibration values. In Chapter 4, the construction of a simple curve number watershed model with reasonably good performance and few input data needs offers a possible flow simulation tool for medium to highly impervious watersheds at small scales. Moreover, the stormwater management pathways along with cost-benefit assessment using green stormwater management practices serve as a first step to determine effectiveness of certain green practices at the watershed scale. It provides insights and help identify future research needs to fill gaps in our understanding of green stormwater management practices and how they affect ecosystem services. HIGH-VOLUME RAINFALL IMPACTS AND ADAPTATION IN THE U.S. MID-ATLANTIC UNDER CLIMATE CHANGE AND URBANIZATION by Ibraheem Muhammad Pasha Khan Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2018 Advisory Committee: Professor Klaus Hubacek, Chair Professor Laixiang Sun Professor Matthew C Hansen Professor Glenn E Moglen Associate Professor Kaye L Brubaker, Dean’s Representative © Copyright by Ibraheem Muhammad Pasha Khan 2018 Dedication In loving memory of my inspiring father (Maj (retd.) Shahzad Pasha Khan), mother (Neelofar Khanam), sister (Saira Shahzad), and uncle (Abid Khan Barki). ii Acknowledgements There is a long list of people who have directly and indirectly helped and guided me to make this research work possible. I sincerely thank everyone. The guidance, support, and encouragement of my advisor Dr. Klaus Hubacek is central to this accomplishment. I like to express my sincere gratitude to Dr. Glenn Moglen for his coaching at several important junctures of my research, Dr. Kaye Brubaker for several brainstorming sessions to improve my work and providing me with the great opportunity to work on GISHydroNXT project which had a reinforcing effect on my research, and Dr. In-Young Yeo for guiding me into the program. I also like to sincerely thank my dissertation committee members Dr. Laixiang Sun and Dr. Matthew Hansen for their insights to improve my work. A special thanks to Kate Majchrzak from Maryland Department of Environment for providing stormwater infrastructure information, John Riverson and Ryan Murphy from Paradigm Environmental for helpful discussion on use of SUSTAIN model, and Dr. Linda Macri for academic writing guidance. Last but not least, my dissertation work would not have been possible without unconditional love, care, and encouragement of my brothers (Jahanzaib Pasha Khan and Akbar Farooq Pasha Khan), sister (Kashifa Shahzad), and nephews (Capt. Usman Khan Barki, Ayat-ur-Rehman Khan Barki, and Omer Khan Barki). iii Table of Contents Dedication ..................................................................................................................... ii Acknowledgements ...................................................................................................... iii Table of Contents ......................................................................................................... iv List of Tables ............................................................................................................... vi List of Figures ............................................................................................................. vii List of Abbreviations ................................................................................................. viii Chapter 1: Introduction ................................................................................................. 1 1. Climate change and urbanization .......................................................................... 1 2. Urban ecosystem services ..................................................................................... 4 3. Urban stormwater management ............................................................................ 9 4. Research questions .............................................................................................. 11 5. Dissertation structure and contribution ............................................................... 12 Chapter 2: Future storm events frequency and runoff ................................................ 15 1. Introduction ......................................................................................................... 15 2. Methods and data ................................................................................................ 19 2.1. Experimental overview ................................................................................ 19 2.2. Geographical area ........................................................................................ 20 2.3. Observed/historic precipitation: event frequency analysis .......................... 21 2.4. Future precipitation ...................................................................................... 22 2.5. Surface runoff .............................................................................................. 26 2.6. Cost estimation............................................................................................. 31 3. Results and discussion ........................................................................................ 32 3.1. Historic/observed capture depth events frequency (1981-2015) ................. 32 3.2. Future capture depth event frequency (2016-2035) ..................................... 34 3.3. Runoff scenarios .......................................................................................... 36 3.4. Cost magnitudes ........................................................................................... 36 4. Conclusions ......................................................................................................... 37 Chapter 3: Soil erosion and sediment yield ................................................................ 40 1. Introduction ......................................................................................................... 40 2. Methods and data ................................................................................................ 44 2.1. Experimental overview ................................................................................ 44 2.2. Soil erosion estimation ................................................................................. 46 2.3. Sediment yield estimation ............................................................................ 48 2.4. Soil conservation valuation .......................................................................... 49 3. Results and discussion ........................................................................................ 50 3.1. Rainfall erosivity .......................................................................................... 50 3.2. Soil erosion rates .......................................................................................... 52 3.3. Sediment yield ............................................................................................. 53 3.4. Soil conservation benefits ............................................................................ 56 4. Conclusions ......................................................................................................... 57 Chapter 4: Stormwater management and ecosystem services .................................... 60 1. Introduction ......................................................................................................... 60 2. Methods and data ................................................................................................ 65 iv 2.1. Experimental overview ................................................................................ 65 2.2. Watershed modeling .................................................................................... 67 2.3. Curve number watershed model (CWM) overview ..................................... 69 2.4. SUSTAIN setup and datasets ....................................................................... 73 2.5. Quantification of ecosystem services........................................................... 76 3. Results and discussion ........................................................................................ 78 3.1. CWM skill test ............................................................................................. 78 3.2. Current portfolio .......................................................................................... 80 3.3. Benefit assessment of green infrastructure .................................................. 83 4. Conclusions ......................................................................................................... 84 Chapter 5: Conclusions .............................................................................................. 87 1. Major findings ..................................................................................................... 87 2. Future research .................................................................................................... 91 Appendix A ................................................................................................................. 93 Appendix B ................................................................................................................. 98 Appendix C ............................................................................................................... 106 Appendix D ............................................................................................................... 134 Bibliography ............................................................................................................. 140 v List of Tables Table 1.1. Classification and functions of urban ecosystem services correspond to regional/local level ........................................................................................................ 8 Table 2.1. Climate models used in this study ............................................................. 24 Table 2.2. Historical and forecasted Curve Number estimates at the county level .... 31 Table 2.3. Capture depths from historic/observed data, by county ............................ 33 Table 2.4. Mann-Kendall test statistical significance for trend in number of events . 34 Table 2.5. Cost percent change relative to current estimates under climate change, and climate change urbanization scenarios........................................................................ 37 Table 3.1. Rainfall erosivity estimates and adjustment factors .................................. 51 Table 3.2. CF values based on CMIP5 climate models .............................................. 53 Table 3.3. Erosive power change for each percent of precipitation (2016_2050 – 1981_2015) ................................................................................................................. 53 Table 3.4. Empirical and mathematical model based SDR values ............................. 54 Table 3.5. Soil conservation cost dollar values from ERS-USDA ............................. 57 Table 4.1. BMP management practices in SUSTAIN ................................................ 69 Table 4.2. USGS gauges for validation ...................................................................... 73 Table 4.3. PBIAS results for study sites ..................................................................... 80 Table A.1. Rainfall depth values for 2-year, 24-hour ................................................. 93 Table A.2. Anderson Lookup Table ........................................................................... 93 Table B.1. Land use conversion codes...................................................................... 102 Table B.2. Total annual soil erosion (2016-2050) [units: tons/watershed] ............... 105 Table C.1. Urban stormwater infrastructure locations .............................................. 132 Table C.2. Montgomery county watershed infrastructure drainage area .................. 133 Table D.1. MD BMP eras cost estimation sheet ....................................................... 134 vi List of Figures Figure 1.1. Developed area trend .................................................................................. 2 Figure 1.2. Total population trend ................................................................................ 2 Figure 2.1. Capture depth concept diagram ................................................................ 18 Figure 2.2. Conceptual diagram for future storm runoff ............................................ 19 Figure 2.3. Study area watersheds with Land Use Land Cover (MDP, 2010) ........... 21 Figure 2.4. Historical storm events count ................................................................... 22 Figure 2.5. Observational (1981-2015) and climate models-based (2016-2035) scaled [CF based] observations event count .......................................................................... 27 Figure 2.6. Historical and future storm events count for Howard county .................. 28 Figure 2.7. Historical and future storm events count for Montgomery county .......... 28 Figure 2.8. Historical and future storm events count for Anne Arundel county ........ 29 Figure 2.9. Historical and future storm events count for Prince George’s county ..... 29 Figure 2.10. Curve Number trend lines at the county level ........................................ 30 Figure 2.11. Runoff under climate and urbanization scenarios [CF based] for 2016- 2035 exceeding d85 of 1981-2015 ............................................................................... 32 Figure 3.1. Percent change in rainfall erosive power of total and d85-99 events .......... 53 Figure 3.2. SDR estimates based on equation 3.6....................................................... 54 Figure 3.3. Total erosion (tons/watershed) ................................................................. 56 Figure 3.4. Percent change in erosion of d85-99 events ................................................ 56 Figure 3.5. Annual soil erosion reduction cost based on climate models ................... 57 Figure 4.1. Curve number watershed model (CWM) conceptual basis ...................... 72 Figure 4.2. Routing architecture for BMPs ................................................................. 74 Figure 4.3. CWM validation using simulated daily flow values ................................ 79 Figure 4.4. Montgomery county watershed Current Portfolio (1981-2015) ............... 81 Figure 4.5. Montgomery county watershed SMPs...................................................... 83 Figure A.1. Curve number map (1973) ....................................................................... 95 Figure A.2. Curve number map (1990) ....................................................................... 96 Figure A.3. Curve number map (1997) ....................................................................... 96 Figure A.4. Curve number map (2002) ....................................................................... 97 Figure A.5. Curve number map (2010) ....................................................................... 97 Figure B.1. Sediment delivery ratio model concept diagram ................................... 104 Figure C.1. SUSTAIN’s multiple scales of application [source: US EPA (2009)] .. 128 Figure C.2. HRU file format ..................................................................................... 128 Figure C.3. CNT-based cost benefits outputs ........................................................... 130 Figure C.4. Urban stormwater infrastructure locations ............................................ 131 Figure D.1. MD BMP Eras (Pre 2010 and Post 2010).............................................. 138 Figure D.2. MD structural investment ...................................................................... 139 vii List of Abbreviations AR5 Assessment Report 5 ATP Adaptation Tipping Point BMP Best Management Practice CBP Chesapeake Bay Program CC Climate Change CC+Urb Climate Change and Urbanization CDF Cumulative Distribution Function CF Change Factor CMIP5 Coupled Model Intercomparison Project 5 CN Curve Number CNT Center of Neighborhood Technology CWM Curve number Watershed Model DAtotal Total Drainage Area ESD Environmental Site Design GCM General Circulation Model GI Green Infrastructure HRU Hydrologic Response Unit HSG Hydrologic Soil Group IPCC Intergovernmental Panel on Climate Change LCLU Land Cover Land Use MDP Maryland Department of Planning NCDC National Climatic Data Center NOAA National Oceanic and Atmospheric Administration NRCS Natural Resources Conservation Service RUSLE Revised Universal Soil Loss Equation SDR Sediment Delivery Ratio SMPs Stormwater Management Pathways MD SWM Maryland Stormwater Manual SUSTAIN System for Urban Stormwater Treatment and Analysis Integration USDA US Department of Agriculture USGS US Geological Survey US EPA U.S. Environmental Protection Agency viii Chapter 1: Introduction 1. Climate change and urbanization Ecosystems provide valuable services for the well-being of humans (Constanza et al. 1997) such as protection from floods & storms (Martin and Watson, 2016). These services are affected by climate and land cover change (MEA, 2005; Nelson et al., 2009). According to the IPCC AR5 (Kirtman et al., 2013), the frequency, intensity, and magnitude of extreme precipitation events are expected to increase between 2010 and 2035. Meanwhile, impervious land cover, one of the strongest indicators of urbanization (Arnold and Gibbons, 1996) is expected to increase in several counties of Maryland between 2006 and 2030, based on forecasts derived from estimations of population and employment growth (Jantz et al., 2014). They estimated that the current rate of increase in urban population may result in 66 percent of the population living in cities by 2050 (United Nations, 2015). Similarly, historical time series and future projections of developed area (Figure 1.1) and population (Figure 1.2) show increasing trends in the U.S. Mid-Atlantic region. Both climate change and urbanization influence (Brauman et al., 2007) and amplify effects on hydrological services (Hejazi and Moglen, 2008). In addition, service levels of the current stormwater management infrastructure, which is based on historical analysis, requires adaptation strategies by incorporating anticipated climate change and urbanization. 1 Figure 1.1. Developed area trend1 Figure 1.2. Total population trend2 Climate change and urbanization are the main drivers affecting changes to hydrological processes (Nelson et al., 2009). Atmospheric and Ocean General Circulation Models, agree on an increase in intensity and magnitude of precipitation and seasonality of storms in the 21st century (Kirtman et al., 2013). Accurate prediction of climate change is difficult due to its non-stationary nature even while using current state-of-the-art climate models (Pielke, 2009). In addition, urbanization induces stressors such as flooding, low flows, and nutrient loadings impacting watershed functions (Sun and Lockaby, 2012). One of the strongest indicators of urbanization is impervious land cover and has been identified as an indirect marker of 1 https://www.chesapeakebay.net/who/projects-archive/land_use_workgroup 2 https://planning.maryland.gov/MSDC/Pages/projection/projectionsbytopic.aspx 2 environmental quality (Arnold and Gibbons, 1996). Climate change and urbanization each separately influence hydrologic behavior, and effects are amplified when both are present (Hejazi and Moglen, 2008). Definition and quantification of anticipated extreme climate events are necessary for risk management (Sura 2014; Kodra and Ganguly 2014; Zwiers et al 2013; Coumou and Rahmstorf 2012; Smith 2011; Meehl et al 2000; Easterling et al. 2000). A combination of the following methods are used to study extreme events: statistical (extreme value theory), empirical-physical (empirical knowledge and physical reasoning), numerical modeling (integrating general circulation models), and non-Gaussian stochastic (stochastic theory to study extreme events) (Sura 2014; Sura 2011). The overarching goal of this dissertation is to study change in the mean state of the climate (Meehl et al., 2000; Katz and Brown 1992) which causes periodicity in extreme events (Smith, 2011). Nevertheless, frequency and intensity of infrequent extreme events (100-year event) are mostly researched with a focus on higher percentiles (greater than 95th or 99th percentiles) of climatological probability distributions. However, asymmetry of the extreme events distribution3 is anticipated to increase along with periods of climate extreme (Kodra and Ganguly 2014; Smith 2011) causing rise in frequency and intensity of precipitation events. Therefore, it is interesting to assess the frequency of smaller frequent storm events, such as in excess of 85, 90, 95, and 99 percent by volume, as they have the potential to cause flooding 3 https://www.nature.com/articles/srep05884/figures/1 3 and soil erosion due to excess runoff. These capture depths are defined as percent of total rainfall when all event depths are ranked and cumulated. Urbanization alters hydrological response due to extreme precipitation (DeFries and Eshleman, 2004; Sun and Lockaby, 2012) and affects ecosystem services (Brauman et al., 2007). O’Driscoll et al. (2010) and Sun and Lockaby (2012) provided an exhaustive list of studies showing hydrological implications of increased urbanization in Washington DC and the Maryland metropolitan areas. An overall increase in high flow events is evident in most of these studies. In addition, variability in surface runoff is often analyzed with regard to spatial distribution, patterns, and intensity of urban development (Lee and Heaney, 2003; Shuster et al., 2005; Mejía and Moglen, 2009; Mejía and Moglen, 2010a; Mejía and Moglen, 2010b; Yang et al., 2010; Yang et al., 2011). 2. Urban ecosystem services According to the Millennium Ecosystem Assessment (2005), ecosystems provide various services that are beneficial for human well-being. These services include, but are not limited to biodiversity, supply of clean water, livable climate, food production, and flood resistance (Daily, 1997). The Millennium Ecosystem Assessment (2005) and The Economics of Ecosystems and Biodiversity (2010) classified these services, and associated functions, into four broad categories: provisioning (food and fresh water), regulating (local climate and air quality, waste water treatment, moderation of extreme events, nutrient recycling, pollination, carbon 4 sequestration and storage), cultural (recreation and mental health, spiritual experience), and supporting (habitats for species). However, urban ecosystem services are site specific and vary from one city to another (Bolund and Hunhammar, 1999) and their benefits to humans depend on socio-economic characteristics and proximity of the population (Elmqvist et al., 2013). Influx of human population to urban areas is expected to increase due to economic growth and demographic changes (Grimm et al., 2008) and therefore ecosystem services in cities and in surrounding areas require maintenance and regulation due to high intensity of demand and use (De Groot et al, 2002; Elmqvist et al, 2015). Urban ecosystems consist of built as well as ecological infrastructure and provide valuable ecosystem services (Elmqvist et al., 2013). The benefits enjoyed by people in an urban area are sometimes provided outside of its boundary, such as climate regulation in a city due to forested area around it (Serna-Chavez et al., 2014). It is therefore important to distinguish between “ecology of cities” rather than “ecology in cities” and analyze services at multiple scales (Niemelä et al, 2011; Hubacek and Kronenberg, 2013; Jansson, 2013). However, temporal quantification of these services termed as “flow of ecosystem services” (Chan et al., 2006; Fisher et al., 2011; Bagstad et al., 2013) requires spatially explicit assessment from provisioning to benefiting areas (Serna-Chavez et al., 2014). It is necessary to track production, flows, and delivery of various ecosystem services to understand their spatiotemporal configurations and dynamics. 5 Regulating services such as moderation of extreme events, flood mitigation, and climate regulation are severely affected by land conversion (Stürk et al., 2015). Conversion of land to urban areas increases imperviousness and creates water quantity and quality related problems such as water yield, sediment transportation, nutrient retention, moderation of climate, and runoff mitigation (Sun and Lockaby, 2012; Schueler, 1994; O’Driscoll et al., 2010). In essence, spatial configurations affect provision and flow of different ecosystem services (Serna-Chavez et al., 2014; Chan et al., 2006; Fisher et al., 2011; Bagstad et al., 2013) which are mainly caused by land cover change (Stürk et al., 2015). In addition to land conversion, climate change is one of the main driver to affect the supply of ecosystem services (MEA, 2005; Nelson et al., 2009; Jiang et al., 2016). Natural ecosystem services and functions are socially, economically, and aesthetically valuable (Daily et al. 2000). Unlike traditional economic assessment of goods and services, valuation of ecosystem’s benefits (or losses) received tremendous attention in recent years (De Groot et al, 2002; MA, 2005; Maler et al., 2008) especially in urban areas (Hubacek and Kronenberg, 2013; Elmqvist et al., 2013; Gomez-Baggethun and Barton, 2013; Haase et al, 2014; Posner et al, 2016). Some of the urban ecosystem services are provided in Table 1.1 along with their description. ‘Services’ are benefits enjoyed, consumed, and provided for the well-being of humans (Boyed & Banzhaf, 2007); accounting of these services is performed using biophysical, economical, socio-cultural values, health values, environmental justice, and insurance values (Constanza et al., 1997; De Groot et al., 2002; Barton et al, 6 2012; Burkhard et al., 2012; Elmqvist et al., 2013; Constanza et al., 2014). In addition, provision of these services in urban areas causes synergies and trade-offs, and requires loss assessment (Rodriguez et al, 2006; Haase et al., 2012; Kain et al., 2016) for urban planning purposes. According to De Groot et al. (2002) and Willemen (2010), ‘ecosystem services’ can be defined as the ‘ecosystem’s capacity’ to provide benefits for human satisfaction. Some of these benefits are mainly affected by high flow hydrological processes in the form of increased pollutants and sediment transport to streams and flooding (Bolund and Hunhammar, 1999; Brauman et al., 2007; Elmqvist et al., 2013). The uninterrupted availability of these urban ecosystem services is at risk due to extreme climate events and urbanization (Foley et al., 2005; Grimm et al., 2008; Farrugia et al., 2013; Lawler et al, 2014). Furthermore, interactions and feedback mechanisms among different urban ecosystem services may support and impair each other (Rodríguez et al., 2006; Haase et al., 2012). Therefore, impact assessment of extreme climate events and urbanization on services helps in decision making for resource management and sustainability. These assessments at small scale (~3 km2) are representative of a planned housing subdivision where land development frequently takes place. In addition, stormwater management takes place at smaller scales to mitigate impacts on ecosystem (such as soil erosion and sedimentation) due to extreme climate events. 7 Table 1.1. Classification and functions of urban ecosystem services correspond to regional/local level Ecosystem functions Ecosystem service type Description Citations Percolation and regulation of Runoff mitigation Soil and vegetation infiltrate Villarreal and Bengtsson (2005) runoff and river discharge and percolate water during heavy and/or prolonged precipitation events Photosynthesis, shading, and Urban temperature regulation Trees and other urban Bolund and Hunhammar (1999) evapotranspiration vegetation provide shade, create humidity and block wind Absorption of sound waves by Noise reduction Absorption of sound waves by Aylor (1972); Ishii (1994); Kragh vegetation and water vegetation barriers, especially (1981) thick vegetation Dry deposition of gases and Air purification Absorption of pollutants by Escobedo and Nowak (2009); Jim et particulate matter urban vegetation in leaves, al. (2009); Chaparro and Terradas stems, and roots (2009); Escobedo et al. (2011) Physical barrier and absorption Moderation of environmental Storm, flood, and wave Danielsen et al. (2005); Costanza et al. of kinetic energy extremes buffering by vegetation barriers; (2006b) heat absorption during severe heat waves; intact wetland areas buffer river flooding Removal or breakdown of xenic Waste treatment Effluent filtering and nutrient Vauramo and Setälä (2010) nutrients fixation by urban wetlands Ecosystems with recreational Recreation Urban green areas provide Chiesura (2004); Maas et al. (2006) values opportunities for recreation, meditation, and relaxation Human experience of Cognitive development Allotment gardening as Barthel et al. (2010); Groening ecosystems preservation of socio-ecological (1995); Tyrväinen et al. (2005) knowledge Ecosystems with aesthetic Aesthetic benefits Urban parks in sight from Tyrväinen et al (1997); Cho et al. values houses (2008); Troy and Grove (2008) Source: Modified from Gómez-Baggethun and Barton (2013) and Elmqvist et al. (2013) 8 3. Urban stormwater management Stormwater mitigation practices include use of gray (such as retention basins) and green (also known as low impact developments such as bio-retention, bio-swale, raingardens, infiltration trenches, and green roofs) infrastructure. Anticipated change in climate is likely to increase the frequency of more common storm events (such as 2-year events). In an urban watershed, this may cause more instances of flash floods and affect proper functioning of ecosystem services by inducing more nutrient transport, soil erosion, sedimentation, and channel incision. A commonly-sought engineering goal is to reduce runoff flows to pre- development levels. To mitigate the impacts of flood events, best management practices are employed to regulate post-development flows. These practices vary from one region to another with municipalities often weighing the relative magnitudes of stormwater infrastructure costs and maintenance with benefits of stormwater mitigation. Hydrological implications of climate change and increased urbanization lead to instances of flooding and damage to ecosystem services (Yang et al., 2011; Verburg et al., 2012). In urban areas, conventional stormwater management infrastructure is installed to regulate surface runoff (US EPA, 2009). As an alternative, green infrastructure not only offers runoff regulation but also provides additional benefits such as urban cooling, improved air quality, noise reduction, and aesthetics (Derkzen et al., 2015; Farrugia et al., 2013; Oberndorfer et al., 2007). Several studies highlighted effectiveness of green infrastructure for stormwater management due to its decentralized approach (Davis and McCuen, 2005; Shuster et al., 2008; Ahiablame et al., 2012; Dietz, 2007; Jayasooriya and Ng, 2014). Moreover, economically sound strategies require retrofitting of newly proposed structures to 9 current capacity of runoff mitigation through ‘adaptive management’, optimal spatial placement, and cost-effectiveness. Thus, it is crucial to test the resilience of the existing stormwater management infrastructure under anticipated future climate and urbanization. In addition, it is also necessary to analyze green stormwater management and how these systems affect ecosystem services. 10 4. Research questions The overarching goal of this dissertation is to assess service levels of current stormwater management practices under changing climate and urbanization, and estimate their possible impacts on ecosystem services. The specific questions proposed to attain this goal are as follows: a. What is the likely frequency of storm events in excess of current values of d 485 , d90, d95, and d99 due to climate change b. How do these storm events, combined with urbanization, affect surface runoff volumes? c. How much change is expected in the erosive power and erosion rates from storm events (all and d85-d99) in excess of current values with increase in each percent of precipitation in the near future? d. How do we obtain representative calibration values for the sediment delivery ratio model to estimate sediment yield from small-scale (~3km2) urban watersheds? e. Is it possible to obtain good flow simulation accuracy for stormwater assessment by using a simple Curve Number Watershed Model (CWM), developed in this dissertation, for small-scale urban catchments? f. Can green infrastructure provide added benefits to retain/replace/mitigate increased runoff and erosion effects under climate change and urbanization at watershed scale? 4 Rainfall percentile by volume. More details are in Chapter 2. 11 5. Dissertation structure and contribution This dissertation consists of five chapters. Below is a summary of each chapter along with its contribution: Chapter 25 provides a review on extreme precipitation events and identifies a research gap of less studied smaller frequent storm events (such as events in excess of 85 percent by volume). These storm events have the potential to cause flooding and soil erosion. In this context, change in stormwater quantity due to climate change and urbanization is a timely research topic. This chapter presents an analytical approach to estimate rainfall percentiles by volume (precipitation capture depth) and results in excess of current values of d85, d90, d95, and d99 and associated runoff under historical and future climate. A climate change study approach and the finding of likely increases in future high volume storm events in this chapter are valuable planning-level information for municipal stormwater management. The use of an analytical approach to estimate precipitation capture depths is relevant to flooding concerns at small watershed scales (~3 km2). Estimates of changes in water volume and resultant on-site infrastructure costs can help stakeholders and managers in planning for flood mitigation and protection of ecosystem services. Also, the use of capture depth percentiles in this study, such as d85, d90, d95, and d99, have the potential to serve as meaningful hydrologic indicators for future stormwater management planning. 5 Chapter 2 based paper “Future Storm Frequency and Runoff in Small U.S. Mid-Atlantic Watersheds Evaluated Using Capture Depth” is currently under second review in Journal of Sustainable Water in the Built Environment 12 Chapter 36 provides background on possible future erosivity and soil erosion rates and identifies a research gap of less focus on such assessments at small scale urban watersheds under climate change and urbanization. Also, this chapter identifies modeling limitations in sediment yield assessment. High volume of stormwater can cause erosion which affects stream channel and banks, wildlife habitat, properties, and deposits sediment in downstream areas. This chapter presents possible changes in rainfall erosivity and soil erosion rates in relation to each percent change in precipitation under climate change. Also, this chapter details sediment yield at the watershed outlet using the sediment delivery ratio model which require calibration. It also proposes a new procedure to estimate representative calibration values for the sediment delivery ratio model, at small scale urban watersheds using a sigmoid curve -based approach. Finally, it discusses soil conservation cost valuation. In ecosystem service impact assessment, the likely higher erosion rates and sediment yield in the future are a point of concern. The findings in this chapter are relevant for effective land use planning. Also, the approach in this study to estimate representative calibration values for the sediment delivery ratio model, at small scale (~3 km2) urban watersheds, is valuable for ungauged sites. This approach of representative calibration values has the potential for use in place of average or theoretical calibration values to run sediment delivery model. Chapter 4 includes development of a curve number watershed model with few input data needs and easy setup for examination of small scale urban watersheds as an alternative to complex hydrologic models. Moreover, this chapter provides a cost comparison of 6 Chapter 3 is based on a paper that is ready for submission to journal Sustainability 13 conventional and green stormwater management practices in the form of various pathways under climate change and climate change plus urbanization. These pathways also include benefits of green practices to explore their effectiveness at the watershed scale. This chapter provides a synthesis of green stormwater management practices on how they affect ecosystem services and major limitations and future research needs. The construction of a simple curve number watershed model in this study with good performance and few input data needs offers a possible flow simulation tool for medium to highly impervious watersheds at small scales (~3 km2). Moreover, the stormwater management pathways along with cost-benefit assessment using green stormwater management practices in this chapter serves as a first step to determine effectiveness of certain green practices at the watershed scale. This chapter identifies future research needs to fill gaps in our understanding of the use of green stormwater management practices and how they affect ecosystem services. Chapter 5 summarizes the major findings that are identified in Chapters 2 – 4, and describes the contributions of this dissertation. Future research needs are also detailed based on key outcomes and limitations found in this dissertation. 14 Chapter 2: Future storm events frequency and runoff 1. Introduction Climate change and urbanization affect hydrologic processes (Nelson et al., 2008) in the form of extreme precipitation (Trenberth, 2011; Donat et al., 2017), increased runoff (Shuster et al., 2005), flooding (Kundzewicz, 2003), and elevated nonpoint source pollution (Schueler, 1994). According to the Intergovernmental Panel on Climate Change (IPCC) Assessment Report 5 (Kirtman et al., 2013), the frequency, intensity, and magnitude of extreme precipitation events have increased and are projected to continue to rise as a result of climate change. These findings align with similar findings of the Third National Climate Assessment report focusing on the continental US (Georgakakos et al., 2014). Infrastructure and people in urban areas are highly vulnerable to extreme precipitation events (Mishra et al., 2015). Urban and exurban settlement densities and urban land use are projected to increase for 2010-2050 in the northeast US (Grimm et al., 2008; Passel and Cohn, 2008; Brown et al., 2014). These land use changes, coupled with continued growth of existing urban areas, are expected to result in an expansion of vulnerable urban regions (Wear, 2011). Urbanization alters the hydrologic response to extreme precipitation events (DeFries and Eshleman, 2004; Sun and Lockaby, 2012). Therefore, it is necessary to investigate the implications of growing urban areas, in addition to more extreme precipitation events, on hydrologic processes. O’Driscoll et al. (2010) and Sun and Lockaby (2012) provide an exhaustive list of studies showing hydrologic implications of urbanization in Washington, DC, and its surrounding Maryland metropolitan areas. An overall increase in high flow events is evident in most of these studies. In other studies, variation in surface runoff is analyzed with regard 15 to the spatial distribution, patterns, and intensity of urban development (Lee and Heaney, 2003; Shuster et al., 2005; Jacobson, 2011; Mejía and Moglen, 2009; Mejía and Moglen, 2010a; Mejía and Moglen, 2010b; Yang et al., 2010; Yang et al., 2011). One of the strongest indicators of urban development is impervious land cover (Arnold and Gibbons, 1996) which is projected to increase in several Maryland counties between 2006 and 2030 based on land use forecasts (Jantz et al., 2014) and Maryland Department of Planning zoning maps (MDP, 2016). Altered space-time rainfall patterns affect the watershed hydrology and can increase runoff (US EPA, 2009; Mejía and Moglen, 2010a; Yao et al., 2016). Stormwater management infrastructure is built on the underlying assumption of a stationary climate (Rosenberg et al., 2010). However, anticipated increases in frequency and intensity of extreme precipitation events due to climate change requires an assessment of the current stormwater management infrastructure and its ability to be effective for future climate (Moore et al., 2016; Gersonius et al., 2012). Adaptation of stormwater infrastructure to climate change employs either top-down or bottom-up approaches (Jones and Preston, 2011). Top-down refers to a predict-then-adapt method. Bottom-up is a socioeconomic approach (Gersonius et al., 2012) that incorporates local stakeholders’ experience of risk and adaptation in a changing climate. While both approaches have merit (Jones and Preston, 2011), a top-down approach is applied in this chapter to provide a range of possible future storm frequencies and runoff scenarios for stormwater management, using the capture depth approach. These results are used to estimate the costs of implementing one particular stormwater management practice, as an example of how the predictions might be applied in adaptation. 16 Capture depth is a recommended measure for onsite stormwater management aimed at maintaining pre-development hydrology with respect to volume (NRC, 2009). Capture depth, dX, is defined as the depth, that accounts for x percent of total precipitation when event depths are ranked and cumulated over a substantial period of record. This event-based approach is intended for estimating stormwater volumes for quality management; it is a depth-weighted percentile rather than the simple ranked event percentile that is often applied in stormwater design (e.g. MD SWM vol.1, 2009). The definition of dX is illustrated in Figure 2.1 (Data used in this Figure are from PRISM, discussed below). In contrast, for overbank flood protection, hydrologic indicators of annual maximum probabilities, such as 2- and 10-year, 24-hour storm depths, are used (MD SWM vol., 2009). Annual maximum statistics do not reflect the number of times an event of a particular magnitude might occur in a given year, only the probability of exceedance. For water quality, the cumulative effects of events are of concern, rather than the likelihood of property damage or infrastructure failure in a given year. PRISM data (Di Luizo et al., 2008) show an increase in annual maximum precipitation in Prince George’s County in the last decade (Figure 2.2a). Moreover, PRISM-based d95 (2011-2015) and NOAA ATLAS 14 (Bonnin et al., 2006) 2-year, 24-hour precipitation depths demonstrate recent increases in hydrologic extremes 17 Figure 2.1. Capture depth concept diagram (Figure 2.2b). These extremes are anticipated to increase, as conceptualized in Figure 2.2c (blue for current and red for future probability), requiring the inclusion of smaller, frequent storms for local stormwater management. These events have the potential to cause soil erosion due to excess runoff, which are of interest in this study. In this regard, a distinctive definition of somewhat frequent (d85-d99 events) and infrequent (greater than 99 percent capture depth) storm events is used for this analysis. Several capture depth values: 85, 90, 95, and 99 percent, are investigated in this study, as a guide to designing stormwater infrastructure. This study aims to answer the following two questions for the US Mid- Atlantic Region: 1) What is the likely frequency of storm events in excess of current values of d85, d90, d95, and d99 due to climate change? And, 2) How do these storm events, combined with urbanization, affect surface runoff volumes? 18 Figure 2.2. Conceptual diagram for future storm runoff Note: CDF range of 0.85-1.0 is not linearly related to storm size probabilities of 85-99 percent. 2. Methods and data 2.1. Experimental overview In this study, we examine the hydrologic impacts of climate change and urbanization on small scale watersheds (3 km2) in central Maryland. One watershed was selected in each of four counties (Howard, Montgomery, Anne Arundel, and Prince George’s) in the most urbanized part of Maryland. The selected study watersheds are representative of average precipitation and urbanization conditions (a curve number model for each county is provided below) in each county. Also, rainfall depth values for the 2-year, 24-hour event in these watersheds are similar to county level values (provided in Appendix A). Rainfall depth values are obtained from NOAA Atlas 14 (Bonnin et al. 2006). County level land use and soils data are used to quantify future trends in urbanization (equations 2.5-2.8 provided 19 below). Storm events are characterized by the total daily precipitation. Events are assumed to be independent, without consideration of multi-day events. The capture depths d85, d90, d95, and d99 are calculated from observed historical data. For observed historical and simulated future time series, the number of events per year in excess of d85, d90, d95, and d99 are determined. For runoff estimation using the NRCS Curve Number (CN) method, county-level CN aggregate values are computed at several historical time steps. Historical CN values are regressed with time generating models for each county to forecast future CN values (described further below). Time series of historic and future runoff are generated and summarized using cumulative distribution functions (CDFs). Finally, climate model-based volumetric estimates in each county are used to determine total storage cost using enhanced bioretention. Costs are provided for both climate change and climate change plus urbanization scenarios. 2.2. Geographical area The State of Maryland is undergoing rapid urbanization (Weber et al., 2006; MDP, 2016). According to the Maryland Department of Planning (MDP, 2010), 27 percent of Maryland’s land area is developed. As a result of this urbanization, increased flooding is a safety concern, with stormwater management practices implemented to address this issue. Maryland may be broadly divided into three distinct physiographic regions: The Appalachian/Ridge and Valley region, the Piedmont plateau, and the Coastal Plain. Land development has occurred mainly in the Piedmont region, followed by the Coastal Plain. The 20 selected study watersheds lie in the Piedmont and Coastal Plain regions. Location and land use information for the study watersheds, are provided in Figure 2.3. Figure 2.3. Study area watersheds with Land Use Land Cover (MDP, 2010) 2.3. Observed/historic precipitation: event frequency analysis The historic/observed daily precipitation data used in this investigation, at a spatial resolution of 2.5-arc minutes (~4 km), are obtained from daily PRISM maps. The PRISM daily precipitation product improves on the NCDC’s station-based data by taking into account physiographic factors such as local terrain and water bodies (Daly et al., 2008; Di Luizo et al., 2008). 21 Point time series of precipitation are extracted at watershed centroids from the gridded PRISM data. PRISM data are used to estimate capture depths (d85, d90, d95, and d99) for the historical period 1981-2015. The number of storm events (24-hour precipitation totals) in excess of d85, d90, d95, and d99 are counted for each year of the historical time period (1981- 2015) and presented as cumulative sum in Figure 2.4. Figure 2.4. Historical storm events count 2.4. Future precipitation Scale mismatch has been a longstanding problem when downscaling climate model outputs to the regional or local scale for hydrologic impact assessment (Schmidli et al. 2006; Chen et al. 2011b). Future precipitation datasets are primarily available at coarse resolution on a global to regional extent, which limits their usefulness in simulation of physical processes, such as local flooding and soil erosion, at the urban catchment scale (Willems et 22 al., 2012). It is common to apply statistical downscaling of simulated data at a global to regional scale, using a variety of methods, for local scale impact analysis (Fowler et al, 2007). The accuracy of such assessments depends on the choice of the general circulation model (GCM) and downscaling methods (e.g. Wilby and Harris, 2006; Chen et al., 2011a). Moreover, the accuracy varies among regions. To test the accuracy of climate models for a specific region (the Mid-Atlantic in our case), the region’s monthly mean, variance, standard deviation, and skewness of model precipitation are computed and compared with observational data. This process helps select models with reasonable accuracy to reproduce historical climate. Unfortunately, there is no universal method available which can, of itself, explain all variance at the local scale. Therefore, several models and downscaling methods exist (e.g., Minville 2008; Chen et al. 2011a; Chen et al. 2011b). In this study, we follow Moore et al. (2016) who applied the CF method to baseline precipitation. Many downscaling methods are available, each with strengths and weaknesses (Wilby and Harris, 2006; Piani et al., 2009; Piani et al., 2010; Themeβl et al., 2011; Gudmundsson et al., 2012; Hempel et al., 2013). In this study, the most widely applied simple change factor (CF) or delta change method is used to statistically downscale future data using percent change (ratio of future to historical mean time series) from different GCMs (Diaz-Nieto and Wilby, 2005; Marauan et al., 2010; Chen et al. 2011; Sunyer et al., 2012; Gudmundsson, 2012). In addition, future precipitation times series are calculated using an advanced empirical downscaling method, Multivariate Adaptive Constructed Analogs (MACA), to assess the differences that exist between the CF and MACA approaches. Abatzoglou and Brown (2012) prepared these projections using bias-corrected Coupled Model Intercomparison Project (CMIP5) models. The CF scaling method is simple and easy to 23 implement. However, because it simply scales an observed time series using a calculated ratio, it is inherently limited in its ability to predict variability of extreme precipitation events for future time periods. By contrast, the MACA method reflects differences among model predictions of inter-and intra-annual variability of future precipitation, and provides for the inference of changed storm event frequency. The use of several GCMs is preferred to obtain an ensemble of possible outcomes. The model-derived precipitation estimates in this study are obtained using multiple models. CMIP5 contains bias corrected climate and hydrology projections from General Circulation Models (GCMs). Details of the GCMs used in this study are provided in Table 2.1. Empirically downscaled projections of daily precipitation, based on the MACA method were acquired. The main steps in the MACA method involve correction of bias in GCMs using observational data, computing an epoch adjustment, and construction of analogs. A detailed description of the MACA method is available in Abatzoglou and Brown (2012). Table 2.1. Climate models used in this study Source: https://pcmdi.llnl.gov/mips/cmip5/availability.html Modeling Center Model Institution CCCma CanESM2 Canadian Center For Climate Modeling and Analysis GFDL- NOAA GFDL Geophysical Fluid Dynamics Laboratory ESM2G IPSL-CM5- IPSL Institut Pierre Simon Laplace MR NCC NorESM1-M Norwegian Climate Centre For analysis of projected future precipitation, statistical transformation techniques work well in preserving the precipitation mean, but the frequencies of wet and dry periods are difficult 24 to quantify. Therefore, the CF method is used to transform precipitation values from historical to future time periods. A scaling factor is determined by taking the ratio of the mean of daily bias-corrected CMIP5 model projection for future and historical periods. This process is repeated for all models separately. 𝑃𝑃𝑖𝑖𝑖𝑖(𝑓𝑓𝑓𝑓𝑓𝑓) = 𝑠𝑠 × 𝑃𝑃𝑖𝑖𝑖𝑖(𝑜𝑜𝑜𝑜𝑜𝑜) 2.1 𝑀𝑀���𝑀𝑀��𝑀𝑀��𝑀𝑀�𝑀𝑀�𝑓𝑓�𝑓𝑓��𝑓𝑓�𝑓𝑓�𝑓𝑓�𝑓𝑓�𝑠𝑠 = 2.2 𝑀𝑀���𝑀𝑀��𝑀𝑀�𝑀𝑀��𝑀𝑀�ℎ�𝚤𝚤�𝑜𝑜�𝑓𝑓�𝑜𝑜�𝑓𝑓��𝚤𝚤𝚤𝚤�𝚤𝚤��𝚤𝚤 where 𝑃𝑃𝑖𝑖𝑖𝑖(𝑓𝑓𝑓𝑓𝑓𝑓) (i=day and j=year) is the daily future precipitation (2016-2035), 𝑃𝑃𝑖𝑖𝑖𝑖(𝑜𝑜𝑜𝑜𝑜𝑜) is the daily historical precipitation (1996-2015), and s is the ratio of averaged model future (2016- 2035) to average model historical (1996-2015) time series. An implicit assumption in applying the CF method to GCMs for estimation is that model bias is consistent between historical and future periods. The CF method (Eq. 2.1) is computed using all models (listed in Table 2.1) and then individually applied to the PRISM time series to create future precipitation time series. MACA future time series data are obtained using study area centroids. MACA projections of all climate models starting in 2016 are individually concatenated with historical PRISM data from 1981-2015. Times series of daily (24-hour) precipitation are obtained from four models participating in the CMIP5 program (Table 2.1). Individual precipitation events (24-hour precipitation values) in excess of d85, d90, d95, and d99 (these percentile depth values are only computed once using historic data and remain fixed for both historic and future storm events count) are counted for each year of the future time period, for both the CF and the MACA future time series. CF-based results are presented in the form of sums in excess of each capture depth in Figure 2.5. MACA-based results are 25 presented with event counts in each capture depth range for each year in Figures 2.6, 2.7, 2.8, and 2.9. 2.5. Surface runoff The Natural Resources Conservation Service (NRCS) CN method (SCS, 1986, NEH, 2004) is used for surface runoff estimation. (𝑃𝑃 − 0.2𝑆𝑆)2 𝑄𝑄 = 2.3 (𝑃𝑃 + 0.8𝑆𝑆) 1000𝑆𝑆 = � − 10� ∙ (25.4) 2.4 𝐶𝐶𝐶𝐶 where Q is direct runoff (mm), P is rainfall (mm), S is storage (mm), and CN is the curve number. For small watersheds, surface runoff from a given storm is assumed to be independent of previous rainfall events (McCuen, 1989). This assumption is true for highly impervious watersheds (70 percent or greater) and may result in underestimation of runoff due to antecedent soil moisture conditions (Garen and Moore, 2005; Grimaldi et al., 2013). Soil wetness achieved during the first day of a multi-day event may result in more runoff volume, but is unaccounted for in the CN Method (SCS, 1986, NEH, 2004). The historical CN is obtained using MDP 2010 land use/land cover data and hydrologic soil groups (A, B, C, and D) data (NRCS, 2017). A 2-digit classification system is used to prepare historical CN (details provided in Appendix A). Land use data for 2010, at approximately 30 m spatial resolution, prepared by the Maryland Department of Planning (MDP). Low density residential and transportation land use types are included in the 2010 land use map to obtain improved estimates of total land urbanized. These data provide the initial land use characterization corresponding to 2010 conditions. SSURGO (NRCS, 2017) soils data at 30 26 m spatial resolution are used along with land use data for 1973, 1990, 1997, 2002, and 2010 to compute the historical CN for each of the years using a 2-digit classification (CN county level maps are provided in Appendix A). For future CN estimation, historical CN values at county levels (presented in Table 2.2) are regressed from multiple snapshots in time (1973, 1990, 1997, 2002, and 2010) to generate linear models of CN change for each county (Howard, Montgomery, Anne Arundel, and Prince George’s) as shown in Figure 2.10. Linear models for future CN estimation are provided below for each county. Figure 2.5. Observational (1981-2015) and climate models-based (2016-2035) scaled [CF based] observations event count 27 Figure 2.6. Historical and future storm events count for Howard county Figure 2.7. Historical and future storm events count for Montgomery county 28 Figure 2.8. Historical and future storm events count for Anne Arundel county Figure 2.9. Historical and future storm events count for Prince George’s county 29 Howard: 𝐶𝐶𝐶𝐶 = 0.0441 (𝑦𝑦𝑀𝑀𝑦𝑦𝑦𝑦) − 15.4 R2 = 0.80 2.5 Montgomery: 𝐶𝐶𝐶𝐶 = 0.039 (𝑦𝑦𝑀𝑀𝑦𝑦𝑦𝑦) − 2.77 R2 = 0.91 2.6 Anne Arundel: 𝐶𝐶𝐶𝐶 = 0.0664 (𝑦𝑦𝑀𝑀𝑦𝑦𝑦𝑦) − 63.2 R2 = 0.98 2.7 Prince George’s: 𝐶𝐶𝐶𝐶 = 0.0589 (𝑦𝑦𝑀𝑀𝑦𝑦𝑦𝑦) − 44.7 R2 = 0.92 2.8 The slope in equations 2.5-2.8 are a proxy for the rate of urbanization within each county. Figure 2.10. Curve Number trend lines at the county level Surface runoff, Q (mm), from events greater than d85 was computed, using equations 2.3 and 2.4, under both: 1) Climate Change (CC) using historical CN, and 2) Climate Change plus Urbanization scenarios (CC+Urb) using future CN from Eqns. 2.5-2.8, for all CMIP5 models. The CDF of Q values was estimated for each county/model combination as follows: 1. Each precipitation event in the modeled (2016-2035) time series is used as input to the NRCS CN runoff model. A single value of CN for year 2010 and 2035 is used for both the CC and CC+Urb scenarios, respectively. 30 2. The CDF of each county’s Q time series (Figure 2.11) was calculated using a programmed function available in the Seaborn library (https://seaborn.pydata.org/generated/seaborn.kdeplot.html) Table 2.2. Historical and forecasted Curve Number estimates at the county level County 1973 1990 1997 2002 2010 2035 Howard 71.7 71.7 72.6 72.8 73.3 74.2 Montgomery 74.1 74.9 75.1 75.0 75.7 76.6 Anne Arundel 67.7 69.0 69.6 69.7 70.1 71.9 Prince George’s 71.4 72.3 73.3 73.0 73.5 75.1 2.6. Cost estimation The variation in runoff presented earlier is more instructive if presented in the form of a change in cost magnitudes. Cost estimates can be based on best management practices (BMP) construction or per acre treated (Brown and Schueler, 1997; Houle et al., 2013; Mateleska, 2016). To illustrate the magnitude of anticipated runoff changes, the maximum runoff volume is determined from observational data for the years 1996 to 2015 and CF-based precipitation for 2016 to 2035 of all climate models in each county. The general cost function formula for enhanced bioretention from Mateleska (2016) is used for illustration. 31 Figure 2.11. Runoff under climate and urbanization scenarios [CF based] for 2016-2035 exceeding d85 of 1981-2015 𝑇𝑇𝚤𝚤 = 𝑉𝑉𝑜𝑜 × 𝐵𝐵𝑀𝑀𝑃𝑃𝚤𝚤 × 𝐹𝐹𝚤𝚤 2.9 where Tc is total cost, Vs is the volume storage in m3, BMPc is a cost estimate in $/m3 (adjusted to 2018 dollars), and Fa is an adjustment factor (a value of two is used for new BMPs in developed areas). Total runoff volume is computed by converting Q (mm) for each county/model to a value (m3) (by applying the area of each watershed). Maximum runoff values during historical and future periods are used to develop the cost estimates presented. 3. Results and discussion 3.1. Historic/observed capture depth events frequency (1981-2015) Precipitation capture depth values calculated from the observed data for the 35-year historical period (1981-2015) are provided in Table 2.3. Values of d85, d90, d95, and d99 are 32 similar across all counties, which is to be expected based on their proximity to one another. For each county, the number of events in excess of d85, d90, d95, and d99 are shown as a cumulative sum in Figure 2.4. Events in excess of d85 consistently occurred across all counties with a frequency of 1-2 times per year since 1981, with a noticeable increase from 2000 onwards. Similar patterns are observed for events in excess of d90 with an increase in the last decade for all counties. Events of greater than and equal to d95 magnitude have started to occur regularly in recent years. Events greater than d99 are observed across all counties only after 2005 with an evident rainfall signature in years 2006, 2011, and 2012. These are related to Hurricanes Ernesto, Irene, and Sandy respectively; Maryland counties experienced direct landfall from these hurricanes or tropical storm conditions. It is important to emphasize that (a) dX represent cumulative depth measures, not simply frequency measures; and (b) events greater than d85 generally cause flooding in small urban watersheds. Increasing trends are visually noticeable after year 2000 (Figure 2.4). The Mann-Kendall trend test (Mann, 1945; Kendall, 1975; Gilbert, 1987) is used to determine if a monotonic trend exists in the occurrence of these events during the historical time period. Mann-Kendall trend test results for the historical time period (1981-2015) are presented in Table 2.4 for all counties. Statistically significant increasing trends (p < 0.05) in the number of events greater than d85, d90, and d95 were observed in all counties. Table 2.3. Capture depths from historic/observed data, by county County Number of Events (non-zero) d85 d90 d95 d99 Howard 4527 40.9 48.4 62.2 144.4 Montgomery 4634 38.2 45.7 61.9 104.3 Anne Arundel 4558 41.2 49.5 65.5 108.0 Prince George’s 4474 40.1 46.3 62.8 125.5 33 Table 2.4. Mann-Kendall test statistical significance for trend in number of events per year exceeding specified capture depths, 1981-2015 (values in italics are significant at 10 percent or less) p-value County d85 d90 d95 d99 Howard 0.022 0.011 0.096 0.365 Montgomery 0.005 0.009 0.044 0.6 Anne Arundel 0.007 0.0003 0.004 0.63 Prince George’s 0.001 0.001 0.025 0.435 3.2. Future capture depth event frequency (2016-2035) The climate models generally agree that an increase in total precipitation is likely in the future. Daily precipitation values in all counties, based on the CF approach using various climate models, are predicted to increase one to five percent in the future period. Figure 2.5 shows, by county, the number of events in each defined capture depth range and each model, along with the observation-based values (full time series are not shown, because they are proportional to historic time series, under this downscaling approach). The IPSL-CM5A-MR (Dufrense et al., 2013) and NorESM1-M (Bentsen et al., 2013) models forecast a noticeable change in the number of events in excess of d85, d90, and d95 across all counties, especially for Prince George's county. Both models show an increase in occurrence of events exceeding d85. On the other hand, events exceeding d99 occurred with about same frequency as observed during 2005-2015 in these two models. Future frequency, based on the MACA approach, of precipitation events greater than d85 is analyzed using cumulative trend lines. In Howard County, both the CanESM2 (Chylek et al., 2011) and NorESM1-M models show a decreasing trend in the number of these events after year 2025 (Figure 2.6). In contrast, the 34 IPSL-CM5A-MR model shows a decreasing trend following year 2015 and an increasing trend after 2030. The GFDL-ESM2G (Dunne et al., 2012) model projection shows approximately the same number of events greater than d85 during 2016-2035 as observed during 1990-2015. For Montgomery County, the NorESM1-M model shows a decreasing trend in frequency of events greater than d85 after year 2015 (Figure 2.7). The IPSL-CM5A- MR and GFDL-ESM2G models show a decreasing trend after 2020. However, IPSL-CM5A- MR shows an increase in frequency of events greater than d85 following year 2025 whereas GFDL-ESM2G initially shows an increasing trend after year 2020 with no trend during 2025- 2035. The CanESM2 model shows a mixed response with both increases and decreases in frequency of events greater than d85 throughout the period 2016-2035. For Anne Arundel County (Figure 2.8), the CanESM2, IPSL-CM5A-MR, and GFDL-ESM2G models show a decrease in frequency of events greater than d85 during 2016-2020, whereas each of these models gives a mixed trend response during 2021-2035. The NorESM1-M shows a decrease for events greater than d90 until 2030. For Prince George's County (Figure 2.9), all models show a decrease in frequency of events greater than d85 during 2015-2020 and an increase afterwards until 2025. Only the IPSL-CM5A-MR and GFDL-ESM2G models show a steady increase in frequency of events greater than d90 and d95 in later years. It is inferred that IPSL- CM5A-MR and GFDL-ESM2G models show an increase in frequency of d90 events during 2020-2035. However, both of these models show the same frequency of events greater than d99 in the future as observed during 2005-2015. 35 3.3. Runoff scenarios Figure 2.11 shows the frequency distributions of NRCS CN-based runoff, Q, corresponding to the different future scenarios using the CF downscaling approach. The future CN estimates are incorporated in the CC+Urb scenario, as described in the Methods and Data section. Insets in each subplot of Figure 2.11 show the full extent of the Q CDF for both historic observations (1981-2015) and future climate models (2016-2035). The lower portion of each subplot shows an enlarged view of the 80-100 percent quantiles of the CDF in order to highlight differences in this range of interest. For all models, subtle differences exist between the future and the observational CDFs. The expected runoff increase from all models under the CC scenario lies between a range of 5-10 mm across all counties. Under the CC scenario, future and historical runoff distributions are very similar. Under the CC+Urb scenario in Figure 2.11, the enlarged subplots show a subtle difference between the historic observational and model-based CDFs across all counties. The climate models are in general agreement, showing modest runoff differences. A runoff increase, relative to the CC scenario, of about 5 mm is projected for all study watersheds. 3.4. Cost magnitudes The climate model-based total costs are used along with observational estimates to determine percent change during 2016-2035 relative to the 1996-2015 period. Cost estimates for each county under CC, and CC+Urb scenarios are provided in Table 2.5. The models CanESM2 and GFDL-ESM2G show small changes in cost under the CC scenario, with smaller values relative to the observational period encountered in Howard, Anne Arundel, and Prince George's Counties. On the other hand, IPSL-CM5A-MR and NorESM1-M show 36 about a six percent increase in stormwater management costs relative to the observational time period. It is important to emphasize here that main objective behind presentation of these numbers is to highlight how stormwater management costs would vary in the near future. Under the CC+Urb scenario, CanESM2 and GFDL-ESM2G show an increase of one to five percent in stormwater management costs. Moreover, the models IPSL-CM5A-MR and NorESM1-M show a cost increase of about eight to thirteen percent across all counties. Cost changes in Howard and Montgomery seem to be mainly affected by climate change, rather than urbanization. However, Anne Arundel and Prince George's county estimates project a similar impact of climate change and urbanization on cost change. Table 2.5. Cost percent change relative to current estimates under climate change, and climate change urbanization scenarios IPSL-CM5A- GFDL- County CanESM2 NorESM1-M MR ESM2G Howard -1.1 6.6 1.2 6.2 Montgomery 1.8 7.6 2.3 6.9 Anne Arundel -1.3 7.5 1.4 7.0 Prince George's -1.0 8.5 -0.9 6.0 Howard 1.5 9.4 3.9 8.9 Montgomery 4.6 10.4 5.0 9.7 Anne Arundel 4.7 13.9 7.6 13.5 Prince George's 3.8 13.5 3.8 11.0 4. Conclusions Observational data shows an increasing trend in the frequency of events exceeding the d85 magnitude during the historical period 1981-2015, with a noticeable increase from 2000 onwards. Coincidently, the frequency of d99 events increased from 2000 onward relative to 37 Climate Change Climate Change and Urbanization occurrence during time period 1981-2000. In the CF-based future projections, both IPSL- CM5A-MR and NorESM1-M models show an increase in the frequency of d85 events during the future period (2016-2035) while d99 events occurred with about the same frequency as observed during 2005-2015. In the case of MACA-based projections, a similar relationship is found with IPSL-CM5A-MR and GFDL-ESM2G during the future period. An increase in the occurrence of d85 and d99 events is predicted for the future with a similar frequency as in the observational record. Curve number-derived runoff distributions during the future period based on the CF method from all models and during the historic period are similar, under the climate change scenario. The climate models show general agreement, with modest differences in runoff. In contrast, under the climate change and urbanization scenario, based on the CF method, all models show a slight increase in runoff values across all counties, indicating that land use change is more influential than climate change in modulating runoff in this region. Stormwater management costs based on CF method projections of the IPSL-CM5A- MR and NorESM1-M models show about a six percent increase relative to the observed record. Under the climate change and urbanization scenario, both these models project an increase of about eight to thirteen percent across all counties. Cost changes in Howard and Montgomery counties are predominantly from climate change rather than urbanization. However, Anne Arundel and Prince George's county estimates project a similar impact from climate change and urbanization on cost changes. The study approach and findings are valuable planning-level information for municipal stormwater management. Our use of an analytical approach to estimate precipitation capture 38 depths is relevant to flooding concerns at small watershed scales (~3 km2). Estimates developed in this study of changes in water volume and resultant on-site infrastructure costs can help stakeholders and managers in planning for flood mitigation and protection of ecosystem services. Capture depth percentiles such as d85, d90, d95, and d99, serve as meaningful hydrologic indicators for future stormwater management planning. 39 Chapter 3: Soil erosion and sediment yield 1. Introduction Soil is an invaluable resource (Amundson et al., 2015; Montgomery, 2012) and important foundation of society (Nikiforoff, 1959). It is at the juncture where atmosphere, hydrosphere, biosphere, and pedosphere meet and interact (Singer and Warkentin, 1996). Human beings rely on it for their livelihood (Minami, 2009), sustainable land management (Kumar and Hundal, 2016), maintaining carbon stocks (Amundson et al., 2015), and food production (Pimentel and Burgess, 2013). Therefore, the conservation of soil is viewed as a critical regulating ecosystem service provided by terrestrial ecosystems (Li et al., 2017; Nelson et al., 2009). Sustainability and productive capacity of soils are at risk due to climate change (Li and Fang, 2016; Biasutti and Seager, 2015; Routschek et al., 2014; Zhang and Nearing, 2005; Nearing et al., 2004), intensified agriculture (Pimentel and Burgess, 2013), land use change (Mullan, 2013; Yang et al., 2003), and associated accelerated soil erosion (Amundson et al., 2015; Pimentel and Burgess, 2013; Lal, 2003; Lal, 2001; Pimentel et al., 1995). Soil erosion occurs due to water and wind energy (Montanarella, 2015; Pimentel and Burgess, 2013; Blanco and Lal, 2010). The direct impacts of climate change on soil erosion are in the form of rainfall amount, rainfall intensity, and rainfall spatio-temporal distribution (Bangash et al., 2013; Nearing et al., 2004; Pruski and Nearing, 2002). On the other hand, the indirect impacts include increase in the temperature (Nearing et al., 2004), and socio- economic factors (Li and Fang, 2016). Collectively, the climate change impacts on soil erosion are in the form of changes of rainfall, vegetation, and crop management (Li and Fang, 40 2016). According to a detailed review provided by Li and Fang (2016): i) Onsite problems caused by soil erosion consist of deterioration of soil physical, chemical, and biological properties (Lal et al., 2000), loss of nutrients (Lal, 2003), reduction of agricultural productivity (Lal, 1998), and cropland loss (Pimentel, 2006). ii) Offsite problems consist of fluvial sediment deposition, reservoir sedimentation, and channel silting (Mullan, 2013). The frequency and intensity of extreme precipitation events is likely to increase in the United States (Georgakakos et al., 2014; Wuebbles et al., 2014; Vose et al., 2014; Peterson et al., 2013; Kunkel et al., 2013; Diamond et al., 2013; Groisman et al., 2012). However, the relative contribution of extreme events to total erosion (and erosive power) is uncertain and less studied (García-Ruiz et al., 2015; Boardman, 2006). Precipitation droplets contain energy, which is transmitted to exposed soil particles resulting in splash and sheet erosion (Pimentel et al., 1995). However, the energy of these droplets is expected to alter due to anticipated climate change (Pruski and Nearing, 2002; Nearing, 2001). According to Nearing (2001) and Nearing et al. (2004), an expected increase in erosion rate (the ratio of increase in erosion to annual rainfall increase) is on the order of 1.7 in the United States with overall increase in large parts of the Mid-Atlantic states (time intervals of analysis are 2040-2059 and 2080-2099). However, the exact magnitude of change in the erosivity power of these events under climate change is uncertain due to levels of uncertainty of climate models and precipitation downscaling methods (Biasutti and Seager, 2015). A wide variety of erosion prediction models are available, categorized as empirical, conceptual, physically- (process-) based at different spatial and temporal scales (Morgan and Nearing, 2016; Merritt et al., 2003; Toy et al., 2002). These models are represented at three 41 levels of simplification: black-box, grey-box, and white-box (Morgan and Nearing, 2016). Black-box models are the simplest types of models whereas grey-box are more complex than black-box models. White-box models include more detailed information about erosion processes relative to the other models. According to Morgan and Nearing (2016), no soil erosion model exists which can be exactly categorized as white-box. However, there are few process-based models (such as WEPP (Nearing et al., 1989), GUEST (Rose et al., 1997), LISEM (De Roo et al., 1998), and SHE (Wicks and Bathurst, 1996)) which are somewhere between the grey- and white-box category. Formulation of process-based complex models is still underway (Favis-Mortlock, 2013), but from a practical point of view, the Universal Soil Loss Equation (USLE)/Revised Universal Soil Loss Equation (RUSLE) is still a preferred choice (Cao et al., 2015; Gessesse et al., 2015). Both USLE/RUSLE predict long-term average annual soil loss associated with sheet and rill erosion (Renard et al., 1991). Some major limitations of the predictive capability of RUSLE are non-inclusion of gully erosion, mass movement, and the deposition of the sediment in the modeled area (Merritt et al., 2003). Due to limited predictive capability (long-term average annual soil loss) of RUSLE, it tends to over-predict small annual soil losses and under-predict large annual soil losses (Risse et al., 1993). An exhausting list of field and watershed scale soil erosion and sediment yield models along with their mechanics, scales of implementation, and shortcomings are provided in a review by Pandey et al. (2016). Empirical relationships are also available to estimate the total sediment load using a relationship between drainage area and sediment delivery factor (Walling, 1983). In addition, scaling, regionalization, or global relationships are also available to estimate sediment yield for large catchments (McCuen, 2017; De Vente et al., 42 2007; Syvitski and Millman, 2007; Walling, 1983) but these relationships lack the rigor of physically-based models (Pandey et al., 2016). According to De Vente et al. (2007), use of drainage area alone as means to estimate sediment yield is problematic and information regarding climate, land use, soil, topography, and dominant erosion processes is key in obtaining good estimates of sediment yield. Another limitation is to know where sediment originates and its transportation downhill (De Vente et al., 2013). Models with GIS integration are well equipped to obtain spatial heterogeneity in topography, soil properties, land use characteristics, and hydro meteorological drivers at cell level to estimate soil erosion and total sediment load with reasonable accuracy (Pandey et al., 2016). However, these models are an attempt to simulate physical processes and require validation using observations. Sediment yield observations for model calibration are scarce and mostly unavailable (Merritt et al., 2003). Therefore, physically based models with calibration parameters require sensitivity analysis for reasonable estimates (Pandey et al., 2016). In this research study, a spatially explicit physically based approach is used, based on Borselli et al. (2008) at spatial resolution of elevation raster to compute erosion and sediment delivery at cell level. It employs a connectivity index to track total sediment reaching at the watershed outlet. For calibration, it uses a sigmoid curve based approach between sediment delivery ratio (SDR) and hydraulic connectivity (Hamel et al., 2017; Hamel et al., 2015; Vigiak et al., 2012; Jamshidi et al., 2014; Cavalli et al., 2013; Borselli et al., 2008). However, the results can vary considerably depending on the used calibration parameters. Therefore, in the absence of gauged data on sediment load we need representative calibration parameter values to parametrize the model. Concerning soil erosion rates and sediment yield, this study aims to answer the following two questions for small watersheds in the US Mid-Atlantic 43 region: 1) How much change is expected in the erosivity power and erosion rates from storm events (all and d85-d99) in excess of current values with increase in each percent of precipitation in the near future? and 2) How do we obtain representative calibration values for the SDR model to estimate sediment yield from small scale (~3km2) urban watersheds? 2. Methods and data 2.1. Experimental overview The small-scale watersheds in Howard, Montgomery, Anne Arundel, and Prince George’s county are selected which are representative of trends in precipitation and urbanization at county level. (More details on the counties were presented in experimental overview in chapter 2). Soil erosion is estimated for these watersheds using RUSLE as shown in equation 3.1. Gross erosion is a sum of different erosion types: sheet/rill, stream bank, landslides, wind, gully, and instream deposition. RUSLE is limited in its scope as it is only able to estimate sheet/rill erosion. Inclusion of other erosion types increases estimation complexity and also requires field measurement for validity of results. Therefore, current assessment is limited to sheet/rill erosion types. Total soil erosion is directly associated with rainfall erosivity. Rainfall erosivity has a direct relationship with change in total rainfall. This change can either occur in the form of increase in number of precipitation (wet) days or increase in average precipitation per wet day. According to Pruski and Nearing (2002), increase in average precipitation per wet day causes more erosion compared to number of wet days. In this research, daily precipitation totals are used, to estimate annual mean erosion instead of monthly or annual values because 44 estimates based on daily data capture intraannual variability (Segura et al., 2014). PRISM daily precipitation values are disaggregated to hourly time scale, in this dissertation, and hence can lead to possible underestimation of total erosion estimates (details are provided in Data and Methods section). Since precipitation intensity is of great significance in total erosion estimates, annual erosivity values computed in this study using Brown & Foster (1987) are compared to isoerodent maps of USDA for east coast of US (Renard et al, 1997) and GlobalR dataset (Panagos et al., 2017) to compute an adjustment factor for all events. This adjustment factor is then used to adjust annual erosivity of d85-99 events (more detail provided in Data and Methods section). Climate change scenarios in chapter 3 are constructed using CMIP5 data from 1981-2050. So the CF values are based on the ratio of mean of 2016-2050 to mean of 1981-2015and then used to scale 35 years of historical time series (1981-2015). SDR model (Borselli et al., 2008) computes soil loss at pixel level, using RUSLE, and delivery ratio for transport of eroded soil from each cell to the watershed outlet. Total sediment load at the watershed outlet is sensitive to the delivery ratio. The delivery ratio for ungauged sites leads to use of either empirical or mathematical models. In this research work, a drainage area-based empirical relationship is used in conjunction with the SDR model calibration approach. According to Borselli et al. (2008), the relationship between SDR and hydraulic connectivity is assumed to be similar to a sigmoid curve. Therefore, in order to find a saturation level of the delivery ratio, the SDR model is executed for a range of floating calibration values of kb between 1 and 4. The representative SDR values where the delivery ratio begins to saturate are compared to empirical relation based values. Furthermore, these 45 SDR values are compared to Chesapeake Bay Program (CBP) community model 5.3 based SDR estimates for Piedmont region. 2.2. Soil erosion estimation Estimation of erosion and overland sediment retention is computed using spatially explicit annual soil loss and delivery of those nutrients downstream (Borselli et al., 2008; Cavalli et al., 2013; López-vicente et al., 2013; Sougnez et al., 2011). Following Renard et al. (1997), RUSLE is used for annual soil loss estimation (equation 3.1). RUSLE has same formula as USLE, but has several improvements in determining factors: New isoerodent maps, time-varying approach for soil erodibility, new equation to reflect slope length and steepness, and new conservation-practice values. 𝐴𝐴𝑓𝑓 = 𝑅𝑅𝑅𝑅𝑅𝑅𝑆𝑆𝐶𝐶𝑃𝑃 3.1 where R is rainfall erosivity (MJ × mm (ha × hr)-1), K is soil erodibility (ton × ha × hr (MJ × ha × mm)-1); LS refers to slope length-gradient factor, C refers to crop-management factor, and P refers to the support practice factor. R is sensitive to I30 (30-minute rainfall intensity). Event based R is computed following Brown and Foster (1987) and the mathematical form is shown in equations 3.2 and 3.3. 𝑅𝑅 = 0.29[1 − 0.72𝑀𝑀𝑒𝑒𝑒𝑒−0.05𝑖𝑖𝑓𝑓 𝑎𝑎𝑎𝑎𝑎𝑎]𝑃𝑃𝑓𝑓𝐼𝐼24𝑚𝑚𝚤𝚤𝑚𝑚 3.2 𝑛𝑛 1 𝑅𝑅𝑦𝑦 = �(𝑅𝑅𝐶𝐶 𝑓𝑓 )𝑘𝑘 3.3 𝑘𝑘=1 where Re is rainfall erosivity, Ry is annual erosivity, Pt is total precipitation, I24max is maximum rainfall intensity during a 24-hour duration, n is number of events in a year, and 46 N= total number of years. Since PRISM precipitation is in total daily values and therefore requires disaggregation to fine grain temporal resolution for intensity estimates at hourly time scale. Precipitation disaggregation is performed using design storm for the study site. Design storm is an approach to view the time-varying intensity of rainfall during an event. NRCS has defined three typical rainfall distributions for various regions of United States. These distributions are dimensionless and typically 24-hour in length. Antecedent Rainfall Condition (ARC) for each distribution specifies dry to wet conditions. In this study, an ARC value of 2 is used for 2-year and 24-hour recurrence interval storm design. An automated GIS toolset, GISHydro (http://gishydro.eng.umd.edu/) with a database for the state of Maryland designed for hydrologic analyses is used to generate design storm distributions. The time interval of all distributions is 0.1 hour: these values are aggregated to hourly time scale. Hourly fractions are then used to distribute PRISM daily total precipitation values over 24- hour duration. Design storm distributions for all watersheds are provided in Appendix B. K factor is an index to define cohesive or bonding character of a soil type, which represents its susceptibility to erosion (Renard et al., 1997). Soil survey information regarding soil susceptibility is assembled by USDA in the form of county level shapefiles. These maps are extracted from the SSURGO database using soil explorer extension in ArcGIS. County level rock free maps are extracted, clipped to study site watershed extent, then exported in the form of raster maps at 30m spatial resolution. LS is obtained using method described in Desmet & Govers (1996), which requires elevation data. Elevation data are obtained from National Elevation Dataset at 30m spatial resolution and clipped to the extent of watershed. The formula to compute LS is provided in equation 3.4 47 (𝐴𝐴 + 𝐷𝐷2)𝑚𝑚+1 𝑚𝑚+1 𝑅𝑅𝑆𝑆 = 𝑆𝑆 𝑖𝑖−𝑖𝑖𝑛𝑛 − 𝐴𝐴𝑖𝑖−𝑖𝑖𝑛𝑛 𝑖𝑖 𝑖𝑖 𝐷𝐷𝑚𝑚+2 3.4 . 𝑒𝑒𝑚𝑚𝑖𝑖 . (22.13)𝑚𝑚 C factor is the ratio of soil loss from land cropped under specified conditions to corresponding loss under tilled. For example, a C value of 0.15 means that the erosion will be reduced to 15 percent of the amount that would occur under normal fallow conditions. These conservation management values are obtained from the USDA RUSLE database. P factor is a ratio of soil loss with contouring to straight row farming. These support practice factor values are also obtained from USDA RUSLE database. Each landuse type is assigned C and P values. MDP 2010 land use map is used for current climate and CC scenario. For CC+Urb, CBP LCLU projection map for 2025 is obtained from https://chesapeake.usgs.gov/phase6/map/, and converted to MDP LCLU categories. The conversion Table with explanation is provided in Appendix B. 2.3. Sediment yield estimation RUSLE provides erosion rate for each pixel, and the delivery of eroded soil downstream is computed using SDR. Empirical (USDA, 1983; US EPA, 2010; MDE, 2010) and spatially-explicit (Borselli et al., 2008) models are available for SDR computation. Empirical relationship is based on CBP community model 5.3. In this study, both approaches are used to compute SDR values for each study site. It is pertinent to determine convergence or divergence of SDR values from both approaches. Empirical relation (USDA,1983) (equation 3.5) and the spatially-explicit model (Borselli et al., 2008) (equation 3.6) are provided below. 48 𝑆𝑆𝐷𝐷𝑅𝑅 = 0.417762 × 𝐴𝐴−0.134958 − 0.127097 3.5 𝑆𝑆𝐷𝐷𝑅𝑅 𝑆𝑆𝐷𝐷𝑅𝑅 𝑚𝑚𝚤𝚤𝑚𝑚𝑖𝑖 = 1 + exp �𝐼𝐼𝐶𝐶 − 𝐼𝐼𝐶𝐶 𝑜𝑜 𝑖𝑖 3.6 𝑘𝑘 � Total export = ∑𝑝𝑝𝑖𝑖𝑚𝑚𝑓𝑓𝚤𝚤 𝑖𝑖 𝑈𝑈𝑆𝑆𝑅𝑅𝑈𝑈𝑖𝑖 × 𝑆𝑆𝐷𝐷𝑅𝑅𝑖𝑖 3.7 where 𝑆𝑆𝐷𝐷𝑅𝑅𝑚𝑚𝚤𝚤𝑚𝑚 is the maximum SDR with average value of 0.8, 𝐼𝐼𝐶𝐶𝑜𝑜and 𝑘𝑘 are calibration parameters. Additional details of connectivity index (IC) and its relationship with SDR are provided in Appendix B. Total yield at the watershed outlet is computed using equation 3.7. In addition, SDR using Edge of Field (EoF) approach is also calculated by substituting mean downstream flow length in equation 3.5. Percent change in erosivity and sediment load is calculated using following equation 3.8 where 𝑀𝑀𝑓𝑓 is for 2016-2050 and 𝑀𝑀ℎ is for 1981-2015. �𝑀𝑀 − 𝑀𝑀 � percent Erosivity (d85_99)𝑐𝑐ℎ𝑦𝑦𝑎𝑎𝑎𝑎𝑀𝑀 = 𝑓𝑓 ℎ × 100 3.8 𝑀𝑀ℎ 2.4. Soil conservation valuation According to Economic Research Service (ESR), soil conservation benefits could be viewed in dollar values reflecting what people, businesses, and government agencies would be willing to pay for 1-ton reduction in soil erosion (Hansen and Ribaudo, 2008). ERS provided these estimates at the county level and advised that these levels represent a lower- bound of benefits. Fourteen erosion categories are listed in the ERS report with twelve of them applicable to erosion by water. Water erosion types include: irrigation, marine recreational fishing, freshwater fisheries, marine fisheries, flood damages, road drainage ditches, municipal and industrial water use, municipal water treatment, steam power plants, 49 water based recreation, navigation, and reservoir services. Soil erosion estimates in this study are for sheet and rill erosion. Therefore, the selected water erosion types for soil conservation benefits include flood damage, road drainage ditches, municipal water treatment, and soil productivity. All values are in 2000 dollars and therefore adjusted to 2018 dollar values. The total benefits are calculated using equation 3.9 (Hansen and Ribaudo, 2008). 12 𝑇𝑇𝑀𝑀𝑇𝑇𝑦𝑦𝑀𝑀_𝑏𝑏𝑀𝑀𝑎𝑎𝑀𝑀𝑏𝑏𝑏𝑏𝑇𝑇𝑖𝑖 = � (𝑤𝑤𝑦𝑦𝑇𝑇𝑀𝑀𝑦𝑦_𝑀𝑀𝑦𝑦𝑀𝑀𝑠𝑠𝑏𝑏𝑀𝑀𝑎𝑎_𝑣𝑣𝑦𝑦𝑀𝑀𝑣𝑣𝑀𝑀𝑖𝑖,𝑖𝑖 × 𝑤𝑤𝑦𝑦𝑇𝑇𝑀𝑀𝑦𝑦_𝑀𝑀𝑦𝑦𝑀𝑀𝑠𝑠𝑏𝑏𝑀𝑀𝑎𝑎_𝑇𝑇𝑀𝑀𝑎𝑎𝑠𝑠𝑖𝑖) 3.9 𝑖𝑖=1 where subscript i is for county and j for water-related benefit categories. Total benefits only include flood damage, road drainage ditches, municipal water treatment, and soil productivity. 3. Results and discussion 3.1. Rainfall erosivity According to the RUSLE handbook (Renard et al., 1997), a minimum cumulative rainfall of 12.7 mm is the lower threshold for erosive rainfall events. Therefore, annual erosivity is presented as the sum of energy possessed by all storm events that qualify as erosive rainfall events greater than a rainfall depth of 12.7 mm. The total number of events are about 1000 for all study area watersheds during 1981-2015. The annual erosivity estimates, based on Brown and Foster (1987), for these events are around 750 (MJ × mm × (ha × hr × yr)-1), whereas both isoerodent maps of USDA and GlobalR dataset have erosivity values between 3000-3500 (MJ × mm × (ha × hr × yr)-1) for all watersheds. This shows that erosivity values estimated using Brown and Foster (1987) are underestimated by a factor of 4.5 for all study sites. This could largely be associated with the coarse temporal resolution of 50 precipitation data (PRISM daily totals) and its disaggregation to hourly rate using design storm. Total number of events for d85-99 are about 100 and have adjusted annual erosivity of about 1150 (MJ × mm × (ha × hr × yr)-1) across all study sites. In comparison to total annual erosivity, d85-99 events possess around 30 percent of total annual erosive energy. Table 3.1. Rainfall erosivity estimates and adjustment factors ỻ Total Adjusted Total Annual *Annual County events Erosivity Erosivity Adjustment events Annual (>12.7mm) Factor (d85-99) Erosivity (d85-99) Howard 1048 759 3393 4.47 99 1185 Montgomery 988 655 3075 4.69 101 1126 Anne Arundel 1078 765 3571 4.66 99 1279 Prince George’s 1027 733 3294 4.49 97 1184 ỻ 1981-2015; * USDA (1987) and GlobalR dataset (Panagos et al., 2017) Rainfall erosivity units: (MJ × mm)/(ha × hr × yr) The percent change in erosive power of all events (>12.7 mm) and d85-99 events during 2016- 2050 relative to the historical period 1981-2015 is presented in Figure 3.1. Except CanESM2, all models show a possible increase between 2-18 percent in the erosive power of all events. Both GFDL-ESM2g and IPSL-CM5a model based erosivity change is around 4 percent in all study sites. Moreover, the NorESM1-m model shows about 18 percent change in erosivity in all watersheds. On the other hand, percent change in erosivity of d85-99 events is anticipated to be between 0-30 percent in all study sites. 51 3.2. Soil erosion rates According to Nearing et al. (2004), erosion rate may change about 1.7 percent for each 1 percent change in precipitation during the latter half of 21st century due to change in total rainfall. Rainfall erosive power and erosion percent change estimates of total events in this research show likelihood of even higher values of about 2 percent increase for each 1 percent precipitation change during 2016-2050 as shown by CF values in Table 3.2 and percent change values in Table 3.3 (percent change in erosion values are the same as provided in Table 3.2). For d85-99 events, GFDL-ESM2g, IPSL-CM5a, and NorESM1-m models show an increase of about 3.5-5 percent for each 1 percent change in precipitation in all study sites during 2016-2050, except for Anne Arundel county watershed. The different response of the Anne Arundel watershed is likely attributable to the lower number of events, even after scaling using CF for 2016-2050, within d85-99. Total soil erosion under the CC+Urb scenario is considerably lower than CC scenario in all study watersheds (total soil erosion estimates are provided in Appendix B under CC and CC+Urb scenarios). It is mainly due to projected conversion of natural land into built-up areas and open urban spaces. In Howard county watershed, deciduous forest and pastureland is converted into low-, medium-density residential, and open urban land. In Montgomery county watershed, pastureland and deciduous/evergreen forests are converted into either agricultural breeding, built-up areas, or open urban land. In Anne Arundel county watershed, evergreen and mixed forests are converted into low-density residential and open urban land. In Prince George’s county watershed, bare ground and deciduous forest converted into low-, medium-, and high-density residential areas and open urban land. 52 Table 3.2. CF values based on CMIP5 climate models County CanESM2 GFDL-ESM2g IPSL-CM5a NorESM1-m Howard 0.99 1.02 1.01 1.08 Montgomery 1.0 1.03 1.0 1.08 Anne Arundel 0.99 1.02 1.01 1.08 Prince George’s 0.98 1.01 1.03 1.08 Figure 3.1. Percent change in rainfall erosive power of total and d85-99 events Table 3.3. Erosive power change for each percent of precipitation (2016_2050 – 1981_2015) County CanESM IPSL-CM5A- GFDL- NorESM1-2 MR ESM2G M Howard -2.4 2.3 3.3 2.2 Montgomery 0.0 2.0 0.0 2.3 Anne Arundel -2.5 2.2 3.1 2.2 Prince George's -2.2 3.1 2.1 2.3 Howard 2.8 4.0 5.1 3.6 Montgomery 0.0 3.7 0.0 3.5 Anne Arundel 0.9 1.0 -0.2 2.6 Prince George's -0.6 4.9 3.5 3.7 3.3. Sediment yield The percent change in maximum SDR values computed with floating values of calibration parameter kb are shown in Figure 3.2. CBP model computes SDR by taking the 53 Total events Total events (d85-99) (> 12.7 mm) ratio of the sediment load at the edge of stream (EoS) to EoF load generated in the watershed. SDR based on empirical relation (total drainage area and mean downstream flow length) in equation 3.5 and spatially explicit model (equation 3.6) are provided in Table 3.4. SDR (DrainageAreatotal and EoF) is between 25th and 75th percentile of estimates by CBP Model 5.3 for Piedmont region. Percent change in SDRmax is around 5 percent for calibration parameter kb value of 3.5 for all study sites. It is noteworthy that saturation of max SDR factor is similar to estimates by equation 3.5 using total Drainage Area. SDR using mean downstream flow length distance is relatively smaller than other estimates (still within the 75th percentile of CBP’s estimates). Slight variation in SDR (EoF) results is possibly related to use of single mean downstream flow length instead of individual flow lengths from each land use to stream edge. Figure 3.2. SDR estimates based on equation 3.6 Table 3.4. Empirical and mathematical model based SDR values County SDR (kb)* SDR (DAtotal) SDR (EoF) Howard 0.29 (3.5) 0.29 0.26 Montgomery 0.29 (3.5) 0.27 0.25 Anne Arundel 0.28 (3.5) 0.28 0.25 Prince George’s 0.29 (3.5) 0.28 0.22 54 The sediment load generated due to < d85, and d85-99 during the historical period and percent change in load due to d85-99 events compared to the future period 2016-2050 are provided in Figure 3.3 and 3.4 respectively. Change in load due to d85-99 events exhibit similar magnitudes as observed for erosivity in Figure 3.1 in all study sites. Total sediment load generated due to d85-99 events is around 35 percent and events less than d85 account for about 55 percent of total load across all study sites. It means that infrequent events with depth greater than d99 contribute 10 percent of total load generated in all of these watersheds. The anticipated increase in sediment load, generated due to d85-99 events, ranges from 1-10 percent. It means that based on projection of NorESM1-m model, these events may contribute up to 45 percent of the total load in the future. Total sediment load under CC+Urb scenario is considerably less in all study watersheds due to aforementioned projected conversions of natural land into built-up areas and open urban spaces. 55 Figure 3.3. Total erosion (tons/watershed) Figure 3.4. Percent change in erosion of d85-99 events 3.4. Soil conservation benefits The soil conservation cost dollar values at county level from ERS-USDA (2008) are provided in Table 3.5. These numbers can be viewed as cost estimates that would be required to reduce soil erosion by 1 ton. Total annual soil erosion reduction cost for all study area watersheds is mostly higher than historical values, as shown in Figure 3.5 (y-axis is logarithmic) under all climate change scenarios. The models GFDL-ESM2g and NorESM1-m are showing 5-28 percent increase in cost compared to current management cost. 56 Table 3.5. Soil conservation cost dollar values from ERS-USDA Flood Road Municipal Source Damage Drainage Water Soil Productivity Ditches Treatment *ERS-USDA (2008) 0.77 0.2 0.26 1.27 ỻ 2018 $ 1.13 0.29 0.39 1.86 adjustment * Year 2000 dollar value; ỻ CPI calculator (https://data.bls.gov/cgi-bin/cpicalc.pl) Figure 3.5. Annual soil erosion reduction cost based on climate models 4. Conclusions CMIP5 climate models show an increase of about 1-8 percent in mean precipitation for 2016-2050. Therefore, a change in total rainfall is expected in the form of increase in average precipitation per wet day and number of wet days. On average, the erosive power of extreme rainfall events is expected to increase, during 2016-2050 relative to 1981-2015, by a factor of 2 and higher of all events for each 1 percent increase in total precipitation except in Anne Arundel county. For d85-99 events, an increase of about 3.5-5 is expected except in Anne 57 Arundel county. All models except CanESM2 show consensus on an increase of erosive power of extreme events for all watersheds except in Anne Arundel county. Under CC scenario, erosivity is primary driver in changing soil erosion in all study watersheds. However, under CC+Urb scenario a significant decrease in soil erosion is expected. The decrease in soil erosion can be mainly attributed to conversion of exposed soil to sealed surface. The main finding and point of concern is change in average magnitudes of erosivity and soil erosion rates (for each percent of change in precipitation) during 2016-2050 which are expected to be higher based on our estimates in small scale watersheds in U.S. Mid- Atlantic region. RUSLE has limited predictive capability due to estimation of only long-term annual average soil loss. So when we use a range of extreme precipitation events in determining their relative contribution in total soil loss and sediment load, the estimates reflect lower end of predicted output. In reality, the soil loss and hence total sediment load could be larger. According to Kinnell et al. (2010), the soil loss estimates from more process-oriented model such as WEPP have a similar performance as USLE/RUSLE. Inclusion of gully erosion, mass movement, and the deposition of the sediment in total soil loss require weighted adjustments. Expansion of gullies due to climate change would contribute increased erosion that is not reflected in the RUSLE modeling performed here. Therefore, inclusion of gully erosion and sediment deposition can possibly improve our sediment load estimates at the watershed outlet and is clearly a topic of future research. The spatially explicit physically-based GIS models for soil erosion estimation are effective in integration of spatial heterogeneity in topography, inclusion of soil properties, 58 land use, and hydro-meteorological information at grid cell or watershed-scale. Sediment transport is much easier to track using parametrization and routing on cell-to-cell basis. However, the main challenge is the calibration of such models especially in the absence of gauged dataset. This study shows that the use of SDR saturation approach to determine kb is promising. However, this study only focuses on small scale (~3km2) urban watersheds. Therefore, kb calibration approach requires more testing under diverse land use conditions and bigger watersheds. Soil erosion leads to loss of nutrients and cropland, and to downstream sedimentation issues. Apart from direct impact of extreme rainfall to cause soil erosion, organic carbon stored in soil is also released. From an ecosystem services point of view, soil erosion affects carbon storage and sequestration which impacts climate regulation. Carbon storage and sequestration estimates require information about four carbon pools (aboveground biomass, belowground biomass, soil, and dead organic matter). The valuation in this study shows that total annual soil conservation cost is mostly higher for all study watersheds but these estimates do not include carbon storage benefits. The soil productivity cost estimates used in this research only include loss of topsoil. Therefore, the soil conservation valuation provided in Figure 3.4 shows lower end of estimates. The conversion of natural land (forest, floodplain wetlands, and other wetlands) into sealed surface, as observed in this research, results in less erosion and sedimentation due to less exposed soil surface. However, it leaves urban watersheds more prone to flooding and release of carbon stored. 59 Chapter 4: Stormwater management and ecosystem services 1. Introduction Numerous studies assess the resilience of existing urban stormwater infrastructure in a built environment under projected climate change and urbanization (Miller and Hutchins, 2017; Saraswat et al., 2016; Gersonius et al., 2012; Rosenberg et al., 2010; Hamin and Gurran, 2008). However, a major limitation is the understanding of how use of green stormwater management practices affects ecosystem services (Prudencio and Null, 2018). In addition, use of climate projections presents some serious challenges such as direction, and magnitude of climate change (Willems et al., 2012; Kwadijk et al., 2010; Dessai and Slujis, 2007; Cox and Stephenson, 2007) and their subsequent use for hydrological assessment (Teutschbein and Seibert, 2012; Fowler et al., 2007; Wood et al, 2004). Green stormwater management practices are effective to control runoff volume and also provide ecosystem services (Dhakal and Chealier, 2016; Vogel et al., 2015). A variety of stormwater management studies providing or addressing a sub-category of ecosystem services exists in literature: provisioning (Gittleman et al., 2017; Ackerman, 2012), regulating (Ishimatsu et al., 2017; Klimas et al., 2016; Chen et al., 2014; Dohetry et al., 2014; Berland and Hopton, 2014), cultural (Attwater and Derry, 2017; Kati and Jari, 2016), and social (Kopecká et al., 2017; Hassall, 2014). According to a detailed review by Prudencio and Null (2018), the concentration of stormwater management research related to ecosystem services focus on the parcel size, but lack studies at the watershed scale. In addition, they also find a lack of, robust metrics for ecosystem services quantification studies in the developing world, integration of physical and engineering with social science, and overlapping green 60 stormwater management terminologies, and they detect barriers in implementation (such as lack of financial and political support). The use of the triple bottom line approach (US EPA, 2015) which includes environmental (improved air quality, habitat restoration), economic (job creation, development, and increased property values), and social benefits (recreational opportunities and reductions in crime) is a needed direction to obtain a holistic view of green stormwater management practices benefits as a stormwater guide for local governments (Environmental Finance Center-UMD, 2017). Some of the major limitations of such a proposed holistic view are the lack of valuation data which often requires benefit transfer approach (Johnston et al., 2015; Freeman, 2003), modeling limitations (Jayasooriya and Ng, 2014), knowledge gap regarding benefits and effectiveness of single or a combination of different practices at watershed-scale (Connop et al., 2016; Vogel et al., 2015), and rarely quantified social benefits (Prudencio and Null, 2018). In this study, we explore modeling limitations and benefits of green stormwater management practices to understand how specific practices affect ecosystem services at watershed scale. Climate adaptation strategies are constructed using either predictive top-down or bottom-up approaches (Dessai and Slujis, 2007; Carter et al., 2007). Top-down approaches, even though widely applied for climate scenarios impact assessment (Carter et al., 2007; Adger et al., 2007), are limited due to strong reliance on climate projections, which require downscaling for regional or local hydrologic impact assessment (Willems et al., 2012; Chen et al. 2011b; Schmidli et al. 2006). Bottom-up approaches focus more on adaptive capacity and adaptation measures, are more independent from climate projections (Kwadijk et al., 2010), but still have constraints such as lengthy assessment time, complexity to compare all drivers, and greater reliance on judgment than quantitative results, and thus are often 61 considered to be more promising than they can deliver (Patt et al., 2005; Füssel, 2007). In addition, adaptation under uncertainty in climate, technology, and socio-economic and political factors trouble decision makers (Kwakkel et al., 2016; Haasnoot et al., 2013). Therefore, capacity and effectiveness of stormwater management systems under changing climate through adaptive measures is an active and growing research domain (Moore et al., 2016). The adaptation tipping points approach (ATPs) offers an alternative to deal with climate change vulnerability (Kwakkel et al., 2016; Gersonius et al., 2015; Haasnoot et al., 2013; Kwadijk et al., 2010). In comparison to top-bottom approaches where a strategy is devised at the end, ATP enhances the ability of the system to deal with future change. ATP approach informs on durability of current measures and timing of when adaptation measures are needed (Kwadijk et al., 2010). In this study, the concept of ATPs is utilized to construct stormwater management adaptation pathways (SMPs) under changing climate and urbanization using climate stress and optimal cost considerations. ATP approach is used to enhance the system by providing enough stormwater infrastructure to be effective under future climate to control set targets (such as reduction of 80percent, 90percent, or 100percent surface runoff). The target reduction suggestion can be made through stakeholder’s involvement or thresholds set by EPA. However, a threshold of 99 percent or greater is used as part of experiment in this research study (more details are provided in experimental overview section). The enhancement of stormwater infrastructure resilience, under climate change and urbanization, requires adaptation strategies such as gray, green, and combined gray and green 62 infrastructure (Moore et al., 2016). The ‘conventional gray infrastructure approach’ focuses only on quickly removing stormwater from urban areas whereas from a ‘sustainable management perspective’ it is seen as a resource (stormwater as a resource for groundwater replenishment, rainwater collection and storage for use at home, market place, and landscape irrigation) with benefits for society and environment (Goulden et al., 2018; Fletcher et al., 2015; Barbosa et al., 2012; Roy et al., 2008; Mitchell, 2006). Sustainable stormwater management is possible with implementation of green practices in urban areas (Fletcher et al., 2015; Benedict and McMohan, 2006). Green infrastructure (GI) refers to low-impact development (LID), sustainable urban drainage systems (SUSD), water-sensitive urban design (WSUD), and low-impact urban design and development (LIUDD) (Fletcher et al., 2015; Elliott and Trowsdale, 2007). Apart from the primary objective of flood protection (Demuzene et al., 2014), GI also improves air quality (Jayasooriya et al., 2017) and human health (Tzoulas et al., 2007), preserves ecosystems (Benedict and McMohan, 2006), and reduces the urban heat island effect (Gill et al., 2007). The potential of GI to mitigate flooding, even partially, led to consideration of GI as an integral component of climate change adaptation (Gaffin et al., 2012). However, the categorization and synthesis of GI practices with ecosystem services is less studied (Prudencio and Null, 2018). With realized benefits of GI, more research is needed to assess incorporation of structural and non- structural (for example, bioretention as structural and sheetflow to conservation area or residential downspout disconnection as non-structural) GI practices with conventional practices, and cost of GI compared to conventional engineering approaches for climate change adaptation (Moore et al., 2016). Therefore, this research explored stormwater management pathways (SMPs) as part of adaptation under climate change and climate 63 change plus urbanization scenarios. These pathways include gray, green, and gray combined with green infrastructures. The anticipated increase in urban population (United Nations, 2018), extreme weather events (Georgakakos et al., 2014; Kirtman et al., 2013), and urbanization (Jiang and O’Neill, 2017; Brown et al., 2014) require sustainable management of stormwater as a resource for potential use (such as household use and landscape irrigation) in cities (Barbosa et al., 2012; Sundberg et al., 2004). This requires an integrated urban water management modeling approach (Kirshen et al., 2018; Zhou, 2014; Bahri, 2012). However, the major shortcomings of urban catchment modeling for integrated assessment are complex urban physical systems and limited data of existing stormwater infrastructure (Salvadore et al., 2016). The ideal path forward, in carving adaptation strategies, is through detailed space-time analysis of existing stormwater infrastructure with use of an urban hydrologic model with reasonable accuracy under anticipated change in climate and urbanization. But due to data availability constraints, it is assumed in this study that existing stormwater infrastructure has the capacity to store all of stormwater generated under recent climate (1981-2015) with 99 percent or greater effectiveness to reduce runoff. Therefore, due to the complexity of urban catchment models and the difficulty to acquire existing stormwater infrastructure information, this chapter aims to: 1) Is it possible to obtain good flow simulation accuracy for stormwater assessment by using a simple Curve Number watershed model (CWM), developed in this dissertation, for small-scale urban catchments?, and 2) Can green infrastructure provide added benefits to retain/replace/mitigate increased runoff and erosion effects under climate change and urbanization at watershed scale? 64 2. Methods and data 2.1. Experimental overview A small-scale watershed in Montgomery county is selected for stormwater assessment and all four watersheds are selected for flow simulation model assessment (watershed selection criteria is same as explained in chapter 2 and 3). It is assumed in this study that current stormwater infrastructure is 100 percent effective in retaining surface runoff under the current climate (1981-2015). A BMP space-time analysis is set up in this regard for each of the adaptation pathways (more detail is provided later): i) gray, ii) green, and iii) gray and green infrastructure. SUSTAIN model (more details of this model are provided under data and methods section) is calibrated and optimized to produce about 99 percent or greater effectiveness in terms of storing surface runoff and providing cost effective solutions. Each model simulation is performed with 1000 iterations using NSGAII optimization technique (Deb et al., 2002) to generate cost effective solutions. The choice of BMP types for SUSTAIN model simulations is based on inspection of spatial geodatabase of stormwater infrastructure in Montgomery county (courtesy of Maryland Department of Environment). The 100 percent effectiveness of BMPs during the time period 1981-2015 is termed as “Current Portfolio”. Two scenarios are used to stress test effectiveness of current portfolio under: i) climate change (CC), and ii) climate change and urbanization (CC+Urb). To account for percent change in precipitation in future, change factor (CF) based percent change in precipitation (provided in chapter 2) is used to scale historical (1981-2015) PRISM time series. Use of daily average precipitation for flow generation can average out high flow events. Therefore, a design storm approach (details provided in chapter 3) is used to 65 disaggregate precipitation from daily to hourly time scale. Disaggregation is based on NOAA’s Atlas 14 precipitation depth dataset (Bonnin et al., 2006). Disaggregated time series are used along with conceptualized and proposed curve number watershed model (detailed under methods) to generate aggregate HRUs (Hydrologic Response Units). Curve number watershed model based flow values are validated across four small scale urban watersheds (description of study sites provided in chapter 2) with varying impervious areas. The main assumption is that this model performs well for small scale medium to highly impervious watersheds and therefore require testing over watersheds with low to high imperviousness. Under CC+Urb scenario, CF based percent change in precipitation is used in conjunction with future land use projections for 2025. Finally, total stormwater storage cost, for each selected solution, under each adaptation pathway (gray, green, and gray&green) and under both CC and CC+Urb scenarios are provided in terms of percent change in cost from current to future portfolios. Under gray and green pathways, only wet ponds and enhanced bioretention infrastructures are used, respectively. However, the combined gray and green pathways contain infiltration trenches, enhanced bioretention, dry pond, and wet pond BMP types. The selection of these BMPs is based on inspection of existing infrastructure in the study site (more details are provided in Appendix C). Assessment points for all cases are watershed outlets. Results are presented in the form of Stormwater Management Pathways (SMPs) where each path has one to many relationships with each adaptation option (gray, green, and gray&green) to provide all possible solutions. Each pathway is a combination of different stormwater management practices and it changes over time, during optimization in SUSTAIN model, to adapt to 66 increase in surface runoff under future changes. A co-benefit analysis of GI is also carried out for a gray&green solution under climate change and adaptation. 2.2. Watershed modeling Stormwater management requires watershed modeling, BMP modeling, and cost effective solutions (Sun et al., 2016). A major challenge, as part of decision-making, is the selection of the best combination of BMPs that are cost effective (Lee et al., 2012). The EPA integrated a range of relevant tools to analyze BMPs at the watershed-scale into a System for Urban Stormwater Treatment and Analysis INtegration (SUSTAIN) to provide a decision- support system (Shoemaker et al., 2012; Shoemaker et al., 2009). SUSTAIN is an integration of hydrologic, hydraulic, water quality simulations, and BMP cost effectiveness package (Chen et al., 2014; Shoemaker et al., 2009). It provides best combination of stormwater management practices among many options available. According to Shoemaker et al. (2009), the GIS version of SUSTAIN consists of a framework manager, a post-processor, and five simulation modules (land module, BMP module, conveyance module, and an optimization module). However, it requires ESRI’s ArcGIS 9.3 to support framework functionality. Alternatively, a non-GIS version of SUSTAIN is available in scripting format and has the same set of tools/modules except interactive feature of the GIS version. Both versions support the same set of management practices (Table 4.1). More detail on input data requirements, and parametrization of SUSTAIN is provided in data and methods section. The hydrologic simulation in SUSTAIN is either implemented internally, using EPA’s integrated stormwater management model (SWMM7) , or externally using a model of 7 https://www.epa.gov/water-research/storm-water-management-model-swmm 67 preference. However, calibration and validation of SWMM model requires tuning of a large number of input parameters. It is ideal but very time expansive to setup a hydrologic model with medium/high input complexity, such as SWMM, especially for more than one study site. According to Moriasi et al. (2007), medium to high complexity models do not necessarily guarantee good results at daily to sub-daily time scales. As part of external hydrologic modeling, a user can define its own continuous time series whether it is generated using HSPF (Hydrologic Simulation Program - Fortran) or any other model. In any case, it has to satisfy the required input format to run SUSTAIN. SUSTAIN model requires land use definition in the form of HRUs. HRU is the smallest spatial unit of a hydrologic model where response of similar land use, soils, and slope are lumped together (Kalcic et al., 2015). The watersheds in this study are at small spatial scales (~ 3km2) and have a degree of imperviousness of 40percent or higher. It means that factors (such as disconnected impervious areas [DCIA] and other unknowns) which can modulate total surface runoff volume at the outlet of a watershed may have little impact. Unknowns are basically factors which aren’t parametrized such as water intercepted by plants, taken up by soil, and becomes part of swale. Developed curve number watershed model is simple and based on Curve Number approach. It employs initial abstraction from precipitation. Whereas in real world the abstraction is based on many other considerations such as how much water is intercepted by plants, taken up by soil, or becomes part of swale flow. In short, the abstraction by each land use type is modeled in detail which leads to more complex models. The calibration and validation of those models is time expensive and don’t necessarily guarantee very good results. A full length comparison between different models and their performance is provided in Moriasi et al. (2007). 68 Table 4.1. BMP management practices in SUSTAIN BMP Class BMP Type Bioretention Wet pond Cistern Dry pond Class A Infiltration trench Green roof Porous pavement Rain barrel Regulator Class B Swale Class C Conduit Buffer strip Area strip 2.3. Curve number watershed model (CWM) overview An array of models is available for assessment of stormwater management and economic aspects of adaptation scenarios (Jayasooriya and Ng, 2014). These models differ in terms of their attributes such as potential use, temporal resolution and scale, catchment and drainage representation with spatial scale consideration, runoff generation, flow routing, and management practices (Elliott and Trowsdale, 2007). A common feature among all of these models is the hydrologic modeling which is either conceptual or physically based (Refsgaard, 1997) with further division into lumped, distributed, hydrologic response unit (HRU), grid- based spatially distributed, and urban hydrologic element depending on spatial description of the catchment (Salvadore et al., 2015; Arnold et al., 1998; Refsgaard, 1996). Apart from process and scale concerns, a major challenge is the modeling of spatially-distributed hydrological processes in urban areas (Salvadore et al., 2015) such as precipitation, evapotranspiration, depression storage-overland flow-and runoff, stormwater drainage 69 systems and combined sewer systems, retention basins and stormwater management systems, infiltration and subsurface processes, and direct/indirect ground recharge (Fletcher et al., 2013). The intrinsically complex nature of urban hydrologic modeling is dealt with through parameter reduction (Krebs et al., 2014; Dotto et al., 2011) and selection of physically meaningful parameters (Vieux and Bedient, 2004). In addition, the major shortcoming of urban catchment modeling is the complex physical system which results in more complexity in model calibration (WEF, 2012) and can be simplified with use of theoretical process interactions (Salvador et al., 2015). In view of urban modeling complexity and limitations, in this chapter a parsimonious Curve Number watershed model (CWM) is conceptualized and developed using GIS environment and finally discretized into Hydrologic Response Unit (HRUs) with similar land use and soil characteristics. The conceptualization of CWM follows three basic functions of a catchment according to Wagener et al. (2007): i) partition of precipitation into different flowpaths (interception, infiltration, percolation, runoff etc.), ii) storage of water in different parts of the catchment, and iii) release of water from catchment in the form of channel flow. The CWM is conceptualized to produce HRUs using land use maps, soil maps, hourly precipitation, and NRCS CN method (details of NRCS CN method are already provided in chapter 2). A schematic view of CWM is provided in Figure 4.1. CWM offers low input data complexity and is specifically applicable for small urban watersheds. Since the main limitation of this model is to generate base flow so the flow output from CWM over larger areas with diverse land use types may lead to higher errors. A 2-digit classification system is used to convert each land use type and associated hydrological soil groups (HSG) into a CN map (more detail on 2-digit classification system is provided in chapter 2). There may be 70 instances of more than one pixel with the same CN. In such cases, the surface runoff generated from the same CN pixels are added together to produce an aggregated response of similar CN pixels at 1-acre of each HRU. So the output of CWM is expressed as in-acre/hour for 1-acre of each HRU. The validity of the model can be affected by non-inclusion of factors such as disconnected impervious areas and unknowns, is assessed in this study using USGS stream gauge data. Firstly, the base flow is removed from observed flow data as described in Arnold and Allen (1999). Base flow filter passes over observed streamflow data three times (forward, backward, and forward). Daily base flow values are generated using an online SWAT BFlow generator (http://www.envsys.co.kr/~swatbflow/), which follows the method described in Arnold and Allen (1999). Filtered surface runoff and base flow equations (4.1 and 4.2) are provided below. Daily base flow time series values are filtered from USGS streamflow time series before comparison with CWM-based flow time series. 71 Figure 4.1. Curve number watershed model (CWM) conceptual basis 1 + 𝛽𝛽 𝑞𝑞𝑓𝑓 = 𝛽𝛽𝑞𝑞𝑓𝑓−1 + 4.1 2 × (𝑄𝑄𝑓𝑓 − 𝑄𝑄𝑓𝑓−1) 𝑏𝑏𝑓𝑓 = 𝑄𝑄𝑓𝑓 − 𝑞𝑞𝑓𝑓 4.2 where qt refers to filtered surface runoff at daily time step (t), Q is original streamflow, β is filter parameter (0.925), and bt is baseflow. Observed streamflow, of mostly small watersheds (~ 3-12 km2), is scaled using a Drainage Area-Ratio (Gianfagna et al., 2015) method. Scaling of streamflow data from large watersheds is only performed if small watersheds are not located within proximity of study sites. A list of all USGS stream gauges along with their respective drainage area used during validation for all four small watersheds are provided in Table 4.2. It is important to highlight that CWM model skill is tested over four study sites whereas SMPs and portfolios are only constructed for Montgomery county watershed. The 72 CWM model testing over multiple sites is necessary due to the underlying assumption of its applicability for impervious areas. Table 4.2. USGS gauges for validation County USGS gauge number Drainage Area (km2) Howard 01593450 6.39 01593370 3.1 01644375 3.49 01643395 3.03 Montgomery 01644390 11.6 01647850 7.09 01649150 2.69 Anne Arundel 01589500 12.87 Prince George’s 01653600 102.3 01651000 127.9 2.4. SUSTAIN setup and datasets The modeling framework of SUSTAIN can be employed at multiple scales for watershed planning. It offers a tiered approach to optimize best solutions at multiple scales to prepare an aggregated response (SUSTAIN application scales schematic provided in Appendix C). Non-GIS version of SUSTAIN requires set of inputs into multiple cards in the script file. An input script for “gray and green” flex approach under CC scenario for the Montgomery county watershed is provided in Appendix C. The cards in script file control the simulation process, define inputs (BMP site characteristics and parameters, and various pollutants related parameters), land use definition, land to BMP routing network, BMP site routing network (complete details of inputs are available in script provided in Appendix C). The routing architecture used in this research work is provided in Figure 4.2. 73 Figure 4.2. Routing architecture for BMPs SUSTAIN’s input data requirement is application specific. Since this study only focuses on the water quantity aspect of urban hydrology, water quality inputs and parameters are filled with dummy values (as shown in script file provided in Appendix C). For HRUs, a land use raster map, a soil raster map, and hourly precipitation time series are needed. A set of inputs includes: i) the ratio of maximum to mean velocity under typical flow conditions (CRRAT), monthly Potential Evapotranspiration (PET), definition of individual HRUs, an approximate drainage area to be treated by each BMP, BMP site parameters (such as width, length, orifice height, exit type, weir height and width etc.), BMP site bottom soil and vegetation characteristics (such as infiltration method, soil depth, soil porosity, soil field capacity, soil wilting point, average value of soil capillary suction), infiltration method, BMP site initial moisture content, land to BMP routing network, optimization control with selection of technique and output type options, BMP cost functions, assessment point and evaluation factors. The input values for Montgomery county watershed are provided in the script file in Appendix C. 74 MDP land use 2010 and SSURGO soil dataset raster maps at 30 m spatial resolution are clipped to the extent of the watershed. PRISM daily precipitation totals are disaggregated to hourly using design storm for the study site (as described in chapter 3). CWM based HRUs are generated using land use, soil, and hourly precipitation time series. ASCII files are generated for each HRU as required in specific format for SUSTAIN model run as shown in Figure provided in Appendix C. External land simulation, like performed in this research using CWM (section 2.3), requires hourly data in following order: label column, followed by year, month, date, hour (0-23), minute (0-59), surface runoff volume in in-acre per time step, ground water recharge volume in in-acre per time step, and number of pollutant loading per time step. The drainage area of each HRU is also computed to provide as an input in card 790 (land to BMP routing network) of SUSTAIN script file. The study area watershed is about 40 percent impervious therefore a typical value of 1.5 is used for CRRAT. Monthly PET is computed using Thornthwaite method (equation 4.3) which requires mean monthly temperature for the months of January to December. 𝑃𝑃𝑈𝑈𝑇𝑇𝑖𝑖 = 1.6 × (10𝑇𝑇𝑖𝑖/𝐽𝐽)𝚤𝚤 4.3 where 𝑐𝑐 = 0.000000675𝐽𝐽3 − 0.0000771𝐽𝐽2 + 0.01792𝐽𝐽 + 0.49239, 𝐽𝐽 = ∑12𝑖𝑖=1(𝐼𝐼𝑖𝑖), and 𝐼𝐼𝑖𝑖 = (𝑇𝑇 /5)1.514𝑖𝑖 . Temperature is expressed in Celsius degrees. This method is a function of monthly temperature values, and latitude (degrees). The output is in cm and converted into in/day for model control card 700 in SUSTAIN input script. An approximate drainage area to be treated by each BMP is provided. Approximate values for BMP site parameters, bottom soil and vegetation characteristics are provided in cards 725, 735, and 745. Holtan infiltration method (Holtan, 1961) is used with growth index inputs shown in equation 4.4 (GRI = 75 growth index of vegetation, A = infiltration capacity, Sa = available storage in surface layer, fc = constant final infiltration rate). 𝐹𝐹 = 𝐺𝐺𝑅𝑅𝐼𝐼 × 𝐴𝐴 × 𝑆𝑆1.4𝚤𝚤 + 𝑏𝑏𝚤𝚤 4.4 Monthly growth index values are only required for Bioretention BMP and therefore provided in card 745. All water quality related cards are provided with place holders as they have no effect on results. NSGAII, which is known as a fast and elitist (high speed) genetic algorithm to find best possible solutions, optimal cost in current case, from given population (Deb et al., 2002), is used with 1000 iterations. Following Mateleska (2016), BMP cost values are provided for $/ft3 of storage volume in 2018-dollar value. At the end in card 815, watershed outlet is selected as assessment point with 0-100 as target values for evaluation. 2.5. Quantification of ecosystem services The metrics used in this research to quantify ecosystem services from GI practices (infiltration trench and bioretention) are runoff reduction (regulating service), groundwater recharge (provisioning service), and reduced water treatment (regulating service). The estimates are obtained using an online GI screening tool (http://greenvalues.cnt.org/national/calculator.php) developed by the center of neighborhood technology (CNT). The ground recharge quantification is based on a simple estimation to provide an approximation following Walton (1965) and Locke et al. (2005). The details of the method are as follows: i) the difference between discharge leaving green and conventional infrastructure site is computed, ii) the potential amount of infiltration (average annual rainfall/specified rainfall) for specified rainfall depth (which is 2.4 inches and corresponds to total surface storage volume of 164.6 ac-ft [it is based on selected solution 76 from optimized results and more results are provided below]) is estimated using annual average rainfall (provided for the Montgomery county watershed), and 3) finally, 62.5 percent of potential infiltration is then assumed (based on groundwater recharge studies8) to become ground water recharge. The dollar value used for groundwater recharge is $86.429 per acre-foot which is adjusted for 2018-dollar value using online CPI calculator10. The cost estimates are based on Schicht et al. (1976). The online tool computes reduced water treatment benefits using financial reports, by MWRD11 (2004), and simply uses $29.94 per acre feet of runoff reduced using vegetation filtration strips. The total runoff volume is computed using NRCS CN method which is consistent with SUSTAIN model settings used in this study. In addition, the CNT online tool requires pre- development land cover, runoff-reduction goal, conventional development, and green improvements. CNT based experiment setup parameters and costs-benefit sheets are provided in Appendix C. Equations 4.5 and 4.6 for lifecycle cost/benefit estimation and net present value of costs are provided below. 𝑅𝑅𝐶𝐶𝚤𝚤 = 𝐶𝐶 × 𝑅𝑅𝚤𝚤 + 𝐴𝐴𝑀𝑀𝚤𝚤 × 𝑎𝑎 − 𝐴𝐴𝐵𝐵 × 𝑎𝑎 4.5 𝑇𝑇 𝐶𝐶𝑀𝑀𝑠𝑠𝑇𝑇 𝐶𝐶𝑃𝑃𝑉𝑉 = � 𝑓𝑓 𝑓𝑓 4.6 (1 + 𝑦𝑦) 𝑓𝑓=0 where 𝑅𝑅𝐶𝐶𝚤𝚤 is lifecycle cost including benefits,𝐶𝐶 is construction cost, 𝑅𝑅𝚤𝚤 is number of times component replacement cost, 𝐴𝐴𝑀𝑀𝚤𝚤 is annual maintenance cost, 𝑎𝑎 is total number of years, 𝐴𝐴𝐵𝐵 is annual benefits, 𝐶𝐶𝑃𝑃𝑉𝑉 is net present value, and 𝑦𝑦 is discount rate. The detailed cost sheets 8 http://greenvalues.cnt.org/national/benefits_detail.php#groundwater-recharge 9 https://www.sws.uiuc.edu/pubdoc/RI/ISWSRI-83.pdf 10 https://data.bls.gov/cgi-bin/cpicalc.pl 11 https://mwrd.org/irj/go/km/docs/documents/MWRD/internet/Departments/Finance/docs/CAFR/CAFR2004.pdf 77 used in CNT online green value calculator are available at center of neighborhood technology website12. 3. Results and discussion 3.1. CWM skill test According to guidelines published by ASCE (1993), the evaluation criteria of watershed models are grouped into continuous hydrographs and single events. Continuous hydrographs contain an evaluation measure Dv (modeled vs. observed) which is similar to percent bias (PBIAS). The mathematical form of PBIAS is provided in equation 4.7. ∑𝑛𝑛𝑖𝑖=1(𝑂𝑂𝑏𝑏𝑠𝑠𝑖𝑖 − 𝑆𝑆𝑏𝑏𝑆𝑆𝑖𝑖)𝑃𝑃𝐵𝐵𝐼𝐼𝐴𝐴𝑆𝑆 = ∑𝑛𝑛 × 100 4.7 𝑖𝑖=1𝑂𝑂𝑏𝑏𝑠𝑠𝑖𝑖 For model evaluation, McCuen et al. (2006) and Moriasi et al. (2007) suggested the use of graphical analysis as a supplementary test along with other measures such as PBIAS. PBIAS test values and percent of impervious area for each study site are presented in Table 4.3. The CWM works reasonably well, even with limited input data, to generate daily mean flows for small urban watersheds. Peak flow assessment is limited due to use of design approach for precipitation disaggregation. It can be inferred that CWM performance is directly associated with percent impervious area, such as seen for the Montgomery county watershed. Apart from CWM model output, PBIAS error can also occur due to a couple of factors associated with USGS streamflow gauge sites: 1) size of gauged watershed used for comparison (because gauges with larger drainage area possibly contain flows from diverse land use types), and 2) the proximity of streamflow gauge to study area watershed. The diffuse runoff 12 http://greenvalues.cnt.org/national/cost_detail.php 78 losses due to DCIA and other unknowns are main contributors to higher errors in CWM- based flow estimations. USGS stream gauges within close proximity of the study site in Prince George’s county have large drainage area and therefore the drainage area scaling approach results in large discrepancies. In addition, study sites with impervious areas of less than 30 percent, such as the watershed in Anne Arundel county, also render errors in flow estimation from CWM, because of limited applicability of this model to highly impervious areas. SUSTAIN base flow output at daily time step is validated using scaled observed daily streamflow time series and provided below in Figure 4.3. To address model limitation, model output under boundary condition is adjusted using percent bias in the output. This involves scaling of CWM-based Howard Montgomery Anne Arundel Prince George’s Figure 4.3. CWM validation using simulated daily flow values 79 Table 4.3. PBIAS results for study sites County Impervious area (percent) PBIAS Howard 42.4 28.7 Montgomery 43.4 6.1 Anne Arundel 28.5 49.8 Prince George’s 40.5 19.5 Performance metrics for PBIAS (Moriasi et al., 2007): < 10 percent (very good), 10-15 percent (good), and 15-25 percent (satisfactory) flow values by multiplying it with the PBIAS. The CWM model output for Montgomery county watershed shows a 6 percent underestimation compared to observation. Therefore, a multiplicative modification (Yang et al., 2007) is applied to precipitation values to account for 1.06 PBIAS under all scenarios. HRUs for CC scenario are constructed using percent change in precipitation from CMIP5 models used in this study (list of models provided in chapter 2). For CC+Urb scenario, HRUs for the future are constructed using CF-based precipitation time series in CWM model along with the zoning scenario of 2025 for land use (detailed in chapter 3). 3.2. Current portfolio The cost-effectiveness curve with all and best (subset of all solutions) solutions for Montgomery county watershed is shown in Figure 4.4. This curve represents current portfolio with gray and green adaptation pathway. Each point on the BMP cost-effectiveness curve represents minimum cost for reduction of corresponding target runoff with an optimal combination of BMPs. So any point along the curve represents an aggregated response of a BMP-scenario over time and space in relation to decision variables (storage depth, and BMP cost). All solutions contain 1000 possible combinations (iterations are set to 1000 in SUSTAIN model) of stormwater management practices whereas best solutions are best 80 optimization scenarios for several reduction targets. Also, selected solution could be any depending on state or county regulations and availability of funds for stormwater management infrastructure construction. In this research, a target of 99 percent or greater reduction is used. With 99 percent effectiveness to reduce runoff, the selected solution under gray and green adaptation pathway costs $32 million. Similar adaptation pathways are generated individually for gray and GI. The cost for gray and green pathways, with 99 percent effectiveness to reduce runoff, under current portfolio is around $32.5 and $37 million, respectively. Figure 4.4. Montgomery county watershed Current Portfolio (1981-2015) The percent cost change in current portfolio under climate change and climate change plus urbanization are presented in Figure 4.5. Current portfolio is based on current climate and land use. Percent change in cost under climate change and climate change plus urbanization is based on their percent difference from current portfolio. Therefore, the 81 portfolios in Figure 4.5 only show percent change in flood reduction cost. The purpose of various pathways is to examine stormwater management from an economic perspective. The pathways are numbered for convenience to reference them in discussion. Each pathway is determined by computing the difference in cost compared to “Current portfolio”. For example, Pathway 1 for climate change scenario is determined by computing percent change in green infrastructure cost under current climate ($37 million) and future climate ($42 million) which shows a change of about 13 percent. Pathway 3 has optimal cost under climate change scenario portfolio. Pathways 2, 5, and 8 show that use of gray infrastructure under climate change and urbanization scenario is cost optimal for runoff reduction purposes. The use of gray infrastructure is inexpensive compared to GI for runoff reduction and eventually flood protection. Pathway 3 shows decrease in cost, under climate change, because cost of combined green and gray infrastructure is less compared to cost of only GI. Pathways 3, 6, and 9 have very high costs under climate change and urbanization scenario due to use of GI for runoff reduction. The existing stormwater management infrastructure in Montgomery county watershed is a combination of gray and green practices (As explained in experimental overview, the selection of these BMP is based on inspection of existing infrastructure in the study site and more details are provided in Appendix C). Pathway 9 shows that continued use of same practices may lead to 80 percent more cost, compared to current portfolio, under climate change and urbanization scenario. The benefits of GI for runoff reduction and its secondary benefits (improved air quality, human health, preservation of ecosystem, reduction of urban heat-island effect, soil conservation, and ground water replenishment) are realized (Moore et al., 2016). However, the cost of GI is very high from only a runoff reduction and 82 flood protection perspective and non-inclusion of ecosystem services benefits such as ground recharge and water quality treatment benefits. Figure 4.5. Montgomery county watershed SMPs 3.3. Benefit assessment of green infrastructure Several studies (Keesstra et al., 2018; Demuzere et al., 2014; Farrugia et al., 2013; Benedict and McMahon, 2012; Foster et al., 2011; Schilling and Logan, 2008) have shown that added benefits of GI can potentially offset their higher costs of construction and maintenance by not only providing flood protection but also additional ecosystem services. But according to Nordman et al. (2018), all GI practices are not necessarily effective in this regard. The SUSTAIN model is well suited to provide least cost options through optimization but is limited to account for added benefits of GI. A wide variety of stormwater management tools are available which incorporate GI (Jayasooriya and Ng, 2014). Amongst many 83 stormwater management tools, the Center of Neighborhood Technology’s (CNT) green values calculator is the most comprehensive resource to obtain cost-benefits estimates of stormwater GI. As part of GI benefits assessment, infiltration trench and bioretention practices are used in CNT calculator. The total surface storage volume of selected solution (provided in Figure 4.4) is 164.6 ac-ft (obtained from SUSTAIN output file) and total area covered by these practices based on selected solution is 1-acre. The net present value (NPV) of both GI practices, to provide added benefits of ground replenishment and reduced treatment benefits, is negligible compared to total life-cycle construction and maintenance costs. This could likely be due to: i) oversimplified method for groundwater recharge estimation, ii) lack of inclusion of native vegetation types with/around bioretention/raingarden and filtration strips, iii) the area covered by these green practices compared to watershed size, and iv) CNT ineffectiveness for small scale watershed studies. 4. Conclusions Stormwater management requires hydrologic modeling for flow estimation. However, most of the advanced hydrologic models often include difficulty of calibration and validation due to medium to high input complexity of these models. According to Gan et al. (1997), a simple model structure is fine as long as model output is reasonably good. In this research study, a low input data complexity Curve Number Watershed Model (CWM) is conceptualized and implemented. The GIS environment is used to generate Hydrologic Response Units (HRUs) for flow routing. The implementation of CWM requires low input data complexity and can be easily setup for any small urban watershed. The major limitations of CWM model are use of disaggregated daily precipitation totals (using the design storm 84 approach) and NRCS CN method for runoff estimation. The CWM-based flow estimation error is higher (positive values of PBIAS means model’s underestimation), due to both of these factors, and leads to considerable underestimation of flow values for watersheds with less than 40 percent imperviousness as observed in the case of Anne Arundel county watershed. Additional source of error is unaccounted Disconnected Impervious Area (DCIA) which also contributes to the underestimation of flows. In addition to structure and process, the validation of CWM requires scaling of streamflow data from nearby gauged watersheds. This can also add to the uncertainty of model performance. To account for overall model uncertainty, estimated flow time series from CWM is scaled using PBIAS estimate. Importantly, use of CWM for any other study site should be limited to medium to high impervious small scale (~3km2) urban watersheds. The reasonably good performance of CWM with simple structure and few input data needs offers a potential flow simulation tool at small scale for medium to highly impervious watersheds. Under climate change and urbanization, the ideal path forward is to conduct existing BMP space-time analysis to test the effectiveness of existing stormwater management infrastructure for future conditions. However, the acquisition of existing BMP dataset can be difficult at times due to state or county regulations or in the absence of sufficient funds to request it for research purposes. This study presents an alternative approach to construct SMPs in the absence of existing BMP technical specifications. The underlying assumption is that the existing infrastructure is effective to store surface runoff under current climate and urbanization conditions. For this purpose, SUSTAIN model is setup with a set of BMPs to obtain 99 percent or greater effectiveness to reduce runoff under existing climate and urbanization. It is evident from SMPs under both climate change and climate change plus urbanization portfolio that continuing use of current BMP types in Montgomery county watershed may incur very high costs in future. Given the current set of assumptions and study 85 limitations, the secondary benefits of infiltration trench and bioretention practices are negligible to offset the higher costs of construction and maintenance during their life-cycle. The main factors behind negligible benefits of GI practices is likely due to oversimplification of ground recharge quantification method, very small spatial extent of GI practices compared to total watershed area, and ineffectiveness of tool for small scale watershed studies. Moreover, Pathway 8 is cost optimal solution with use of only gray infrastructure, in addition to current green and gray, for future flows reduction. Therefore, current assessment shows that it is necessary to explore: i) other modeling tools and metrics for quantification of GI benefits, ii) use of other GI practices (such amended soil and landscaping using native vegetation) along with gray infrastructure, and iii) sensitivity tests of “GI/watershed area” and several discounts rates in Montgomery county watershed to meet water quantity and quality targets with higher NPV compared to conventional stormwater management approaches. The optimization of a set of BMPs to obtain runoff reduction target solutions and construction of SMPs offers a framework for cost-effective flood protection. However, the benefits assessment, as conducted in this research, required use of an online tool CNT which operates using a single storm rainfall. According to a detailed review of the stormwater literature by Prudencio and Null (2018), regulating services are most often quantified and less attention given to other three services (provisioning, cultural, and supporting). Currently, some of the major limitations in this regard are: i) the integration of GI benefits within stormwater management modeling environment to obtain optimized solutions inclusive of these co-benefits, ii) inclusion of more GI practices, such as amended soils and landscaping with native vegetation in modeling tools (Nordman et al., 2018; Jayasooriya and Ng, 2014), and iii) apart from economic and environmental benefits, the modeling tools should also include social benefits (such as human health index) of GI (US EPA, 2015). 86 Chapter 5: Conclusions 1. Major findings The increase in total precipitation, population, and urbanization is evident in the U.S. Mid-Atlantic region. An increase in precipitation and developed land not only leads to a higher likelihood of urban flooding but also affects provisioning and regulating urban ecosystem services. This dissertation is interdisciplinary in nature and contributes to our understanding of likely future storm frequency and associated runoff, changes in the soil erosion rates and sediment yield, and stormwater management pathways under anticipated climate change and urbanization in small U.S. Mid-Atlantic watersheds. Climate change analysis in this dissertation contributes to our understanding of characteristic change in future frequency of heavy precipitation storm events and associated runoff. The soil conservation assessment adds to our understanding of likely higher soil erosion rates in the near future at urban catchment scale. This dissertation uses a stormwater management pathways approach, develop a simple flow simulation model, and provides a novel understanding of how green stormwater management practices affect ecosystem services in small-scale watersheds. The findings in this dissertation are at the core of land use planning and municipal stormwater management required at the implementation level. Climate change analysis in Chapter 2 of this dissertation shows an increasing trend in the frequency of events exceeding the d85 magnitude during the historical period 1981-2015. The climate models used in this analysis show an increase in the frequency of these events and a slight increase in runoff, under the climate change and urbanization scenario, during the future period 2016-2035. The presentation of results using a set of climate models and 87 statistical downscaling approaches provides an envelope of possible changes in future frequency and associated runoff. Scale mismatch has been a longstanding problem when downscaling climate model outputs to the regional or local scale for hydrologic impact assessment (Schmidli et al. 2006; Chen et al. 2011b). Therefore, several models and downscaling methods exist (e.g., Minville 2008; Chen et al. 2011a; Chen et al. 2011b) and are used as a guide in this dissertation to statistically downscale climate model outputs. This dissertation uses a systematic sequence to select climate models with reasonable accuracy to reproduce historical climate (for the U.S. Mid-Atlantic region in our case), and choice of downscaling methods. The climate change assessment study approach and findings in this dissertation provide valuable information for municipal stormwater management. Our use of an analytical approach to estimate precipitation capture depths is relevant to flooding concerns for small watersheds (3 km2). Capture depth is a better measure than a “n-year” precipitation because it provides count of events exceeding certain volume in each year. Estimates developed in this study of changes in water volume and resultant on-site infrastructure costs can help stakeholders and managers in planning for flood mitigation and protection of ecosystem services. Capture depth percentiles such as d85, d90, d95, and d99, have the potential to serve as meaningful hydrologic indicators for future stormwater management planning. In this regard, the rainfall percentile by volume approach provided in this dissertation is a timely topic for research. A thorough investigation of available flood control metrics is needed to learn more about the likely change in quantity of stormwater under climate change and urbanization scenarios. The outcome of such assessment is key to guide counties and municipalities regarding stormwater management design and runoff reduction thresholds. 88 Chapter 3 of this dissertation contributes to our understanding of possible increases in soil erosion rates in the future in small-scale urban watersheds in the U.S. Mid-Atlantic region. On average, the erosivity power of extreme rainfall events is expected to increase based on results in this dissertation, during 2016-2050 relative to 1981-2015, by a factor of 2 and higher compared to total power of all storm events for each unit increase in total precipitation. Erosive power and soil erosion rates are also provided in this dissertation to further our understanding of relative magnitude of d85-99 events. We find an increase in average magnitudes of erosivity and soil erosion rates (for each unit change in precipitation) during 2016-2050. These estimates are a point of concern and invaluable for effective land use planning, and sustainable ecosystem resources. The revised universal soil loss equation (RUSLE) is used for soil erosion estimation and subsequently the sediment delivery ratio (SDR) model to determine sediment yield at each watershed outlet. RUSLE has limited predictive capability due to estimation of only long- term annual average soil loss. When we use a range of extreme precipitation events in determining their relative contribution in total soil loss and sediment load, the estimates provide a lower end of predicted output. In reality, the soil loss, and hence total sediment load, could be larger. This information is useful for soil conservation planning. In addition, a SDR model calibration approach is proposed and used in this dissertation. This approach has the potential for use as an alternative to averaged or theoretical calibration values to run the SDR model. It is fair to say that RUSLE predicts sediment yield as a product of rainfall erosive power and soil erodibility within watershed boundary (perceived as system boundary). That is why paving over the soil reduces erosion. Therefore, it may not reflect sediment yield due to increased peak flows encountering stream beds and banks downstream. 89 Chapter 4 of this dissertation, stormwater management requires hydrologic modeling for flow estimation. However, most of the advanced hydrologic models often suffer from difficulty of calibration and validation due to medium to high input complexity of these models. According to Gan et al. (1997), a simple model structure is fine as long as model output is reasonably good. In this dissertation, I developed a low input data complexity Curve Number Watershed Model (CWM) and implemented the model in a GIS environment. Some of the limitations of the CWM include use of disaggregated daily precipitation totals (using the design storm approach), NRCS CN method for runoff estimation, and unaccounted Disconnected Impervious Area (DCIA). But its performance is promising for watersheds with greater than 40 percent imperviousness. The reasonably good performance of the CWM with its simple structure and few input data needs offers a possible flow simulation tool at small scales for medium- to highly-impervious watersheds. This dissertation provides a novel understanding of how green stormwater management practices affect ecosystem services in small watersheds. Given the current set of assumptions and study limitations, the secondary benefits of infiltration trenches and bioretention practices used in this dissertation are negligible to offset the higher life-cycle costs of construction and maintenance. More research is needed with regards to secondary benefits of green stormwater management practices. The effectiveness of green practices in small watersheds is a much needed direction. In addition, stormwater management pathways are constructed in this dissertation. Ideally, the information of existing stormwater infrastructure is used to test system resilience under future changes. However, the acquisition of existing infrastructure information is difficult at times due to state or county regulations, or in the 90 absence of sufficient funds, to request it for research purposes. This dissertation presents an alternative approach, in absence of data, to construct adaptation pathways. 2. Future research Urban watersheds are habitat for an increasing number of human population, and they are degraded by past and present land use policies. It is important to sustain valuable ecosystem services provided by watersheds such as flood protection, clean drinking water, waste treatment, corridors for safe recreation and utilities, aquatic and terrestrial habitats for wildlife, and increased property values. More research is needed to quantify how future erosion rates and sediment totals, due to heavy precipitation events and land use change, lead to decline in topsoil depth, reduction in soil fertility, and decline in available water capacity. According to a detailed review of the available literature on ‘green stormwater management practices and their effect on ecosystem services’ by Prudencio and Null (2018), regulating services are most often quantified and less attention given to the other three ecosystem services (provisioning, cultural, and supporting) from a stormwater management perspective. Needed future research directions are: i) the integration of green infrastructure benefits within stormwater management modeling environment to obtain optimized solutions inclusive of these co-benefits, ii) inclusion of more green infrastructure practices, such as amended soils and landscaping with native vegetation in modeling tools (Nordman et al., 2018; Jayasooriya and Ng, 2014), and iii) apart from economic and environmental benefits, modeling tools should also include social benefits (such as human health) of green infrastructure (US EPA, 2015). 91 Green infrastructure practices for stormwater management are preferred due to their economic, environmental, and social aspects (US EPA, 2015). Maryland’s stormwater management act of 2007 (Act) proposed use of environmental site design (ESD) to the maximum extent practicable (MEP). Title 4, subtitle 201.1(B) of the act defines ESD as “…using small-scale stormwater management practices, nonstructural techniques, and better site planning to mimic natural hydrologic runoff characteristics and minimize the impact of land development on water resources.” Since 1982, the stormwater management law has undergone several revisions in pollutant target control and runoff capture thresholds. The most recent amendment was implemented in year 2010. A brief synopsis of stormwater management infrastructure and capital investment for pre- and post-2010 eras in central Maryland is provided in Appendix D. Based on the growing need for transition from conventional infrastructure to green infrastructure or environmental site design to maximum extent practicable due to climate change and urbanization, more research is needed on the flexibility and adaptability of existing stormwater infrastructure to future conditions. 92 Appendix A Table A.1. Rainfall depth values for 2-year, 24-hour County Rainfall depth Rainfall depth (county level) (watershed level) Howard 3.21 3.22 Montgomery 3.1 3.14 Anne Arundel 3.22 3.17 Prince George’s 3.18 3.18 Mann-Kendall Trend Test 𝑛𝑛−1 𝑛𝑛 𝑆𝑆 = � � 𝑠𝑠𝑎𝑎𝑎𝑎(𝑒𝑒𝑖𝑖 − 𝑒𝑒𝑘𝑘) 𝑘𝑘−1 𝑖𝑖−𝑘𝑘+1 𝑆𝑆 − 1⎧= 𝑏𝑏𝑏𝑏 𝑆𝑆 > 0 ⎪ �𝑉𝑉𝐴𝐴𝑅𝑅(𝑆𝑆) 𝑍𝑍𝑀𝑀𝑀𝑀 = 0 𝑏𝑏𝑏𝑏 𝑆𝑆 = 0 ⎨ 𝑆𝑆 + 1 ⎪= 𝑏𝑏𝑏𝑏 𝑆𝑆 < 0 ⎩ �𝑉𝑉𝐴𝐴𝑅𝑅(𝑆𝑆) where S is number of positive differences minus number of negative differences, ZMK is trend statistics, and VAR is variance. A positive (negative) value of ZMK indicates increasing (decreasing) trend in data. 2-Digit Classification System Table A.2. Anderson Lookup Table LU code Classification Hyd_a Hyd_b Hyd_c Hyd_d Imp LUCat 10 Urban 61 75 83 87 0.38 u 11 Low Density Residential 54 70 80 85 0.25 u 12 Medium Density Residential 61 75 83 87 0.38 u 93 13 High Density Residential 77 85 90 92 0.65 u 14 Commercial 89 92 94 95 0.85 u 15 Industrial 81 88 91 93 0.72 u 16 Institutional 69 80 86 89 0.5 n 17 Extractive 77 86 91 94 0.11 n 18 Open Urban Land 39 61 74 80 0.11 n 20 Agriculture 67 78 85 89 0 n 21 Cropland 67 78 85 89 0 n 22 Pasture 39 61 74 80 0 n 23 Orchards 32 58 72 79 0 n 24 Feeding Operations 59 74 82 86 0 n 25 Row Crops 67 78 85 89 0 n 40 Forest 30 55 70 77 0 f 41 Deciduous Forest 30 55 70 77 0 f 42 Evergreen Forest 30 55 70 77 0 f 43 Mixed Forest 30 55 70 77 0 f 44 Brush 30 48 65 73 0 f 50 Water 100 100 100 100 0 s 60 Wetlands 100 100 100 100 0 s 70 Barren Land 77 86 91 94 0.5 n 71 Beaches 77 86 91 94 0 n 72 Bare Exposed Rock 77 86 91 94 1 n 73 Bare Ground 77 86 91 94 0.5 n 80 Transportation 83 89 92 94 0.75 n 191 Large Lot Agricultural 67 78 85 89 0.15 n 192 Large Lot Forest 30 55 70 77 0.15 f 241 Feeding Operations 59 74 82 86 0.1 n 242 Agricultural Buildings 59 74 82 86 0.1 n 94 LU = Land use Hyd = Hydrologic Soil Group Imp = Imperviousness LUCat = Land use categories u = urban; n = nil; f = forest; s = storage The CN maps are prepared using 2-digit classification system using Anderson look up Table. These Tables are provided for fair, good, and poor hydrologic conditions. In this research, good hydrologic condition lookup Table is used and provided above. Each cell in CN map is assigned a number from lookup Table depending on overlapping land use category and soil type of that cell. Curve Number (CN) County Level Maps Figure A.1. Curve number map (1973) 95 Figure A.2. Curve number map (1990) Figure A.3. Curve number map (1997) 96 Figure A.4. Curve number map (2002) Figure A.5. Curve number map (2010) 97 Appendix B Howard RAINFALL DISTRIBUTION: rtp2-24 0.1 0.0000 0.0012 0.0024 0.0037 0.0049 0.0061 0.0073 0.0085 0.0097 0.0110 0.0122 0.0134 0.0146 0.0158 0.0170 0.0183 0.0195 0.0207 0.0219 0.0231 0.0244 0.0256 0.0268 0.0280 0.0292 0.0304 0.0317 0.0329 0.0341 0.0353 0.0365 0.0377 0.0390 0.0402 0.0414 0.0426 0.0438 0.0450 0.0463 0.0475 0.0487 0.0499 0.0511 0.0524 0.0536 0.0548 0.0560 0.0572 0.0584 0.0597 0.0609 0.0621 0.0633 0.0645 0.0657 0.0670 0.0682 0.0694 0.0706 0.0718 0.0731 0.0756 0.0781 0.0806 0.0831 0.0856 0.0882 0.0907 0.0932 0.0957 0.0982 0.1007 0.1033 0.1058 0.1083 0.1108 0.1133 0.1158 0.1184 0.1209 0.1234 0.1259 0.1284 0.1309 0.1335 0.1360 0.1385 0.1410 0.1435 0.1460 0.1486 0.1531 0.1577 0.1623 0.1668 0.1714 0.1760 0.1805 0.1851 0.1897 0.1943 0.1988 0.2034 0.2080 0.2125 0.2171 0.2214 0.2258 0.2301 0.2344 0.2388 0.2469 0.2551 0.2632 0.2714 0.2795 0.2970 0.3144 0.3423 0.3862 0.5000 0.6138 0.6577 0.6856 0.7030 0.7205 0.7286 0.7368 0.7449 0.7531 0.7612 0.7656 0.7699 0.7742 0.7786 0.7829 0.7875 0.7920 0.7966 0.8012 0.8057 0.8103 0.8149 0.8195 0.8240 0.8286 0.8332 0.8377 0.8423 0.8469 0.8514 0.8540 0.8565 0.8590 0.8615 0.8640 0.8665 0.8691 0.8716 0.8741 0.8766 0.8791 0.8816 0.8842 0.8867 0.8892 0.8917 0.8942 0.8967 0.8993 0.9018 0.9043 0.9068 0.9093 0.9118 0.9144 0.9169 0.9194 0.9219 0.9244 0.9269 0.9282 0.9294 0.9306 0.9318 0.9330 0.9343 0.9355 0.9367 0.9379 0.9391 0.9403 0.9416 0.9428 0.9440 0.9452 0.9464 0.9476 0.9489 0.9501 0.9513 0.9525 0.9537 0.9550 0.9562 0.9574 0.9586 0.9598 0.9610 0.9623 0.9635 0.9647 0.9659 0.9671 0.9683 0.9696 0.9708 0.9720 0.9732 0.9744 0.9756 0.9769 0.9781 0.9793 0.9805 0.9817 0.9830 0.9842 0.9854 0.9866 0.9878 0.9890 0.9903 0.9915 0.9927 0.9939 0.9951 0.9963 0.9976 0.9988 1.0000 98 Montgomery RAINFALL DISTRIBUTION: rtp2-24 0.1 0.0000 0.0012 0.0024 0.0037 0.0049 0.0061 0.0073 0.0085 0.0097 0.0110 0.0122 0.0134 0.0146 0.0158 0.0170 0.0183 0.0195 0.0207 0.0219 0.0231 0.0244 0.0256 0.0268 0.0280 0.0292 0.0304 0.0317 0.0329 0.0341 0.0353 0.0365 0.0377 0.0390 0.0402 0.0414 0.0426 0.0438 0.0450 0.0463 0.0475 0.0487 0.0499 0.0511 0.0524 0.0536 0.0548 0.0560 0.0572 0.0584 0.0597 0.0609 0.0621 0.0633 0.0645 0.0657 0.0670 0.0682 0.0694 0.0706 0.0718 0.0731 0.0756 0.0781 0.0806 0.0831 0.0856 0.0882 0.0907 0.0932 0.0957 0.0982 0.1007 0.1033 0.1058 0.1083 0.1108 0.1133 0.1158 0.1184 0.1209 0.1234 0.1259 0.1284 0.1309 0.1335 0.1360 0.1385 0.1410 0.1435 0.1460 0.1486 0.1531 0.1577 0.1623 0.1668 0.1714 0.1760 0.1805 0.1851 0.1897 0.1943 0.1988 0.2034 0.2080 0.2125 0.2171 0.2214 0.2258 0.2301 0.2344 0.2388 0.2469 0.2551 0.2632 0.2714 0.2795 0.2970 0.3144 0.3423 0.3862 0.5000 0.6138 0.6577 0.6856 0.7030 0.7205 0.7286 0.7368 0.7449 0.7531 0.7612 0.7656 0.7699 0.7742 0.7786 0.7829 0.7875 0.7920 0.7966 0.8012 0.8057 0.8103 0.8149 0.8195 0.8240 0.8286 0.8332 0.8377 0.8423 0.8469 0.8514 0.8540 0.8565 0.8590 0.8615 0.8640 0.8665 0.8691 0.8716 0.8741 0.8766 0.8791 0.8816 0.8842 0.8867 0.8892 0.8917 0.8942 0.8967 0.8993 0.9018 0.9043 0.9068 0.9093 0.9118 0.9144 0.9169 0.9194 0.9219 0.9244 0.9269 0.9282 0.9294 0.9306 0.9318 0.9330 0.9343 0.9355 0.9367 0.9379 0.9391 0.9403 0.9416 0.9428 0.9440 0.9452 0.9464 0.9476 0.9489 0.9501 0.9513 0.9525 0.9537 0.9550 0.9562 0.9574 0.9586 0.9598 0.9610 0.9623 0.9635 0.9647 0.9659 0.9671 0.9683 0.9696 0.9708 0.9720 0.9732 0.9744 0.9756 0.9769 0.9781 0.9793 0.9805 0.9817 0.9830 0.9842 0.9854 0.9866 0.9878 0.9890 0.9903 0.9915 0.9927 0.9939 0.9951 0.9963 0.9976 0.9988 1.0000 99 Anne Arundel RAINFALL DISTRIBUTION: rtp2-24 0.1 0.0000 0.0012 0.0024 0.0037 0.0049 0.0061 0.0073 0.0085 0.0097 0.0110 0.0122 0.0134 0.0146 0.0158 0.0170 0.0183 0.0195 0.0207 0.0219 0.0231 0.0244 0.0256 0.0268 0.0280 0.0292 0.0304 0.0317 0.0329 0.0341 0.0353 0.0365 0.0377 0.0390 0.0402 0.0414 0.0426 0.0438 0.0450 0.0463 0.0475 0.0487 0.0499 0.0511 0.0524 0.0536 0.0548 0.0560 0.0572 0.0584 0.0597 0.0609 0.0621 0.0633 0.0645 0.0657 0.0670 0.0682 0.0694 0.0706 0.0718 0.0731 0.0756 0.0781 0.0806 0.0831 0.0856 0.0882 0.0907 0.0932 0.0957 0.0982 0.1007 0.1033 0.1058 0.1083 0.1108 0.1133 0.1158 0.1184 0.1209 0.1234 0.1259 0.1284 0.1309 0.1335 0.1360 0.1385 0.1410 0.1435 0.1460 0.1486 0.1531 0.1577 0.1623 0.1668 0.1714 0.1760 0.1805 0.1851 0.1897 0.1943 0.1988 0.2034 0.2080 0.2125 0.2171 0.2214 0.2258 0.2301 0.2344 0.2388 0.2469 0.2551 0.2632 0.2714 0.2795 0.2970 0.3144 0.3423 0.3862 0.5000 0.6138 0.6577 0.6856 0.7030 0.7205 0.7286 0.7368 0.7449 0.7531 0.7612 0.7656 0.7699 0.7742 0.7786 0.7829 0.7875 0.7920 0.7966 0.8012 0.8057 0.8103 0.8149 0.8195 0.8240 0.8286 0.8332 0.8377 0.8423 0.8469 0.8514 0.8540 0.8565 0.8590 0.8615 0.8640 0.8665 0.8691 0.8716 0.8741 0.8766 0.8791 0.8816 0.8842 0.8867 0.8892 0.8917 0.8942 0.8967 0.8993 0.9018 0.9043 0.9068 0.9093 0.9118 0.9144 0.9169 0.9194 0.9219 0.9244 0.9269 0.9282 0.9294 0.9306 0.9318 0.9330 0.9343 0.9355 0.9367 0.9379 0.9391 0.9403 0.9416 0.9428 0.9440 0.9452 0.9464 0.9476 0.9489 0.9501 0.9513 0.9525 0.9537 0.9550 0.9562 0.9574 0.9586 0.9598 0.9610 0.9623 0.9635 0.9647 0.9659 0.9671 0.9683 0.9696 0.9708 0.9720 0.9732 0.9744 0.9756 0.9769 0.9781 0.9793 0.9805 0.9817 0.9830 0.9842 0.9854 0.9866 0.9878 0.9890 0.9903 0.9915 0.9927 0.9939 0.9951 0.9963 0.9976 0.9988 1.0000 100 Prince George’s RAINFALL DISTRIBUTION: rtp2-24 0.1 0.0000 0.0012 0.0024 0.0037 0.0049 0.0061 0.0073 0.0085 0.0097 0.0110 0.0122 0.0134 0.0146 0.0158 0.0170 0.0183 0.0195 0.0207 0.0219 0.0231 0.0244 0.0256 0.0268 0.0280 0.0292 0.0304 0.0317 0.0329 0.0341 0.0353 0.0365 0.0377 0.0390 0.0402 0.0414 0.0426 0.0438 0.0450 0.0463 0.0475 0.0487 0.0499 0.0511 0.0524 0.0536 0.0548 0.0560 0.0572 0.0584 0.0597 0.0609 0.0621 0.0633 0.0645 0.0657 0.0670 0.0682 0.0694 0.0706 0.0718 0.0731 0.0756 0.0781 0.0806 0.0831 0.0856 0.0882 0.0907 0.0932 0.0957 0.0982 0.1007 0.1033 0.1058 0.1083 0.1108 0.1133 0.1158 0.1184 0.1209 0.1234 0.1259 0.1284 0.1309 0.1335 0.1360 0.1385 0.1410 0.1435 0.1460 0.1486 0.1531 0.1577 0.1623 0.1668 0.1714 0.1760 0.1805 0.1851 0.1897 0.1943 0.1988 0.2034 0.2080 0.2125 0.2171 0.2214 0.2258 0.2301 0.2344 0.2388 0.2469 0.2551 0.2632 0.2714 0.2795 0.2970 0.3144 0.3423 0.3862 0.5000 0.6138 0.6577 0.6856 0.7030 0.7205 0.7286 0.7368 0.7449 0.7531 0.7612 0.7656 0.7699 0.7742 0.7786 0.7829 0.7875 0.7920 0.7966 0.8012 0.8057 0.8103 0.8149 0.8195 0.8240 0.8286 0.8332 0.8377 0.8423 0.8469 0.8514 0.8540 0.8565 0.8590 0.8615 0.8640 0.8665 0.8691 0.8716 0.8741 0.8766 0.8791 0.8816 0.8842 0.8867 0.8892 0.8917 0.8942 0.8967 0.8993 0.9018 0.9043 0.9068 0.9093 0.9118 0.9144 0.9169 0.9194 0.9219 0.9244 0.9269 0.9282 0.9294 0.9306 0.9318 0.9330 0.9343 0.9355 0.9367 0.9379 0.9391 0.9403 0.9416 0.9428 0.9440 0.9452 0.9464 0.9476 0.9489 0.9501 0.9513 0.9525 0.9537 0.9550 0.9562 0.9574 0.9586 0.9598 0.9610 0.9623 0.9635 0.9647 0.9659 0.9671 0.9683 0.9696 0.9708 0.9720 0.9732 0.9744 0.9756 0.9769 0.9781 0.9793 0.9805 0.9817 0.9830 0.9842 0.9854 0.9866 0.9878 0.9890 0.9903 0.9915 0.9927 0.9939 0.9951 0.9963 0.9976 0.9988 1.0000 101 CBP to MDP LCLU conversion Table B.1. Land use conversion codes LCLU CBP (zoning keys) LCLU MDP (description) 1 Commercial 14 Commercial 2 Residential 191/192 Large lot subdivisions 3 Mixed - Zoning district that allows mix of any other categories 4 Forest 41/42/43 Deciduous & Evergreen & Mixed 5 Scrub 44 Brush 6 Farmland 23/242/25 Orchards/Vineyards/Horticulture & Agriculture & Row and garden crops 7 Barren 70/71/72/73 Barren & Beaches & Bare exposed & ground 8 Water 50 Water 9 Wetlands 60 Wetlands 21 Developed Open Space 11-16 Low/Medium/High/Commercial/Industrial/Institutional 22 Low-Intensity Residential 11 Low-Density Residential 23 Medium-Intensity Residential 12 Medium-Density Residential 24 High-Intensity Residential 13 High-Density Residential 1. Conversion of CBP categories 2, 4, 6, and 7 require visual inspection of current MDP land use in each watershed. These conversions play a critical role in “Biophysical Tables” for RUSLE because total erosion estimates can change significantly under different LCLU types. Also, impervious percent is estimated based on 2-digit classification system using Anderson lookup Table as described in chapter 2. 102 2. CBP class ‘21’ overlaps with MDP’s LU categories 11-16. It will be erroneous to classify all of developed land, which include low-high density residential areas, to open space. Therefore, during conversion of CBP to MDP, all cells belonging to 11-16 categories are excluded from conversion. 3. Except 11-16 and 50, all categories have potential to change. 103 Sediment Delivery Ratio (SDR) module According to Borselli et al. (2008), SDR module first computes connectivity index. The mathematical form of connectivity index is provided below: 𝐷𝐷 𝐼𝐼𝐶𝐶 = 𝑀𝑀𝑀𝑀𝑎𝑎 ( 𝑓𝑓𝑝𝑝10 ) 𝐷𝐷𝑑𝑑𝑛𝑛 The conceptual approach of the SDR module is provided below: Figure B.1. Sediment delivery ratio model concept diagram Source: http://data.naturalcapitalproject.org/nightly-build/invest-users-guide/html/sdr.html The SDR is computed at each pixel by taking into account upslope area and downslope flow path. The formulae for both upslope area and downslope flow path are provided below: 𝐷𝐷𝑓𝑓𝑝𝑝 = 𝐶𝐶̅𝑆𝑆̅√𝐴𝐴 Where 𝐶𝐶̅ is the average C factor (C factor from RUSLE equation) of the upslope contributing area, 𝑆𝑆̅ is the average slope gradient of the upslope contributing area, and A is the upslope contributing area in m2. 104 𝑀𝑀 𝐷𝐷 𝑖𝑖𝑑𝑑𝑛𝑛 = � 𝐶𝐶 𝑖𝑖 𝑖𝑖 𝑆𝑆𝑖𝑖 Where 𝑀𝑀𝑖𝑖 is the length of the flow path along the ith cell, 𝐶𝐶𝑖𝑖𝑦𝑦𝑎𝑎𝑀𝑀 𝑆𝑆𝑖𝑖 are the C factor and the slope gradient of the ith cell, respectively. Total Soil Erosion under CC and CC+Urb Table B.2. Total annual soil erosion (2016-2050) [units: tons/watershed] County CanESM2 IPSL-CM5A- GFDL-MR ESM2G NorESM1-M Howard 5678 6086 6009 6841 Montgomery 2164 2252 2154 2515 Anne Arundel 184 197 194 221 Prince George's 288 310 320 356 Howard 1511 1619 1599 1820 Montgomery 274 285 273 319 Anne Arundel 11 12 12 13 Prince George's 87 94 97 107 105 Climate Climate Change and Change Urbanization Appendix C SUSTAIN script: Climate Change scenario c----------------------------------------------------------------------------------------- c c SUSTAIN: System for Urban Stormwater Treatment and Analysis INtegration c Version 1.2 - May 2013 c c Designed and maintained by: c Tetra Tech, Inc. c 10306 Eaton Place, Suite 340 c Fairfax, VA 22030 c (703) 385-6000 c c NOTE: The line starting with the letter c and followed by space or - is a comment line. c There must be no comment line in between the data lines c The input text field must be a continuous string without any space in between the characters c c-------------------------------------------------------------------------------------------- c700 Model Controls c c LINE1 = Land simulation control (0-external,1-internal), c Land output directory (containing land output timeseries), c external land timeseries data must be in this order; flow (in./timestep), groundwater recharge (in./timestep), pollutant 1 (lb/acre/timestep), pollutant2, ... c Mixed land use output file name (for internal control), c PreDeveloped land use output file name (for internal control) c LINE2 = Start date of simulation (Year Month Day) c LINE3 = End date of simulation (Year Month Day) c LINE4 = Land Timeseries timestep (Min), 106 c BMP simulation timestep (Min), c CRRAT = The ratio of max velocity to mean velocity under typical flow conditions (value of 1.0 or greater) c Model output control (0-the same timestep as land time series; 1-hourly), c Model output directory c LINE5 = ET Flag (0-constant monthly ET,1-daily ET from the timeseries,2-calculate daily ET from the daily temperature data), c Climate time series file path (required if ET flag is 1 or 2), c Latitude (Decimal degrees) required if ET flag is 2 c LINE6 = Monthly ET rate (in/day) if ET flag is 0 OR c Monthly pan coefficient (multiplier to ET value) if ET flag is 1 OR c Monthly variable coefficient to calculate ET values c 0 D:\BMPs\SUSTAIN\HRUs\Montgomery\FutureClimate 1996 1 1 2015 12 31 60 15 1.5 1 D:\BMPs\SUSTAIN\Outputs\BMPconfig&opt\Montgomery\FutureClimate\Green&Gray 0 0.013 0.131 0.700 1.964 3.430 5.107 5.131 5.154 3.583 1.883 0.826 0.157 c-------------------------------------------------------------------------------------------- c705 Pollutant Definition c c POLLUT_ID = Unique pollutant identifier (Sequence number same as in land output time series) c POLLUT_NAME = Unique pollutant name c MULTIPLIER = Multiplying factor used to convert the pollutant load to lbs (external control) c SED_FLAG = The sediment flag (0-not sediment,1-sand,2-silt,3-clay,4-total sediment) c if = 4 SEDIMENT will be splitted into sand, silt,and clay based on the fractions defined in card 710. c SED_QUAL = The sediment-associated pollutant flag (0-no, 1-yes) c if = 1 then SEDIMENT is required in the pollutant list c SAND_QFRAC = The sediment-associated qual-fraction on sand (0-1), only required if SED_QUAL = 1 c SILT_QFRAC = The sediment-associated qual-fraction on silt (0-1), only required if SED_QUAL = 1 c CLAY_QFRAC = The sediment-associated qual-fraction on clay (0-1), only required if SED_QUAL = 1 c 107 c POLLUT_ID POLLUT_NAME MULTIPLIER SED_FLAG SED_QUAL SAND_QFRAC SILT_QFRAC CLAY_QFRAC 1 TSS 1 0 0 0.000 0.000 0.000 c-------------------------------------------------------------------------------------------- c710 LAND USE DEFINITION (required if land simulation control is external) c c LANDTYPE = Unique land use definition identifier c LANDNAME = land use name c IMPERVIOUS = Distinguishes pervious/impervious land unit (0-pervious; 1-impervious) c TIMESERIESFILE = File name containing input timeseries c SAND_FRAC = The fraction of total sediment from the land which is sand (0-1) c SILT_FRAC = The fraction of total sediment from the land which is silt (0-1) c CLAY_FRAC = The fraction of total sediment from the land which is clay (0-1) c c LANDTYPE LANDNAME IMPERVIOUS TIMESERIESFILE SAND_FRAC SILT_FRAC CLAY_FRAC 1 LowDensityResidential_imp 1 HRU11_6.txt 0 0 0 2 MediumDensityResidential_perv 0 HRU12_8.txt 0 0 0 3 MediumDensityResidential_imp 1 HRU12_15.txt 0 0 0 4 MediumDensityResidential_imp 1 HRU12_19.txt 0 0 0 5 HighDensityResidential_perv 0 HRU13_16.txt 0 0 0 6 HighDensityResidential_imp 1 HRU13_22.txt 0 0 0 7 HighDensityResidential_imp 1 HRU13_24.txt 0 0 0 8 Commercial_imp 1 HRU14_25.txt 0 0 0 9 Commercial_imp 1 HRU14_27.txt 0 0 0 10 Industrial_perv 0 HRU15_20.txt 0 0 0 11 Industrial_imp 1 HRU15_23.txt 0 0 0 12 Industrial_imp 1 HRU15_26.txt 0 0 0 13 Institutional_imp 1 HRU16_13.txt 0 0 0 14 Institutional_imp 1 HRU16_18.txt 0 0 0 15 OpenUrbanLand_imp 1 HRU18_4.txt 0 0 0 16 Cropland_perv 0 HRU21_12.txt 0 0 0 17 Cropland_perv 0 HRU21_17.txt 0 0 0 108 18 Cropland_perv 0 HRU21_21.txt 0 0 0 19 Pasture_perv 0 HRU22_5.txt 0 0 0 20 Pasture_perv 0 HRU22_14.txt 0 0 0 21 DeciduousForest_perv 0 HRU41_1.txt 0 0 0 22 DeciduousForest_perv 0 HRU41_7.txt 0 0 0 23 DeciduousForest_perv 0 HRU41_9.txt 0 0 0 24 EvergreenForest_perv 0 HRU42_2.txt 0 0 0 25 EvergreenForest_perv 0 HRU42_10.txt 0 0 0 26 MixedForest_perv 0 HRU43_3.txt 0 0 0 27 MixedForest_perv 0 HRU43_11.txt 0 0 0 28 Brush_perv 0 HRU44_0.txt 0 0 0 29 Water 0 HRU50_28.txt 0 0 0 c-------------------------------------------------------------------------------------------- c712 Aquifer INFORMATION c c AquiferID = Unique Aquifer identifier c AquiferNAME = Aquifer name c Initial Storage = Initial Storage (ac-ft) c RecessionCoef = Recession Coefficient (1/hr) c SeepageCoef = Seepage Coefficient (1/hr) c c AquiferID AquiferNAME InitialStorage RecessionCoefSeepageCoef c-------------------------------------------------------------------------------------------- c713 Aquifer Pollutant Background Concentration c c AquiferID = Unique Aquifer identifier as in c712 c Ci = Background concentration for pollutant i (mg/l) c Where i = 1 to N (N = Number of QUAL from card 705) c c AQUIFER_ID QUALC1 QUALC2 ... QUALCN c-------------------------------------------------------------------------------------------- c714 FTable for BMP Class A, B, and C 109 c c FTABLE_ID = Unique FTable identifier (continuous string) c FLOW_LENGTH = Flow length (ft) c BED_SLOPE = Longitudinal bed slope (ft/ft) c NUM_RECORD = Number of layers in the FTable c c DEPTH = Water depth (ft) c SURFACE_AREA = Water surface area at the given depth (acre) c VOLUME = Storage volume at the given depth (ac-ft) c FLOW_WEIR = Overflow or weir outflow rate at the given depth (cfs) c FLOW_ORIFICE = Channel flow or orifice outflow rate at the given depth (cfs) c c FTABLE_ID FLOW_LENGTH BED_SLOPE NUM_RECORD c DEPTH SURFACE_AREA VOLUME FLOW_WEIR FLOW_ORIFICE c-------------------------------------------------------------------------------------------- c715 BMP SITE INFORMATION c c BMPSITE = Unique BMP site identifier c BMPNAME = BMP template name or site name c BMPTYPE = Unique BMP Types (must use the exact same keyword) c (BIORETENTION,WETPOND,CISTERN,DRYPOND,INFILTRATIONTRENCH,GREENROOF,POROUSPAVEMENT,RAINBARR EL,REGULATOR,SWALE,CONDUIT,BUFFERSTRIP,AREABMP) c DArea = Total Drainage Area in acre c NUMUNIT = Number of BMP structures c DDAREA = Design drainage area of the BMP structure (acre) c PreLUType = Predevelopment land use type (for external land simulation option) c AquiferID = Unique Aquifer ID, 0 --- no aquifer (for external land simulation option) c FTableFLG = FTable flag, 0 = no, 1 = yes (for BMP Class A, B, and C) c FTABLE_ID = Unique FTable identifier (continuous string) as in card 714 c 110 c BMPSITE BMPNAME BMPTYPE DArea NUMUNIT DDAREA PreLUType AquiferID FTableFLG FTABLE_ID 1_1 InfiltrationTrench1 INFILTRATIONTRENCH 14.455 0 0 1 0 0 No 1_2 BioRetention1 BIORETENTION 24.684 0 0 1 0 0 No 1_3 DryPond1 DRYPOND 113.642 0 0 1 0 0 No 1_4 WetPond1 WETPOND 643.154 0 0 1 0 0 No 1 Junction JUNCTION 29.353 1 0 1 0 0 No 2 Junction JUNCTION 0.000 1 0 1 0 0 No 3 Conduit CONDUIT 0 1 0 1 0 0 No c-------------------------------------------------------------------------------------------- c720 Point Source Definition c c point source timeseries data must be in this order; flow (in.ac/timestep), pollutant 1 (lbs/timestep), pollutant2, ... c c POINTSOURCE = Unique point source identifier c DESCRIPTION = Point source description (a continuous string without any space) c BMPSITE = BMP site identifier in card 715 c MULTIPLIER = Multiplier applied to the timeseries file (flow and pollutants). It will be in addition to the pollutant multiplier in card 705 c TIMESERIESFILE = File name containing input timeseries c SAND_FRAC = The fraction of total sediment which is sand (0-1) c SILT_FRAC = The fraction of total sediment which is silt (0-1) c CLAY_FRAC = The fraction of total sediment which is clay (0-1) c c POINTSOURCE DESCRIPTION BMPSITE MULTIPLIER TIMESERIESFILE SAND_FRAC SILT_FRAC CLAY_FRAC c-------------------------------------------------------------------------------------------- c721 Tier-1 Watershed Outlets Definition c c BMPSITE = BMP site (watershed outlet) identifier in card 715 c NUMBREAKS = Number of break points on the cost-effectiveness curve c CECurveFile = CECurve_Solutions file for the project cost (sorted cost value) of each break point 111 c c BMPSITE NUMBREAKS CECurveFile c-------------------------------------------------------------------------------------------- c722 Tier-1 Watershed Timeseries Definition c c BMPSITE = BMP site (watershed outlet) identifier in card 721 c BREAKPOINTID = Unique break point id on cost-effectiveness curve c (0 for initial, -1 for PreDev, and -2 for PostDev condition) c MULTIPLIER = Multiplier applied to the timeseries file c TIMESERIESFILE = Timeseries output file corresponding to the breakpoint id c c BMPSITE BREAKPOINTID MULTIPLIER TIMESERIESFILE c-------------------------------------------------------------------------------------------- c723 Pump Curve (applies if PUMP_FLG is ON in card 725) c c PUMP_CURVE = The unique name of pump curve (continuous string without space) c NUM_RECORD = Number of points on the curve c c DEPTH = Depth (ft) c FLOW = Pumping flow rate (cfs) c c PUMP_CURVE NUM_RECORD c DEPTH FLOW c-------------------------------------------------------------------------------------------- c725 CLASS-A BMP Site Parameters (required if BMPSITE is CLASS-A in card 715) c c BMPSITE = Class A BMP dimension group identifier in card 715 c WIDTH = Basin bottom width (ft) c LENGTH = Basin bottom length (ft) / diameter (ft) for rain barrel or cistern c OHEIGHT = Orifice Height (ft) c DIAM = Orifice Diameter (in) c EXTP = Exit Type (1 for C=1,2 for C=0.61, 3 for C=0.61, 4 for C=0.5) 112 c RELTP = Release Type (1-Cistern, 2-Rain barrel, 3-others) c PEOPLE = Number of persons (Cistern Option) c DDAYS = Number of dry days (Rain Barrel Option) c WEIRTP = Weir Type (1-Rectangular,2-Triangular) c WEIRH = Weir Height (ft) c WEIRW = (weir type 1) Weir width (ft) c THETA = (weir type 2) Weir angle (degrees) c ET_MULT = multiplier to PET c PUMP_FLG = pump option (0-OFF, 1-ON) c DEPTH_ON = water Depth (ft) at which the pump is started c DEPTH_OFF = water Depth (ft) at which the pump is stopped c PUMP_CURVE = The unique name of pump curve (continuous string without space) c c BMPSITE WIDTH LENGTH OHEIGHT DIAM EXITYPE RELEASETYPE PEOPLE DDAYS WEIRTYPE WEIRH WEIRW THETA ET_MULT PUMP_FLG DEPTH_ON DEPTH_OFF PUMP_CURVE 1_1 10 20 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1_2 10 30 0 0 1 3 0 0 1 1 5 0 1 0 0 0 1_3 25 80 0 0 1 3 0 0 1 5 5 0 1 0 0 0 1_4 25 80 0 0 1 3 0 0 1 5 5 0 1 0 0 0 c-------------------------------------------------------------------------------------------- c730 Cistern Control Water Release Curve (applies if release type is cistern in card 720) c c BMPSITE = Class A BMP dimension group identifier in card 715 c Flow = Hourly water release per capita from the Cistern Control (ft3/hr/capita) c c BMPSITE FLOW 1_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 113 1_3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1_4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c-------------------------------------------------------------------------------------------- c735 CLASS B BMP Site DIMENSION GROUPS c c BMPSITE = BMP Site identifier in card 715 c WIDTH = basin bottom width (ft) c LENGTH = basin bottom Length (ft) c MAXDEPTH = Maximum depth of channel (ft) c SLOPE1 = Side slope 1 (ft/ft) c SLOPE2 = Side slope 2 (ft/ft) (1-4) c SLOPE3 = Side slope 3 (ft/ft) c MANN_N = Manning 's roughness coefficient c ET_MULT = multiplier to PET c c BMPSITE WIDTH LENGTH MAXDEPTH SLOPE1 SLOPE2 SLOPE3 MANN_N ET_MULT c-------------------------------------------------------------------------------------------- c740 BMP Site BOTTOM SOIL/VEGITATION CHARACTERISTICS c c BMPSITE = BMPSITE identifier in c715 c INFILTM = Infiltration Method (0-Green Ampt, 1-Horton, 2-Holtan) c POLROTM = Pollutant Routing Method (1-Completely mixed, >1-number of CSTRs in series) c POLREMM = Pollutant Removal Method (0-1st order decay, 1-kadlec and knight method ) c SDEPTH = Soil Depth (ft) c POROSITY = Soil Porosity (0-1) c FCAPACITY = Soil Field Capacity (ft/ft) c WPOINT = Soil Wilting Point (ft/ft) c AVEG = Vegetative Parameter A (0.1-1.0) (Empirical), only required for Holtan infiltration method c FINFILT = Soil layer infiltration rate (in/hr) c UNDSWITCH = Consider underdrain (1), Do not consider underdrain (0) 114 c UNDDEPTH = Depth of storage media below underdrain (ft) c UNDVOID = Fraction of underdrain storage depth that is void space (0-1) c UNDINFILT = Background infiltration rate, below underdrain (in/hr) c SUCTION = Average value of soil capillary suction along the wetting front, value must be greater than zero (in), only required for Green-Ampt infiltration method c IMDMAX = Difference between soil porosity and initial moisture content, value must be greater than or equal to zero (a fraction), only required for Green-Ampt infiltration method c MAXINFILT = Maximum rate on the Horton infiltration curve (in/hr), only required for Horton infiltration method c DECAYCONS = Decay constant for the Horton infiltration curve (1/hr), only required for Horton infiltration method c DRYTIME = Time for a fully staurated soil to completely dry (day), only required for Horton infiltration method c MAXVOLUME = Maximum infiltration volume possible (in), only required for Horton infiltration method c c BMPSITE INFILTM POLROTM POLREMM SDEPTH POROSITY FCAPACITY WPOINT AVEG FINFILT UNDSWITCHUNDDEPTH UNDVOID UNDINFILT SUCTION IMDMAX MAXINFILT DECAYCONS DRYTIME MAXVOLUME 1_1 2 1 0 6 0.4 0.28 0.11 0 0 0 0 0 0 3 0.3 3 4 7 0 1_2 2 1 0 1 0.4 0.3 0.15 0.6 0.5 0 0 0 0 3 0.3 3 4 7 0 1_3 2 1 0 0 0 0.3 0.15 0 0 0 0 0 0 3 0.3 3 4 7 0 1_4 2 1 0 1.5 0.4 0.25 0.15 0.6 0.5 0 0.5 0.5 0.5 3 0.3 3 4 7 0 c-------------------------------------------------------------------------------------------- c745 BMP Site HOLTAN GROWTH INDEX c c HOLTAN EQUATION: F = GI * AVEG * (Computed Available Soil Storage)^1.4 + FINFILT c c BMPSITE = BMPSITE identifier in card 715 c GIi = 12 monthly values for GI in HOLTAN equation c Where i = jan, feb, mar ... dec c 115 c BMPSITE jan feb mar apr may jun jul aug sep oct nov dec 1_1 0 0 0 0 0 0 0 0 0 0 0 0 1_2 0.55 0.6 0.65 0.85 0.95 1 1 1 1 0.95 0.75 0.6 1_3 0 0 0 0 0 0 0 0 0 0 0 0 1_4 0 0 0 0 0 0 0 0 0 0 0 0 c-------------------------------------------------------------------------------------------- c747 BMP Site Initial Moisture Content c c BMPSITE = BMP Site identifier in card 715 c WATDEP_i = initial surface water depth (ft) c THETA_i = initial soil moisture (ft/ft) c c BMPSITE WATDEP_i THETA_i 1_1 0.000 0.000 1_2 0.000 0.150 1_3 0.000 0.150 1_4 0.000 0.150 c-------------------------------------------------------------------------------------------- c750 Class-C Conduit Parameters (required if BMPSITE is CLASS-C in card 715) c c BMPSITE = BMP site identifier in card 715 c INLET_NODE = BMP Id at the entrance of the conduit c OUTLET_NODE = BMP Id at the exit of the conduit c LENGTH = Conduit length (ft) c MANNING_N = Manning's roughness coefficient c INLET_IEL = Invert Elevation at the entrance of the conduit (ft) c OUTLET_IEL = Invert Elevation at the exit of the conduit (ft) c INIT_FLOW = Initial flow in the conduit (cfs) c INLET_HL = Head loss coefficient at the entrance of the conduit c OUTLET_HL = Head loss coefficient at the exit of the conduit c AVERAGE_HL = Head loss coefficient along the length of the conduit c 116 c BMPSITE INLET_NODE OUTLET_NODE LENGTH MANNING_N INLET_IEL OUTLET_IEL INIT_FLOW INLET_HL OUTLET_HL AVERAGE_HL 3 1 2 2267.44 0.14 0 0 0 0 0 0 c-------------------------------------------------------------------------------------------- c755 Class C Conduit Cross Sections c c LINK = BMP site identifier in card 715 c TYPE = Conduit Type (rectangular, circular...) c GEOM1 = Geometric cross-sectional property of the conduit c GEOM2 = Geometric cross-sectional property of the conduit c GEOM3 = Geometric cross-sectional property of the conduit c GEOM4 = Geometric cross-sectional property of the conduit c BARRELS = Number of Barrels in the conduit c c LINK TYPE GEOM1 GEOM2 GEOM3 GEOM4 BARRELS 3 DUMMY 0 0 0 0 0 c-------------------------------------------------------------------------------------------- c760 Irregular Cross Sections c c Format of transect data follows: c NC nLeft nRight nChannel c X1 name nSta xLeftBank xRightBank 0 0 0 xFactor yFactor c GR Elevation Station ... c c-------------------------------------------------------------------------------------------- c761 BufferStrip BMP Parameters (required if BMPTYPE is BUFFERSTRIP in card 715) c c BMPSITE = BMP site identifier in card 715 c Width = BMP width (ft) c FLength = flow length (ft) c DStorage = Surface depression storage (in) c SLOPE = Overland slope (ft / ft) 117 c MANNING_N = Overland Manning's roughness coefficient c POLREMM = Pollutant Removal Method (0-1st order decay, 1-kadlec and knight method) c ET_MULT = multiplier to PET c c BMPSITE Width FLength DStorage SLOPE MANNING_N POLREMM ET_MULT c-------------------------------------------------------------------------------------------- c762 Area BMP Parameters (required if BMPTYPE is AREABMP in card 715) c c BMPSITE = BMP site identifier in card 715 c Area = BMP area (ft2) c FLength = flow length (ft) note: area width = area / flow length c DStorage = Surface depression storage (in) c SLOPE = Overland slope (ft / ft) c MANNING_N = Overland Manning's roughness coefficient c SAT_INFILT = Saturated infiltration rate (in/hr) c POLREMM = Pollutant Removal Method (0-1st order decay, 1-kadlec and knight method) c DCIA = Percentage of Directly Connected Impervious Area (0-100) c TOTAL_IMP_DA = Total Impervious Drainage Area (acre) c c BMPSITE Area FLength DStorage SLOPE MANNING_N SAT_INFILT POLREMM DCIA TOTAL_IMP_DA c-------------------------------------------------------------------------------------------- c765 BMP SITE Pollutant Decay/Loss rates c c BMPSITE = BMP site identifier in card 715 c QUALDECAYi = First-order decay rate for pollutant i (hr^-1) c Where i = 1 to N (N = Number of QUAL from TIMESERIES FILES) c c BMPSITE QUALDECAY1 QUALDECAY2 ... QUALDECAYN 1_1 0.0100 1_2 0.0100 1_3 0.0100 118 1_4 0.0100 2 0.0100 c-------------------------------------------------------------------------------------------- c766 Pollutant K' values (applies when pollutant removal method is kadlec and knight method in card 740) c c BMPSITE = BMP site identifier in card 715 C K 'i = Constant rate for pollutant i (ft/yr) c Where i = 1 to N (N = Number of QUAL from card 705) c c BMPSITE QUALK'1 QUALK'2 ... QUALK'N 1_1 3280 1_2 3280 1_3 3280 1_4 49200 2 0 c-------------------------------------------------------------------------------------------- c767 Pollutant C* values (applies when pollutant removal method is kadlec and knight method in card 740) c c BMPSITE = BMP site identifier in card 715 c C*i = Background concentration for pollutant i (mg/l) c Where i = 1 to N (N = Number of QUAL from card 705) c c BMPSITE QUALC*1 QUALC*2 ... QUALC*N 1_1 12 1_2 12 1_3 12 1_4 30 2 0 c-------------------------------------------------------------------------------------------- c770 BMP Underdrain Pollutant Percent Removal (applies when underdrain is on in card 740) c c BMPSITE = BMPSITE identifier in card 715 119 c QUALPCTREMi = Perecent Removal for pollutant i through underdrain (0-1) c Where i = 1 to N (N = Number of QUAL from TIMESERIES FILES) c c BMPSITE QUALPCTREM1 QUALPCTREM2 ... QUALPCTREMN 1_1 0.1 1_2 0.1 1_3 0.1 1_4 0.85 c-------------------------------------------------------------------------------------------- c775 Sediment General Parameters (required if pollutant type is sediment in card 705) c c BMPSITE = BMP site identifier in card 715 c BEDWID = Bed width (ft) - this is constant for the entire simulation period c BEDDEP = Initial bed depth (ft) c BEDPOR = Bed sediment porosity c c BMPSITE BEDWID BEDDEP BEDPOR 1_1 0 0 0.5 1_2 0 0 0.5 1_3 0 0 0.5 1_4 0 0 0.5 2 0 0 0.5 c-------------------------------------------------------------------------------------------------------------------------- c780 Sand Transport Parameters (required if pollutant type is sediment in card 705) c c BMPSITE = BMP site identifier in card 715 c D = Effective diameter of the transported sand particles (in) c W = The corresponding fall velocity in still water (in/sec) c RHO = The density of the sand particles (lb/ft3) c KSAND = The coefficient in the sandload power function formula c EXPSND = The exponent in the sandload power function formula c 120 c BMPSITE D W RHO KSAND EXPSND 1_1 0 0 2.65 0 0 1_2 0 0 2.65 0 0 1_3 0 0 2.65 0 0 1_4 0 0 2.65 0 0 2 0 0 2.65 0 0 c-------------------------------------------------------------------------------------------------------------------------- c785 Silt Transport Parameters (required if pollutant type is sediment in card 705) c c BMPSITE = BMP site identifier in card 715 c D = Effective diameter of the transported silt particles (in) c W = The corresponding fall velocity in still water (in/sec) c RHO = The density of the silt particles (lb/ft3) c TAUCD = The critical bed shear stress for deposition (lb/ft2) c TAUCS = The critical bed shear stress for scour (lb/ft2) c M = The erodibility coefficient of the silt particles (lb/ft2/day) c c BMPSITE D W RHO TAUCD TAUCS M 1_1 0 0 2.65 10000000000 10000000000 0 1_2 0 0 2.65 10000000000 10000000000 0 1_3 0 0 2.65 10000000000 10000000000 0 1_4 0 0 2.65 10000000000 10000000000 0 2 0 0 2.65 10000000000 10000000000 0 c-------------------------------------------------------------------------------------------------------------------------- c786 Clay Transport Parameters (required if pollutant type is sediment in card 705) c c BMPSITE = BMP site identifier in card 715 c D = Effective diameter of the transported clay particles (in) c W = The corresponding fall velocity in still water (in/sec) c RHO = The density of the silt/clay particles (lb/ft3) c TAUCD = The critical bed shear stress for deposition (lb/ft2) c TAUCS = The critical bed shear stress for scour (lb/ft2) 121 c M = The erodibility coefficient of the clay particles (lb/ft2/day) c c BMPSITE D W RHO TAUCD TAUCS M 1_1 0 0 2.65 10000000000 10000000000 0 1_2 0 0 2.65 10000000000 10000000000 0 1_3 0 0 2.65 10000000000 10000000000 0 1_4 0 0 2.65 10000000000 10000000000 0 2 0 0 2.65 10000000000 10000000000 0 c-------------------------------------------------------------------------------------------------------------------------- c790 LAND TO BMP ROUTING NETWORK (required for external land simulation control in card 700) c c UniqueID = Identifies an instance of LANDTYPE in SCHEMATIC c LANDTYPE = Corresponds to LANDTYPE in c710 c AREA = Area of LANDTYPE in ACRES c DS = UNIQUE ID of DS BMP (0 - no BMP, add to end) c c UniqueID LANDTYPE AREA DS 1 1 5.559 1_3 2 2 181.918 1_4 3 3 2.891 1_3 4 4 17.791 1_3 5 5 87.401 1_3 6 6 4.225 1_2 7 7 10.230 1_2 8 8 113.198 1_4 9 9 14.455 1_1 10 10 136.995 1_4 11 11 0.444 1_4 12 12 1.111 1_4 13 13 9.340 1_2 14 14 0.889 1_2 15 15 5.334 1_4 122 16 16 17.124 1_4 17 17 2.223 1_4 18 18 0.222 1_4 19 19 89.402 1_4 20 20 11.786 1_4 21 21 59.379 1_4 22 22 0.222 1_4 23 23 23.796 1_4 24 24 11.564 1 25 25 0.222 1 26 26 3.113 1 27 27 1.111 1 28 28 5.782 1 29 29 7.561 1 c-------------------------------------------------------------------------------------------- c795 BMP Site ROUTING NETWORK c c BMPSITE = BMPSITE identifier in card 715 c OUTLET_TYPE = Outlet type (1-total, 2-weir, 3-orifice or channel, 4-underdrain) c DS = Downstrem BMP site identifier in card 715 (0 - no BMP, add to end) c c BMPSITE OUTLET_TYPE DS 2 1 0 1_1 1 1_4 1_2 1 1_4 1_3 1 1 1_4 1 1 1 1 3 3 3 2 c-------------------------------------------------------------------------------------------- c800 Optimization Controls c 123 c Technique -- Optimization Techniques c 0 = no optimization c 1 = Scatter Search c 2 = NSGAII c Option -- Optimization options c 0 = no optimization c 1 = specific control target and minimize cost c 2 = generate cost effectiveness curve c StopDelta -- Criteria for stopping the optimization iteration c in dollars($), meaning if the cost not improved by this criteria, stop the search (for Option 1) c MaxRuns -- Maximum number of iterations c NumBest -- Number of best solutions for output (for Option 1) c c Technique Option StopDelta MaxRuns NumBest 2 2 0 1000 2 c-------------------------------------------------------------------------------------------- c805 BMP Cost Functions c Cost ($) = ((LinearCost)*Length^(LengthExp) + (AreaCost)*Area^(AreaExp) + (TotalVolumeCost)*TotalVolume^(TotalVolExp) c (MediaVolumeCost) * SoilMediaVolume ^ (MediaVolExp) + (UnderDrainVolumeCost) * UnderDrainVolume ^ (UDVolExp) c + (ConstantCost)) * (1+PercentCost/100) c c BMPSITE = BMP site identifier in card 715 c LinearCost = Cost per unit length of the BMP structure ($/ft) c AreaCost = Cost per unit area of the BMP structure ($/ft^2) c TotalVolumeCost = Cost per unit total volume of the BMP structure ($/ft^3) c MediaVolumeCost = Cost per unit volume of the soil media ($/ft^3) c UnderDrainVolumeCost = Cost per unit volume of the under drain structure ($/ft^3) c ConstantCost = Constant cost ($) c PercentCost = Cost in percentage of all other cost (percent) c LengthExp = Exponent for linear unit c AreaExp = Exponent for area unit c TotalVolExp = Exponent for total volume unit 124 c MediaVolExp = Exponent for soil media volume unit c UDVolExp = Exponent for underdrain volume unit c c BMPSITE LinearCost AreaCost TotalVolumeCost MediaVolumeCost UnderDrainVolumeCost ConstantCost PercentCost LengthExp AreaExp TotalVolExp MediaVolExp UDVolExp 1_1 0 13.08 0 0 0 0 0 1 1 1 1 1 1_2 0 16.19 0 0 0 0 0 1 1 1 1 1 1_3 0 0 7.12 0 0 0 0 1 1 1 1 1 1_4 0 0 7.12 0 0 0 0 1 1 1 1 1 c-------------------------------------------------------------------------------------------- c810 BMP SITE AdjusTable Parameters c c BMPSITE = BMP site identifier in card 715 c VARIABLE = Variable name (must use the exact same keyword) c LENGTH --- BMP length, c NUMUNIT --- number of units, c WEIRH --- weir height, c SDEPTH --- soil media depth, c DCIA --- directly connected impervious area for area BMP type, c MAXDEPTH --- maximum surface storage depth for swale, c CECURVE --- cost-effectiveness curve for Tier-1 solution c FROM = From value in the range c TO = To value in the range c STEP = Increment step c c BMPSITE VARIABLE FROM TO STEP 1_1 NUMUNIT 0 200 1 1_2 NUMUNIT 0 250 1 1_3 NUMUNIT 0 500 1 1_4 NUMUNIT 0 500 1 c-------------------------------------------------------------------------------------------- c814 Predeveloped Timeseries at Assessment Point for Flow Duration Curve 125 c c BMPSITE = BMP site identifier in card 715 if it is an assessment point c NumBins = Number of bins for flow duration curve c PreDevFlag = Pre-developed timeseries option (1-internal,2-external) c PreDevFile = Pre-developed timeseries file path for external option c The timeseries file format (AssessmentPoint_ID Year Month Day Hour Minute Flow_cfs) c The first line is skipped (comment line) and data start from the second line in the required format. c c BMPSITE NumBins PreDevFlag PreDevFile c-------------------------------------------------------------------------------------------- c815 Assessment Point and Evaluation Factor c c BMPSITE -- BMP site identifier in card 715 if it is an assessment point c FactorGroup -- Flow or pollutant related evaluation factor group c -1 = flow related evaluation factor c # = pollutant ID in card 705 c FactorType -- Evaluation Factor Type (negative number for flow related and positive number for pollutant related) c -1 = AAFV Annual Average Flow Volume (ft3/yr) c -2 = PDF Peak Discharge Flow (cfs) c -3 = FEF Flow Exceeding frequency (#times/year) c -4 = FDC Flow Duration Curve (sum of sorted flow difference with pre-developed condition, cfs) c 1 = AAL Annual Average Load (lb/yr) c 2 = AAC Annual Average Concentration (mg/L) c 3 = MAC Maximum #days Average Concentraion (mg/L) c FactorVal1 -- if FactorType = 3 (MAC): Maximum #Days; c -- if FactorType = -3 (FEF): Threshold (cfs) c -- if FactorType = -4 (FDC): Low flow limit (cfs) c -- all other FactorType : -99 c FactorVal2 -- if FactorType = -3 (FEF): Minimum inter-exceedance time (hr) c if = 0 then daily running average flow exceeding frequency c if = -1 then daily average flow exceeding frequency c otherwise minimum inter-exceedance time for simulation interval 126 c -- if FactorType = -4 (FDC): High flow limit (cfs) c -- all other FactorType : -99 c CalcMode -- Evaluation Factor Calculation Mode c -99 for Option 0 (card 800): no optimizaiton c 1 = percent percent of value under existing condition (0-100) c 2 = S scale between pre-develop and existing condition (0-1) c 3 = V absolute value in the unit as shown in FactorType (third block in this card) c TargetVal1 -- Target value for evaluation factor calculation mode c -99 for Option 0 (card 800): no optimizaiton c Target value for minimize cost Option 1 (card 800) c Lower limit of target value for cost-effective curve Option 2 (card 800) c TargetVal2 -- Target value for evaluation factor calculation mode c -99 for Option 0 (card 800): no optimizaiton c -99 for Option 1 (card 800): minimize cost c Upper limit of target value for cost-effective curve Option 2 (card 800) c Factor_Name -- Evaluation factor name (user specified without any space), e.g. FlowVolume or SEDIMENT c c BMPSITE FactorGroup FactorType FactorVal1 FactorVal2 CalcMode TargetVal1 TargetVal2 FactorName 2 -1 -3 4 0 1 0 100 FEF c-------------------------------------------------------------------------------------------- 127 SUSTAIN application scales and HRU file format Figure C.1. SUSTAIN’s multiple scales of application [source: US EPA (2009)] Figure C.2. HRU file format 128 Cost computation and optimization (cost-effectiveness) in SUSTAIN The BMP cost function is provided in card 805 in above script and the details of how cost minimization and optimization is performed is given below. Following text is from US EPA (2009) report: “SUSTAIN - A Framework for Placement of Best Management Practices in Urban Watersheds to Protect Water Quality report” page 3-89 to 3-90: Optimization SUSTAIN provides two optimization options: (1) cost minimization, and (2) cost- effectiveness curve. In the cost-minimization option, the optimization search process identifies the near-optimal solutions that meet the user-specified management targets. With the cost-effectiveness curve option, the optimization process reveals all the cost-effective solutions within the user-specified management target range. Cost-Effectiveness Curve Option Under the cost-effectiveness curve option, the search aims at identifying the cost-effective solutions within the specified management target range. The multi objective problem can be expressed as follows: Minimize n ΣCost(BMPi ) and i=1 Minimize EF where BMPi = a set of BMP configuration decision variables associated with location i and EF = the management evaluation factor (EF) at one given assessment point According to Mateleska (2016): “Each unit cost per storage value represents the capital cost of construction/installation of the BMP and includes a 35percent design/engineering/contingency (D & E) cost. This 35percent fixed percentage of the total construction cost follows a general “rule of thumb,” often used 129 by consulting firms…..The unit costs do not include the cost of purchasing any land, nor does it include any O&M costs (which is discussed in more detail in a subsequent section). Therefore, each unit cost in based on the CRWA’s 2010 values was calculated by multiplying the relevant BMP cost by 1.35.” CNT experiment setup parameters and costs sheet Lot information Size of Lot 825.3 acres Predevelopment (percent) Impervious area: 45; Lan in good condition: 35; Lawn in fair condition: 19; Water: 1 Runoff Reduction Goal Select a goal: Custom; Precipitation capture depth: 2.4; Volume captured over: Whole site Figure C.3. CNT-based cost benefits outputs Conventional Development Roof size (ft2): 10000; Number of Parking spots: 25 130 Green Improvements Rain Garden (ft2): 14000; Vegetation Filter Strips: 10,30,12,0.35 Advanced Options Life Cycle: 30 BMPs selected for Montgomery county watershed assessment The BMPs selected for assessment are based on existing stormwater management practices currently employed in the watershed. The BMP drainage area, class, and types are provided in following Figures. This information is courtesy of Maryland Department of Environment’s Stormwater Management Geodatabase. Figure C.4. Urban stormwater infrastructure locations 131 Table C.1. Urban stormwater infrastructure locations 132 There are multiple BMP types shown in above Figure. These BMPs are combined into corresponding categories in SUSTAIN model. The excel sheet with details of aggregation is provided below. Table C.2. Montgomery county watershed infrastructure drainage area 133 Appendix D MD BMP Structural Practices Cost Estimation Table D.1. MD BMP eras cost estimation sheet *Average Annual BMP Code Code Description County cost ($/Imp acre treated/yr) FBIO Bioretention H,M,A,P 4380 FSND Sand Filter H,M,A,P 4390 FUND Underground Filter P 5009 IBAS Infiltration Basin H,M,A,P 4280 (w/o sand) ITRN Infiltration Trench H,M,A,P 4280 (w/o sand) MIDW Dry Well A MILS Landscape infiltration M 4280 (w/o sand) ODSW Dry Swale H,M,A,P 3427 (same as Bioswale) PMED Micropool Extended Detention Pond H,A PMPS Multiple Pond System A PPKT Pocket Pond A PWED Extended Detention Structure, H,M,A,P 2225 (same as wet Wet pond) PWET Retention Pond (Wet Pond) H,M,A,P 2225 WSHW Shallow Marsh H,M,A,P XDED Extended Detention Structure, Dry H,M,A,P 3596 XDPD Detention Structure (Dry Pond) H,M,A,P 3596 XOGS Oil Grit separator H,M,A,P The drainage area treatment and investment in structural and ESD practices in state of Maryland are computed for pre 2010 and post 2010. For drainage area computation, stormwater management geo database of 2015 and 2016 is provided by MDE for state 134 of Maryland. This database includes shapefiles (covering urban, restoration, and inspected BMPs) and associated attribute Tables with information of each practice (courtesy of the Maryland Department of the Environment’s Stormwater Management Program1 and NPDES Municipal Separate Storm Sewer System2). BMP class, types, and built date fields in the database are used to determine pre and post 2010 drainage area treatment trend. For structural and ESD investment estimates, built dates from geo database and cost estimates from spreadsheet by King and Hagan (King and Hagan, 2011) is used. Average annual cost ($/impervious acre treated/year) estimates, provided in 2011, are adjusted for 2018 dollar values. 1https://mde.maryland.gov/programs/Water/StormwaterManagementProgram/Pages/i ndex.aspx 2https://mde.maryland.gov/programs/Water/StormwaterManagementProgram/Pages/s torm_gen_permit.aspx  Average annual costs include: 1. Includes cost of site discovery, surveying, design, planning, permitting, etc. which, for various BMPs tend to range from 10percent to 40percent of BMP construction costs. 2. Includes routine annual maintenance costs, average annual intermittent maintenance costs, and county implementation costs. 3. Initial BMP costs, including preconstruction, construction, and land costs, are amortized over 20 years at 3percent to arrive at annualized initial costs. 135 "The Environmental Protection Agency’s (EPA) National Pollutant Discharge Elimination System (NPDES) stormwater regulations were published in 1990. Phase I of these regulations require large urban jurisdictions to control pollution in stormwater to the maximum extent practicable (MEP). For permitting purposes, municipalities with populations of greater than 250,000 are considered “large” and those with populations of between 100,000 and 250,000 as “medium.” Municipalities with less than 100,000 are handled separately under Phase II NPDES stormwater rules discussed separately. After receiving applications from Phase I municipalities in 1991 and 1992, MDE began issuing NPDES municipal stormwater permits in 1993. These permits are updated every five years. The following provides information on the most current activities on the Phase I permits." MDE has developed a web mapping application called "StormwaterPrint" (http://mde.maryland.gov/programs/Water/Stormprint/Pages/index.aspx). "StormwaterPrint is an interactive map depicting the location and type of stormwater best management practices (BMPs) in Maryland as reported to the Maryland Department of the Environment (MDE) by local agencies. Drainage area to a practice, when available, is also included. The accuracy and completeness of the information may only be achieved over time through continual updates. MDE believes sharing this information with the public is valuable and will ultimately help improve the quality of this product in the future. MDE makes no warranties of merchantability or fitness, expressed or implied, for the use of information obtained here. The requester (user) acknowledges and accepts all limitations, including the fact that all data, 136 information, and maps are dynamic and in a constant state of maintenance, correction and/or update. " MD BMP Stormwater Management Eras Based on growing need of transition from conventional infrastructure to GI or ESD to MEP due to climate change and urbanization, a comparative analysis of BMP structures and capital investment is carried out in this research for pre- and post-2010 eras. The structural, ESD, Alternate, and Other practices for stormwater treatment during pre and post 2010 Eras for individual counties in Maryland are presented in Figure (MD BMP Eras (Pre 2010 and Post 2010)). The drainage area units are in acres and converted to hectares. Some practices date back to 1950s therefore use of Pre 2010 is more appropriate to assess continuation and transition of structural practices to ESD. Pre 2010, stormwater treatment and storage of about 70 percent in Howard and Prince George’s county and 85percent in Montgomery and Anne Arundel county is carried out using structural practices. Maryland’s predominate stormwater management eras from pre 2010 to post 2010 (current) shows more focus on ESD compared to structural practices. Since ESD implementation to MEP is more recent phenomenon, the total drainage area treated using these practices is relatively small and still in the implementation phase. The investment in ESD practices in Howard and Anne Arundel county after year 2010 is about 10-15 percent of total structural capital during pre-2010 era as shown in Figure (MD structural investment). All of ESD practices have unified sizing criteria across state of Maryland to replicate hydrology for “woods in good condition”. These 137 practices are small scale stormwater management practices integrated into the urban landscape. The ESD investment after year 2010 shown in Figure (MD structural investment) includes practices to minimize imperviousness, rooftop disconnections in commercial and residential development, green roofs, permeable pavements, rainwater harvesting, landscape infiltration, dry wells, swales, residential landscape planters, residential raingardens, microscale treatment systems. Detailed cost sheets used to prepare cost investment charts are provided in Appendix C. Pre 2010 Post 2010 Howard Montgomery Anne Arundel Prince George’s Figure D.1. MD BMP Eras (Pre 2010 and Post 2010) 138 Figure D.2. MD structural investment 139 Bibliography Abatzoglou, J. T., & Brown, T. J. (2012). A comparison of statistical downscaling methods suited for wildfire applications. International Journal of Climatology, 32(5), 772–780. https://doi.org/10.1002/joc.2312 Ackerman, K. (2012). Urban agriculture: opportunities and constraints. In Metropolitan Sustainability (pp. 118-146). Adger, W. N., Agrawala, S., Mirza, M. M. Q., Conde, C., o’Brien, K., Pulhin, J., ... & Takahashi, K. (2007). Assessment of adaptation practices, options, constraints and capacity. Climate change, 717-743. Ahiablame, L. M., Engel, B. A., & Chaubey, I. (2012). Effectiveness of low impact development practices: Literature review and suggestions for future research. Water, Air, and Soil Pollution, 223(7), 4253–4273. doi:10.1007/s11270-012- 1189-2. Amundson, R., Berhe, A. A., Hopmans, J. W., Olson, C., Sztein, A. E., & Sparks, D. L. (2015). Soil and human security in the 21st century, 348(6235). https://doi.org/10.1126/science.1261071 Arnold, C. and Gibbons, J. (1996). “Impervious surface coverage: the emergence of a key environmental indicator.” J. Am. Planning Assoc. 62, 243–258. Arnold, J. G., & Allen, P. M. (1999). Automated methods for estimating baseflow and ground water recharge from streamflow records 1. JAWRA Journal of the American Water Resources Association, 35(2), 411-424. Arnold, J. G., Srinivasan, R., Muttiah, R. S., & Williams, J. R. (1998). Large area hydrologic modeling and assessment part I: model development 1. JAWRA Journal of the American Water Resources Association, 34(1), 73-89. ASCE Task Committee on Definition of Criteria for Evaluation of Watershed Models of the Watershed Management Committee, Irrigation and Drainage Division. (1993). Criteria for evaluation of watershed models. Journal of Irrigation and Drainage Engineering, 119(3), 429-442. 140 Assessment, M. E. (2005). Living beyond our means: natural assets and human well- being. Attwater, R., & Derry, C. (2017). Achieving resilience through water recycling in peri-urban agriculture. Water, 9(3), 223. Aylor, D. (1972). Noise Reduction by Vegetation and Ground. The Journal of the Acoustical Society of America, 1(2), 197–205. doi:10.1121/1.1912830. Bagstad, K. J., Semmens, D. J., Waage, S., & Winthrop, R. (2013). A comparative assessment of decision-support tools for ecosystem services quantification and valuation. Ecosystem Services, 5, 27–39. doi:10.1016/j.ecoser.2013.07.004 Bahri, A. (2012). Integrated urban water management. The Background Papers. Global Water Partnership (GWP) Technical Committee (TEC), Stockholm, Sweden. Bangash, R. F., Passuello, A., Sanchez-canales, M., Terrado, M., López, A., Elorza, F. J., … Schuhmacher, M. (2013). Science of the Total Environment Ecosystem services in Mediterranean river basin: Climate change impact on water provisioning and erosion control. Science of the Total Environment, The, 458– 460, 246–255. https://doi.org/10.1016/j.scitotenv.2013.04.025 Barbosa, A. E., Fernandes, J. N., & David, L. M. (2012). Key issues for sustainable urban stormwater management. Water research, 46(20), 6787-6798. Barthel, S., Folke, C., Colding, J. (2010). Social–ecological memory in urban gardens-retaining the capacity for management of ecosystem services. Global Environmental Change, 20, 255–265. Barton, D., Lindhjem, H., Magnussen, K., & Holen, S. (2012). Valuation of Ecosystem Services From Nordic Watersheds-from awareness raising to policy support. VALUESHED). Report to the Nordic Council of Ministers. TemaNord, 506. Benedict, M. A., & McMahon, E. T. (2012). Green infrastructure: linking landscapes and communities. Island Press. 141 Benedict, M. A., & McMahon, E. T. (2006). Green infrastructure. Island, Washington, DC. Bentsen, M., Bethke, I., Debernard, J. B., Iversen, T., Kirkevåg, A., Seland, Ø., ... & Kristjánsson, J. E. (2013). The Norwegian earth system model, NorESM1-M— Part 1: Description and basic evaluation of the physical climate. Geosci. Model Dev, 6(3), 687-720. Berland, A., & Hopton, M. E. (2014). Comparing street tree assemblages and associated stormwater benefits among communities in metropolitan Cincinnati, Ohio, USA. Urban forestry & urban greening, 13(4), 734-741. Biasutti, M., & Seager, R. (2015). Projected changes in US rainfall erosivity, 2945– 2961. https://doi.org/10.5194/hess-19-2945-2015 Blanco, H., & Lal, R. (2010). Soil and water conservation. Principles of Soil Conservation and Management. Springer. Boardman, J. (2006). Soil erosion science: Reflections on the limitations of current approaches. Catena, 68(2-3), 73-86. Bolund, P., and Hunhammar, S. (1999). Ecosystem services in urban areas. Ecological Economics, 29(2), 293–301. doi:10.1016/S0921-8009(99)00013-0 Bonnin, G. M., Martin, D., Lin, B., Parzybok, T., Yekta, M., and Riley, D. (2006). “Precipitation-frequency atlas of the United States.” National Weather Service, Silver Spring, MD. Borselli, L., Cassi, P., & Torri, D. (2008). Prolegomena to sediment and flow connectivity in the landscape: a GIS and field numerical assessment. Catena, 75(3), 268-277. Boyd, J., & Banzhaf, S. (2007). What are ecosystem services? The need for standardized environmental accounting units. Ecological Economics, 63(2-3), 616–626. doi:10.1016/j.ecolecon.2007.01.002 Brauman, K. A., Daily, G. C., Duarte, T. K., & Mooney, H. A. (2007). The nature and value of ecosystem services: an overview highlighting hydrologic services. 142 Annu. Rev. Environ. Resour., 32(1), 67–98. doi:10.1146/annurev.energy.32.031306.102758 Brown, L. C., & Foster, G. R. (1987). Storm erosivity using idealized intensity distributions. Transactions of the ASAE, 30(2), 379-0386. Brown, W., and T. Schueler. (1997). The economics of stormwater BMPs in the mid- Atlantic region: final report. Center for Watershed Protection, Silver Spring, MD. Brown, D. G., C. Polsky, P. Bolstad, S. D. Brody, D. Hulse, R. Kroh, T. R. Loveland, and A. Thomson, 2014: Ch. 13: Land Use and Land Cover Change. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program, 318-332. https://doi.org/10.7930/J05Q4T1Q Burkhard, B., Kroll, F., Nedkov, S., & Müller, F. (2012). Mapping ecosystem service supply, demand and budgets. Ecological Indicators, 21, 17–29. doi:10.1016/j.ecolind.2011.06.019 Cao, L., Zhang, K., Dai, H., & Liang, Y. (2015). Modeling interrill erosion on unpaved roads in the loess plateau of China. Land Degradation & Development, 26(8), 825-832. Carter, T., Jones, R. N., Lu, X., Bhadwal, S., Conde, C., Mearns, L., ... & Zurek, M. (2007). New assessment methods and the characterisation of future conditions. Cavalli, M., Trevisani, S., Comiti, F., & Marchi, L. (2013). Geomorphometric assessment of spatial sediment connectivity in small Alpine catchments. Geomorphology, 188, 31-41. Chan, K. M. A., Shaw, M. R., Cameron, D. R., Underwood, E. C., & Daily, G. C. (2006). Conservation planning for ecosystem services. PLoS Biology, 4(11), 2138–2152. doi:10.1371/journal.pbio.0040379 Chen, J., Brissette, F. P., & Leconte, R. (2011a). Uncertainty of Downscaling Method in Quantifying the Impact of Climate Change on Hydrology. Journal of Hydrology, 401(3-4), 190-202. 143 Chen, J., Brissette, F. P., Poulin, A., & Leconte, R. (2011b). Overall uncertainty study of the hydrologic impacts of climate change for a Canadian watershed. Water Resources Research, 47(12). https://doi.org/10.1029/2011WR010602 Chen, C. F., Sheng, M. Y., Chang, C. L., Kang, S. F., & Lin, J. Y. (2014). Application of the SUSTAIN model to a watershed-scale case for water quality management. Water, 6(12), 3575-3589. Chen, Y., Day, S. D., Wick, A. F., & McGuire, K. J. (2014). Influence of urban land development and subsequent soil rehabilitation on soil aggregates, carbon, and hydraulic conductivity. Science of the Total Environment, 494, 329-336. Chiesura, A. (2004). The role of urban parks for the sustainable city. Landscape and Urban Planning, 68, 129–138. Cho, S.-H., Poudyal, N. C., & Roberts, R. K. (2008). Spatial analysis of the amenity value of green open space. Ecological Economics, 66 (2–3), 403–416. Chylek, P., Li, J., Dubey, M. K., Wang, M., & Lesins, G. (2011). Observed and model simulated 20th century Arctic temperature variability: Canadian earth system model CanESM2. Atmospheric Chemistry and Physics Discussions, (8), 22893-22907. Connop, S., Vandergert, P., Eisenberg, B., Collier, M. J., Nash, C., Clough, J., & Newport, D. (2016). Renaturing cities using a regionally-focused biodiversity- led multifunctional benefits approach to urban green infrastructure. Environmental Science & Policy, 62, 99-111. Costanza, R., Wilson, M., Troy, A., Voinov, A., Liu, S., D'Agostino, J. (2006b). The Value of New Jersey's Ecosystem Services and Natural Capital. Gund Institute for Ecological Economics, University of Vermont for New Jersey Department of Environmental Protection. (July 2006). Costanza, R., d’Arge, R., de Groot, R., Faber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., & van der Belt, M. (1997). The value of the world’s ecosystems and natural capital. Nature 387, 253–260. 144 Costanza, R., de Groot, R., Sutton, P., van der Ploeg, S., Anderson, S. J., Kubiszewski, I., … Turner, R. K. (2014). Changes in the global value of ecosystem services. Global Environmental Change, 26(1), 152–158. doi:10.1016/j.gloenvcha.2014.04.002 Coumou, D., & Rahmstorf, S. (2012). A decade of weather extremes. Nature Climate Change, 2(7), 1–6. doi:10.1038/nclimate1452 Cox, P., & Stephenson, D. (2007). A changing climate for prediction. Science, 317(5835), 207-208. Daily, G. (1997). Nature's services: societal dependence on natural ecosystems. Island Press. Daily GC, Söderqvist T, & Aniyar S et al. (2000). The Value of Nature and the Nature of Value. Science 289, 395–396. Daly, C., Halbleib, M., Smith, J. I., Gibson, W. P., Doggett, M. K., Taylor, G. H., … Pasteris, P. P. (2008). Physiographically sensitive mapping of climatological temperature and precipitation across the conterminous United States. International Journal of Climatology, 28(15), 2031–2064. https://doi.org/10.1002/joc.1688 Danielsen, F., Sorensen, M.K., Olwig, M.F., Selvam, V., Parish, F., Burgess, N.D., Hiraishi, T., Karunagaran, V.M., Rasmussen, M.S., Hansen, L.B., Quarto, A., Suryadiputra, N. (2005). The Asian tsunami: a protective role for coastal vegetation. Science 310, 643. Davis, A. P., & McCuen, R. H. (2005). Stormwater Management for Smart Growth. Deb, K., Pratap, A., Agarwal, S., & Meyarivan, T. A. M. T. (2002). A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE transactions on evolutionary computation, 6(2), 182-197. De Groot, R. S., Wilson, M. A., & Boumans, R. M. J. (2002). A typology for the classification, description and valuation of ecosystem functions, goods and services. Ecological Economics, 41(3), 393–408. doi:10.1016/S0921- 8009(02)00089-7 145 De Roo, A. P. J., & Jetten, V. G. (1999). Calibrating and validating the LISEM model for two data sets from the Netherlands and South Africa. Catena, 37(3-4), 477- 493. De Vente, J., Poesen, J., Arabkhedri, M., & Verstraeten, G. (2007). The sediment delivery problem revisited. Progress in Physical Geography, 31(2), 155-178. De Vente, J., Poesen, J., Verstraeten, G., Govers, G., Vanmaercke, M., Van Rompaey, A., ... & Boix-Fayos, C. (2013). Predicting soil erosion and sediment yield at regional scales: where do we stand?. Earth-Science Reviews, 127, 16- 29. DeFries, R., & Eshleman, K. N. (2004). Land-use change and hydrologic processes: a major focus for the future. Hydrologic Processes, 18(AUGUST), 2183–2186. https://doi.org/10.1002/hyp.5584 Demuzere, M., Orru, K., Heidrich, O., Olazabal, E., Geneletti, D., Orru, H., ... & Faehnle, M. (2014). Mitigating and adapting to climate change: Multi- functional and multi-scale assessment of green urban infrastructure. Journal of environmental management, 146, 107-115. Derkzen, M. L., van Teeffelen, A. J. A., & Verburg, P. H. (2015). REVIEW: Quantifying urban ecosystem services based on high-resolution data of urban green space: An assessment for Rotterdam, the Netherlands. Journal of Applied Ecology, 52(4), 1020–1032. doi:10.1111/1365-2664.12469 Desmet, P. J. J., & Govers, G. (1996). A GIS procedure for automatically calculating the USLE LS factor on topographically complex landscape units. Journal of soil and water conservation, 51(5), 427-433. Dessai, S., & van der Sluijs, J. P. (2007). Uncertainty and climate change adaptation: A scoping study (Vol. 2007). Copernicus Institute for Sustainable Development and Innovation, Department of Science Technology and Society. Dhakal, K. P., & Chevalier, L. R. (2016). Urban stormwater governance: the need for a paradigm shift. Environmental management, 57(5), 1112-1124. 146 Dietz, M. E. (2007). Low impact development practices: A review of current research and recommendations for future directions. Water, Air, and Soil Pollution, 186(1-4), 351–363. doi:10.1007/s11270-007-9484-z Di Luzio, M., Johnson, G. L., Daly, C., Eischeid, J. K., & Arnold, J. G. (2008). Constructing retrospective gridded daily precipitation and temperature datasets for the conterminous United States. Journal of Applied Meteorology and Climatology, 47(2), 475–497. https://doi.org/10.1175/2007JAMC1356.1 Diamond, H. J., Karl, T. R., Palecki, M. A., Baker, C. B., Bell, J. E., Leeper, R. D., ... & Goodge, G. (2013). US Climate Reference Network after one decade of operations: Status and assessment. Bulletin of the American Meteorological Society, 94(4), 485-498. Doherty, J. M., Miller, J. F., Prellwitz, S. G., Thompson, A. M., Loheide, S. P., & Zedler, J. B. (2014). Hydrologic regimes revealed bundles and tradeoffs among six wetland services. Ecosystems, 17(6), 1026-1039. Donat, M. G., Lowry, A. L., Alexander, L. V., O’Gorman, P. A., & Maher, N. (2017). Addendum: More extreme precipitation in the world’s dry and wet regions. Nature Climate Change, 7(2), 154–158. https://doi.org/10.1038/nclimate3160 Dotto, C. B., Kleidorfer, M., Deletic, A., Rauch, W., McCarthy, D. T., & Fletcher, T. D. (2011). Performance and sensitivity analysis of stormwater models using a Bayesian approach and long-term high resolution data. Environmental Modelling & Software, 26(10), 1225-1239. Dufresne, J. L., Foujols, M. A., Denvil, S., Caubel, A., Marti, O., Aumont, O., ... & Bony, S. (2013). Climate change projections using the IPSL-CM5 Earth System Model: from CMIP3 to CMIP5. Climate Dynamics, 40(9-10), 2123-2165. Dunne, J. P., John, J. G., Adcroft, A. J., Griffies, S. M., Hallberg, R. W., Shevliakova, E., ... & Krasting, J. P. (2012). GFDL’s ESM2 global coupled climate–carbon earth system models. Part I: Physical formulation and baseline simulation characteristics. Journal of Climate, 25(19), 6646-6665. Easterling, D. R., Evans, J. L., Groisman, P. Y., Karl, T. R., Kunkel, K. E., & Ambenje, P. (2000). Observed variability and trends in extreme climate events: A brief review. Bulletin of the American Meteorological Society, 81(3), 417– 425. doi:10.1175/1520-0477(2000)081<0417:OVATIE>2.3.CO;2 147 Elliott, A. H., & Trowsdale, S. A. (2007). A review of models for low impact urban stormwater drainage. Environmental modelling & software, 22(3), 394-405. Elmqvist, T., Fragkias, M., Goodness, J., Güneralp, B., Marcotullio, P. J., McDonald, R. I., ... & Wilkinson, C. (Eds.). (2013). Urbanization, biodiversity and ecosystem services: challenges and opportunities: a global assessment. Springer. Elmqvist, T., Setälä, H., Handel, S. N., van der Ploeg, S., Aronson, J., Blignaut, J. N., … de Groot, R. (2015). Benefits of restoring ecosystem services in urban areas. Current Opinion in Environmental Sustainability, 14, 101–108. doi:10.1016/j.cosust.2015.05.001 Environmental Finance Center, UMD. (2017). Holistically analyzing the benefits of green infrastructure. Guidance for local governments. EPA. (2009). Technical guidance on implementing the stormwater runoff requirements for federal projects under section 438 of the energy independence and security act. EPA 841-B-09-001 Escobedo, F.J., Nowak, D.J. (2009). Spatial heterogeneity and air pollution removal by an urban forest. Landscape and Urban Planning 90, 102–110. Escobedo, F.J., Kroeger, T.,Wagner, J.E. (2011). Urban forests and pollution mitigation: analyzing ecosystem services and disservices. Environmental Pollution, 159, 2078–2087. Farrugia, S., Hudson, M. D., & McCulloch, L. (2013). An evaluation of flood control and urban cooling ecosystem services delivered by urban green infrastructure. International Journal of Biodiversity Science, Ecosystem Services & Management, 9(2), 136-145. Favis-Mortlock, D. (2013). Non‐Linear Dynamics, Self‐Organization and Cellular Automata Models. Environmental Modelling: Finding Simplicity in Complexity, 45-67. Fisher, B., Turner, R. K., Burgess, N. D., Swetnam, R. D., Green, J., Green, R. E., … Others. (2011). Measuring, modeling and mapping ecosystem services in the 148 eastern arc mountains of Tanzania. Progress in Physical Geography, 35(5), 595–611. doi:10.1177/0309133311422968 Fletcher, T. D., Shuster, W., Hunt, W. F., Ashley, R., Butler, D., Arthur, S., ... & Mikkelsen, P. S. (2015). SUDS, LID, BMPs, WSUD and more–The evolution and application of terminology surrounding urban drainage. Urban Water Journal, 12(7), 525-542. Fletcher, T. D., Andrieu, H., & Hamel, P. (2013). Understanding, management and modelling of urban hydrology and its consequences for receiving waters: A state of the art. Advances in water resources, 51, 261-279. Foley, J. A., DeFries, R., Asner, G. P., … Snyder, P. K. (2005). Global Consequences of Land Use. Science, 309(5734), 570–574. doi:10.1126/science.1111772 Foster, J., Lowe, A., & Winkelman, S. (2011). The value of green infrastructure for urban climate adaptation. Center for Clean Air Policy, 750, 1-52. Fowler, H. J., Blenkinsop, S., & Tebaldi, C. (2007). Linking climate change modelling to impacts studies: recent advances in downscaling techniques for hydrologic modelling. International journal of climatology, 27(12), 1547-1578. Freeman III, A. M., Herriges, J. A., & Kling, C. L. (2014). The measurement of environmental and resource values: theory and methods. Routledge. Füssel, H. M. (2007). Adaptation planning for climate change: concepts, assessment approaches, and key lessons. Sustainability science, 2(2), 265-275. Gaffin, S. R., Rosenzweig, C., & Kong, A. Y. (2012). Adapting to climate change through urban green infrastructure. Nature Climate Change, 2(10), 704. Gan, T. Y., Dlamini, E. M., & Biftu, G. F. (1997). Effects of model complexity and structure, data quality, and objective functions on hydrologic modeling. Journal of Hydrology, 192(1-4), 81-103. García-Ruiz, J. M., Beguería, S., Nadal-Romero, E., González-Hidalgo, J. C., Lana- Renault, N., & Sanjuán, Y. (2015). A meta-analysis of soil erosion rates across the world. Geomorphology, 239, 160-173. 149 Garen, D. C., & Moore, D. S. (2005). Curve number hydrology in water quality modeling: Uses, abuses, and future directions. JAWRA Journal of the American Water Resources Association, 41(2), 377-388. Georgakakos, A., P. Fleming, M. Dettinger, C. Peters-Lidard, Terese (T.C.) Richmond, K. Reckhow, K. White, and D. Yates. 2014: Ch. 3: Water Resources. Climate Change Impacts in the United States: The Third National Climate Assessment, J. M. Melillo, Terese (T.C.) Richmond, and G. W. Yohe, Eds., U.S. Global Change Research Program, 69-112. https://doi.org/10.7930/ J0G44N6T. Gersonius, B., Nasruddin, F., Ashley, R., Jeuken, A., Pathirana, A., & Zevenbergen, C. (2012). Developing the evidence base for mainstreaming adaptation of stormwater systems to climate change. Water Research, 46(20), 6824–6835. https://doi.org/10.1016/j.watres.2012.03.060 Gersonius, B., Ashley, R., Jeuken, A., Pathinara, A., & Zevenbergen, C. (2015). Accounting for uncertainty and flexibility in flood risk management: comparing R eal‐I n‐O ptions optimisation and A daptation T ipping P oints. Journal of Flood Risk Management, 8(2), 135-144. Gessesse, B., Bewket, W., & Bräuning, A. (2015). Model‐based characterization and monitoring of runoff and soil erosion in response to land use/land cover changes in the Modjo watershed, Ethiopia. Land degradation & development, 26(7), 711-724. Gianfagna, C. C., Johnson, C. E., Chandler, D. G., & Hofmann, C. (2015). Watershed area ratio accurately predicts daily streamflow in nested catchments in the Catskills, New York. Journal of Hydrology: Regional Studies, 4, 583-594. Gilbert RO. (1987). Statistical Methods for Environmental Pollution Monitoring. Van Nostrand Reinhold, New York. Gill, S. E., Handley, J. F., Ennos, A. R., & Pauleit, S. (2007). Adapting cities for climate change: the role of the green infrastructure. Built environment, 33(1), 115-133. Gittleman, M., Farmer, C. J., Kremer, P., & McPhearson, T. (2017). Estimating stormwater runoff for community gardens in New York City. Urban ecosystems, 20(1), 129-139. 150 Gómez-Baggethun, E., & Barton, D. N. (2013). Classifying and valuing ecosystem services for urban planning. Ecological Economics, 86, 235–245. doi:10.1016/j.ecolecon.2012.08.019 Goulden, S., Portman, M. E., Carmon, N., & Alon-Mozes, T. (2018). From conventional drainage to sustainable stormwater management: Beyond the technical challenges. Journal of environmental management, 219, 37-45. Grimaldi, S., Petroselli, A., & Romano, N. (2013). Green‐Ampt Curve‐Number mixed procedure as an empirical tool for rainfall–runoff modelling in small and ungauged basins. Hydrological processes, 27(8), 1253-1264. Grimm, N. B., Grimm, N. B., Faeth, S. H., Golubiewski, N. E., Redman, C. L., Wu, J., … Briggs, J. M. (2008). Global change and the ecology of cities. Science (New York, N.Y.), 319(5864), 756–760. https://doi.org/10.1126/science.1150195 Groening, G. (1995). School garden and kleingaerten: for education and enhancing life quality. Acta Horticulturae, 391, 53–64. Groisman, P. Y., Knight, R. W., & Karl, T. R. (2012). Changes in intense precipitation over the central United States. Journal of Hydrometeorology, 13(1), 47-66. Gudmundsson, L., Bremnes, J. B., Haugen, J. E., & Engen-Skaugen, T. (2012). Technical Note: Downscaling RCM precipitation to the station scale using statistical transformations – a comparison of methods. Hydrology and Earth System Sciences, 16(9), 3383–3390. https://doi.org/10.5194/hess-16-3383-2012 Haase, D., Larondelle, N., Andersson, E., Artmann, M., Borgström, S., Breuste, J., … Elmqvist, T. (2014). A quantitative review of urban ecosystem service assessments: Concepts, models, and implementation. Ambio, 43(4), 413–433. doi:10.1007/s13280-014-0504-0 Haasnoot, M., Kwakkel, J. H., Walker, W. E., & ter Maat, J. (2013). Dynamic adaptive policy pathways: A method for crafting robust decisions for a deeply uncertain world. Global environmental change, 23(2), 485-498. Hamel, P., Falinski, K., Sharp, R., Auerbach, D. A., Sánchez-Canales, M., & Dennedy-Frank, P. J. (2017). Sediment delivery modeling in practice: 151 Comparing the effects of watershed characteristics and data resolution across hydroclimatic regions. Science of the Total Environment, 580, 1381-1388. Hamel, P., Chaplin-Kramer, R., Sim, S., Mueller, C. (2015). A new approach to modeling the sediment retention service (InVEST 3.0): Case study of the Cape Fear catchment, North Carolina, USA. Sci. Total Environ. 166–177. Hamin, E. M., & Gurran, N. (2009). Urban form and climate change: Balancing adaptation and mitigation in the US and Australia. Habitat international, 33(3), 238-245. Hansen, LeRoy and Ribaudo, Marc, Economic Measures of Soil Conservation Benefits: Regional Values for Policy Assessment (2008). USDA Technical Bulletin No. 1922. Available at SRN: https://ssrn.com/abstract=2014840 Hassall, C. (2014). The ecology and biodiversity of urban ponds. Wiley Interdisciplinary Reviews: Water, 1(2), 187-206. Hejazi, M. I., & Moglen, G. E. (2008). The effect of climate and land use change on flow duration in the Maryland Piedmont region. Hydrological Processes: An International Journal, 22(24), 4710-4722. Hempel, S., Frieler, K., Warszawski, L., Schewe, J., & Piontek, F. (2013). A trend- preserving bias correction – The ISI-MIP approach. Earth System Dynamics, 4(2), 219–236. https://doi.org/10.5194/esd-4-219-2013 Holtan, H. N. (1961). A concept for infiltration estimates in watershed engineering. Agricultural Research Service, US Dept. of Agriculture, ARS 41-51, October. Houle, J. J., Roseen, R. M., Ballestero, T. P., Puls, T. A., and Sherrard Jr, J. (2013). Comparison of maintenance cost, labor demands, and system performance for LID and conventional stormwater management. Journal of environmental engineering, 139(7), 932-938. Hubacek, K., & Kronenberg, J. (2013). Synthesizing different perspectives on the value of urban ecosystem services. Landscape and Urban Planning, 109(1), 1– 6. doi:10.1016/j.landurbplan.2012.10.010 152 Ishii, M. (1994). Measurement of road traffic noise reduced by the employment of low physical barriers and potted vegetation. Inter-noise, 29–31, 595–597. Ishimatsu, K., Ito, K., Mitani, Y., Tanaka, Y., Sugahara, T., & Naka, Y. (2017). Use of rain gardens for stormwater management in urban design and planning. Landscape and Ecological Engineering, 13(1), 205-212. Jacobson, C. R. (2011). Identification and quantification of the hydrological impacts of imperviousness in urban catchments: A review. Journal of Environmental Management, 92(6), 1438–1448. https://doi.org/10.1016/j.jenvman.2011.01.018 Jakob Themeßl, M., Gobiet, A., & Leuprecht, A. (2011). Empirical-statistical downscaling and error correction of daily precipitation from regional climate models. International Journal of Climatology, 31(10), 1530–1544. https://doi.org/10.1002/joc.2168 Jamshidi, R., Dragovich, D., & Webb, A. A. (2014). Distributed empirical algorithms to estimate catchment scale sediment connectivity and yield in a subtropical region. Hydrological Processes, 28(4), 2671-2684. Jansson, Å. (2013). Reaching for a sustainable, resilient urban future using the lens of ecosystem services. Ecological Economics, 86, 285-291. Jantz, C., Drzyzga, S., & Maret, M. (2014). Calibrating and Validating a Simulation Model to Identify Drivers of Urban Land Cover Change in the Baltimore, MD Metropolitan Region. Land, 3(3), 1158–1179. https://doi.org/10.3390/land3031158 Jayasooriya, V. M., Ng, A. W. M., Muthukumaran, S., & Perera, B. J. C. (2017). Green infrastructure practices for improvement of urban air quality. Urban Forestry & Urban Greening, 21, 34-47. Jayasooriya, V. M., & Ng, A. W. M. (2014). Tools for modeling of stormwater management and economics of green infrastructure practices: a review. Water, Air, & Soil Pollution, 225(8), 2055. Jiang, C., Li, D., Wang, D., & Zhang, L. (2016). Quantification and assessment of changes in ecosystem service in the Three-River Headwaters Region, China as a 153 result of climate variability and land cover change. Ecological Indicators, 66, 199–211. doi:10.1016/j.ecolind.2016.01.051 Jiang, L., & O’Neill, B. C. (2017). Global urbanization projections for the Shared Socioeconomic Pathways. Global Environmental Change, 42, 193-199. Jim, C.Y., Wendy, Y., Chen, W.Y. (2009). Ecosystem services and valuation of urban forests in China. Cities, 26, 187–219. Johnston, R. J., Rolfe, J., Rosenberger, R. S., & Brouwer, R. (2015). Introduction to benefit transfer methods. In Benefit Transfer of Environmental and Resource Values (pp. 19-59). Springer, Dordrecht. Jones, R. N., & Preston, B. L. (2011). Adaptation and risk management. Wiley Interdisciplinary Reviews: Climate Change, 2(2), 296-308. Kain, J.-H., Larondelle, N., Haase, D., Rodríguez-Rodríguez, D., & Kaczorowska, A. (2016). Land use scenarios exploring local consequences for supply and demand of urban ecosystem services in Stockholm year 2050. Ecological Indicators. doi:10.1016/j.ecolind.2016.02.062 Kalcic, M. M., Chaubey, I., & Frankenberger, J. (2015). Defining Soil and Water Assessment Tool (SWAT) hydrologic response units (HRUs) by field boundaries. International Journal of Agricultural and Biological Engineering, 8(3), 69-80. Kati, V., & Jari, N. (2016). Bottom-up thinking—Identifying socio-cultural values of ecosystem services in local blue–green infrastructure planning in Helsinki, Finland. Land Use Policy, 50, 537-547. Katz, R. W., & Brown, B. G. (1992). Extreme events in a changing climate: Variability is more important than averages. Climatic Change, 21(3), 289–302. doi:10.1007/BF00139728 Keesstra, S., Nunes, J., Novara, A., Finger, D., Avelar, D., Kalantari, Z., & Cerdà, A. (2018). The superior effect of nature based solutions in land management for enhancing ecosystem services. Science of the Total Environment, 610, 997- 1009. 154 Kendall, M. G. (1975). Rank Correlation Methods, Griffin, London (1975). Kinnell, P. I. A. (2010). Event soil loss, runoff and the Universal Soil Loss Equation family of models: A review. Journal of hydrology, 385(1-4), 384-397. Kirshen, P., Aytur, S., Hecht, J., Walker, A., Burdick, D., Jones, S., ... & Mather, L. (2018). Integrated urban water management applied to adaptation to climate change. Urban climate, 24, 247-263. Kirtman, B., Power, S. B., Adedoyin, J. a., Boer, G. J., Bojariu, R., Camilloni, I., … Wang, H. J. (2013). Near-term Climate Change: Projections and Predictability. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, 953–1028. https://doi.org/10.1017/CBO9781107415324.023 Klimas, C., Williams, A., Hoff, M., Lawrence, B., Thompson, J., & Montgomery, J. (2016). Valuing ecosystem services and disservices across heterogeneous green spaces. Sustainability, 8(9), 853. Kodra, E., & Ganguly, A. R. (2014). Asymmetry of projected increases in extreme temperature distributions. Scientific Reports, 4, 5884. doi:10.1038/srep05884 Kopecká, M., Szatmári, D., & Rosina, K. (2017). Analysis of urban green spaces based on sentinel-2A: case studies from Slovakia. Land, 6(2), 25. Kragh, J. (1981). Road traffic noise attenuation by belts of trees. Journal of Sound and Vibration 74, 235–241. Krebs, G., Kokkonen, T., Valtanen, M., Setälä, H., & Koivusalo, H. (2014). Spatial resolution considerations for urban hydrological modelling. Journal of hydrology, 512, 482-497. Kumar, K., & Hundal, L. S. (2016). Soil in the city: Sustainably improving urban soils. Journal of environmental quality, 45(1), 2-8. Kundzewicz, Z. W. (2003). Extreme precipitation and floods in the changing world. Water Resources Systems-Hydrological Risk, Management and Development, (281), 32–39. 155 Kunkel, K. E., Karl, T. R., Brooks, H., Kossin, J., Lawrimore, J. H., Arndt, D., ... & Emanuel, K. (2013). Monitoring and understanding trends in extreme storms: State of knowledge. Bulletin of the American Meteorological Society, 94(4), 499-514. Kwadijk, J. C., Haasnoot, M., Mulder, J. P., Hoogvliet, M. M., Jeuken, A. B., van der Krogt, R. A., ... & de Wit, M. J. (2010). Using adaptation tipping points to prepare for climate change and sea level rise: a case study in the Netherlands. Wiley Interdisciplinary Reviews: Climate Change, 1(5), 729-740. Kwakkel, J. H., Haasnoot, M., & Walker, W. E. (2016). Comparing Robust Decision- Making and Dynamic Adaptive Policy Pathways for model-based decision support under deep uncertainty. Environmental Modelling & Software, 86, 168- 183. Lal, R. (Eds.). (1998). Soil quality and soil erosion. CRC press. Lal, R., Ahmadi, M., & Bajracharya, R. M. (2000). Erosional impacts on soil properties and corn yield on Alfisols in central Ohio. Land degradation & development, 11(6), 575-585. Lal, R. (2001). Soil degradation by erosion. Land degradation & development, 12(6), 519-539. Lal, R. (2003). Soil erosion and the global carbon budget, 29, 437–450. https://doi.org/10.1016/S0160-4120(02)00192-7 Lawler, J. J., Lewis, D. J., Nelson, E., Plantinga, A. J., Polasky, S., Withey, J. C., … Radeloff, V. C. (2014). Projected land-use change impacts on ecosystem services in the United States. Proceedings of the National Academy of Sciences of the United States of America, 111(20), 7492–7. doi:10.1073/pnas.1405557111 Lee, J. G., Selvakumar, A., Alvi, K., Riverson, J., Zhen, J. X., Shoemaker, L., & Lai, F. H. (2012). A watershed-scale design optimization model for stormwater best management practices. Environmental Modelling & Software, 37, 6-18. Lee, J. G., & Heaney, J. P. (2003). Estimation of Urban Imperviousness and its Impacts on Storm Water Systems. Journal of Water Resources Planning and 156 Management, 129(5), 419–426. https://doi.org/10.1061/(ASCE)0733- 9496(2003)129:5(419) Li, Y., Zhang, L., Qiu, J., Yan, J., & Wan, L. (2017). Spatially explicit quantification of the interactions among ecosystem services. Landscape Ecology, 32(6), 1181– 1199. https://doi.org/10.1007/s10980-017-0527-6 Li, Z., & Fang, H. (2016). Earth-Science Reviews Impacts of climate change on water erosion: A review. Earth Science Reviews, 163, 94–117. https://doi.org/10.1016/j.earscirev.2016.10.004 Locke, R. A., & Meyer, S. C. (2005). Kane County water resources investigations: interim report on shallow aquifer potentiometric surface mapping. Contract Report; 2005-04. López-Vicente, M., Poesen, J., Navas, A., & Gaspar, L. (2013). Predicting runoff and sediment connectivity and soil erosion by water for different land use scenarios in the Spanish Pre-Pyrenees. Catena, 102, 62-73. Maas, J., Verheij, R.A., de Groenewegen, P.P., Vries, S., Spreeuwenberg, P. (2006). Green space, urbanity, and health: how strong is the relation? Journal of Epidemiology and Community Health, 60, 587–592. Mäler, K.-G., S. Aniyar, and A ° Jansson. 2008. Accounting for ecosystem services as a way to understand the requirements for sustainable development. Proceedings of the National Academy of Sciences of the United States of America 105: 9501–9506. Mann, H. B. (1945). Nonparametric Tests Against Trend. Econometrica, 13(3), 245– 259. https://doi.org/10.2307/1907187 Martin, T. G., & Watson, J. E. M. (2016). Intact ecosystems provide best defence against climate change. Nature Climate Change, 6(2), 122–124. doi:10.1038/nclimate2918 Maraun, D., Wetterhall, F., Chandler, R. E., Kendon, E. J., Widmann, M., Brienen, S., … Thiele-Eich, I. (2010). Precipitation downscaling under climate change: Recent developements to bridge the gap between dynamical models and the end 157 user. Reviews of Geophysics, 48(2009RG000314), 1–38. https://doi.org/10.1029/2009RG000314.1.INTRODUCTION Maryland Department of Planning Land Use Land Cover Data 2010. https://planning.maryland.gov/Pages/OurWork/landuse.aspx Maryland Department of Planning Generalized Zoning Data (extrapolated from local zoning codes) 2016. https://catalog.data.gov/dataset/zoning-districts- generalized-zoning-data-this-data-set-offers-jurisdiction-zoning-boundaries-a Mateleska, K. (2016). Methodolgy for developing cost estimates for Opti-Tool. US Environmental Protection Agency. https://www3.epa.gov/region1/npdes/stormwater/ma/green-infrastructure- stormwater-bmp-cost-estimation.pdf McCuen, R. H., (1989). Hydrologic Analysis and Design. Prentice Hall, Englewood Cliffs, NJ. McCuen, R. (2017). Hydrologic analysis and design (Fourth ed.). Hoboken, NJ: Pearson Higher Education. McCuen, R. H., Knight, Z., & Cutter, A. G. (2006). Evaluation of the Nash–Sutcliffe efficiency index. Journal of Hydrologic Engineering, 11(6), 597-602. MDE (Maryland Department of Environment). (2010). Water Quality Analysis of Sediment in Middle Patuxent River, Howard County, Maryland. http://www.mde.state.md.us/programs/water/TMDL/ApprovedFinalTMDLs/Do cuments/www.mde.state.md.us/assets/document/MidPatx_Sed_WQA_091010_ final.pdf Meehl, G. A., Karl, T., Easterling, D. R., Changnon, S., Pielke Jr, R., Changnon, D., ... & Mearns, L. O. (2000). An introduction to trends in extreme weather and climate events: observations, socioeconomic impacts, terrestrial ecological impacts, and model projections. Bulletin of the American Meteorological Society, 81(3), 413-416. Mejía, A. I., & Moglen, G. E. (2009). Spatial Patterns of Urban Development from Optimization of Flood Peaks and Imperviousness-Based Measures, Journal of Hydrologic Engineering, ASCE, 14(4): 416-424. 158 Mejía, A. I., & Moglen, G. E. (2010a). Spatial distribution of imperviousness and the space-time variability of rainfall, runoff generation, and routing. Water Resources Research, 46(7). https://doi.org/10.1029/2009WR008568 Mejía, A. I., & Moglen, G. E. (2010b). Impact of the spatial distribution of imperviousness on the hydrologic response of an urbanizing basin. Hydrological Processes, 24(23), 3359–3373. https://doi.org/ 10.1002/hyp.7755 Merritt, W. S., Letcher, R. A., & Jakeman, A. J. (2003). A review of erosion and sediment transport models, 18, 761–799. https://doi.org/10.1016/S1364- 8152(03)00078-1 Miller, J. D., & Hutchins, M. (2017). The impacts of urbanisation and climate change on urban flooding and urban water quality: A review of the evidence concerning the United Kingdom. Journal of Hydrology: Regional Studies, 12, 345-362. Minami, K. (2009). Soil and humanity: Culture, civilization, livelihood and health. Soil science and plant nutrition, 55(5), 603-615. Minville, M., Brissette, F., & Leconte, R. (2008). Uncertainty of the impact of climate change on the hydrology of a nordic watershed. Journal of Hydrology, 358(1– 2), 70–83. https://doi.org/10.1016/j.jhydrol.2008.05.033 Mishra, V., Ganguly, A. R., Nijssen, B., & Lettenmaier, D. P. (2015). Changes in observed climate extremes in global urban areas. Environmental Research Letters, 10(2), 24005. https://doi.org/10.1088/1748-9326/10/2/024005 Mitchell, V. G. (2006). Applying integrated urban water management concepts: a review of Australian experience. Environmental management, 37(5), 589-605. Montanarella, L. (2015). The importance of land restoration for achieving a land degradation–neutral world. In Land Restoration (pp. 249-258). Montgomery, D. R. (2012). Dirt: the erosion of civilizations. University of California Press. Moore, T. L., Gulliver, J. S., Stack, L., & Simpson, M. H. (2016). Stormwater management and climate change: vulnerability and capacity for adaptation in 159 urban and suburban contexts. Climatic Change, 491–504. https://doi.org/10.1007/s10584-016-1766-2 Morgan, R. P. C., & Nearing, M. (Eds.). (2016). Handbook of erosion modelling. John Wiley & Sons. Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D., & Veith, T. L. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE, 50(3), 885- 900. Mullan, D. (2013). Soil erosion under the impacts of future climate change: Assessing the statistical signi fi cance of future changes and the potential on-site and off- site problems. Catena, 109, 234–246. https://doi.org/10.1016/j.catena.2013.03.007 Nikiforoff, C. C. (1959). Reappraisal of the soil. Science, 129(3343), 186-196. NOAA-NCDC (2003). Data Documentation for data set 3240, Summary of the Month Cooperative. National Climatic Data Center, National Oceanic and Atmospheric Administration. ftp://ftp.ncdc.noaa.gov/pub/data/hourly_precip- 3240/dsi3240.pdf National Research Council. (2009). Urban stormwater management in the United States. National Academies Press. Nations, U. (2015). World Population Prospects: The 2015 Revision. United Nations Economic and Social Affairs (Vol. XXXIII). doi:10.1007/s13398-014-0173-7.2 Nations, U. (2018). 2018 Revision of World Urbanization Prospects: https://www.un.org/development/desa/publications/2018-revision-of-world- urbanization-prospects.html NRCS (Natural Resources Conservation Service). (2004). Chapter 10 - Hydrologic Soil-Cover Complexes. Part 630 Hydrology, National Engineering Handbook. 20pp. Washington, DC: Natural Resources Conservation Service. NRCS (Natural Resources Conservation Service). (2017). “Description of SSURGO Database.” 160 (accessed October 30, 2017). Nearing, M. A., Foster, G. R., Lane, L. J., & Finkner, S. C. (1989). A process-based soil erosion model for USDA-Water Erosion Prediction Project technology. Transactions of the ASAE, 32(5), 1587-1593. Nearing, M. A. (2001). Potential changes in rainfall erosivity in the US with climate change during the 21st century. Journal of Soil and Water Conservation, 56(3), 229-232. Nearing, M. A., Pruski, F. F., & O'neal, M. R. (2004). Expected climate change impacts on soil erosion rates: a review. Journal of soil and water conservation, 59(1), 43-50. Nelson, K. C., Palmer, M. a, Pizzuto, J. E., Moglen, G. E., Angermeier, P. L., Hilderbrand, R. H., Hayhoe, K. (2009). Forecasting the combined effects of urbanization and climate change on stream ecosystems: from impacts to management options. The Journal of Applied Ecology, 46(1), 154–163. https://doi.org/10.1111/j.1365-2664.2008.01599.x Niemelä, J., Breuste, J. H., Guntenspergen, G., McIntyre, N. E., Elmqvist, T., & James, P. (Eds.). (2011). Urban ecology: patterns, processes, and applications. OUP Oxford. Nordman, E. E., Isely, E., Isely, P., & Denning, R. (2018). Benefit-cost analysis of stormwater green infrastructure practices for Grand Rapids, Michigan, USA. Journal of Cleaner Production, 200, 501-510. Nuzzo, R. (2014). Statistical errors. Nature, 506(7487), 150. Oberndorfer, E., Lundholm, J., Bass, B., Coffman, R. R., Doshi, H., Dunnett, N., … Rowe, B. (2007). Green Roofs as Urban Ecosystems: Ecological Structures, Functions, and Services. BioScience, 57(10), 823. doi:10.1641/B571005 O’Driscoll, M., Clinton, S., Jefferson, A., Manda, A., & McMillan, S. (2010). Urbanization Effects on Watershed Hydrology and In-Stream Processes in the Southern United States. Water, 2(3), 605–648. https://doi.org/10.3390/w2030605 161 Pandey, A., Himanshu, S. K., Mishra, S. K., & Singh, V. P. (2016). Catena Physically based soil erosion and sediment yield models revisited. Catena, 147, 595–620. https://doi.org/10.1016/j.catena.2016.08.002 Passel, J. S., & Cohn, D. V. (2008). U.S. Population Projections: 2005 – 2050. Pew Research Center: Social & Demographic Trends, 1–49. Retrieved from http://pewhispanic.org/reports/report.php?ReportID=85 Patt, A., Klein, R. J., & de la Vega-Leinert, A. (2005). Taking the uncertainty in climate-change vulnerability assessment seriously. Comptes Rendus Geoscience, 337(4), 411-424. Peterson, T. C., Heim Jr, R. R., Hirsch, R., Kaiser, D. P., Brooks, H., Diffenbaugh, N. S., ... & Katz, R. W. (2013). Monitoring and understanding changes in heat waves, cold waves, floods, and droughts in the United States: state of knowledge. Bulletin of the American Meteorological Society, 94(6), 821-834. Piani, C., Haerter, J. O., & Coppola, E. (2009). Statistical bias correction for daily precipitation in regional climate models over Europe. Theoretical and Applied Climatology, 99(1-2), 187–192. https://doi.org/10.1007/s00704-009-0134-9 Piani, C., Weedon, G. P., Best, M., Gomes, S. M., Viterbo, P., Hagenmann, S., & Haerter, J. O. (2010). Statistical bias correction of global simulated daily precipitation and temperature for the application of hydrologic models. Journal of Hydrology, 395, 199–215. Pielke, R. A. (May 2009). “Collateral damage from the death of stationary”, GEWEX WCRP. Pimentel, D., Harvey, C., Resosudarmo, P., Sinclair, K., Kurz, D., McNair, M., ... & Blair, R. (1995). Environmental and economic costs of soil erosion and conservation benefits. Science, 267(5201), 1117-1123. Pimentel, D. (2006). Soil erosion: a food and environmental threat. Environment, development and sustainability, 8(1), 119-137. Pimentel, D., & Burgess, M. (2013). Soil erosion threatens food production. Agriculture, 3(3), 443-463. 162 Posner, S., Verutes, G., Koh, I., Denu, D., & Ricketts, T. (2016). Global use of ecosystem service models. Ecosystem Services, 17, 131–141. doi:10.1016/j.ecoser.2015.12.003 Prudencio, L., & Null, S. E. (2018). Stormwater management and ecosystem services: a review. Environmental Research Letters, 13(3), 033002. Pruski, F. F., & Nearing, M. A. (2002). Runoff and soil-loss responses to changes in precipitation: A computer simulation study. Journal of soil and Water Conservation, 57(1), 7-16. Refsgaard, J. C. (1997). Parameterisation, calibration and validation of distributed hydrological models. Journal of hydrology, 198(1-4), 69-97. Renard, K. G., Foster, G. R., Weesies, G. A., & Porter, J. P. (1991). RUSLE: Revised universal soil loss equation. Journal of soil and Water Conservation, 46(1), 30- 33. Renard, K. G., Foster, G. R., Weesies, G. A., McCool, D. K., & Yoder, D. C. (1997). Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE) (Vol. 703). Washington, DC: United States Department of Agriculture. Risse, L. M., Nearing, M. A., Laflen, J. M., & Nicks, A. D. (1993). Error assessment in the universal soil loss equation. Soil Science society of America journal, 57(3), 825-833. Rodríguez, J. P., Beard, T. D., Bennett, E. M., Cumming, G. S., Cork, S. J., Agard, J., … Peterson, G. D. (2006). Trade-offs across space, time, and ecosystem services. Ecology and Society, 11(1). doi:10.2307/2390206 Rose, C. W., Coughlan, K. J., Ciesiolka, C. A. A., & Fentie, B. (1997). Program GUEST (Griffith University Erosion System Template). A New Soil Conservation Methodology and Application to Cropping Systems in Tropical Steeplands, KJ Coughlan and CW Rose, eds., ACIAR Technical Report, 34-38. Rosenberg, E. A., Keys, P. W., Booth, D. B., Hartley, D., Burkey, J., Steinemann, A. C., & Lettenmaier, D. P. (2010). Precipitation extremes and the impacts of 163 climate change on stormwater infrastructure in Washington State. Climatic Change, 319–349. https://doi.org/10.1007/s10584-010-9847-0 Routschek, A., Schmidt, J., & Kreienkamp, F. (2014). Catena Impact of climate change on soil erosion — A high-resolution projection on catchment scale until 2100 in Saxony / Germany. Catena, 121, 99–109. https://doi.org/10.1016/j.catena.2014.04.019 Roy, A. H., Wenger, S. J., Fletcher, T. D., Walsh, C. J., Ladson, A. R., Shuster, W. D., ... & Brown, R. R. (2008). Impediments and solutions to sustainable, watershed-scale urban stormwater management: lessons from Australia and the United States. Environmental management, 42(2), 344-359. Salvadore, E., Bronders, J., & Batelaan, O. (2015). Hydrological modelling of urbanized catchments: A review and future directions. Journal of hydrology, 529, 62-81. Saraswat, C., Kumar, P., & Mishra, B. K. (2016). Assessment of stormwater runoff management practices and governance under climate change and urbanization: An analysis of Bangkok, Hanoi and Tokyo. Environmental Science & Policy, 64, 101-117. Schicht, R., Adams, J. R., & Stall, J. B. (1976). Water resources availability, quality, and cost in northeastern Illinois. Illinois State Water Survey. Schilling, J., & Logan, J. (2008). Greening the rust belt: A green infrastructure model for right sizing America's shrinking cities. Journal of the American Planning Association, 74(4), 451-466. Schmidli, J., Frei, C., & Vidale, P. L. (2006). Downscaling from GCM precipitation: A benchmark for dynamical and statistical downscaling methods. International Journal of Climatology, 26(5), 679–689. https://doi.org/10.1002/joc.1287 Schueler, T. R. (1994) The Importance of Imperviousness. Watershed Protection Techniques, 1(3), 100-111. SCS (Soil Conservation Service). (1986). “Urban hydrology for small watersheds.” Technical Release 55, 2nd ed., Washington, DC. 164 Segura, C., Sun, G., McNulty, S., & Zhang, Y. (2014). Potential impacts of climate change on soil erosion vulnerability across the conterminous United States. Journal of Soil and Water Conservation, 69(2), 171-181. Serna-Chavez, H. M., Schulp, C. J. E., Van Bodegom, P. M., Bouten, W., Verburg, P. H., & Davidson, M. D. (2014). A quantitative framework for assessing spatial flows of ecosystem services. Ecological Indicators, 39, 24–33. doi:10.1016/j.ecolind.2013.11.024 Shoemaker, L., Riverson, J., Alvi, K., Zhen, J. X., & Murphy, R. (2012). Report on Enhanced Framework (SUSTAIN) and Field Applications to Placement of BMPs in Urban Watersheds. US Environmental Protection Agency, Washington, DC. EPA/600/R-11/144, 2012. Available at https://cfpub. epa. gov/si/si_public_record_report. cfm. Shoemaker, L., Riverson, J., Alvi, K., Zhen, J. X., Paul, S., & Rafi, T. (2009). SUSTAIN: a framework for placement of best management practices in urban watersheds to protect water quality. Washington, DC: National Risk Management Research Laboratory, Office of Research and Development, US Environmental Protection Agency. Shuster, W. D., Bonta, J., Thurston, H., Warnemuende, E., & Smith, D. R. (2005). Impacts of impervious surface on watershed hydrology: A review. Urban Water Journal, 2(4), 263–275. https://doi.org/10.1080/15730620500386529 Shuster, W. D., Morrison, M. A., & Webb, R. (2008). Front-loading urban stormwater management for success: A perspective incorporating current studies on the implementation of LIDR. Cities and the Environment (CATE), 1(2), 1–15. Singer, M. J., & Warkentin, B. P. (1996). Soils in an environmental context: an American perspective. Catena, 27(3-4), 179-189. Smith, M. D. (2011). The ecological role of climate extremes: Current understanding and future prospects. Journal of Ecology, 99(3), 651–655. doi:10.1111/j.1365- 2745.2011.01833.x Sougnez, N., van Wesemael, B., & Vanacker, V. (2011). Low erosion rates measured for steep, sparsely vegetated catchments in southeast Spain. Catena, 84(1-2), 1- 11. 165 Stürck, J., Schulp, C. J. E., & Verburg, P. H. (2015). Spatio-temporal dynamics of regulating ecosystem services in Europe – The role of past and future land use change. Applied Geography, 63, 121–135. doi:10.1016/j.apgeog.2015.06.009 Sun, Y., Tong, S., & Yang, Y. J. (2016). Modeling the cost-effectiveness of stormwater best management practices in an urban watershed in Las Vegas Valley. Applied Geography, 76, 49-61. Sun, G., & Lockaby, B. G. (2012). Water quantity and quality at the urban–rural interface. Urban–rural interfaces: Linking people and nature, (urbanruralinter), 29-48. Sundberg, C., Svensson, G., & Söderberg, H. (2004). Re-framing the assessment of sustainable stormwater systems. Clean technologies and environmental policy, 6(2), 120-127. Sunyer, M. A., Madsen, H., & Ang, P. H. (2012). A comparison of different regional climate models and statistical downscaling methods for extreme rainfall estimation under climate change. Atmospheric Research, 103, 119–128. https://doi.org/10.1016/j.atmosres.2011.06.011 Sura, P. (2011). A general perspective of extreme events in weather and climate. Atmospheric Research, 101(1-2), 1–21. doi:10.1016/j.atmosres.2011.01.012 Sura, P. (2014). Encyclopedia of Natural Resources: Air Climate : Extreme Events Climate : Extreme Events, (December), 37–41. doi:10.1081/E-ENRA- 120047635 Syvitski, J. P., & Milliman, J. D. (2007). Geology, geography, and humans battle for dominance over the delivery of fluvial sediment to the coastal ocean. The Journal of Geology, 115(1), 1-19. TEEB. (2010). The economics of ecosystems and biodiversity. Ecological and Economic Foundations. Earthscan London and Washington. Teutschbein, C., & Seibert, J. (2012). Bias correction of regional climate model simulations for hydrological climate-change impact studies: Review and evaluation of different methods. Journal of Hydrology, 456, 12-29. 166 Thornthwaite, C. W. (1948). An approach toward a rational classification of climate. Geographical review, 38(1), 55-94. Toy, T. J., Foster, G. R., & Renard, K. G. (2002). Soil erosion: processes, prediction, measurement, and control. John Wiley & Sons. Trenberth, K. E. (2011). Changes in precipitation with climate change. Climate Research, 47(1–2), 123–138. https://doi.org/10.3354/cr00953 Troy, A., & Grove, J. M. (2008). Property values, parks, and crime: A hedonic analysis in Baltimore, MD. Landscape and Urban Planning, 87 (3), 233–245. Tyrväinen, L., Pauleit, S., Seeland, K., de Vries, S. (2005). Benefits and uses of urban forests and trees. In: Konijnendijk, C., Nilsson, K., Randrup, T., Schipperijn, J. (Eds.), Urban Forests and Trees. Springer. Tzoulas, K., Korpela, K., Venn, S., Yli-Pelkonen, V., Kaźmierczak, A., Niemela, J., & James, P. (2007). Promoting ecosystem and human health in urban areas using Green Infrastructure: A literature review. Landscape and urban planning, 81(3), 167-178. USDA (United States Department of Agriculture). (1983). Sediment Sources, Yields, and Delivery Ratios. In National Engineering Handbook, Section 3, Sedimentation. Washington, D.C: United States Department of Agriculture, Natural Resources Conservation Service. US EPA. (2015). Tools Strategies and Lessons Learned from EPA Green Infrastructure Technical Assistance Projects. EPA 832-R-15-016 US EPA. (2010). Chesapeake Bay phase 5.3 community watershed model. EPA 903S10002-CBP/TRS-303-10. Vauramo, S., Setälä, H. (2011). Decomposition of labile and recalcitrant litter types under different plant communities in urban soils. Urban Ecosystems 14, 59–70. Verburg, P. H., Koomen, E., Hilferink, M., Pérez-Soba, M., & Lesschen, J. P. (2012). An assessment of the impact of climate adaptation measures to reduce flood risk on ecosystem services. Landscape Ecology, 27(4), 473–486. doi:10.1007/s10980-012-9715-6 167 Vieux, B. E., & Bedient, P. B. (2004). Assessing urban hydrologic prediction accuracy through event reconstruction. Journal of Hydrology, 299(3-4), 217- 236. Vigiak, O., Borselli, L., Newham, L.T.H., Mcinnes, J., Roberts, A.M. (2012). Comparison of conceptual landscape metrics to define hillslope-scale sediment delivery ratio. Geomorphology 138, 74–88. Villarreal, E. L., & Bengtsson, L. (2005). Response of a Sedum green-roof to individual rain events. Ecological Engineering, 25(1), 1–7. doi:10.1016/j.ecoleng.2004.11.008 Vogel, J. R., Moore, T. L., Coffman, R. R., Rodie, S. N., Hutchinson, S. L., McDonough, K. R., ... & McMaine, J. T. (2015). Critical review of technical questions facing low impact development and green infrastructure: A perspective from the Great Plains. Water Environment Research, 87(9), 849- 862. Vose, R. S., Applequist, S., Bourassa, M. A., Pryor, S. C., Barthelmie, R. J., Blanton, B., ... & Easterling, D. R. (2014). Monitoring and understanding changes in extremes: Extratropical storms, winds, and waves. Bulletin of the American Meteorological Society, 95(3), 377-386. Wagener, T., Sivapalan, M., Troch, P., & Woods, R. (2007). Catchment classification and hydrologic similarity. Geography compass, 1(4), 901-931. Walling, D. E. (1983). The sediment delivery problem. Journal of hydrology, 65(1-3), 209-237. Walton, W. C. (1965). Ground-water recharge and runoff in Illinois (Vol. 58, p. 55). Champaign, IL: Illinois State Water Survey. Wear, D. N. (2011). Forecasts of county-level land uses under three future scenarios: a technical document supporting the Forest Service 2010 RPA Assessment. Gen. Tech. Rep. SRS-141. Gen. Tech. Rep., SRS-141, 41. Weber, T., Sloan, A., & Wolf, J. (2006). Maryland’s Green Infrastructure Assessment: Development of a comprehensive approach to land conservation. 168 Landscape and Urban Planning, 77(1–2), 94–110. https://doi.org/10.1016/j.landurbplan.2005.02.002 Weber, T., Sloan, A., & Wolf, J. (2006). Maryland’s Green Infrastructure Assessment: Development of a comprehensive approach to land conservation. Landscape and Urban Planning, 77(1–2), 94–110. https://doi.org/10.1016/j.landurbplan.2005.02.002 W. E. F. (2012). Design of Urban Stormwater Controls. ASCE. Wicks, J. M., & Bathurst, J. C. (1996). SHESED: a physically based, distributed erosion and sediment yield component for the SHE hydrological modelling system. Journal of Hydrology, 175(1-4), 213-238. Wilby, R. L., & Harris, I. (2006). A framework for assessing uncertainties in climate change impacts: Low-flow scenarios for the River Thames, UK. Water Resources Research, 42(2). https://doi.org/10.1029/2005WR004065 Willemen, L., 2010. Mapping and modelling multifunctional landscapes. Ph.D. Thesis. Wageningen, 152 pp. Willems, P., Arnbjerg-Nielsen, K., Olsson, J., & Nguyen, V. T. V. (2012). Climate change impact assessment on urban rainfall extremes and urban drainage: Methods and shortcomings. Atmospheric Research, 103, 106–118. https://doi.org/10.1016/j.atmosres.2011.04.003 Wood, A. W., Leung, L. R., Sridhar, V., & Lettenmaier, D. P. (2004). Hydrologic implications of dynamical and statistical approaches to downscaling climate model outputs. Climatic change, 62(1-3), 189-216. Wuebbles, D., Meehl, G., Hayhoe, K., Karl, T. R., Kunkel, K., Santer, B., ... & Goodman, A. (2014). CMIP5 climate model analyses: climate extremes in the United States. Bulletin of the American Meteorological Society, 95(4), 571-583. Yang, D., Kanae, S., Oki, T., Koike, T., & Musiake, K. (2003). Global potential soil erosion with reference to land use and climate changes. Hydrological processes, 17(14), 2913-2928. 169 Yang, G., Bowling, L. C., Cherkauer, K. a., Pijanowski, B. C., & Niyogi, D. (2010). Hydroclimatic Response of Watersheds to Urban Intensity: An Observational and Modeling-Based Analysis for the White River Basin, Indiana. Journal of Hydrometeorology, 11(1), 122–138. https://doi.org/10.1175/2009JHM1143.1 Yang, G., Bowling, L. C., Cherkauer, K. a., & Pijanowski, B. C. (2011). The impact of urban development on hydrologic regime from catchment to basin scales. Landscape and Urban Planning, 103(2), 237–247. https://doi.org/10.1016/j.landurbplan.2011.08.003 Yang, J., Reichert, P., Abbaspour, K. C., & Yang, H. (2007). Hydrological modelling of the Chaohe Basin in China: Statistical model formulation and Bayesian inference. Journal of Hydrology, 340(3-4), 167-182. Yao, L., Wei, W., & Chen, L. (2016). How does imperviousness impact the urban rainfall-runoff process under various storm cases? Ecological Indicators, 60, 893–905. https://doi.org/10.1016/j.ecolind.2015.08.041 Zhang, X. C., & Nearing, M. A. (2005). Impact of climate change on soil erosion, runoff, and wheat productivity in central Oklahoma, 61, 185–195. https://doi.org/10.1016/j.catena.2005.03.009 Zhou, Q. (2014). A review of sustainable urban drainage systems considering the climate change and urbanization impacts. Water, 6(4), 976-992. Zwiers, F. W., Alexander, L. V, Hegerl, G. C., Knutson, T. R., Kossin, J. P., Naveau, P., … Zhang, X. (2013). Climate Science for Serving Society. doi:10.1007/978- 94-007-6692-1 170