xvEPA
United States
Environmental Protection EPA/600/R-14/329
Agency ARS/309819
October 2014
Representing Green
Infrastructure Management
Techniques in Arid and
Semi-arid Regions:
Software Implementation
and Demonstration using
the AGWA/KINEROS2
Watershed Model
RESEARCH AND DEVELOPMENT
-------
-------
Representing Green Infrastructure
Management Techniques in Arid
and Semi-arid Regions: Software
Implementation and Demonstration
using the AGWA/KINEROS2
Watershed Model
Yoganand Korgaonkar1,1. Shea Burns1, D. Phillip Guertin1, David C.
Goodrich2, Carl L. Unkrich2, Jane E. Barlow1, and William G. Kepner3
1University of Arizona, School of Natural Resources and the Environment, Tucson, AZ
2USDA-Agricultural Research Service, Southwest Watershed Research Center, Tucson, AZ
3U.S. Environmental Protection Agency, Office of Research and Development, Las Vegas, NV
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
-------
Acknowledgements
This project was funded through the U.S. Environmental Protection Agency (EPA) Office of
Water via the Office of Research and Development (ORD).
We would like to acknowledge the key reviewers of this report for their valuable suggestions.
Specifically, our thanks go to Dr. Kevin E. Lansey, Department Head Department of Civil
Engineering and Engineering Mechanics, University of Arizona, Tucson, AZ; William Bunch,
ORISE Fellow, Ecosystems Protection Program, Aquatic Resource Protection and
Accountability Unit, U.S. EPA Region 8, Denver, CO; and Dr. Laura M. Norman, Research
Physical Scientist, Western Geographic Science Center, U.S. Geological Survey, Tucson, AZ.
This report has been subjected to the EPA/ORD and U.S. Department of
Agriculture/Agricultural Research Service (USDA/ARS) peer and administrative review
processes and has been approved for publication. The Automated Geospatial Watershed
Assessment (AGWA) tool was jointly developed by EPA/ORD, USDA/ARS, and the University
of Arizona (EPA/600/C-13/148 and ARS/296053).
in
-------
IV
-------
Table of Contents
1.0 Abstract 1
2.0 Introduction 3
3.0 Literature Review 5
3.1 Modeling Approaches to GI Practices 5
3.2 Green Infrastructure Practices 7
A. Bioretention 7
B. Permeable Pavements 7
C. Rainwater Harvesting 8
4.0 Objectives and Scope 9
5.0 AGWA and KINEROS2 13
6.0 Design and Development 17
7.0 Testing 25
7.1 Lot Scale 25
7.2 Subdivision Scale 29
8.0 Case Study 31
9.0 Limitations and Issues 39
10.0 Conclusions 41
11.0 References 43
-------
VI
-------
List of Figures
Figure 1: Study Area Map Showing Gauge Locations, Infiltration Measurement
Locations, and Watershed Boundaries; Area in Upper Right of Urban
Watershed Drains Directly to Watershed Outlet Through an Underground
Culvert; Runoff from Remaining Area is Routed Along Streets (Background
Image Courtesy USGS Earth Resources Observation and Science Center)
(from Kennedy et al., 2013) 10
Figure 2: AGWA Workflow 15
FigureS: KINEROS2 Elements 15
Figure 4: KINEROS2 Urban Element Components 15
Figure 5: Lot Scale Representation of the KINEROS2 Urban Element 25
Figure 6: Hydrographs of Lot Scale Testing Scenarios, Illustrating Effective
Hydraulic Conductivities Versus Theoretical, Steady-State Hydraulic
Conductivities 28
Figure 7: Hydrographs of Lot Scale Testing Scenarios, Comparing Modeled
Outflow to Theoretical Steady-State Outflow 29
Figure 8: Simulated Versus Observed Event Runoff and Peak Flows (n = 47)
for July 2005 Through September 2006 for the La Terraza Subdivision 30
Figure 9: La Terraza Subdivision and Flow Routed Towards Outlet.
(Sierra Vista, AZ) 32
Figure 10: KINEROS2 Representation of a Pre-Development Parcel 34
Figure 11: KINEROS2 Representation of Parcel ID 7 for Post-Development
without GI Practices 34
Figure 12: KINEROS2 Representation of Parcel ID 7 for Post-Development
with all GI Practices 34
Figure 13: Percent Change in Infiltration for Post-Development with and
without GI as Compared to Pre-Development 35
Figure 14: Percent Change in Runoff for Post-Development with and without GI
Practices as Compared to Pre-Development 36
Figure 15: Comparison of Flow Accumulation for Post-Development with and
without GI Practices 37
vn
-------
Vlll
-------
List of Tables
Table 1: Description of Fields in the Flow Routing Table 18
Table 2: Description of Fields in the Parameterization Table 18
Table 3: Description of Fields in the Retention Basin Designs Table 20
Table 4: Description of Fields in the Permeable Pavement Designs Table 20
Table 5: Description of Fields in the Rainwater Harvesting Design Table 21
Table 6: Description of Fields in the Placement Plans table 21
Table 7: Description of Fields in the Simulations Table 22
Table 8: Description of the Lot Scale Verification Scenarios 26
Table 9: Volume Balances of the Lot Scale Verification Scenarios 26
Table 10: Effective Versus Steady State, Weighted Hydraulic Conductivities for the
Lot Scale Verification Scenarios 27
Table 11: Parameter Values for Pre-Development and Post-Development Simulations 31
IX
-------
-------
Acronyms and Abbreviations
AGWA
ARS
BMP
BMPDSS
CGP
CN
EPA
ESRI
GI
CIS
HEC-HMS
HSPF
IHACRES
IMP
KINEROS2
LID
L-THIA-LID
NRCS
ORD
PC
PICP
SCS
SSURGO
SUSTAIN
SWAT
SWMM
USDA
Automated Geospatial Watershed Assessment Tool
Agricultural Research Service
Best Management Practice
Best Management Practice Decision Support System
Connecting Grid Pavers
Curve Number
U.S. Environmental Protection Agency
Environmental Systems Research Institute
Green Infrastructure
Geographic Information System
Hydrologic Engineering Center's Hydrologic Modeling System
Hydrologic Simulation Program - Fortran
Integrated Hydrologic Analysis - Center for Resources and
Environmental Studies
Integrated Management Practice
Kinematic Runoff and Erosion Model
Low Impact Development
Long-term Hydrologic Impact Assessment - Low Impact
Development
Natural Resources Conservation Service
Office of Research and Development
Pervious Concrete
Permeable Interlocking Concrete Pavers
Soil Conservation Service
Soil Survey Geographic Data Base
System for Urban Storm Water Treatment and Analysis Integration
Soil and Water Assessment Tool
Storm Water Management Model
U.S. Department of Agriculture
XI
-------
Xll
-------
1.0 Abstract
Increasing urban development in the arid and semi-arid regions of the southwestern
United States has led to greater demand for water from a region of limited water resources which
has fundamentally altered the hydrologic response of developed watersheds. Green Infrastructure
(GI) practices are being widely adopted to mitigate the impacts of development on water quantity
and quality. However, Geographic Information System (GlS)-based watershed tools that operate
from the lot-to-sub division-to-watershed level for rapid GI planning assessments are lacking.
The Automated Geospatial Watershed Assessment (AGWA) tool was modified to allow the
design and placement of a small set of GI practices in order to simulate urban hydrology with
and without GI features. This software development effort was undertaken to take advantage of
the advanced, physically-based infiltration algorithms and geometric flexibility of the Kinematic
Runoff and Erosion (KINEROS) 2 watershed model. The resulting software provides an up-to-
date GIS GI assessment framework that automatically derives model parameters from widely
available spatial data. It is also capable of manipulating GI features and simulating at the lot
scale within a graphical interface to conveniently view and compare simulation results with and
without GI features. These features distinguish the approach presented herein from existing GI
hydrology tools. The AGWA GI software was then tested at the lot level with and without GI
features to ensure programming integrity and hydrologically sound results. Further testing was
conducted at the subdivision level without GI features as high-resolution rainfall-runoff
observations were available from a subdivision in Sierra Vista, Arizona. This testing also
confirmed programming integrity and the capability to realistically simulate urban hydrology. A
set of case study simulations was then conducted for the Sierra Vista subdivision with various
combinations of the implemented GI features. Results indicate that the resulting software was
robust at the lot, subdivision, and small watershed level and it can realistically represent and
simulate storm runoff responses for the selected GI features. The AGWA GI tool offers a
foundation for the incorporation of a broader array of GI features.
-------
-------
2.0 Introduction
Urbanization has numerous effects on a watershed as it replaces vegetation and pervious
open areas with impervious surfaces such as roofs, driveways, parking lots, and roads. The
introduction of impervious surfaces has significant impacts on watershed hydrology, especially
in regard to drastic reductions in infiltration of rainfall, resulting in increased runoff volumes,
peak discharges, and higher energy releases. Increased runoff results in lower groundwater
recharge and base flows in humid regions (Leopold, 1968; Makepeace et al., 1995; Rose and
Peters, 2001). In semi-arid regions urbanization will increase runoff, as in humid regions, but can
also result in increased groundwater recharge by concentrating runoff in areas with higher
infiltration capacity, such as ephemeral alluvial channels (Scanlon et al., 2003; Goodrich et al.,
2004).
Soil compaction and lower infiltration also result in lower soil storage volume, and rapid
soil saturation with rainfall events (Booth and Jackson, 1997; Dunne and Leopold, 1978;
Kennedy et al., 2013). Larger runoff volume increases downstream discharge and flood
magnitudes (Norman et al., 2010). Urbanization may increase small floods by a factor of 10 or
more depending on the percentage of paved area (Hollis, 1975). Higher peak discharge and
runoff volumes can significantly alter stream morphology (White and Greer, 2006). Bank erosion
and down-cutting of stream beds due to higher discharge volumes results in wider and deeper
streams (Booth, 1991; Hammer, 1972). Research suggests that watersheds with as little as 10 to
20 percent impervious area have the potential to increase stream instability (Bledsoe and Watson,
2001).
Nonpoint source pollution is the major cause of urban water quality problems. Storm
water runoff collects and concentrates contaminants as it flows over impervious areas
(Characklis and Wiesner, 1997; Russ and Russ, 2002). Commonly known pollutants include
heavy metals such as copper, lead, zinc and iron; suspended solids; fecal coliform bacteria;
nutrients in the form of nitrogen and phosphorus; and hydrocarbons (Bedan and Clausen, 2009;
Duda et al., 1982; Norman et al., 2008a).
Traditional storm water management techniques involve transporting the water away
from urban areas as quickly as possible; reducing lag times, and increasing runoff volume and
peak flows (Booth et al., 2002; Hammer, 1972; Hollis, 1975; Hood et al., 2007; Leopold, 1968).
Sustainable planning for urban growth has also been used as a technique which can reduce
downstream impacts by decreasing growth in erosion or high-flow hot spots (Norman et al.,
2008b). These techniques have numerous impacts on downstream hydrology. New storm water
management approaches focus on handling and treating water at the source and reducing
downstream impacts. These so-called Low Impact Development (LID) Best Management
Practices (BMP) or Green Infrastructure (GI) practices aim at managing rainwater at the site
level before it reaches channels.
The Department of Environmental Resources of Prince George's County, Maryland,
pioneered LID to mitigate the urbanization impact of increasing impervious surfaces (County,
1999). As opposed to traditional storm water management practices, LID aims to preserve the
pre-development hydrology using a variety of cost effective on-site design techniques that store,
infiltrate, evaporate, and detain runoff. The overall goal is to encourage source control practices
-------
to manage storm water. Prince George's County introduced a new concept of Integrated
Management Practices (IMP), that include many LID controls such as bioretention cells or
basins, dry wells, filter strips, vegetated buffers, level spreaders, grassed swales, rain barrels,
cisterns, and infiltration trenches. LID controls help reduce the need for more traditional storm
water management techniques such as curb-and-gutter systems or large detention basins.
Following the guidelines set forth by Prince George's County, there have been multiple
implementations of LID practices in urban developments. One example of a large scale housing
development was the Jordan Cove Urban Watershed project in Waterford, Connecticut that
included monitoring water quality and quantity during and after construction (http://
jordancove.uconn.edu). The Jordan Cove Urban Watershed National Monitoring Program
Project was a ten-year study funded through the Connecticut Department of Environmental
Protection by the U.S. Environmental Protection Agency's (EPA) Section 319 National
Monitoring Program (Clausen, 2007). This study employed a paired-watershed approach,
wherein, one watershed was developed using traditional development practices (referred to as the
"traditional" watershed and a second watershed was developed using LID practices (referred to
as the "BMP" watershed). The study's goal was to determine water quality and quantity benefits
of urban residential storm water and pollution prevention BMPs. This project focused on
assessing the cumulative impact of LID controls such as rain gardens, cul-de-sacs, grassed
swales, porous paver roads, shared pervious driveways, and permanent open spaces. The project
was successful in maintaining pre-development runoff peaks and volumes using LID controls.
Runoff volume was observed to have decreased by 74 percent in the BMP watershed as
compared to the traditional watershed.
LID BMPs have been implemented and evaluated all around the world. Much effort has
been put into the modeling of these practices to aid in decision making with respect to design,
cost, efficiency, and effectiveness (Ahiablame et al., 2012a; Dietz, 2007; Elliott and Trowsdale,
2007; USEPA, 2000). Models that simulate LID practices include the Storm Water Management
Model (SWMM) (Rossman and Supply, 2010), Long-term Hydrologic Impact Assessment -
Low Impact Development (L-THIA-LID) model (Ahiablame et al., 2012b), System for Urban
storm water Treatment and Analysis Integration (SUSTAIN) (Lee et al., 2012), Hydrologic
Simulation Program - Fortran (HSPF) (Bicknell et al., 2001), and BMP Decision Support
System (BMPDSS) (Cheng et al., 2009).
The goal of this study is to design and develop a decision support tool to assist in the
planning and application of GI practices in urban developments and integrate it into a
Geographic Information System (GIS) framework. As the majority of GI/LID work has been
developed and applied in humid landscapes this study will focus on arid and semi-arid
watersheds. The AGWA GI software will serve as the decision tool and be applicable at the lot,
subdivision, and small watershed scales and utilize several of the features of the KINEROS2
rainfall-runoff and erosion model that are well suited to arid and semi-arid watersheds (Goodrich
etal. 2012).
-------
3.0 Literature Review
The literature review will discuss several modeling systems that have been developed to
estimate and assess the hydrologic and water quality impacts of urbanization and the use of
various GI/LID features. A small subset of GI features, commonly used in arid and semi-arid
regions, was selected for incorporation into the modeling tool for this study. A brief review of
these GI features and a sampling of how they have been used is also presented.
3.1 Modeling Approaches to GI Practices
Various models can be used to simulate and evaluate the hydrological effects of multiple
green infrastructure combinations at different scales, varying from lots, to neighborhoods, to
watersheds. A review paper compared ten models for low impact urban storm water drainage
based on uses, temporal and spatial resolution and scale, catchment representation, runoff
generation, flow routing, contaminant treatment, green infrastructure practices, and user
interfaces (Elliott and Trowsdale, 2007). Numerous studies have used these models to evaluate
the performance of green infrastructure practices in various areas (Ahiablame et al., 2012b;
Ahiablame et al., 2013; Brander et al., 2004; Damodaram et al., 2010; Gilroy and McCuen, 2009;
Jia et al., 2012; Kronaveter et al., 2001; Lee et al., 2012; Loucks et al., 2004; Williams and Wise,
2006). A discussion of the most widely used models and a subset of studies follows.
The Long-Term Hydrologic Impact Assessment-Low Impact Development (L-THIA-
LID) model uses the Soil Conservation Service Curve Number (SCS-CN) method to simulate
runoff and infiltration behavior of LID practices (Ahiablame et al., 2012b; Ahiablame et al.,
2013). LID features are considered within a sub-catchment using the concept of Hydrologic
Response Units (HRUs) that use an area-weighted average of soil and land cover characteristics
to compute a CN. A lot-level LID screening tool is available for one lot at a time but it cannot
simultaneously describe and simulate multiple lots of different types (e.g. residential next to
commercial) within an urban area. It does not treat runoff-runon from one LID feature into
another. L-THIA GIS (ver. 2013) does provide some spatial functionality by providing
watershed delineation, computing an area-weighted CN for a sub-catchment, and multi-gauge
precipitation data. The GIS generated information can then be imported into the L-THIA
spreadsheet calculator.
Ahiablame et al. (2012b) utilized L-THIA-LID to demonstrate a computational procedure
to assess theoretical impacts of urban developments on pre-development hydrology and analyze
the possible impacts of six LID practices (bioretention, rain barrels and cisterns, green roofs,
open wooded space, porous pavement, and permeable patios) in a residential subdivision in
Lafayette, Indiana. The authors recommend using this procedure as a quick assessment and
screening tool, before proceeding to site-specific data to calibrate and validate the model.
Ahiablame et al. (2013) calibrated and validated the L-THIA-LID model for two urbanized
watersheds near Indianapolis, Indiana. This calibrated model was used to simulate six different
scenarios for retrofitting the urban watershed with rain barrels/cisterns and porous pavements.
Both these studies were successful in the theoretic evaluation of LID practices. However, for
practical applications, the study recommends field studies and detailed calibration and validation.
SWMM5 is a dynamic rainfall-runoff simulation model used for hydrologic and
hydraulic modeling of urban areas using an object oriented framework. It is a relatively
-------
comprehensive model with an atmospheric compartment to track dry deposition, and land
surface, groundwater and transport compartments. It treats snow and snowmelt but its strength is
in hydraulic modeling of the constructed environment. The transport compartment represents a
"network of conveyance elements (channels, pipes, pumps, and regulators) and storage/treatment
units that transport water to outfalls or to treatment facilities" using node and link objects (U.S.
EPA, 2010, p. 33). Within a conduit link, steady, kinematic and dynamic routing options are
available. The components of the transport compartment are modeled with node and link objects.
SWMM5 has the capability to model the capture and retention of rainfall/runoff using various
types of LID practices (Damodaram et al., 2010; Jia et al., 2012).
The non-proprietary version of SWMM5 does not have a GIS interface and attributes of
contributing area objects must be manually entered. The user is responsible for dividing the
overall catchment into sub-catchments, and for identifying their outlet points. The contributing
areas outside the urban environment are modeled in a relatively simple manner as sub-
catchments represented by a non-linear reservoir. Within these sub-catchments, LID's cannot be
implemented at the lot-level of a proposed subdivision and cannot represent LID features in
series with runoff out of one LID flowing into another. Numerous applications of SWMM for
LID assessments have been made. As an example, Aad et al. (2010) modeled rain barrels and
rain gardens using the sub-catchment object in SWMM5 with the Green-Ampt infiltration
equation. The study claimed to be successful in theoretically representing the rain barrels and
rain gardens using SWMM5, but did not provide any comparison to observed data.
The EPA developed SUSTAIN as a decision-support system for the selection and
placement of LID BMPs in urban watersheds (Lee et al., 2012). SUSTAIN is more focused on
the effectiveness of these BMPs in terms of costs and efficiencies. SUSTAIN aggregates
distributed BMPs and focuses on the overall impact of the BMPs at the subdivision/watershed
scale. Like SWMM5 it cannot implement LID BMPs at the lot level. SUSTAIN contains an
optimization module that develops cost-effective BMP placement and selection strategies based
on input parameters such as potential sites, applicable BMP types, and size ranges. This module
performs numerous searches for the optimal combination of BMPs that meet user-defined
decision criteria. SUSTAIN calculates hydrological processes using components that are derived
from version 5 of the Storm Water Management Model (SWMM5). SUSTAIN also provides a
GIS framework to design and place LID BMPs. However, the non-proprietary version of
SUSTAIN is no longer supported and has not been updated to ArcGIS 10.x.
The freely available versions of both SWMM and SUSTAIN do not automatically
estimate model infiltration and hydraulic roughness parameters from commonly available GIS
data layers for soils and land cover. Nor do any of the models discussed above have the
capability to spatially display simulation results and readily difference non-LID vs LID
simulations to aid in targeting LID features.
-------
3.2 Green Infrastructure Practices
A. Bio retention
Bioretention systems are depressions filled with highly permeable soil, and planted with
vegetation. These systems allow storm water to pond and infiltrate, thereby supporting
vegetation growth while achieving storm water retention, pollutant removal, and groundwater
recharge. Smaller scale bioretention systems are also referred to as rain gardens and their design
and effectiveness are more dependent on lot sizes and placement within the watershed.
Bioretention system design is highly dependent on the soil type, site conditions, and land use. A
typical bioretention system would include a sand/soil/organic media for the treatment of
infiltrating runoff, a surface mulch layer, native vegetation, and a depression to allow storm
water ponding. Davis et al. (2009) and Roy-Poirier et al. (2010) review bioretention technology
to address existing design considerations, hydrologic and water quality performance, modeling
efforts, and research needs.
Bioretention systems have been shown to be highly efficient in reducing peak flows,
detaining storm water runoff, increasing infiltration and groundwater recharge in more humid
regions, increasing evapotranspiration, and sustaining native vegetation (Aad et al., 2010; Davis,
2008; DeBusk and Wynn, 2011; Dietz and Clausen, 2005; Dussaillant et al.; 2004; Emerson and
Traver, 2008; Hatt et al., 2009; Heasom et al., 2006; Hoskins and Peterein, 2013; Hunt et al.,
2006; Hunt et al., 2008; James and Dymond, 2012; Jenkins, et al., 2010; Li et al., 2009;
Olszewski and Davis 2013;Sharkey, 2006).
Bioretention basins are also effective for pollutant removal from storm water runoff, by
taking advantage of the chemical, biological, and physical properties of plants, soils, and
microbes (Brown and Hunt, 2011; County, 2002; Davis et al., 2006; Davis, 2007; Davis et al.,
2001; Davis et al. 2003; DeBusk and Wynn, 2011; Dietz and Clausen, 2005; Hatt et al., 2009;
Hsieh and Davis, 2005; Hunt et al., 2006; Hunt et al., 2008; Sharkey, 2006).
B. Permeable Pavements
Permeable pavements are paved surfaces that reduce runoff by allowing infiltration.
These are usually designed as a matrix of concrete paver blocks with voids filled with sand,
gravel, or soil. These voids encourage infiltration of storm water into the underlying soil layer.
Two review papers address the various design considerations for permeable pavements and their
performance in terms of water quality and hydrology (Pratt, 1995; Scholz and Grabowiecki,
2007).
Numerous studies evaluating permeable pavements have illustrated their effectiveness in
increasing infiltration, reducing peak flows and surface runoff volumes, and increasing
groundwater recharge in humid regions (Andersen et al., 1999; Bean et al., 2007a; Bean et al.,
2007b; Booth and Leavitt, 1999; Brattebo and Booth, 2003; Collins et al., 2008; Dreelin et al.,
2006; Gilbert and Clausen, 2006; Fagotto et al., 2000; Rushton, 2001; Schluter and Jefferies,
2002).
Permeable pavements have also been successfully installed to help reduce the
concentrations of storm water pollutants, including heavy metals, nutrients, and hydrocarbons
-------
(Bean et al., 2007b; Booth and Leavitt, 1999; Brattebo and Booth, 2003; Dreelin et al., 2006;
Gilbert and Clausen, 2006; Legret and Colandini, 1999; Fagotto et al., 2000; Rushton, 2001).
C. Rainwater Harvesting
Rainwater harvesting includes the use of rain barrels and cisterns to retain rooftop runoff
for future use. Rain barrels tend to have a storage capacity of less than 0.38 cubic meters (100
gallons) and are usually placed above the ground. Cisterns have a capacity of more than 0.38
cubic meters and can be self-contained, above-ground, or below-ground systems. Rainwater
harvesting system designs are based on the average annual precipitation, drainage area or roof
area, runoff coefficient, expected storage requirements, and expected water use out of the
cisterns (French, 1988; Jones and Hunt, 2010; Shuster and Rhea, 2013; Thurston et al.,
2008). Various research papers have used analytical models to aid in the sizing of rainwater
harvesting systems based on the above parameters (Abdulla and Al-Shareef, 2009; Ghisi et al.,
2007; Guo and Baetz, 2007; Jennings et al., 2013; Sample et al., 2012; Ward et al., 2010; Ward
etal., 2012; Waterfall 2004).
Various studies have analyzed the use of rain barrels to reduce storm water runoff (Aad et
al. 2010; Boers and Ben-Asher, 1982; Fewkes, 2000; French, 1988; Ghisi et al., 2007; Guo and
Baetz, 2007; Jennings et al., 2013; Jones and Hunt, 2010; Kim and Yoo, 2009; Sands and
Chapman, 2003;Shuster et al., 2008; Shuster and Rhea, 2013; Thurston et al., 2008; Trieu et al.,
2001; Ward et al., 2010). The results from these studies indicate a decrease in the runoff with the
implementation of rain barrels/cisterns. However, to have a significant impact on storm water
management, the studies recommend the implementation of a large number of rain
barrels/cisterns to obtain a cumulative effect on the capture of roof runoff.
-------
4.0 Objectives and Scope
The scope of this study is to develop publically available prototype software within a GIS
environment. The purpose is to address some of the limitations in the models reviewed above
with emphasis on application to arid and semiarid environments. The prototype will be built
within the Automated Geospatial Watershed Assessment (AGWA; Miller et al., 2007) tool using
the KINEROS2 rainfall-runoff-erosion watershed model (Goodrich et al., 2012). AGWA can
automatically delineate and describe sub-catchments that flow into or lie downstream of
proposed areas for development. Using nationally available topography, soils, and land cover
spatial data layers it can automatically derive initial model parameter estimates for KINEROS2.
KINEROS2 has advanced physically-based infiltration algorithms, runoff-runon routing
capability, and geometric flexibility so that GI features can be realistically represented, and their
hydrologic response behavior can be simulated at the lot-level. The following objectives guided
development and testing of the AGWA GI prototype software.
1. Develop software that can be used to represent a limited set of GI features within the
AGWA/KINEROS2 ArcGIS modeling environment.
2. Enable GI representation, parameterization, and simulation at the lot-level with runoff-
runon capability for multiple GI features within a lot.
3. Model subdivision level GI implementation and response across multiple lots and streets.
4. Verify the hydrologic behavior of GI features at the lot level.
5. Verify the subdivision representation of lots with observed rainfall-runoff observations.
The scope of the project will attempt to develop a tool that strikes a balance between the
L-THIA-LID and SUSTAIN/SWMM models. The AGWA GI tool, like L-THIA-LID, can be
used for rapid screening assessments to evaluate subdivision level GI practices, but do so in a
more complex fashion by representing a hypothetical configuration of lots and streets. In many
arid and semi-arid developments, subsurface storm drainage features are not used due to
infrequent rainfall. Therefore this project did not go to the level of complexity to incorporate
subsurface storm drainage features that are comprehensively treated in SUSTAIN/SWMM.
Incorporation of economics and optimization on the number and placement of GI features, also
treated by SUSTAIN/SWMM is beyond the scope of this project but could be incorporated in the
future.
Only a small number of GI design features, those reviewed above, have been
incorporated into the prototype. However, it should be stressed that the functionality of
KINEROS2 already is capable of simulating infiltrating channels and grass or vegetated swales.
In addition, the two-layer infiltration algorithms in KINEROS2 can simulate the effects of soil
compaction as part of subdivision site preparation or importation of off-site soils for fill material
(Smith et al., 1995). However, it should be stressed that, local post-construction infiltration
measurements should be made for realistic simulations of infiltration.
We tested the implementation of the AGWA GI tool at two scales. At the lot scale
(Objective 4) a typical lot configuration will be constructed with the AGWA GI tools and
-------
subjected to a constant rainfall intensity for a sufficient duration so that an equilibrium or steady
state runoff rate is achieved. In the second case, a constant intensity storm of known total rainfall
depth is applied to the lot with GI features. The storm depth selected (12.5 mm) corresponds to
relatively small event (less than a 1-year return period, 1-hour design storm in southern Arizona
where the test was conducted). The small storm was selected so that the impact of the LID/GI
features would be apparent and not overwhelmed by a large event. Knowing the constant rainfall
intensity and the detention and infiltration properties of the GI and constructed features on the lot
it is possible to independently compute the steady state runoff rate from the lot and ensure that
routing and runoff-runon routing is correctly represented in the code. The overall water balance
of the various components (rainfall, runoff, infiltration, etc.) will also be computed to further
verify that the model is functioning properly.
At the subdivision scale we will test the integrity of the code for multiple lots and for lot-
to-street-to-outlet connectivity and routing capability on an actual watershed in southeastern
Arizona in the City of Sierra Vista (Objective 5). A subdivision within this watershed was
selected for this study because high-quality rainfall-runoff observations are available as well as
detailed watershed characterization data. This watershed, consisting of a natural undeveloped
sub-catchment that drains into the La Terraza subdivision, and associated observations are
discussed in more detail in Kennedy et al. (2013) and Kennedy (2007). A map of the study area,
measurements and instrumentation locations is contained in Figure 1. A primary objective of the
Kennedy et al. (2013) study was validation of the KINEROS2 urban model element discussed in
more detail below. Unfortunately, the La Terraza subdivision does not contain any GI features.
Ideally high-quality rainfall-runoff observations in an arid or semiarid subdivision with a variety
of GI features would be available for more thorough verification of the AGWA GI tool.
EXPLANATION
^ USGS stream gauge
O USDA-ARS rain gauge
o Tension infiltrometer
measurement locations
Watershed boundary
Flowpath direction
Underground culvert
FLAGSTAFF
•
PHOENIX
•
TUCSON
SIERRA'VISTA +
Figure 1: Study Area Map Showing Gauge Locations, Infiltration Measurement Locations, and Watershed Boundaries; Area in
Upper Right of Urban Watershed Drains Directly to Watershed Outlet Through an Underground Culvert; Runoff from
Remaining Area is Routed Along Streets (Background Image Courtesy USGS Earth Resources Observation and
Science Center)(from Kennedy et al., 2013).
10
-------
The intent is that testing at the two scales will provide some level of assurance that the
AGWA GI tool can be used to provide realistic assessments of the effects of GI features at the
lot, subdivision, and small watershed scales. In addition, when site specific data is lacking, the
AGWA GI tool can be confidently used for the relative change assessments discussed in
Goodrich et al. (2012) if it passes the two levels of testing. As noted above a thorough validation
can only be achieved with a more comprehensive set of observations for the site being assessed.
As noted above the GI simulation prototype will be developed to operate in the AGWA
ArcGIS environment and build on many of the core hydrologic and routing capabilities of
KINEROS2. A brief overview of AGWA and KINEROS2 is provided in the following section.
11
-------
12
-------
5.0 AGWA and KINEROS2
The AGWA tool provides a GIS user interface for two hydrologic models - the Kinematic
Runoff and Erosion model (KINEROS2) and the Soil and Water Assessment Tool (SWAT)
(Daniel et al., 2011, Miller et al., 2007). AGWA is a customized toolbar in ArcMap that uses
existing spatial datasets in the form of digital elevation models, land cover maps, soil maps, and
weather data as inputs (Figure 2). These inputs are processed to prepare input parameters for
hydrologic models. The simulation results are quantified and imported back into AGWA for
spatial display and analysis. The interoperability of KINEROS2 and AGWA is described in
Goodrich etal. (2012).
KINEROS2 is a distributed, physically based model which simulates runoff and erosion for
small watersheds. It utilizes kinematic equations to simulate overland flow over rectangular
planar or curvilinear hillslopes and channelized flows through open trapezoidal channels (Figure
3) (Woolhiser, et al., 1990; Goodrich et al., 2012). In addition to the standard overland flow
(planar or curvilinear) and channel modeling elements, KINEROS2 also has an "Urban"
modeling element (Figure 4) which consists of up to six overland flow areas that contribute to
one-half of a paved, crowned street with the following configurations: (1) directly connected
pervious area, (2) directly connected impervious area, (3) indirectly connected impervious area,
(4) indirectly connected pervious area, (5) connecting pervious area, and (6) connecting
impervious area. The "Urban" modeling element represents an abstraction of a typical
subdivision. The La Terraza study illustrated in Figure 1 evaluated the urban element and results
from this study were successful in indicating that KINEROS2 could be used to model urban
residential watersheds with this abstract representation of different surface types and runoff-
runon combinations (Kennedy et al., 2013).
Various case studies that have been reviewed have proven that there are a number of
useful modeling approaches and tools to evaluate how green infrastructure systems will affect
runoff responses. However, very few packages exist that can provide a decision-support system
with spatial, robust, and accurate modeling capabilities. Popular models from these case studies
lack the physical routing of water through the watershed, running continuous simulations,
provisions for erosion modeling, or the use of a spatial tool. The robustness of KINEROS2 and
the GIS interface provided by AGWA creates the option to use these in unison to provide a
powerful modeling platform for GI practices in urban development scenarios.
13
-------
14
-------
Rainfall
Simulation
Results
Channel Flow
Figure 3: KINEROS2 Elements.
Directly
Connected
Pervious
(DCP)
Indirectly
Connected
Impervious
(ICI)
Connecting
Pervious
(CP)
Indirectly
Connected
Pervious
(ICP)
Connecting
Impervious
(Cl)
Directly
Connected
Impervious
pei)
Retention Basin (RBI
Non
Contributing
Area
(NC)
Street (half)
Figure 2: AGWA Workflow.
Figure 4: KINEROS2 Urban Element Components.
15
-------
16
-------
6.0 Design and Development
Based on the study objectives, and the existing AGWA functionality, a modified
workflow was designed to utilize KINEROS2 to simulate urban environments and GI practices.
Limitations of the KINEROS2 model are discussed in Goodrich et al. (2012). The modified
workflow was developed in the .NET Framework using Microsoft Visual Studio 2010. C# and
VB.NET were the programming languages used. Environmental Systems Research Institute
(ESRI) provides an ArcObjects software development kit for the .NET Framework to build
Windows applications with GIS functionalities. With the help of ArcObjects, windows based
forms were developed which could use existing GIS functionalities in ArcMap. The description
for each step in the workflow is given below.
Setup Urban Geodatabase
The Setup Urban Geodatabase form allows the user to provide a location and a name for
the geodatabase, and a "discretization" name. The geodatabase becomes the workspace for
feature classes and tables that are created in subsequent processes and the discretization name is
used to identify them. The user also provides the subdivision parcels and a corresponding road
layer in the form of polygon feature classes. It is necessary that the parcel feature class attribute
table has a column that defines the width of each parcel adjoining the street.
Once the inputs are supplied, the program creates the geodatabase in the workspace
location and copies the two feature classes into it. An "ElementID" column is added to the new
parcel feature class that uniquely identifies each parcel.
Flow Routing
Flow routing is an important step in simulating an urban subdivision as post construction
flow paths are typically different from pre-development topography. KINEROS2 requires the
path that water will follow from the lot to the basin outlet. The Urban element in KINEROS2
assumes all of the rainfall to flow from the lot towards the street. The street is assumed to be
crowned to allow the routing of water along the streets. The flow routing step accepts a routing
name from the user which uniquely identifies the route. The user then draws flow paths on the
parcel feature class using the built-in drawing tools in ArcMap. Once saved, the flow paths are
checked by the software to ensure that all parcels are associated with a flow path, and that they
fall within the boundaries of the parcels. Using these flow paths, the "FROM" and "TO" parcels
are extracted to create a conceptual flow map draining towards the outlet. The program also
generates route identifiers, which will identify the flow when writing the parameter file for
KINEROS2. The flow route is stored in the "FlowRouting_" table (Table 1) in
the urban geodatabase.
17
-------
Table 1: Description of Fields in the Flow Routing Table.
Column
ROUTEJD
FROM_PARCEL
TO_PARCEL
FLOWJD
Description
Identifies Every Line Drawn by the User.
Identifies the Parcel from where the Flow is Initiated.
Identifies the Parcel where the Flow Ends.
Identifies the Sequence of the Flow in Decreasing Order, from Origin to Parcel
Before the Outlet.
Parameterization
The Parameterization step defines KINEROS2 input parameters based on geometry, land
cover, and soils properties for each parcel. Existing AGWA functionality was modified to
parameterize parcel elements, land cover, and soils for subdivisions. The user provides a unique
name for the parameterization and provides inputs to the Element Parameterization form and the
Land Cover and Soils form. The first form defines element parameters, including the parcel
width field from the parcel feature class, house area, driveway area, slope, street width, cross
slope, and grade, all of which can also be defined using fields from the feature classes or with
user defined values. The second form defines land cover and soils parameters including canopy
cover fractions, impervious fractions, pervious fractions, street roughness, and impervious and
pervious interception values. A Soil Survey Geographic Data Base (SSURGO) soil map is
required along with the corresponding database to prepare soil parameters. User-entered values
in both forms are applied to all the parcels in the feature class uniformly.
The user can edit these values outside of AGWA, by editing the parameterization table in
the corresponding geodatabase. The parameters are stored in a table (Table 2) with the name
"_u_parLUT" in the urban geodatabase.
Table 2: Description of Fields in the Parameterization Table.
Column
X
Y
LOT_AREA
LENGTH
WIDTH
SLOPE
HOUSE_AREA
Description
X Coordinate of the Parcel Centroid
Y Coordinate of the Parcel Centroid
Area of the Parcel Generated from Feature Class Geometry
Overland Flow Length of the Parcel Perpendicular to the Street
Calculated Using "WIDTH" and "LOT_AREA"
Width of the Parcel Parallel to the Street (Street Length)
Overland Slope of the Parcel, Perpendicular to the Street
Area of the House on the Parcel Corresponding to the Indirectly
Connected Impervious Area
Units
Square Meters
Meters
Meters
Percent
Square Meters
18
-------
Column
DWAY_AREA
DWAY_LENGTH
DWAY_WIDTH
LANE_WIDTH
CROSS_SLOPE
GRADE
IMP_N
PERV_N
STREET_N
IMPJNT
PERVJNT
CANOPY
CV
KSAT
G
DIST
FOR
ROCK
Description
Area of the Driveway on the Parcel Corresponding to the Directly
Connected Impervious Area
Length of the Driveway Perpendicular to the Street
Width of the Driveway Parallel to the Street
Width of the Street
Lateral Street Slope from Gutter to Street Centerline
Slope of the Street Corresponding to the Flow Direction
Manning's Roughness Coefficient for Impervious Surfaces
Manning's Roughness Coefficient for Pervious Surfaces
Manning's Roughness Coefficient for Street Surfaces
Interception Depth for Impervious Surfaces
Interception Depth for Pervious Surfaces
Canopy Cover Fraction of Surface Covered by Intercepting Canopy
Coefficient of Variation of Saturated Hydraulic Conductivity
Saturated Hydraulic Conductivity
Mean Capillary Drive
Pore Size Distribution Index
Porosity
Volumetric Rock Fraction
Units
Square Meters
Meters
Meters
Meters
Percent
Percent
mm
mm
mm
Green Infrastructure Design and Placement
The Green Infrastructure Design and Placement tool allows users to design and place
retention basins, permeable pavements or rainwater harvesting systems on one or more parcels in
a subdivision.
Retention Basins: A retention basin design requires the width, length and depth (in feet)
in order to calculate the area and volume associated with the retention basin. In addition to the
above dimensions, KINEROS2 requires the hydraulic conductivity of the retention basin in
inches/hour or mm/hour.
Permeable Pavements: Design parameters for permeable pavements can be provided in
the form of length and width or selecting the "Same as driveway area" option. With the "Same as
driveway area" option, AGWA calculates the permeable pavement area from the driveway area
19
-------
in the parcel feature class. A hydraulic conductivity value in inches/hour or mm/hr is also
required.
Rainwater Harvesting: For the design of a rainwater harvesting system, the volume of the
rain barrel can be provided in gallons or cubic meters or can be calculated using height and
diameter (both, in feet or meters) of the rain barrel.
Each of these designs can be saved in the Geodatabase with a unique name. The designs
are saved in the Retention Basin Designs (Table 3), Permeable Pavement Designs (Table 4), and
Rainwater Harvesting Designs (Table 5) tables. The user provides a unique name for the
placement plan. The user also selects a design and applies it to one or more parcels using the
selection tool in ArcMap. Each placement plan is then saved in the
"PlacementPlans_" table (Table 6) in the urban geodatabase.
Table 3: Description of Fields in the Retention Basin Designs Table.
Column
RBJJNITS
RB_NAME
RB_LENGTH
RB_WIDTH
RB_DEPTH
RB_HYDCON
Description
Units of the Design Parameters: ENGLISH/METRIC
Retention Basin Design Name
Length of the Retention Basin Parallel to the Road
Width of the Retention Basin Perpendicular to the Road
Depth of the Retention Basin
Hydraulic Conductivity of the Retention Basin
Units
Feet or Meters
Feet or Meters
Feet or Meters
in/hr or mm/hr
Table 4: Description of Fields in the Permeable Pavement Designs Table.
Column
PPJJNITS
PP_NAME
PP_LENGTH
PP_WIDTH
PP_SAMEASDRIVEWAY
PP_AREA
PP_HYDCON
Description
Units of the Design Parameters: ENGLISH/METRIC
Permeable Pavement Design Name
Length of the Permeable Pavement. -99 Indicates
"Same as Driveway Area"
Width of the Permeable Pavement. -99 Indicates
"Same as Driveway Area"
YES/NO. Indicates if the Permeable Pavement Area is
calculated using the driveway area
Area of the Permeable Pavement. -99 indicates "Same
as Driveway Area"
Hydraulic Conductivity of the Permeable Pavement
Units
Feet or Meters
Feet or Meters
Square Feet or
Square Meters
in/hr or mm/hr
20
-------
Table 5: Description of Fields in the Rainwater Harvesting Design Table.
Column
RHJJNITS
RH_NAME
RHJVOLUME
RH_VOLUME_GAL
RH_DIAMETER
RH_HEIGHT
RH_UTILIZATION
Description
Units of the Design Parameters: ENGLISH/METRIC
Rainwater Harvesting Design Name
Volume of the Rain Barrel
Volume of the Rain Barrel in US Gallons
Diameter of the Rain Barrel. -99 Indicates Volume Provided by User
Height of the Rain Barrel. - 99 Indicates Volume Provided by User
Percentage Daily Utilization of Rainwater
Units
Cubic Feet or
Cubic Meters
Gallons
Feet or Meters
Feet or Meters
Percent
Table 6: Description of Fields in the Placement Plans Table.
Column
ElementID
RB_NAME
PP_NAME
RH_NAME
Description
Uniquely Identifies the Parcel in the Parcel Feature Class
Retention Basin Design Name Applied to the Corresponding Parcel
Permeable Pavement Design Name Applied to the Corresponding Parcel
Rainwater Harvesting Design Name Applied to the Corresponding Parcel
Precipitation
KINEROS2 accepts rainfall data in the form of time-intensity pairs or time-accumulated
depth pairs. AGWA allows the user to provide rainfall data in the form of precipitation frequency
grids, design storm tables, user-defined depths, or user-defined hyetographs. Rainfall is assumed
to be applied uniformly over the entire subdivision area. The user specifies a unique name for the
precipitation file and is stored as ".pre" This functionality remains
unchanged from the original AGWA implementation. More information can be found in the
AGWA Documentation on the AGWA website (^ww.tucson.ars^ag.gov/agwa/ or
Write Input Files
In the "Write Input Files" step, AGWA aggregates all the inputs that were provided in the
preceding steps and prepares files required by the KINEROS2 model. The following files are
required by the KINEROS2 model.
.par: This file is written based on the inputs provided and the parameters generated in the
flow routing, parameterization and BMP placement steps. Parameters are written in "urban"
blocks for each parcel using the sequence generated in the flow routing step.
21
-------
.pre: The precipitation file, which was created in the Precipitation step, is copied from the
precipitation directory into the simulations directory.
lid.fil: This file stores the initial volumes for the retention basin and rainwater harvesting
systems respectively. AGWA assumes zero starting volumes for both.
kin.fil: KINEROS2 control file which directs the model with the input parameter
filename, input precipitation filename, output filename, and duration of the simulation.
file.bat: this is the batch file that executes the KINEROS2 model.
k21id.exe: The KINEROS2 model executable which is copied from the AGWA directory.
The user selects the flow routing table, parameterization, placement plan table,
precipitation file, and provides a unique name for the simulation. AGWA creates a table (Table
7) "Simulations_" in the geodatabase, which stores this information.
AGWA creates a new directory in the workspace with the given simulation name. The lid.fil,
kin.fil, file.bat, precipitation file and ".par" are written and saved in this
directory.
Table 7: Description of Fields in the Simulations Table.
Column
SIMULATION
ROUTING_TABLE
PARAMETERIZATION
PLACEMENT_PLAN
PRECIPITATION
Description
Name of the Simulation
Name of the Routing Table
Name of the Parameterization Stored in the Parameterization Look up Table
Name of the Green Infrastructure Placement Plan Table
Name and Location of the Precipitation File
Execute KINEROS2 Model
In the "Execute KINEROS2 Model" step, the user selects the discretization and an
associated simulation and runs the KINEROS2 model. AGWA executes the file.bat created in
the "Write Input Files" step. A command prompt displays the progress of the simulation and
whether it was successful or if it encountered any errors. The output file (.out), which
summarizes the hydrology for each urban element, is created in the simulations directory by the
model. AGWA imports these values in the next step.
View Results
The "View Results" form allows the user to visualize the results of the KINEROS2
simulation. The user can import results from previously run simulations into AGWA. This step
creates a results table from the simulation output, and joins it to the parcel attribute table. AGWA
allows the user to visualize the output for each parcel in the form of infiltration, runoff and
accumulated runoff volumes. AGWA also allows the user to visualize the absolute/percent
difference between two simulations. Infiltration and runoff volumes results are visualized for
22
-------
each individual parcel. Accumulated runoff, which is comprised of the runoff from each parcel
along with the runoff from the upland parcel, can be visualized along the street.
23
-------
24
-------
7.0 Testing
7.1 Lot Scale
Verification of the Urban element at the lot scale was approached by confirming the
following 1) that event volumes of hydrologic components are balanced properly; and, 2) that the
steady-state runoff rates are as expected. To test both, an element was created representing a
typical lot in the La Terraza subdivision (Figure 5) that was used in six scenarios (Table 8): pre-
development; post-development w/o GI; retention basins; permeable pavements; rainwater
harvesting; and all GI practices. The GI practices for the verification exercise were setup to
illustrate the impact of the practices. Each scenario was simulated using rainfall applied at a
constant intensity of 12.5 mm/hr for 120 minutes. The rainfall intensity and duration were
selected so that the element reached steady-state outflow rates.
Figure 5: Lot Scale Representation of the KINEROS2 Urban Element.
25
-------
Table 8: Description of the Lot Scale Verification Scenarios.
Pre-Development
Post-Development (w/o GI)
Retention Basin
Permeable Pavement
Rainwater Harvesting
All GI Practices
Empty Lot with a Street and Soils Attributes Obtained from the NRCS
SSURGO Soils Spatial Database
Lot with a House Area of 2500 Square Feet and a -12 by 19.5 Foot
Impermeable Driveway (21.76 Square Meters)
Post-Development Parameters with the Addition of a Retention Basin with a
Hydraulic Conductivity of 8.3 in/hr (-210 mm/hr) and Sized with a Surface
Area of Approximately 72 Square Feet and a Depth of ~10 Inches, Yielding a
Retention Capacity of ~60 Cubic Feet (-444 Gallons or 1.68 m3)
Post-Development Parameters with the Conversion of the Driveway to
Permeable Pavement with a Hydraulic Conductivity of 8.3 in/hr (-210 mm/hr)
Post-Development Parameters with a Rainwater Harvesting Feature with a
Capacity of -500 Gallons (1.9 m3)
Post-Development Parameters Along with all of the Above GI Practices
Verifying the water balance is a basic accounting exercise that ensures model inputs
equal model outputs plus any change in storage. In this exercise, a 12.5 mm/hr rainfall event was
applied for a two-hour duration onto a lot size of 0.1933 hectares, yielding a total rainfall of
96.66 m3; this is the model input. Model outputs include interception, infiltration, storage, and
outflow in m3. Table 9 includes a summary of the inputs and outputs for six different
development scenarios. The error term represents the percent difference between the sum of the
inputs and sum of the outputs and storage. For all scenarios, the error is less than 1%.
Table 9: Volume Balances of the Lot Scale Verification Scenarios.
Rain (m3)
Interception (m3)
Infiltration (m3)
Stored (m3)
Outflow (m3)
Error (%)
Pre-
Development
48.33
0
28.76
0
19.73
-0.33
Post-
Development
(w/o GI)
48.33
0
24.90
0.00006
23.76
-0.68
Rainwater
Harvesting
48.33
1.9
24.82
0.00006
21.90
-0.60
Retention
Basin
48.33
0
29.28
0.00006
19.36
-0.64
Permeable
Pavement
48.33
0
25.45
0.00003
23.22
-0.70
A11GI
Practices
48.33
1.9
29.72
0.00003
16.98
-0.56
26
-------
Effective hydraulic conductivity is defined as the rainfall rate minus outflow rate.
Because each scenario reached steady-state outflow rates, the effective hydraulic conductivity
could be compared to the expected steady-state weighted saturated hydraulic conductivity
calculated from the different overland flow areas of the Urban element. The weighted hydraulic
conductivity is calculated by converting the infiltration capacity in mm/hr to m3 for each of the
overland flow areas of the Urban element. Conversion to a volumetric rate is necessary so that
contributing volumes can be subtracted out when overland flow areas that receive input from
upslope (e.g. the retention basin - Figure 4) have higher infiltration capacities than the rainfall
rate.
Table 10: Effective Versus Steady State, Weighted Hydraulic Conductivities for the Lot Scale Verification
Scenarios.
Effective Ks
(mm/hr)
Steady state
weighted Ks
(mm/hr)
Difference
Pre-
Development
7.05
7.01
-0.57%
Post-
Development
(w/o GI)
5.98
5.94
-0.67%
Rainwater
Harvesting
5.98
5.94
-0.67%
Retention
Basin
6.33
6.29
-0.64%
Permeable
Pavement
6.26
6.22
-0.64%
A11GI
Practices
6.61
6.57
-0.61%
27
-------
Lot-scale Outflows
-
Rainfall rate = 12.5 mm/hr
Effective hydraulic conductivity = rainfall rate-outflow rate
7.02 mm/hr - 12.5mm/hr - 5.48 mm/hr
5 95 mm/hr = 12.5mm/hr - 6.55 mm/hr
5.95 mm/hr = 12.5mm/hr - 6.55 mm/hr
6.65 mm/hr = 12.5mm/hr - 5.85 mm/hr
6.09 mm/hr = 12.5mm/hr - 6.41 mm/hr
6.80 mm/hr = 12.5mm/hr - 5.70 mm/hr
— Pre-development
Post-development (w/o Gl)
Retention Basin
Permeable Pavement
Rainwater Harvesting
—All Gl practices
Time (min)
Figure 6: Hydrographs of Lot Scale Testing Scenarios, Illustrating Effective Hydraulic Conductivities Versus
Theoretical, Steady-State Hydraulic Conductivities.
28
-------
Lot-scale Outflows
Rainfall rate = 12.5 mm/hr
Expected steady state outflow - rainfall rate - area weighted saturated hydraulic conductivity
5.49 mm/hr - 12.5mm/hr - 7.01 mm/hi Pre-development
6.56 mm/hr = I2.5mm/hr • 5.94 mm/hr Post-development (w/o 61)
6.56 mm/hr = 12.5mm/hr - 5.94 mm/hr Retention Basin
5.86 mm/hr = 12.5mm/hr - 6.64 mm/hr Permeable Pavement
6,42 mm/hr = 12.5mm/hr • 6.08 mm/hr Rainwater Harvesting
5.72 mm/hr = 12,5mm/hr - 6.78 mm/hr All 51 practices
— — Pre-development steady state outflow
— Post-development (w/o Gl) steady state outflow
- • - - Retention basin steady state outflow
Permeable pavement steadystate outflow
Rainwater harvesting steady state outflow
All Gl steady state outflow
Figure 7: Hydrographs of Lot Scale Testing Scenarios, Comparing Modeled Outflow to Theoretical Steady-State
Outflow.
7.2 Subdivision Scale
Verification of the model for the La Terraza subdivision (Figure 1) was conducted using
observed rainfall and runoff data collected from July 2005 through September 2006 (Kennedy,
2007). Rainfall was measured by four recording rain gauges, with areal average rainfall event
totals ranging from 2 to 35 mm (events less than 2 mm were not used). Runoff both into and out
of the La Terraza subdivision was measured by v-notch weirs. Runoff events that overtopped the
outlet weir were excluded, giving a high-quality data set of 47 events.
The parameter file created by AGWA was modified to incorporate some of the
parameters used by Kennedy (2007) as well as the measured inflows from the adjacent
undeveloped watershed. The altered parameters included the interception and Manning n values,
and street slopes were reduced from 0.02 to 0.01 to better reflect the values measured by
Kennedy. Initial soil saturation values for each event were also obtained from Kennedy.
The total event runoff volumes and peak flow rates for the 47 simulated events compared
to measured values are shown in Figure 8. Both volumes and peaks yielded Nash-Sutcliffe
efficiencies (coefficients of determination) greater than 0.9, with very little tendency to over or
under predict the observed values. This test provides assurances that with high-quality rainfall-
29
-------
runoff observations, the AGWA GI tool can realistically simulate the effects of subdivision scale
development for multiple lots and streets within a larger watershed with upslope contributions
from a natural, undeveloped sub-catchment. Ideally, a development containing GI features with
high-quality rainfall-runoff observations could be located to provide real world testing of the
AGWA GI tool. Until such data becomes available, this test coupled with the successful lot level
testing described in the prior section, provides a measure of confidence in the ability of the
AGWA GI tool to simulate the selected GI features in arid and semiarid areas at the lot,
subdivision, and small catchment scale. GI features were added to lots in the La Terraza
subdivision to illustrate the capability of the AGWA GI tool to simulate the effects at the
subdivision case study level.
Volume (mm)
PaŁk(mm/hr)
Figure 8: Simulated Versus Observed Event Runoff and Peak Flows (n = 47) for July 2005 through September
2006 for the La Terraza Subdivision.
30
-------
8.0 Case Study
As discussed in the scope and objectives section, the La Terraza subdivision in Sierra
Vista, Arizona (Figure 1) was used to demonstrate and test the AGWA GI tool. Sierra Vista is
part of Cochise County and thus, the parcel feature class for this subdivision was obtained from
the Cochise County IT department (http://cochise.az.gov/cochise_gis.aspx?id=6688). Sixty-six
parcels were extracted to create the "terraza_parcels" feature class. A polygon layer, comprising
the streets for La Terraza, was also extracted to create the "terraza_streets" feature class.
For the flow routing phase, the study mimicked the flow of water from Kennedy et al. (2013).
For representation purposes, we draw the flow paths on the parcels. However, KINEROS2
interprets these flow paths as flow that exits the parcel and follows the direction along the road
towards the outlet. In our case, parcels ID 28, 39, and 64 act as terminal parcels before the flow
exits the subdivision. The parameters used in the simulations are listed in Table 11. Three of the
six simulation scenarios from the lot level testing (Figure 8) were simulated for the entire
subdivision: pre-development, post-development without GI practices and post-development
with all GI practices. Each of these scenarios were run for a time period of 275 minutes using
observed rainfall from July 31, 2005 (Kennedy et al., 2013).
Table 11: Parameter Values for Pre-Development and Post-Development Simulations.
Parameter
Parcel Width
House Area
Driveway Width
Driveway Length
Overland Slope
Lane Width
Cross Slope
Grade
Manning's Roughness
Impervious Surfaces
Pervious Surfaces
Streets
Interception
Impervious Surfaces
Pervious Surfaces
Canopy Cover
Soils
PRE-DEVELOPMENT
From "Parcels_dl" Feature Class
0
0
0
2% Rise
From "Streets_dl" Feature Class
2%
1%
0.012
0.020
0.014
1 mm
2 mm
1 mm
SSURGO
POST-DEVELOPMENT
From "Parcels_dl" Feature Class
232.26 m2 (2500 ft2)
3.65m (12 ft)
6. 10m (20 ft)
2% Rise
From "Streets_dl" Feature Class
2%
1%
0.012
0.020
0.014
1 mm
2 mm
1 mm
SSURGO
31
-------
45
90
180 Meters
i i r
162.5 325
I
650 Feet
n
A
La Terraza, Sierra Vista, AZ
Flow Route
Parcels
Streets
SSURGO
Figure 9: La Terraza Subdivision and Flow Routed Towards Outlet. (Sierra Vista, AZ).
32
-------
For the pre-development simulation (Figure 10), each parcel is represented by a directly
connected pervious area draining towards the street. For post-development without GI practices
(Figure 11), the house area is assumed to be equal to the roof size, which is represented by the
indirectly connected impervious area. The driveway is represented by the directly connected
impervious area, and the remaining lot area is represented by the connecting pervious area. For
post-development with GI practices (Figure 12), a retention basin area is added right before the
street. The driveway acts as permeable pavement and the roof area contributes runoff to the
volume that is stored in the rain barrel. In Figure 10 through Figure 12, values represent percent
of lot area.
Figure 13 shows the infiltration output that the AGWA GI tool provides. In this case,
percent change in infiltration for post-development with and without GI practices are compared
to pre-development. Percent change is calculated using the following formula:
(post-development w/o GI - pre-development)/pre-development * 100
(post-development with GI - pre-development)/pre-development * 100
The results indicate an overall increase in infiltration with the addition of GI practices.
This can be observed from the higher number of lighter shaded parcels, indicating lower percent
change when compared to pre-development.
AGWA also provides a spatial view of the runoff results from the simulation (Figure 14).
In this case, percent change in runoff for post-development with and without GI practices are
compared to pre-development. Percent change is calculated using the abovementioned formula.
Without GI practices, there is an increase in runoff from each of the parcels. However,
with all of the GI practices, our results indicate a decrease in runoff when compared to pre-
development.
The lower half of the subdivision showed a visible increase in infiltration volumes
(Figure 13) and a decrease in runoff volumes (Figure 14) as compared to the upper half. This
trend can be explained by the underlying soil survey boundaries depicted in Figure 9. There are
two distinct soil types that split the subdivision into two parts contributing to the difference in
infiltration and runoff volumes. SSURGO captures some of the soil variability within the
subdivision, however some direct measurements of soil properties in post-preparation lots would
help capture more information at a finer resolution that will impact the spatial variability of
changes in hydrology. For example, compaction of the soils can also affect infiltration properties
on prepared lots (Kennedy et al., 2013). Whether direct measurements are available or not, the
AGWA Urban tool allows users to alter soil infiltration characteristics supporting multiple
simulations with a range of soil infiltration parameters which enables the exploration of relative
impacts that lot preparation may have in addition to the application of GI features. For the
purpose of this study, direct measurements of soil properties in post-preparation lots were not
taken or incorporated into the simulations.
Figure 15 shows the third output type, Accumulated Runoff (cubic meters) that the
AGWA GI tool is able to provide. Results indicate an overall reduction in accumulated runoff
with the addition of all GI practices. Parcel ID 28 shows higher flow accumulation as compared
33
-------
to parcels ID 64 and 39 because it accumulates a runoff volume from a larger number of
preceding parcels as can be seen from the flow route.
Directly Connecting Pervious
(DCP)
100%
Street (half)
Figure 10: KINEROS2 Representation of a Pre-Development Parcel.
Indirectly Connected
Impervious (ICI)
13.9%
Connecting Pervious
(CP)
84.7%
Directly
Connected
Impervious
(DCI)
1.4%
Street (half)
Figure 11: KINEROS2 Representation of Parcel ID 7 for Post-Development without GI Practices.
Indirectly Connected
Impervious
(ICI)
13.9%
Connecting Pervious
(CP)
84.4%
Directly
Connected
impervious
(DCI)
1.4%
Retention Basin (RB)
0.3%
Street (half)
Figure 12: KINEROS2 Representation of Parcel ID 7 for Post-Development with all GI Practices.
34
-------
>st-development without Gl
compared to Pre-development
-development with all Gl
compared to Pre-development
Figure 13: Percent Change in Infiltration for Post-Development with and without Gl as Compared to Pre-Development.
35
-------
•10
80
160 Meters
A
» »> Row Route
| | streets
Runoff (% Change)
| [6.45-1231
11 12-31 -1817
^B 1B 17-24 03
^B 24 03 -29 89
^H 29 89 - 35 74
40
80
160 Meters
Post-development without Gl
compared to Pre-development
A
»• >•» Row Route
| | streets
Runoff (% Change)
| M6.38--38.59
\^^\ -39.59--32 81
j^B -32.81 --26 02
|B-26 02 --19 23
^H -19.23--12 44
Post-development with all Gl
compared to Pre-development
Figure 14: Percent Change in Runoff for Post-Development with and without Gl Practices as Compared to Pre-Development.
36
-------
40 80
160 Meters
A
Post-development without
Green Infrastructure practices
**-*»• Flow Route
| | Parcels
| | streets
Accumulated Runoff (m
| | 1 -50
[ 'I 51 - 100
^H 101 - no
^B '51 - 200
^H «" • 2^0
^B 251 • 300
Post-development with all
Green Infrastructure practices
Figure 15: Comparison of Flow Accumulation for Post-Development with and without GI Practices.
37
-------
38
-------
9.0 Limitations and Issues
Limitations of the KINEROS2 model are discussed in Goodrich et al. (2012). It is an
event-based model and will not simulate plant water use, soil water movement between events,
or track snow accumulation and melt. Before simulating an event, it requires an initial estimate
of soil moisture. The event-based version precludes modeling of the changes in soil moisture due
to drainage, evaporation, and plant water use. This could have an impact when attempting to
realistically simulate how water captured by rain harvesting GI features is drawn down for
watering through different weather scenarios. Within an event, this version of KINEROS2 will
not model snowfall or frozen soil conditions, or route subsurface water flow. It can include
circular conduits in the channel network, but flow must remain below capacity (no surcharging
or pressurized flow). The representation of two-soil layer infiltration available in KINEROS2 has
not been implemented within the AGWA GI tool. Unless site specific post-development soils
and infiltration data is available this limitation is not viewed as a major shortcoming for the
AGWAGItool.
While KINEROS2 can compute infiltration and route runoff on planar or curvilinear
overland flow elements the Urban GI element is restricted to a planar surface with one slope
designated for the non-street components and another slope for the one-half street component.
The urban element assumes water flows directly to the street, and the street is assumed to be
crowned to allow independent routing of water on each side of the street. Flow from one lot will
not cross the mid-line of the street to the other half so street runoff is uniquely associated with
one lot. The Urban GI element in KINEROS2 assumes all of the runoff generated will flow from
the back of the lot towards the street. In reality, lot-generated runoff could flow out the back of
the lot or onto adjacent lots.
If high-quality post construction topographic data from lidar were available it would be
possible to further sub-divide a lot into more than one overland flow element coupled with a
Urban GI element to these cases. This is most easily done by altering the KINEROS2 parameter
file outside of the AGWA GIS environment. This limitation is not seen as a major shortcoming
as the primary application envisioned for the AGWA GI tool is for rapid relative change
assessments to evaluate the hydrologic response effects of GI features where minor flow path
deviations should not have a major effect on the overall assessment of the value of adding GI
features. Dead storage, such as a swimming pool or walled yard that effectively traps and holds
runoff, cannot currently be represented in the Urban GI element.
39
-------
40
-------
10.0 Conclusions
The AGWA GI tool was designed and developed to represent retention basins, permeable
pavements and rainwater harvesting systems within the AGWA/KINEROS2 ArcGIS modeling
environment. The "urban" element in KINEROS2 was modified to provide a realistic
representation of individual housing lots and the placement of the GI features noted above. Two
new tools were developed as part of the AGWA GI to spatially prepare parameters for the
KINEROS2 Urban GI model element. The "Flow Routing" tool allows the user to draw the flow
paths on the map, guiding storm water along platted or post-development drainage paths and to
the outlet. This is important as analysis of pre-development topography from nationally or
locally available digital elevation model (DEM) data will not typically result in flow paths
similar to the constructed development. Even in urbanized areas with high-resolution DEM data
on the scale needed to construct 0.3 m (1 foot) contour intervals, accurate flow paths can often be
difficult to discern with automated drainage analysis due to small drainage control features such
as curbs and gutters.
The "GI Design and Placement" tool allows the design and placement of retention basins,
permeable pavements, and rainwater harvesting systems at each lot in a subdivision.
Additionally, various combinations of GI placements can be designed and simulated for an entire
subdivision. The case study highlights the three output types provided by the AGWA GI tool, i.e.
infiltration, runoff, and accumulated runoff. Comparisons using these outputs can be made
between pre-development and post-development with or without GI practices.
The hydrologic behavior of GI features was tested at the lot level by verifying: 1) that
event volumes of hydrologic components were balanced properly; and 2) the steady-state runoff
rate reflected the independently computed effective hydraulic conductivity. Six scenarios (Table
8): Pre-development; Post-development without GI; Retention basins; Permeable pavements;
Rainwater harvesting; and a combination of all GI practices implemented were tested. Each of
the simulated scenarios had water balance errors that were less than 1%. The second verification
showed that simulated and expected effective hydraulic conductivity all agreed within 1%,
resulting in the expected steady-state peak runoff rates. Verification of the model at the
subdivision scale was conducted on the La Terraza subdivision using a high-quality set of
observed rainfall and runoff data. Simulated runoff volumes and peak flow rates yielded high
Nash-Sutcliffe efficiencies (>0.9) and very little bias compared to the observed data. Based on
these tests, the AGWA GI tool performed as expected.
Currently, the AGWA GI tool only allows the design and development of retention
basins, permeable pavements, and rainwater harvesting systems. However, there are many other
practices which are being considered for implementation in the AGWA GI tool. At present the
AGWA GI tool only focuses on hydrology. Some limitations mentioned related to KINEROS2
will be addressed when the continuous version is available which includes plant growth
functionality and biogeochemistry (K2-O2; Massart et al., 2010). Once integrated into AGWA,
the continuous version of KINEROS2 will enable simulation of numerous water quality effects
of GI features. Considering its current capabilities, the AGWA GI tool can be a used to inform
planning decisions related to urban development and storm water management on lot,
subdivision, and small catchment scales. This information will be useful in understanding the
expected differences in storm water runoff between neighboring developments or natural
41
-------
environments. In traditional post-development urban environments, the increase in storm water
runoff can negatively impact downstream natural resources. GI features have the potential to
mitigate those effects by achieving pre-development runoff volumes.
42
-------
11.0 References
Aad, M. A., Suidan, M., and Shuster, W. (2010). Modeling techniques of best management
practices: Rain barrels and rain gardens using EPA SWMM-5. Journal of Hydrologic
Engineering 15(6): 434-443. doi: 10.1061/(ASCE)HE. 1943-5584.0000136.
Abdulla, F. A. and Al-Shareef, A. W. (2009). Roof rainwater harvesting systems for household
water supply in Jordan. Desalination 243(1-3): 195-207. doi: 10.1016/j.desal.2008.05.013.
Ahiablame, L. M., Engel, B. A., and Chaubey, I. (2012a). 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/sl 1270-012-1189-2.
Ahiablame, L. M., Engel, B. A., and Chaubey, I. (2012b). Representation and evaluation of low
impact development practices with L-THIA-LID: An example for site planning. Environment
and Pollution 1(2), pi. doi: 10.5539/ep.vln2pl.
Ahiablame, L. M., Engel, B. A., and Chaubey, I. (2013). Effectiveness of low impact
development practices in two urbanized watersheds: Retrofitting with rain barrel/cistern and
porous pavement. Journal of Environmental Management 119:151-161. doi:
10.1016/j.jenvman.2013.01.019.
Andersen, C., Foster, I, and Pratt, C. (1999). The role of urban surfaces (permeable pavements)
in regulating drainage and evaporation: Development of a laboratory simulation experiment.
Hydrological Processes 13(4): 597-609. doi: 10.1002/(SICI)1099-
1085(199903)13:4<597::AID-HYP756>3.0.CO;2-Q.
Bean, E. Z., Hunt, W. F., and Bidelspach, D. A. (2007a). Field survey of permeable pavement
surface infiltration rates. Journal of Irrigation and Drainage Engineering 133(3): 249-255.
doi: 10.1061/(ASCE)0733-9437(2007)133:3(249).
Bean, E. Z., Hunt, W. F, and Bidelspach, D. A. (2007b). Evaluation of four permeable pavement
sites in eastern North Carolina for runoff reduction and water quality impacts. Journal of
Irrigation and Drainage Engineering: 133(6): 583-592. doi: 10.1061/(ASCE)0733-
9437(2007)133:6(583).
Bedan, E. S. and Clausen, J. C. (2009). Stormwater runoff quality and quantity from traditional
and low impact development Watersheds. Journal of the American Water Resources
Association 45(4): 998-1008. doi: 10.1111/j.l752-1688.2009.00342.x.
Bicknell, B. R., Imhoff, J. C., Kittle, J., Jobes, T. H., and Donigian, A. (2001). HSPF version 12
user's manual. National Exposure Research Laboratory. Office of Research and
Development US Environmental Protection Agency.
Bledsoe, B. P. and Watson, C. C. (2001). Effects of urbanization on channel instability. Journal
of the American Water Resources Association 37(2): 255-270. doi: 10.1111/j. 1752-
1688.2001.tb00966.x.
43
-------
Boers, T. M. and Ben-Asher, J. (1982). A review of rainwater harvesting. Agricultural Water
Management 5(2): 145-158. doi: 10.1016/0378-3774(82)90003-8.
Booth, D. B. (1991). Urbanization and the natural drainage system—impacts, solutions, and
prognoses. The Northwest Environmental Journal 7(1): 93-118.
Booth, D. B. and Jackson, C. R. (1997). Urbanization of aquatic systems: degradation thresholds,
stormwater detection, and the limits of mitigation. Journal of the American Water Resources
Association 33(5): 1077-1090. doi: 10.1111/j. 1752-1688.1997.tb04126.x.
Booth, D. B. and Leavitt, J. (1999). Field evaluation of permeable pavement systems for
improved stormwater management. Journal of the American Planning Association 65(3):
314-325. doi: 10.1080/01944369908976060.
Booth, D. B., Hartley, D., and Jackson, R. (2002). Forest cover, impervious surface area, and the
mitigation of stormwater impacts. Journal of the American Water Resources Association
38(3): 835-845. doi: 10.1111/j.l752-1688.2002.tb01000.x.
Brander, K. E., Owen, K. E., and Potter, K. W. (2004). Modeled impacts of development type on
runoff volume and infiltration performance. Journal of the American Water Resources
Association 40(4): 961-969. doi: 10.1111/j.l752-1688.2004.tb01059.x.
Brattebo, B. O. and Booth, D. B. (2003). Long-term stormwater quantity and quality
performance of permeable pavement systems. Water Research 37(18): 4369-4376. doi:
10.1016/S0043-1354(03)00410-X.
Brown, R. and Hunt, W. (2011). Underdrain configuration to enhance bioretention exfiltration to
reduce pollutant loads. Journal of Environmental Engineering 137(11): 1082-1091. doi:
10.1061/(ASCE)EE. 1943-7870.0000437.
Characklis, G. W. and Wiesner, M. R. (1997). Particles, metals, and water quality in runoff from
large urban watershed. Journal of Environmental Engineering 123(8): 753-759. doi:
10.1061/(ASCE)0733-9372(1997)123:8(753).
Cheng, M., Zhen, J. X., and Shoemaker, L. (2009). BMP decision support system for evaluating
stormwater management alternatives. Frontiers of Environmental Science and Engineering in
China 3(4): 453-463. doi: 10.1007/sll783-009-0153-x.
Clausen, J. C. (2007). Jordan Cove watershed project 2007 final report. Storrs: Department of
Natural Resource Management and Engineering, University of Connecticut.
coyj
Collins, K. A., Hunt, W. F., and Hathaway, J. M. (2008). Hydrologic comparison of four types of
permeable pavement and standard asphalt in eastern North Carolina. Journal of Hydrologic
Engineering 13(12): 1146-1157. doi: 10.1061/(ASCE)1084-0699(2008)13:12(1146).
44
-------
County, P. G. S. and June, M. D. (1999). Low-impact development design strategies: An
integrated design approach. Department of Environmental Resources, Programs and Planning
Division, Prince George's County, Maryland.
County, P. G. S (2002). Bioretention manual. Prince George's County (MD) Government,
Department of Environmental Protection. Watershed Protection Branch, Landover,
Maryland.
Damodaram, C., Giacomoni, M. H., Prakash Khedun, C., Holmes, H., Ryan, A., Saour, W., and
Zechman, E. M. (2010). Simulation of combined best management practices and low impact
development for sustainable stormwater Management. Journal of the American Water
Resources Association 46(5): 907-918. doi: 10.1111/j.l752-1688.2010.00462.x.
Daniel, E. B., Camp, J. V., LeBouef, E. F., Penrod, J. R., Dobbins, J. P., and Abkowitz, M. D.
2011. Watershed Modeling and its Applications: A State-of-the-Art Review. The Open
Hydrology Journal 5:26-50. doi: 10.2174/1874378101105010026.
Davis, A. P., Shokouhian, M., Sharma, H., and Minami, C. (2001). Laboratory study of
biological retention for urban stormwater management. Water Environment Research: 73(1):
5-14. http://www.jstor.org/stable/25045454.
Davis, A. P., Shokouhian, M., Sharma, H., Minami, C., and Winogradoff, D. (2003). Water
quality improvement through bioretention: Lead, copper, and zinc removal. Water
Environment Research: 75(1): 73-82. http://www.jstor.org/stable/25045664
Davis, A. P., Shokouhian, M., Sharma, H., and Minami, C. (2006). Water quality improvement
through bioretention media: Nitrogen and phosphorus removal. Water Environment
Research: 78(3): 284-293. doi: 10.2175/106143005X94376.
Davis, A. P. (2007). Field performance of bioretention: Water quality. Environmental
Engineering Science 24(8): 1048-1064. doi: 10.1089/ees.2006.0190.
Davis, A. (2008). Field performance of bioretention: Hydrology impacts. Journal of Hydrologic
Engineering: 13(2): 90-95. doi: 10.1061/(ASCE)1084-0699(2008)13:2(90).
Davis, A., Hunt, W., Traver, R., and Clar, M. (2009). Bioretention technology: Overview of
current practice and future needs. Journal of Environmental Engineering: 135(3): 109-117.
doi: 10.1061/(ASCE)0733-9372(2009)135:3(109).
DeBusk, K. and Wynn, T. (2011). Storm-water bioretention for runoff quality and quantity
mitigation. Journal of Environmental Engineering: 137(9): 800-808. doi:
10.1061/(ASCE)EE. 1943-7870.0000388.
Dietz, M. E. and Clausen, J. C. (2005). A field evaluation of rain garden flow and pollutant
treatment. Water, Air, and Soil Pollution: 167(1-4): 123-138. doi: 10.1007/sl 1270-005-8266-8.
45
-------
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/sll270-007-9484-z.
Dreelin, E. A., Fowler, L., and Ronald Carroll, C. (2006). A test of porous pavement
effectiveness on clay soils during natural storm events. Water Research: 40(4): 799-805. doi:
10. 1016/j.watres.2005. 12.002.
Duda, A. M., Lenat, D. R., and Penrose, D. L. (1982). Water quality in urban streams: What we
can expect. Water Pollution Control Federation: 54(7): 1139-1147.
"•! 1 i1-' '
Dunne, T., and Leopold, L. B. (1978). Water in environmental planning, WH Freeman, New
York, 818p.
Dussaillant, A., Wu, C., and Potter, K. (2004). Richards equation model of a rain garden. Journal
of Hydrologic Engineering: 9(3): 219-225. doi: 10.1061/(ASCE)1084-0699(2004)9:3(219).
Elliott, A., and Trowsdale, S. (2007). A review of models for low impact urban stormwater
drainage. Environmental Modelling and Software: 22(3): 394-405. doi:
10. 1016/j.envsoft.2005. 12.005.
Emerson, C., and Traver, R. (2008). Multiyear and seasonal variation of infiltration from storm-
water best management practices. Journal of Irrigation and Drainage Engineering: 134(5):
598-605. doi: 10.1061/(ASCE)0733-9437(2008)134:5(598)
Fewkes, A. (2000). Modelling the performance of rainwater collection systems: Towards a
generalized approach. Urban Water: 1(4): 323-333. doi: 10.1016/81462-0758(00)00026-1.
French, R. (1988). Cisterns for water conservation and flood control. Journal of Water Resources
Planning and Management: 114(5): 565-577. doi: 10.1061/(ASCE)0733-
9496(1988)114:5(565).
Ghisi, E., Bressan, D. L., and Martini, M. (2007). Rainwater tank capacity and potential for
potable water savings by using rainwater in the residential sector of southeastern Brazil.
Building and Environment: 42(4): 1654-1666. doi: 10.1016/j.buildenv.2006.02.007.
Gilbert, J. K. and Clausen, J. C. (2006). Stormwater runoff quality and quantity from asphalt,
paver, and crushed stone driveways in Connecticut. Water Research: 40(4): 826-832. doi:
10. 1016/j.watres.2005. 12.006.
Gilroy, K. L. and McCuen, R. H. (2009). Spatio-temporal effects of low impact development
practices. Journal of Hydrology: 367(3-4): 228-236. doi: 10.1016/j.jhydrol.2009.01.008.
Goodrich, D. C., Williams, D. G., Unkrich, C. L., Hogan, J. F., Scott, R. L., Hultine, K. R., Pool,
D., Goes, A. L., and Miller, S. N. (2004). "Comparison of Methods to Estimate Ephemeral
Channel Recharge, Walnut Gulch, San Pedro River Basin, Arizona." Recharge and Vadose
Zone Processes: Alluvial Basins of the Southwestern United States, F.M. Phillips, J.F.
46
-------
Hogan, and B. Scanlon, eds., Water Science and Application 9, American Geophysical
Union, Washington, DC, p. 77-99.
Goodrich, D. C., Burns I. S., Unkrich C. L., Semmens D. J., Guertin D. P., Hernandez M.,
Yatheendradas S., Kennedy J. R., and Levick L. R. (2012) KINEROS2/AGWA: Model Use,
Calibration, and Validation. Transactions of the ASABE: 55(4): 1561-1574.
Guo, Y. and Baetz, B. (2007). Sizing of rainwater storage units for green building applications.
Journal of Hydrologic Engineering: 12(2): 197-205. doi: 10.1061/(ASCE)1084-
0699(2007)12:2(197).
Hammer, T. R. (1972). Stream channel enlargement due to urbanization. Water Resources
Research: 8(6): 1530-1540. doi: 10.1029/WR008i006p01530.
Hatt, B. E., Fletcher, T. D., and Deletic, A. (2009). Hydrologic and pollutant removal
performance of stormwater biofiltration systems at the field scale. Journal of Hydrology:
365(3-4): 310-321. doi: 10.1016/j.jhydrol.2008.12.001.
Heasom, W., Traver, R. G., and Welker, A. (2006). Hydrologic modeling of a bioinfiltration best
management practice. Journal of the American Water Resources Association: 42(5): 1329-
1347. doi: 10.1111/j.l752-1688.2006.tb05304.x.
Hollis, G. E. (1975). The effect of urbanization on floods of different recurrence interval. Water
Resources Research: 11(3): 431-435. doi: 10.1029/WR011i003p00431.
Hood, M. J., Clausen, J. C., and Warner, G. S. (2007). Comparison of stormwater lag times for
low impact and traditional residential development. Journal of the American Water
Resources Association: 43(4): 1036-1046. doi: 10.1111/j.l752-1688.2007.00085.x.
Hoskins, J. andPeterein, J. (2013). Case study of St. Louis, Missouri: Comparison of
bioretention performance to the runoff component of a restored water balance. Journal of
Environmental Engineering: 139(4): 516-521. doi: 10.1061/(ASCE)EE.1943-7870.0000625.
Hsieh, C. and Davis, A. (2005). Evaluation and optimization of bioretention media for treatment
of urban storm water runoff. Journal of Environmental Engineering: 131(11): 1521-1531.
doi: 10.1061/(ASCE)0733-9372(2005)131:11(1521).
Hunt, W., Jarrett, A., Smith, J., and Sharkey, L. (2006). Evaluating bioretention hydrology and
nutrient removal at three field sites in North Carolina. Journal of Irrigation and Drainage
Engineering: 132(6): 600-608. doi: 10.1061/(ASCE)0733-9437(2006)132:6(600).
Hunt, W., Smith, J., Jadlocki, S., Hathaway, J., and Eubanks, P. (2008). Pollutant removal and
peak flow mitigation by a bioretention cell in urban Charlotte, N.C. Journal of Environmental
Engineering: 134(5): 403-408. doi: 10.1061/(ASCE)0733-9372(2008)134:5(403).
James, M. and Dymond, R. (2012). Bioretention hydrologic performance in an urban stormwater
network. Journal of Hydrologic Engineering: 17(3): 431-436. doi: 10.1061/(ASCE)HE.1943-
5584.0000448.
47
-------
Jenkins, J., Wadzuk, B., and Welker, A. (2010). Fines accumulation and distribution in a storm-
water rain garden nine years postconstruction. Journal of Irrigation and Drainage
Engineering: 136(12): 862-869. doi: 10.1061/(ASCE)IR. 1943-4774.0000264.
Jennings, A., Adeel, A., Hopkins, A., Litofsky, A., and Wellstead, S. (2013). Rain Barrel-urban
garden stormwater management performance. Journal of Environmental Engineering: 139(5):
757-765. doi: 10.1061/(ASCE)EE. 1943-7870.0000663.
Jia, H., Lu, Y., Yu, S. L., and Chen, Y. (2012). Planning of LID-BMPs for urban runoff control:
The case of Beijing Olympic village. Separation and Purification Technology: 84: 112-119.
doi: 10.1016/j.seppur.2011.04.026.
Jones, M. P. and Hunt, W. F. (2010). Performance of rainwater harvesting systems in the
southeastern united states. Resources, Conservation and Recycling: 54(10): 623-629. doi:
10.1016/j.resconrec.2009.11.002.
Kennedy, J.R. (2007). "Changes in storm runoff with urbanization: The role of pervious areas in
a semi-arid environment." M.S. thesis, Univ. of Arizona, Tucson, AZ, 111 p.
Kennedy, J., Goodrich, D., and Unkrich, C. (2013). Using the KINEROS2 modeling framework
to evaluate the increase in storm runoff from residential development in a semiarid
environment. Journal of Hydrologic Engineering: 18(6): 698-706. doi:
10.1061/(ASCE)HE. 1943-5584.0000655.
Kim, K. and Yoo, C. (2009). Hydrological modeling and evaluation of rainwater harvesting
facilities: Case study on several rainwater harvesting facilities in Korea. Journal of
Hydrologic Engineering: 14(6): 545-561. doi: 10.1061/(ASCE)HE. 1943-5584.0000030.
Kronaveter, L., Shamir, U., and Kessler, A. (2001). Water-sensitive urban planning: Modeling
on-site infiltration. Journal of Water Resources Planning and Management: 127(2): 78-88.
doi: 10.1061/(ASCE)0733-9496(2001)127:2(78).
Lee, J. G., Selvakumar, A., Alvi, K., Riverson, J., Zhen, J. X., Shoemaker, L., and Lai, F. (2012).
A watershed-scale design optimization model for stormwater best management practices.
Environmental Modelling and Software: 37(0): 6-18. doi: 10.1016/j.envsoft.2012.04.011.
Legret, M. and Colandini, V. (1999). Effects of a porous pavement with reservoir structure on
runoff water: Water quality and fate of heavy metals. Water Science and Technology: 39(2):
111-117. doi: 10.1016/80273-1223(99)00014-1.
Leopold, L. B. (1968). Hydrology for urban land planning: A guidebook on the hydrologic
effects of urban land use. Washington (DC): US Geological Survey.
Li, H., Sharkey, L., Hunt, W., and Davis, A. (2009). Mitigation of impervious surface hydrology
using bioretention in North Carolina and Maryland. Journal of Hydrologic Engineering:
14(4): 407-415. doi: 10.1061/(ASCE)1084-0699(2009)14:4(407).
48
-------
Loucks, E. D., Garrett, N., and Oriel, K. A. (2004). Evaluating the Effectiveness of stormwater
Reduction Practices Using Continuous Simulation. Bridges: 10(40737): 81. doi:
10.1061/40737(2004)81.
Massart, I, Guertin, D.P., Smith, R.E., Goodrich, D.C., Unkrich, C.L., Levick, L., 2010. K2-O2
(KINEROS-Opus) spatially based watershed hydrologic and biogeochemical model. Proc.
2nd Joint Federal Interagency Conference, Las Vegas, NV, June 27 - July 1, 2010, 12
p. irtljK^
Makepeace, D. K., Smith, D. W., and Stanley, S. J. (1995). Urban stormwater quality: Summary
of contaminant data. Critical Reviews in Environmental Science and Technology: 25(2): 93-
139. doi: 10.1080/10643389509388476.
Miller, S. N., Semmens, D. J., Goodrich, D. C., Hernandez, M., Miller, R. C., Kepner, W. G.,
and Guertin, D. P. (2007). The automated geospatial watershed assessment tool.
Environmental Modelling and Software: 22(3): 365-377. doi: 10.1016/j.envsoft.2005.12.004.
Norman, L.M., Guertin, D.P., and Feller, M. (2008a). An approach to prevent nonpoint-source
pollutants and support sustainable development in the Ambos Nogales Transboundary
Watershed. Proceedings of a USGS Workshop on facing tomorrow's challenges along the
US-Mexico border; monitoring, modeling, and forecasting change within the Arizona-Sonora
trnasboundary watersheds: USGS Circular 1322,
Norman, L.M., Guertin, D.P., and Feller, M. (2008b). A coupled model approach to reduce
nonpoint source pollution resulting from predicted urban growth: A case study in the Ambos
Nogales Watershed. Journal of Urban Geography: 29(5): 496-516.
Norman, L.M., Huth, H., Levick, L., Burns, IS., Guertin, D.P., Lara- Valencia, F., and Semmens,
D. (2010). Flood Hazard Awareness and Hydrologic Modeling at Ambos Nogales, US-
Mexico Border. Journal of Flood Risk Management: 3(2) 15 pp.
Olszewski, J. and Davis, A. (2013). Comparing the hydrologic performance of a bioretention cell
with pre-development values. Journal of Irrigation and Drainage Engineering: 139(2): 124-
130. doi: 10. 1061/(ASCE)IR. 1943-4774.0000504.
Fagotto, C., Legret, M., and Le Cloirec, P. (2000). Comparison of the hydraulic behaviour and
the quality of highway runoff water according to the type of pavement. Water Research:
34(18): 4446-4454. doi: 10.1016/80043-1354(00)00221-9.
Pratt, C. (1995). A review of source control of urban stormwater runoff. Journal of the Institution
of Water and Environmental Management: 9(2): 132-139. doi: 10.1 11 1/j. 1747-
6593.1995.tb01602.x.
Rose, S. and Peters, N. E. (2001). Effects of urbanization on streamflow in the Atlanta area
(Georgia, USA): A comparative hydrological approach. Hydrological Processes: 15(8):
1441-1457. doi: 10.1002/hyp.218.
49
-------
Rossman, L. A, and Supply, W. (2010). Storm water management model user's manual, version
5.0. National Risk Management Research Laboratory, Office of Research and Development,
US Environmental Protection Agency.
Roy-Poirier, A., Champagne, P., and Filion, Y. (2010). Review of bioretention system research
and design: Past, present, and future. Journal of Environmental Engineering: 136(9): 878-
889. doi: 10.1061/(ASCE)EE. 1943-7870.0000227.
Rushton, B. T. (2001). Low-impact parking lot design reduces runoff and pollutant loads. Journal
of Water Resources Planning and Management: 127(3): 172-179. doi: 10.1061/(ASCE)0733-
9496(2001)127:3(172).
Russ, T. H. and Russ, S. M. (2002). Site Planning and Design Handbook, McGraw-Hill Inc.,
Boston, MA.
Sample, D., Liu, J., and Wang, S. (2012). Evaluating the dual benefits of rainwater harvesting
systems using reliability analysis. Journal of Hydrologic Engineering: 18(10): 1310-1321.
doi: 10.1061/(ASCE)HE. 1943-5584.0000748.
Sands, K., and Chapman, T. (2003). Rain barrels - truth or consequences. In: Proceedings of the
National Conference on Urban Storm Water: Enhancing Programs at the Local Level.
Chicago, IL, February 17-20, pp 390-394. http://water.epa.gov/polwaste/nps/
stormwater/upload/2003 03 25 NFS natlstormwater03 fullreport.pdf
Scanlon, B. R., Keese, K., Reedy, R.C., Simunek, J., and Andraski, B. J. (2003). Variations in
flow and transport in thick desert vadose zones in response to paleoclimatic forcing (0-90
kyr): Field measurements, modeling, and uncertainties. Water Resources Research: 39(7):
1179. doi: 10.1029/2002WR001604.
Schliiter, W. and Jefferies, C. (2002). Modelling the outflow from a porous pavement. Urban
Water: 4(3): 245-253. doi: 10.1016/81462-0758(01)00065-6.
Scholz, M. and Grabowiecki, P. (2007). Review of permeable pavement systems. Building and
Environment: 42(11): 3830-3836. doi: 10.1016/j.buildenv.2006.11.016.
Sharkey, L. J. (2006). The performance of bioretention areas in North Carolina: A study of water
quality, water quantity, and soil media. Master's thesis.
http://www.lib.ncsu.edu/resolver/1840.16/2062
Shuster, W., Roy, A., Thurston, H., Morrison, M., Taylor, M., and Clagett, M. (2008).
Implementation of retrofit best management practices in a suburban watershed (Cincinnati,
OH) via economic incentives. In: Proceedings of the 11th International Conference on Urban
Drainage, Edinburgh, Scotland, UK, August 31-September 5.
Shuster, W. and Rhea, L. (2013). Catchment-scale hydrologic implications of parcel-level
stormwater management (Ohio, USA). Journal of Hydrology: 485(0): 177-187. doi:
10.1016/j.jhydrol.2012.10.043.
50
-------
Smith, R. E., Goodrich, D. C., Woolhiser, D. A., and Unkrich, C. L. (1995). KINEROS - A
kinematic runoff and erosion model. Chap. 20, Computer Models of Watershed Hydrology,
(Ed. by Singh, V. J.) Water Resources Pub., Highlands Ranch, Colo., pp. 697-732.
Thurston, H. W., Roy, A. H., Shuster, W. D., Morrison, M. A., Taylor, M. A., and Cabezas, H.
(2008). Using economic incentives to manage stormwater runoff in the Shepherd Creek
watershed, Part 1. US Environmental Protection Agency, Washington, DC, EPA/600/R-
08/129.
Trieu, P., Guillozet, P., Galli, J., and Smith, M. (2001). Combined sewer overflow rooftop type
analysis and rain barrel demonstration project. Department of Environmental Programs,
Metropolitan Washington Council of Governments.
http://www.anacostia.net/Archives/download/Rain_Barrel_Report.pdf
USEPA. (2000). Low impact development (LID): A literature review. (EPA-841-B-00-005).
United States Environmental Protection Agency.
http ://water. epa. gov/polwaste/green/upload/lid.pdf
U.S. EPA, 2010. Storm Water Management Models - User's Manual, Version 5, EPA Report
600/R-05/040, 295 p.
Ward S., Memon F. A., and Butler, D. (2010). Rainwater harvesting: Model-based design
evaluation. Water Science and Technology: 61(1): 85-96. doi: 10.2166/wst.2010.783.
Ward, S., Memon, F. A., and Butler, D. (2012). Performance of a large building rainwater
harvesting system. Water Research: 46(16): 5127-5134. doi: 10.1016/j.watres.2012.06.043.
Waterfall, P. (2004). Harvesting rainwater for landscape use. The University of Arizona
Cooperative Extension. 2nd Edition.
http://arizona.openrepository.com/arizona/handle/10150/144824
White, M. D. and Greer, K. A. (2006). The effects of watershed urbanization on the stream
hydrology and riparian vegetation of Los Penasquitos Creek, California. Landscape and
Urban Planning: 74(2): 125-138. doi: 10.1016/j.landurbplan.2004.11.015.
Williams, E. S., and Wise, W. R. (2006). Hydrologic impacts of alternative approaches to storm
water management and land development. Journal of the American Water Resources
Association: 42(2): 443-455. doi: 10.1111/j.l752-1688.2006.tb03849.x.
Woolhiser, D. A., Smith, R., and Goodrich, D. C. (1990). KINEROS: A kinematic runoff and
erosion model: Documentation and user manual US Department of Agriculture, Agricultural
Research Service.
51
-------
52
-------
53
-------
&EPA
United States
Environmental Protection
Agency
Office of Research
and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
EPA/600/R-14/329
ARS/309819
October 2014
www.epa.gov
Please make all necessary changes on the below label, detach
or copy and return to the address in the upper left hand corner.
If you do not wish to receive these reports CHECK HERE Q]
detach, or copy this cover, and return to the address in the
upper left hand corner.
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA PER MIT No. G-35
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |