&EPA
United States
Environmental Protection
Agency
EPA 600/R-13/230 I June 2013 I www.epa.gov/ada
Tracking the Fate of
Watershed Nitrogen:
The "N-Sink" Web Tool and
Two Case Studies
Office of Research and Development
National Risk Management Research Laboratory, Ada, Oklahoma 74820
-------
-------
Tracking the Fate of Watershed Nitrogen:
The "N-Sink" Web Tool and Two Case Studies
by
C. Arnold1, D.Q. Kellogg2, K. J. Forshay3, C. Damon2,
E. H. Wilson1, A. Gold2, E.A. Wentz4, and M.M. Shimizu4
1 University of Connecticut Center for Land Use Education and Research
2 University of Rhode Island Coastal Institute
3 US EPA Office of Research and Development
4 Arizona State University School of Geographical Sciences and Urban Planning
Project Officer
Dr. Kenneth J. Forshay
Ground Water and Ecosystems Restoration Division
National Risk Management Research Laboratory
US EPA Office of Research and Development
Ada, OK 74820
Technical Support Center Director
Dr. David S. Burden
Ground Water and Ecosystems Restoration Division
National Risk Management Research Laboratory
US EPA Office of Research and Development
Ada, OK 74820
¦ Office of Research and Development
National Risk Management Research Laboratory, Ada, Okiahoma 74820
-------
Notice
The U.S. Environmental Protection Agency through its Office of Research and Development funded
and managed the research herein. This document was prepared by both the University of Connecticut
Center for Land Use Education and Research and the University of Rhode Island Coastal Institute
for the U.S. Environmental Protection Agency (U.S. EPA) under CB&I Contract No. EP-C-08-034,
Subcontract No. 824181-000 OP. It has been subjected to the Agency's peer review and has been
approved for publication as an U.S. EPA document. No official endorsement should be inferred.
Mention of trade names or commercial products does not constitute endorsement or recommendation
for use. All research projects making conclusions or recommendations based on environmentally
related measurements and information and funded by the Environmental Protection Agency are
required to comply with the requirements of the Agency Quality Assurance Program. This project
was conducted under an approved Quality Assurance Project Plan.
Contact Information:
Chester Arnold
Associate Director
Center for Land Use Education and Research
University of Connecticut
1066 Saybrook Road, Haddam, CT 06438
Chester. arnold(a)ticonn. edu
860-345-5230
Q Kellogg, Ph.D.
Coastal Institute & Dept. of Natural Resources Science
University of Rhode Island
105 Coastal Institute in Kingston
Kingston, RI 02881
qkelloeg(a)gmail. com
401-874-4866
Kenneth J. Forshay, Ph.D.
Research Ecologist
Robert S. Kerr Environmental Reseach Center
919 Kerr Research Drive
P.O. Box 1198
Ada, OK 74820
forshav. ken(a).epa. gov
580-436-8912
-------
Contents
Notice ii
List of Tables and Figures iv
Abstract v
1.0 Introduction 1
1.1 Purpose 1
1.2 Nitrogen and Coastal Watersheds 1
1.3 Land Use Decision Makers and Geospatial Technology 2
2.0 The N-Sink Tool 4
2.1 Overview 4
2.2 Data Used and Calculations Performed 5
2.3 Scientific Basis for N-Sink 6
2.3.1 Landscape N sinks 6
2.3.1.1 Freshwater Wetlands 7
2.3.1.2Lentic Waterbodies: Ponds, Lakes and Reservoirs 7
2.3.1.3 Lotic Waterbodies: Stream Reaches 8
2.3.2 Landscape N Sources 9
2.3.2.1 Developed Land 9
2.3.2.2 Agricultural Land 9
2.4 Assumptions and Limitations 10
2.5 Tool Description 11
2.6 Future N-Sink improvements 16
3.0 Results: Two Case Studies 17
3.1 Study area 17
3.2 Case Study #1: The Niantic River Watershed, CT 19
3.3 Case Study #2: Hie Saugatucket River Watershed, RI 20
3.4 QA and Model Verification: Comparing N-Sink N Reduction Tool Results to Conventional
Calculations 21
3.4.1 Methods 22
3.4.2 Comparison Results 22
4.0 Next Steps 25
5.0 References 26
Appendix A. Niantic River Watershed Case Study 1 31
Appendix B. Saugatucket River Case Study 2 40
Appendix C. Modelbuilder® charts for N-Sink 50
-------
List of Tables and Figures
Table 1. Databases used in N-Sink 6
Table 2. N-Sink High/Medium/Low N removal designations 7
Table 3. N-Sink estimates of N removal based on width of riparian wetland 7
Table 4.1 Comparison of % N removal by reach along flow path shown in Saugatucket 23
Table 4.2 Comparison of % N removal by reach along flow path show?n in Saugatucket 23
Table 4.3 Comparison of % N removal by reach along flow path shown in Saugatucket 24
Table 4.4 Comparison of % N removal by reach along flow path shown in Saugatucket 24
Figure 1. Conceptual Diagram of N-Sink 4
Figure 2. Screen capture of N-Sink "portal" website 12
Figure 3. Example of stream reach pop-up box details 13
Figure 4. Example of pond/lake pop-up box details 14
Figure 5. N Removal tool, control box detail 14
Figure 6. N Removal tool flow path example 15
Figure 7. Study area map 17
Figure 8. Two case study watersheds 18
Figure 9. Niantic River watershed, CT 19
Figure 10. Saugatucket River watershed, RI 21
Figure C-1 Overall layout of how the final model is constructed 51
Figure C-2. FlowPath sub-model 52
Figure C-3. Land Removal sub-model 53
-------
Abstract
Nitrogen is increasingly being identified as a pollutant of concern in both coastal and inland
waters. In some areas, the majority of the nitrogen loading comes from wastewater treatment
plants and/or combined sewer overflows. However, in less urbanized watersheds nonpoint source
runoff and nitrogen from septic systems are the primary vehicles of nitrogen delivery. In these
areas, watershed land use has a direct relationship with both sources and sinks of nitrogen.
The "N-Sink" tool was created to provide a useful and accessible tool for local land use managers
to explore the relationship of land use in their towns and counties to nitrogen pollution of their
waters. N-Sink uses the best available science on land use/nitrogen interactions, plus widely
available basic datasets for hydrography, soils and land use, to highlight major sources and sinks of
nitrogen within a watershed context.
N-Sink was originally designed as an Arc Map extension for use with desktop software, but during
this project was redesigned as a geospatial web tool using ArcGIS Viewer for Flex. The tool
highlights N sources and sinks within a watershed, and allows the non-technical user to estimate
relative N removal efficiencies from any chosen location within the watershed. The project team
used N-Sink to generate N source/sink information for two case study coastal HUC-12 watersheds,
the Niantic River watershed in southeastern Connecticut and the Saugatucket River watershed in
southern Rhode Island on the west side of Narragansett Bay. Next steps will include using N-Sink-
generated information in educational programs for decision makers within the two watersheds. In
addition, the team will explore the challenges and methods of making N-Sink a nationally available
tool.
-------
-------
1.0
Introduction
1.1 Purpose
Although nitrogen fertilization is vital to
maintain a productive and vigorous food
supply, excess reactive nitrogen (N) released
to the environment causes deleterious effects
on human and ecosystem health. In some
areas, the majority of the nitrogen loading
comes from wastewater treatment plants and/or
combined sewer overflows. However, in less
urbanized watersheds nonpoint source runoff
and nitrogen from septic systems are the
primary vehicles of nitrogen delivery. In these
areas, watershed land use has a direct relation-
ship with both sources and sinks of nitrogen.
In order to manage the widespread N pollution
problem in watersheds it is critical for decision
makers to have a strong technical grasp of the
factors that control N sources, sinks (regions
or areas that can retain N), and pathways along
which nitrogen is moved and transformed
(U.S. EPA 2011). Here we define landscape
N sinks as areas where N cycling processes
such as denitrification, as well as plant and
microbial uptake remove, bury or sequester
N, thereby reducing downstream N transport.
Denitrification removes N permanently, while
burial and sequestration are not permanent but
represent long-term storage of N.
The "N-Sink" nitrogen management tool is
different from other N transportation models
in several ways. First, N-Sink focuses on
sinks rather than sources. Instead of estimat-
ing N loading from land use and runoff data,
N-Sink estimates N attenuation along a flow
path from source to receiving water. This
directs the focus of decision makers on
landscapes that might be valuable to preserve
in the future, rather than just source areas than
need to be addressed in the present. Second,
N-Sink is designed to use widely available
national datasets as its input, rather than rely
on potentially more accurate field derived
data. N-Sink is intended as a decision support
tool for widespread and easy use, not a state-
of-the-art model. N-Sink does, however, use
state-of-the-art techniques when it comes to
the actual use of the tool for decision sup-
port. It is configured as a web tool, relatively
self-explanatory and accessible to anyone with
internet access.
In summary, N-Sink has been designed based
on the best available science, but with the
explicit goals of communicating informa-
tion to decision makers in an accessible and
understandable manner (U.S. EPA 2012). This
paper describes the development to date of this
effort.
1.2 Nitrogen and Coastal Watersheds
Nitrogen pollution is emerging as a major
threat to coastal watersheds, estuaries and
embayments, and the communities within their
watersheds. Nitrogen loading to coastal waters
can spur harmful algal blooms, hypoxia,
decline of eelgrass, and destruction of critical
spawning habitats in coastal waters (Valiela
et al. 1990; Oviatt et al. 1995; Nixon 1995;
Short and Burdick 1996; NRC 2000; Nixon et
al. 2001; Conley et al. 2009). Coastal com-
munities in New England and along the entire
East Coast of the Atlantic [e.g., Long Island
Sound, CT (NYDEC and CTDEP 2000); Cape
Cod, MA (MA DEP 2008); Pawcatuck River,
RI (RI DEM 2010); Christina Reservoir, ME
(ME DEP 2010); Neuse River estuary, NC
(NC DENR 1999; Stow et al. 2011)] have
been forced to address N pollution as a result
of state and federal water quality programs
such as TMDLs (Total Maximum Daily Loads)
that can mandate community investments in N
controls.
-------
The sources, sinks, and conveyance of N are
highly landscape and hydrology dependent. N
export from urban and suburban watersheds is
much higher than from forested watersheds,
although lower than from agricultural water-
sheds (Groffman et al. 2003). High concentra-
tions of nitrate in shallow groundwater and
streams are correlated with agricultural land
use and unsewered residential developments
(Nowicki and Gold, 2008). Sink areas include
wetlands (characterized by hydric soils), reser-
voirs, small-order streams and impoundments
(Groffman et al., 2003). In particular, riparian
wetlands can be a significant sink for N due to
the combination of surface filtering of sedi-
ments, plant and microbial uptake, and subsur-
face denitrification (Gold et al., 2001). Studies
in both urbanizing (Kaushal et al. 2008) and
agricultural (Clausen et al. 2000) watersheds
have demonstrated that riparian restoration can
reduce the delivery of nitrogen to streams.
1,3 Land Use Decision Makers and
Geospatial Technology
The N-Sink tool is based on the premise that
local decision makers require environmental
data that is highly localized, easily accessed
and immediately understandable (Last 1995;
Merry et al. 2008a). Since N sources and
sinks are so closely linked to land use, land
use decision makers are a critical audience
for tools that can translate science into man-
agement-relevant information. The primary
purpose of the work described in this paper is
to incorporate scientific understanding of land
use and nitrogen cycling relationships into
practices that local land-use decision makers
can adopt and act upon to support sustain-
able and healthy communities. To optimize
N control strategies for coastal waters, it is
important for land use managers to recognize
that N delivery is linked to areas that serve as
sources and sinks within their watersheds.
Most land use decisions are made by volun-
teers - elected officials or appointed members
of planning, zoning and appeals boards - most
of whom have little or no training in land
planning or natural resource protection (Arnold
1999; Gold and Joubert 2007). Few have
training in watershed management and water
quality issues (Arnold 1999), and most do not
have the time to become experts on all water-
related topics of concern in their communities
such as storm water, sediment, nonpoint source
pollution, etc. This extends also to the profes-
sionals (e.g. planners, engineers) that support
these volunteers in the towns lucky enough to
have this capacity.
Thus, community decision makers often have
no way in which to factor N pollution into
their land use policies and decisions. A tool
that combines maps and local data related to
N source and sink potential can be extremely
useful in helping land use managers identify
areas that could benefit from source controls,
as well as areas that serve to reduce N delivery
downstream and thus should be targeted for
protection and restoration.
Geospatial tools, based on geographical
information systems (GIS), geographic visu-
alization, exploratory data analysis tools,
and spatial statistical analysis and modeling
(Armstrong 1992), provide new opportunities
to help land use managers improve their access
to, and use of, watershed science to make
wise land use decisions. Remotely sensed
land cover data, placed in a GIS environment
in combination with other natural resource
data, has been used effectively to inform land
use decisions in Connecticut and elsewhere
(Wilson et al., 2011; Arnold et al. 2000).
Merry et al. (2008a and b) found that digital
maps and GIS databases were the preferred
way to view land use change data among
Extension professionals and land use planners.
GIS-based tools have improved the capac-
ity of local decision makers to examine and
compare the cumulative impacts of a variety
of alternative decisions (Cova and Church
1997; Davis and Keller 1997; Thumerer et al.
2000; Tulloch et al. 2003; Lant et al. 2005;
-------
Brown 2006). Combining data management
capabilities, display functions, and modeling
tools generates a more robust and sophisticated
means for problem evaluation and resolution
(Carver 1991).
The increasing "fusion" of geospatial and
internet technologies has opened up an even
wider range of possibilities for decision
support and information-sharing, allowing
the creation of web-based tools that require
no geospatial expertise to use (Dickson et al.,
2011; Dickson and Arnold, 2009; Rozum et al.
2005). For this reason, a web-based platform
is the goal of this project.
-------
2.0
The N-Sink Tool
2.1 Overview
N-Sink estimates the extent of N delivery from
a source area, such as unsewered residential
development, croplands or livestock areas, to a
watershed outlet. Land cover and N movement
relationships taken from the scientific literature
are used to identify N source and sink areas,
and to roughly quantify their impact on the N
budget.
N-Sink uses the particle tracking capabilities
provided by data from NHDPlus v2 to estimate
flow paths from source areas to a watershed
outlet. The focus is on examining the land-
scape N sinks (streams, ponds/lakes, and
wetlands) that are encountered along a flow
path to estimate incremental N removal based
on sink characteristics such as wetland size,
stream reach length, and pond area (Figure 1).
More detail can be found in Section 2.3.
There are a variety of tools currently avail-
able on the web to estimate N delivery from a
watershed. Examples include L-THIA, focused
on the mid-West, (https:iknzineeritig.purdue.
edii'-lthiaA. and NLoad (http: nload.mhl.
eduA. Both of these tools provide important
services to the decision-making community.
N-Sink builds on this foundation and seeks to
improve upon these tools by using flow path
analysis to understand the spatial relationships
between N sources and sinks on the landscape
Other web-based tools generally use a lumped
sum approach to estimate N delivery from
a watershed or user-defined area, summing
contributions from different land use/soils
combinations, but ignoring location within a
hydrologic context. N-Sink's unique approach
allows users to better understand how land-
scape sinks contribute to N removal and to
N-Sink INPUTS
N-Sink CALCULATIONS and MAPPING
NHDPlus v2
Stream reach characteristics:
length, drainage area, flow rate
Pond/lake characteristics:
pond area, drainage area
Starting Location
(Source)
Supplied by user by
clicking on the map
N removal (%) is calculated and stored for
all streams and ponds/lakes;
Surface water sinks are color coded and
drawn as low/medium/high N removal;
Basic statistics (area, land cover) are made
available for all watersheds
NHDPlus v2
Particle tracking capability
Flow path is
generated and
drawn
NLCD
Land cover
Terrestrial component
of flow path
SSURGO
Hydric (wetland) soils
N removal (%) is cumulatively calculated
from each landscape sink encountered
along the flow path, starting with the
terrestrial components (wetlands),
followed by surface water components
(stream reaches and ponds/lakes).
Figure 1. Conceptual diagram of N-Sink
-------
compare N delivery to a receiving water body
from different source locations.
The prototype for N-Sink, created by the
University of Rhode Island in partnership with
Arizona State University and the University
of Connecticut, was supported by both USD A
and EPA. The prototype was originally
planned and designed as an extension to be
used with ArcMap desktop GIS software
(Kellogg et al., 2010). During the current
project, supported by U.S. EPAs Nitrogen
Management Group as part of the Sustainable
and Healthy Communities program, N-Sink
has been transformed into a web-based tool
using ArcGIS Viewer for Flex.
ArcGIS Viewer for Flex is a configurable web
mapping application that can be used to create
customized interactive maps on the web. The
Viewer is built using the ArcGIS API for Flex
and utilizes Adobe's Flash browser plugin. The
N-Sink web tool includes basic Flex "widgets"
that make the web tool easy and efficient to
use. These allow a map reader to control the
visibility of data layers, draw and measure
features on the map and view the map legend.
At the center of the N-Sink web tool is the
Nitrogen Removal widget. This widget allows
a map user to click on any location within a
watershed and retrieve the estimated relative
nitrogen removal efficiencies from that loca-
tion. The data is processed on the back end
of the tool through a geospatial model created
in Modelbuilder for ArcGIS and published as
a geoprocessing service using ArcGIS Server.
See Appendix C.
2.2 Data Used and Calculations
Performed
In the tool, nitrogen removal from sink areas
is based on a wide range of published studies.
Some guiding principles in N-Sink (Kellogg et
al. 2010) are listed below. See also Section 2.3
for a more complete explanation of the scien-
tific basis for N-Sink calculations.
• Riparian zones are more effective sinks with
hydric soils (Lowrance et al. 1997; Gold et
al. 2001) and the extent of the riparian sink
can be related to the width of the vegetated
hydric soils (Mayer et al. 2007);
• The extent of the sink potential of lakes and
reservoirs is higher with longer retention
times and shallower depth (Seitzinger et al.
2006; David et al. 2006). Similarly sized
lakes with relatively smaller drainage areas
will have longer retention times than those
with larger drainage areas; and
• Stream depth is also critical in assigning N
removal potential in stream reaches, with
more N being removed in shallower streams
with longer retention times (Alexander et al.
2007).
N-Sink was developed using widely available
national databases in order to provide broad
applicability. These data are listed in Table 1
and include:
• Topography, hydrography and watershed
boundaries (NHDPlus v2, an integrated geo-
spatial dataset that incorporates the National
Hydrography Dataset (NHD), the National
Elevation Dataset (NED) and the Watershed
Boundary Dataset (WBD). From EPA, with
assistance from USGS)
• Soils (SSURGO, Soil Survey Geographic
Database. Soil maps and tabular data with
information on soil properties and uses,
covering most of the US. From USD A/
NRCS)
• Land cover (National Land Cover Dataset,
Landsat-based land cover data, organized by
class and including descriptive data; multi-
agency consortium)
-------
Table 1. Databases used in N-Sink.
Type of
Data
Source of Data
Resolution of Data
Hydrography,
Topography
and
Watershed
Boundaries
NHDPlus v2(www. horizon-svstems. com/NHDPlus/index.
php). incorporating the National Hydrography Dataset
(NHD: httv://nhd.uses.2ov). National Elevation Dataset
(NED: http://ned.us2s.20v/). and the Watershed Boundary
Dataset (WBD: http://www.nrcs.usda.20v/wps/portal/nrcs/
10 m
main/national/water/watersheds/dataset/)
Soils
Soil Survey Geographic (SSURGO) Database
http: //soils, usda. 20v/survev/2eo2raphv/ssur2o/
1:15,840
Land use
2006 National Land Cover Data (NLCD 2006)
http://www. mrlc. 20v/nlcd 2006. php
30 m
NHDPlus v2 is an integrated suite of geo-
spatial datasets that combine features of the
National Hydrography Dataset (NHD), the
National Elevation Dataset (NED), and the
Watershed Boundary Dataset (WBD). The flow
direction data from NHDPlus v2 are the basis
for particle tracking, which allows the user to
generate a flow path from any chosen point
within a watershed to the watershed's outlet.
This capability supports the Nitrogen Removal
widget.
Ultimately, two types of N delivery informa-
tion will be available to the user. One analysis,
focused on highlighting N sinks and delivery,
calculates the percent reduction of N from
the source to the outlet. This is regardless of
the magnitude of the source, and is meant
primarily to give the user an understanding
of the way N interacts with the various sinks
(wetlands, stream reaches, ponds) along a flow
path, and to allow analysis and comparison of
various locations within the watershed. The
delivery analysis is currently available in the
beta version of the tool delivered with this
report.
The second analysis, focused on loadings, will
estimate the mass of N exported per unit land
area per unit time (kg N/ha/yr) from source
areas, based on land cover and standard data
available in the literature on nutrient export
coefficients (e.g., Beulac and Reckhow 1982).
The user can choose a specific area and either
use a pre-set 1-acre box, or a user-drawn
polygon. The land cover within the area will
be changeable, allowing the user to compare
loadings of existing and proposed land cover.
The loadings analysis is not yet available
on the beta version but will be on the next
version, which the authors hope to have out in
early 2014.
2.3 Scientific Basis for N-Sink
Earlier work on the first (desktop) version of
the N-Sink tool by URI and ASU is described
in detail in Kellogg et al. (2010) and Kellogg
et al. (2011). Excerpts from these two publica-
tions describing the basic science behind the
N-Sink model assumptions are included here
in adapted form.
2.3.1 Landscape N sinks
N-Sink identifies three types of landscape N
sinks: freshwater wetlands, lentic water bodies
(ponds, lakes, or reservoirs), and lotic water
bodies (stream reaches). All landscape N sinks
are characterized as having Low, Medium, or
High potential for N removal, based on esti-
mates calculated for each sink. Because each
type of sink has a different range of N removal
potential, we have currently chosen to create
-------
break points for Low/Medium/High that differ
for the three types (Table 2).
Table 2. N-Sink High/Medium/Low N removal
designations for landscape N sinks
N Sink Type
% N removal
High Medium Low
Riparian
Wetlands
Pond/Lakes/
Reservoirs
Stream Reaches
> 60% 40 to 60% < 40%
> 50% 25 to 50% < 25%
> 30% 15 to 30% < 15%
For example, 30% removal would be con-
sidered High for stream reaches, but Low for
riparian wetlands. This approach recognizes
the inherent characteristics of the different
types of sink that affect critical N removal
factors, such as retention time. The breakpoints
can be changed in future versions.
2.3.1.1 Freshwater Wetlands
Riparian wetlands have been identified by
many researchers as a significant potential sink
for nitrogen. Mayer et al. (2007) performed a
meta-analysis on data available in the scien-
tific literature, based on a wide range of field
studies, to identify trends between riparian N
removal efficiency and riparian buffer width,
surface vs. subsurface flow, and vegetation.
Using widely available GIS data, vegetation
type and surface vs. subsurface processes
cannot be readily identified. However, riparian
land use and soils can be readily identified. It
has been well documented that riparian zones
are most effective as N sinks when undevel-
oped and vegetated, and relatively ineffective
if hydro!ogically altered to bypass the riparian
ecosystem through residential, agricultural or
other types of development (e.g., Carpenter et
al, 1998). Research also suggests that ripar-
ian wetlands, characterized by hydric soils,
act as effective N sinks while riparian areas
with non-hydric soils are less reliable N sinks
(e.g., Lowrance et al., 1997; Gold et al., 2001).
Hydric soils provide conditions that favor
microbial denitri fication: high water table, low
dissolved oxygen, and high soil organic matter
to provide carbon as an electron donor.
N-Sink therefore uses a series of if/then
statements to estimate riparian N removal
efficiency. If land use is developed, then we
assume no removal. If land use is undeveloped
and soils are non-hydric, then we assume no
removal. If riparian land use is undeveloped
and soils are hydric (i.e., wetland soils), then N
removal efficiency is based on the width of the
undeveloped (vegetated) hydric soils (Table 3).
Regression equations provided by Mayer et al.
(2007) guide estimates of N removal effective-
ness in vegetated riparian areas as a function
of wetland width. The selected width classes
recognize commonly used regulatory limits.
These estimates are comparable to estimates
derived from field assessments that are used to
direct management efforts in the Neuse River
watershed (Osmond et al, 2008).
Table 3. N-Sink estimates of N removal based
on width of riparian wetland
Riparian
Hydric
Width
% N
Land Use
Soil Status
(m)a
removal
Developed
0
Vegetated
Non-hydric
0
Hydric
< 5
0
5 to 15
40
15 to 30
60
>30
80
11 Width classes are based on current regulatory practices and
are relevant to local decision-makers
2.3.1.2 Lentic Waterbodies: Ponds, Lakes and
Reservoirs
Ponds, lakes and reservoirs are potentially
large sinks for N because of their long reten-
tion times (Seitzinger et al., 2002, 2006;
Harrison et al., 2009), and hypoxic and anoxic
-------
benthic zones that provide conditions favoring
den Unification. A linear regression analysis
of lake and reservoir data from Seitzinger
et al. (2002), representing a variety of lentic
waterbodies, yields the following relationship
between N removal and the ratio of reservoir
depth, D [m], and residence time, T [years]:
N removal (%) = 79.24 - 33.26 x log10 (D/T)
(1)
Average lake/reservoir depth can be expressed
as volume, V [km3], divided by surface area of
the lake/reservoir, Ar [km2]. Residence time,
T [y], can be expressed as volume, V [km3],
divided by annual discharge, Q . [km3 y1].
Thus,
D/T [my"1] = (V/Ar)/(V/Qyr)xl000
= Qyr/Arxl000
(2)
In this Beta version of N-Sink, we obtain
flow data from NHDPlus v2. GIS data also
provide us with the lake surface area. Due
to occasional water releases, reservoirs can
be viewed as hydrologically distinct from
lakes and ponds. Outflow from reservoirs
can be regulated by dam(s), thereby shifting
residence time from that expected for a natural
lake based on drainage area. Because of the
uncertainty associated with dam manipula-
tion schedules, N-Sink treats reservoirs the
same as lakes when characterizing N removal
potential.
2.3.1.3 Lotic Waterbodies: Stream Reaches
The role of streams in watershed nitrogen
dynamics has been the focus of intensive
research, with early studies formulating the
nutrient spiraling model (e.g., Newbold et al.,
1981), which has since been used to assess
in-stream denitrification (e.g., Royer et al,
2004). The wide range of observed nitrogen
loss rates within streams has spurred research
using both field experiment techniques and
statistical approaches based on spatial data.
The Lotic Intersite Nitrogen Experiment
(LINX) has used N addition and isotopic
analysis to explore the extent to which stream
characteristics - hydrodynamic, chemical, and
metabolic - might explain the wide variation
among streams in nitrogen uptake, removal
and cycling.
Alexander et al. (2000) developed a hybrid
statistical/mechanistic mass-balance model
to estimate N flux in the Mississippi basin
(SPARROW - SPAtially-Referenced
Regression On Watershed attributes), cor-
relating observations of stream N flux with
spatially referenced N sources and physical
characteristics of the landscape and water
bodies. Regression results showed that N loss
rates were inversely related to stream depth
and that much of the N removal in streams
was occurring in lower order reaches. They
concluded that the proximity of N sources to
higher order streams and rivers is an important
factor in N delivery to the Mississippi basin
outlet. Recognizing the variability of stream
function among different regions of the U.S.,
SPARROW has since been developed for
other parts of the country, including New
England (Moore et al., 2004). An important
result of the New England modeling effort
was the lack of statistically significant annual
modeled N reduction for streams with flows
greater than 2.83 m3 s~!, highlighting the
importance of lower-order streams in mitigat-
ing watershed N export.
Alexander et al. (2007) further refined this
New England SPARROW model to investi-
gate and quantify the influence of headwater
streams of the northeastern U.S. on N deliv-
ery to downstream waters. The extent of N
removal and cycling in streams, including the
permanent removal of N via denitrification, is
limited by the extent of interaction with the
stream channel and hyporheic zone, both of
which decrease with increasing stream order
and drainage area. Alexander et al. (2007)
-------
incorporated what was currently known about
N transport to arrive at an expression for the
fraction of N transported along a stream reach
as a function of stream characteristics. We
used this expression as follows:
N (%) removed along stream reach
= (1 -exp (-esiD0S2T))xlOO
(3)
where
0S1 = 0.0513 m d~', 0S2 = -1.319, D = water
depth [m], and T = time of travel [d].
Water depth, D, is expressed as a func-
tion of mean annual stream flow, Q [m3 s"1]
(Alexander et al., 2000):
D = 0.2612 Q03966 (4)
This expression originates from Leopold and
Maddock (1953) with flow data from 112
streams in the South and MidWestern U.S.
Kellogg et al. (2010) used similar methods to
examine the appropriateness of this expression
to estimate stream reach depths in southern
New England and found close agreement
(within 10%) when Q is less than 10 m3 s"1.
N-Sink therefore uses Equation 4 to estimate
stream depth because it is based on a larger
sample of streams that encompasses a wide
range of geographic settings and is in line with
widely-used relationships.
Mean travel time, T [d], along a given stream
reach can be expressed as reach length
[m]/ mean velocity [m d1]. Reach length can
be extracted from the spatial data using simple
GIS tools. In this Beta version of N-Sink, we
get velocity from NHDPlus v2 data.
Both the flow [Q0001E] and velocity
[V0001E] from the NHDPlus v2 dataset have
been adjusted with stream gauge measure-
ments. Both of these variables are in the
Extended Unit Runoff Method (EROM) table.
They are linked back to the flowlines using the
common ID field (McKay et al. 2012).
2.3.2 Landscape N Sources
N-Sink currently recognizes two types of N
sources: unsewered developed land and agri-
culture as row crops. Future versions can be
modified to recognize a more extensive array
of sources and corresponding loads.
2.3.2.1 Developed Land
N-Sink lumps all developed land (e.g., resi-
dential development, institutional, commer-
cial) and assumes that it leaches N at a rate
similar to unsewered medium density residen-
tial development. Future versions of N-Sink
will distinguish between different types of
developed land, such as different densities of
residential development, and different types
of agricultural land. Future versions will also
allow the user to change the N loading rate
from selected areas.
N-Sink assumes that unsewered developed
land contributes 41.7 lb N/acre/yr, from the
combined contributions of septic systems and
lawns, based on the following assumptions:
For one household (or dwelling unit, d.u.),
N input to the septic system is 8.8 lb N/cap/
yr (U.S. EPA, 2002). If we assume an aver-
age of 3 people/household, this comes to
26.4 lb N /d.u./yr. Medium density residential
is typically characterized by V2 acre zoning,
or 2 d.u./acre. Thus, septic system input is
52.8 lb N/ac/yr. We assume that 79% of that
N leaches from the septic system (Gold et al.,
1990), which comes to 41.7 lb/ac/yr. Gold
et al. (1990) also found that fertilized lawns
contributed only a fraction of the total N load.
2.3.2.2 Agricultural Land
N-Sink currently assumes that agricultural
land is cultivated as row crops, with N loading
similar to silage corn. This is a crop that is
common to southern New England and can
contribute N loads comparable to unsewered
residential development. N-Sink assumes
crops are fertilized with manure and that
no cover crop is planted. By assuming no
-------
cover crop, we are presenting a "worst case"
scenario. Cover crops can serve to sequester
excess nutrients present in the soil, and reduce
the leaching of nutrients from the field. N-Sink
uses the average of two years of data presented
in Gold et al. (1990), arriving at a load of
53.7 lb N/ac/yr from agricultural row crops.
N sources are characterized as Low, Medium,
or High in N loading, based on the area and
loading rate. The Low N loading is character-
ized as < 5,000 lb N/yr, Medium is 5,000 to
25,000 lb N/yr, and High is > 25,000 lb N/yr.
Note that the Table of Contents mis-labels
the units as Ib/acre/yr. This will be addressed
in future versions. In particular, we may find
that distinguishing developed lands accord-
ing to intensity (lb/acre/year) rather than total
load (lb/yr) would be more useful to decision
makers.
2.4 Assumptions and Limitations
The goals of broad applicability and ease of
use dictate a trade-off in the precision of the
N delivery estimates, while still incorporating
many complex biogeochemical and hydrologic
relationships otherwise unavailable to deci-
sion makers (Figure 1; Section 2.3). N-Sink
works as a prioritization and visualization
tool, which enables users to understand how
N moves in a given watershed and investigate
the relative N-related impacts of various land
use scenarios. As such, the numeric outputs of
the tool focus on percent removal of nitrogen
from source to receiving water and loading
estimates suitable for planning purposes, such
as comparing N loadings from different loca-
tions or types of development. However, these
estimates are not accurate enough to use in any
applications where the quantity of N is being
tracked, for instance, in documenting compli-
ance with a TMDL.
Assumptions and caveats for N-Sink calcula-
tions include:
• For the sake of simplicity and ease of use
N-Sink does not incorporate all aspects of
watershed nitrogen dynamics. Rather, it
focuses on landscape nitrogen sinks and
their spatial relationship to nitrogen sources
with the goal of informing the decision-
making process at the local level.
• N-Sink assumes that groundwater flow
paths are similar to surface water flow
paths.
• Landscape sinks are currently limited to
riparian wetlands (as characterized by veg-
etated hydric soils), lakes/ponds, and stream
reaches, and do not include site-specific
BMP's, with the exception of restored ripar-
ian wetlands.
• Flow paths are characterized by a terrestrial
component and a surface water component.
If the terrestrial portion of a flow path
intercepts vegetated hydric soils, then N
removal will be estimated from the wetland.
If, however, the surface water portion flows
through a wetland, removal is estimated
based on surface water characteristics, not
as a wetland.
• The surface water component of a flow path
may be initiated by the flow path entering
a stream midway along a stream reach.
Currently N-Sink estimates N removal
based on the entire stream reach, as if the
flow path had entered the stream at the
upgradient end of the reach. Because N
removal estimates within stream reaches
are usually low, this is not likely to cause
significant error. It may be more of an
issue in headwater streams that would have
slower flow and longer retention times
and therefore higher removal estimates.
However, future work plans include adjust-
ing N removal from the first stream reach
encountered based on where the flow path
entered that stream reach.
• Users are strongly encouraged to visit
areas under consideration to enhance their
understanding of the system and confirm the
-------
viability of any important landscape sinks
identified with N-Sink, such as riparian
wetlands.
2.5 Tool Description
The beta version of N-Sink is now available at:
www. edc. uri. edu/nsinkv2/
A basic web portal for the tool, including
background information on the project and
instructions on using the tool, has been created
at:
htto: clear, uconn. edu/proiects/nsink
(Figure 2):
Currently, the following functions are available
on the web-based tool.
1. Basemaps: Topographic, States, Imagery,
and Neutral Gray.
2. Data layers:
a. Nitrogen removal (%) of landscape
sinks, color coded.
b. HUC121 Watersheds, with flow
direction arrows and outlet point.
c. NHDPlus hydrography, including flow
lines (terrestrial and surface water),
water bodies, stream reaches, and
catchments (sub-basins).
d. Land Cover, using NLCD 2006 data.
3. Pop-up boxes with the ability to zoom
to the selected feature and obtain more
detailed information, including (by
feature):
a. HUC12s: Land cover summary as
pie chart, with numeric information
available when hovering the mouse
over pie slices.
b. Stream reaches: Stream name, reach
length (km), local drainage area (km2),
cumulative drainage area (km2), flow
1 HUC12 refers to Hydrologic Units with 12 digit codes, as
defined by USGS.
rate (m3/s), and estimated N removal
(%) (Figure 2).
c. Ponds/Lakes: Pond name, pond
area (m2), local drainage area (km2),
cumulative drainage area (km2), flow
rate (m3/s), estimated N removal (%)
(Figure 3).
4. N removal tool (Figures 4 & 5,
Appendices A & B)
a. The user clicks on the blue dot
(Figure 4) and then clicks on a location
on the map. The tool then generates a
flow path from that location allowing
the user to see the N sinks that the flow
path encounters and to calculate the
relative percentage of source N that
is removed by these sinks (Figure 5).
Currently, the tool uses an arbitrary N
source value of 100 for this calculation.
The team continues to work on the
interface for this tool - the Flex viewer
software currently has many limitations
on output design.
-------
N-Sink: Tracking Nitrogen in the Environment
A collaborative project of the University of Rhode Island, University of Connecticut and the U.S. EPA
Home N-Sink Tool About Aquatic Ecosystems
Contact CLEAR UConn
Go to N-Sink tool
About N-Sink
Help with the Tool
N and Aquatic Systems
N-Sink Partners
About the N-Sink Project
University of Rhode Island
Coastal Institute
The mission of the Coastal Institute Is
to advance knowledge and develop
solutions to environmental problems In
coastal ecosystems.
University of Connecticut Center
for Land Use Education and
Research
CLEAR provides Information, education
and assistance to Connecticut's land
use decision makers, community
organizations and citizens on how to
better protect natural resources while
accommodating economic growth.
U.S. EPA Office of Research and
Development
EPA is forging a path forward to develop
sustainable solutions to the nation's
highest priority science needs.
"N-Sink" is a cutting-edge web tool, usable by anyone with just a little
familiarization.
The N-Sink tool was created to be a useful, easy tool for local land use
managers interested in exploring the relationship between land use and
nitrogen pollution in their waters. N-Sink uses the best available science
on land use/nitrogen interactions, plus widely available basic datasets for
waterway networks, soils and land use, to highlight major sources and
sinks of nitrogen within a watershed context.
Funding for N-Sink v/eb tool development provided by EPA Office of
Research and Development.
Nitrogen (N) is Increasingly being Identified as a pollutant of concern In both coastal and inland
waters. Because N generation has a direct relationship with land use, better management of N
needs to include land use planning and storm water runoff strategies.
Links
Home
Contact Us
CLEAR
CLEAR Staff
UConn CANR
UConn
Contact
Chet Arnold, Associate Director, CLEAR
Email: chester.arnold@uconn.edu
Phone: 860-345-5230
PO Box 70
1066 Saybrook Road
Haddam, CT 06438
lit*" Custom Search
THE
UNIVERSITY
OF RHODE ISLAND
COLLEGE OF
THE ENVIRONMENT
AND LIFE SCIENCES
COASTAL
INSTITUTE
UCONN
UNIVERSITY OP CON NECTICUT
ft CLEAR
Figure 2. Screen capture of N-Sink web portal, currently residing on the UConn CLEAR server.
-------
S
~ ~ ~ ~ ~ n ~
l« I I "
J * I « * .
llMli
5 ^55 &3fl
g CJGE'StCX-BESu)
I i £3 Qill|f|ll
Z
c o>
I I S3 3 |
v
cts
2
CTS
Q)
CD
CtJ
.C
|
5
o
,o
K TO
£ £
Cf5 03
JB
03
O)
r- TO
G .£
03 03
oT "c5
c o
§5
C qj
g.s
w I
w
S> ®
«
3 3
O "O
c c
-S
"ll
s-8
c «
° *
"O 5
03 -Q
-C 03
.g>£:
5 *<—
,e> o
lc
0)
¦c
o
O)
03
03 CL
03 "-
C 03
TO E
03 .£
"*¦"-¦
m w
"3 ?J)
"O —-
03 TO
^ C
TO «
3} 03~
1? TO
O) V
i ^
O 5=
« TO
>< £
O 03
03
& CD
3 03
a-S
o TO
a-fe
0 <13
.03 -S
a. is
5 =>
1 |
Uj o
CO
o
'_
3
OS
-------
*?
(
¦a
CD
CD
TO
C
03
55 cd
* g>
21
TO
£
TO
"O
CD
c
o *~
Q..C
-c
Q>
e
TO
C
~0
C
o
Cl
CD
W
3
-o
¦S?
W ~-
"§ 1
¦3 o
^3 -c
c w
P
I
Si H-~
O O
"O CM
a CD
-R
¦5?io
¦§>1
o
E
CD
-S
TO
c
o _
Q_ CO
-O E
® --S
o «
© ©
"© "o
w
c
TO
¦£
5
o
"O
O) m-
C TO
'1 |
o <0
-C CD
« o>
§
-Q
§-"6
4 §
O -ft
Q. TO
a o
E ™
I S
Uj TO
®
i*.
=J
O)
-------
1
"V
9 I
3 |
* S
u w
5 p o
0) J »
§ o I
III
^ 05
CC *- ®
5-oS
IT)
C
Q>
£
CD
a
CO
Q)
§ -s
-2jl
<5 co
5 °
~ ~o
E c
O 03
4s ^
-c -2
03 3
CL O
I III®
a a q a a a a
r," N tL
1,1 io | g
111* ¦ 'z
Ji liSilii
| I i £ 11111111
i 2 a c 3 ¦ • c xI I
||l§ el|ll^§"!ll f
¦g£i!:r-ciiESM(3czQ.nx s
MtEniis'ii-ee S
-X. z
.£ c
CO ?
Z £
ttch
"O
§
8;
0)
Q)
CO
I
CO
c
.o
¦G
.O
cts
o
Q)
£
CO
-o
J2
0
.c
1
-o
§-
§-
a
CD
f£
"o
-2
1
o
£
O)
a;
cc
S
c
o
0)
¦S
"o
£
% CO
3.
3
OJ
15
-------
2.6 Future N-Sink improvements
During this project period the team was able to
successfully migrate N-Sink from an out-of-
date prototype ArcMap desktop tool to a state-
of-the-art ArcGIS for Flex web tool. However,
the programming work to accomplish this was
extensive, and as a result there are a number of
additions and improvements to the tool still to
be undertaken before we feel that it is ready to
use by the intended target audience. In addi-
tion, the downside of the web tool is that it is
not as customizable as a desktop tool, so more
work remains on crafting final user interface
items. A brief list of this unfinished agenda, in
rough order of priority, appears below:
a. N Loadings estimation tool.
This tool will allow the user to draw
a polygon and get land cover-based
loadings estimates for either existing
or assigned land cover. N-Sink
uses land cover, literature loading
estimates, and the particle tracking
tool to estimate N loadings, in
kg/ha/yr, of that polygon.
b. Automated N H D database updates.
Both CT and RI are due for major
National Hydrography Dataset
(NHD) upgrades in the next year.
In addition, NHD data is routinely
updated. We need to ensure
that N-Sink will automatically
incorporate these updates.
c. Watershed-wide N processing
intensity map
Currently, the N-Sink user can select
any point in the watershed to get
N reduction estimates. The project
team would like to have another,
watershed-wide tool output to relate
these issues. The output would be
a watershed map, colored to show
areas of high, medium, and low N
delivery to the outlet.
d. Website enhancements
As the tool is enhanced and
improved, so should the framework
provided by the overall website.
Future website improvements include
additional information on strategies
to reduce N loadings and examples
of ways to use N-Sink.
-------
3.0
Results: Two Case Studies
17
3.1 Study area
The current version of N-Sink was developed
for the coastal watersheds of Rhode Island and
southeastern Connecticut, from Narragansett
Bay to the Connecticut River area (Figure 7).
The area comprises 19 watersheds at the
USGS classification of Hydrologic Unit
Code 12, which average about 40 square miles
in size, nationally (Seaber et al., 1994 ).
Our initial outputs are from the Niantic River
and Saugatucket River watersheds (Figure 8).
Both of these watersheds have a recognized N
pollution problem, a state approved watershed
management plan, and well-developed citizen
participation in watershed matters. The latter
criterion is particularly important, as the
planned next step for this project will be to
use N-Sink outputs as the basis for educational
programs for watershed stakeholders (see last
section).
We explored both watersheds using N-Sink,
looking at N removal rates for various parts of
the watershed, including sites in close prox-
imity with one another. The results of these
analyses are included as Appendix A (Niantic)
and Appendix B (Saugatucket); these comprise
the case studies. Background information on
the two watersheds appear below.
Q N-Sink
Tracking Nitrogen in the Environment
W. O
'Boston
Cambrrck
Florence
Mat lo p;
Gal skill
Scituate
Holbrook.
Palmer
Northbridge'
^Springfield
Roxborough
{ Mansfield'
Springfield
II I'm
'Vomers Stafford
Kingston)
Wmdsc
Raynham
Tolland'
East Windsor
Providence/
•Lakevill'
Canton
T orringtc
East Brooklyn
Hartford;
DU.tgitss
Litchfield
CONNECTICUT
Buzzards
Barringtc
"West
Wanvick
Glastonbury
'„Plainville
Plainfietil
Harwich
Plymouth
Hyannis
Greenwich
(Westport
iWappingers
fSouthingk
Hamptorr - colchesK
Waterbtiry
Middletomn
Glenham
laugatuck
Brb'dkfi'eld
Vineyard
,Southbury
flarragansett
Highland
r AIRFIELD
Ridgefield
Madison
Saybrook
Montrose
Iridg'eport
•Canaan Westport
Spring
Valley
Southport
Scars dale Rye.
Bayville
impton
Iblbrook
'Jersev/
York
Figure 7. Study area in eastern CT and Rl coast.
-------
18
Hebron
1- _
NEW L-ONdSn"
AV.ySW UJCfi
Volunlownv
Preston
City/
Norwich
\Hope
Valley
Salem
Conanicut
Island
OldLyme
Old Saybrook
Fishers
I slur: a
N-Sirik
Tracking Nitrogen in the Environment
About
N-Sink
Tracking Nitrogen in the Environment
Mo os Up
Warwick'
Sterling
Scotland
Canterbury
Warwick
Plainfleld
Windham
>87 m
•lOVlMl
North j
WAsmF-iGii'ij
Beach Pond
Sims Part
Voluntowh
Norwich
N ewpo
Old Lyme
Old Saybrook
y
Figure 8. Close-up of study area showing Case Study watersheds highlighted in blue: Niantic River
(top) and Saugatucket River (bottom).
-------
3.2 Case Study #1: The Niantic River
Watershed, CT
The Niantic River watershed is a Hydrologic
Unit Code (HUC) 12 drainage basin in
Southeastern Connecticut that covers
31.3 square miles, or approximately 20,000
acres. It is comprised of three sub regional
basins: Latimer Brook, Oil Mill Brook, and
the main stem Niantic River (CT DEEP,
2009) (Figure 9). The watershed includes
portions of four municipalities, including large
portions of the downstream towns of East
Lyme and Waterford, and smaller portions
of the upstream communities of Salem and
Montville, UConn CLEAR's 2010 30-meter
land cover data for the watershed breaks down
as follows for the major categories: 13.5%
developed; 4.5% turf and grass; 3.3% agricul-
ture; 61.1% forest; 4.6% wetland.
Salem
Barnes V£
Reservoir,/
Montville
'Bogue'jBmak^
,ReserVpir^Ti
Latimer
gf Brook
Watershed
Oil Mill Broo®
Watershed
Golden
5K
Niantic River
Watershed
feterford
New Londor
Oswegatehie
V Hills
Amtrak Railroad
Bridge
DEPT OF ENVIRONMENTAL PROTECTION
"MTreRpfoHHECTCUT
NIANTIC RIVER WATERSHED
MANAGEMENT PLAN
NIANTIC RIVER WATERSHED I
AND MAIN TRIBUTARIES \
Mago Point
1314-001
Niantic
Kleinschmidt
Figure 9. Niantic River watershed, showing towns and subbasins. From the watershed plan, CT DEER
19
-------
The River is on the state's impaired waters
list for a variety of reasons, nitrogen pollution
being chief among them. The Niantic River
Watershed Management Plan (CT DEEP,
2009) summarizes (emphasis added):
The Niantic River does not currently meet
state water quality standards because
of high levels of indicator bacteria and
observed degradation of aquatic life.
According to the State of Connecticut's
§303(d) List of Impaired Waters, the
Niantic River is not supporting activities
such as shellfishing and swimming; the
Niantic River's shellfish beds are closed
after rain events of one inch or more. The
§303(d) List of Impaired Waters states
that the water quality of the Niantic River
is not supporting the aquatic life known
to inhabit the estuary. Symptoms of this
condition include, algal blooms, sea-
sonal variations in eelgrass populations,
loss of scallop populations and changes
to the fish communities. These ecologi-
cal changes are thought to be linked to
excessive nutrients, especially nitrogen,
entering the river.
In addition, the Niantic is part of a much larger
area (incorporating almost all of Connecticut)
that is covered by a nitrogen TMDL approved
for Long Island Sound, where nitrogen has
been identified as a major cause of harmful
algal blooms leading to hypoxia (US EPA,
2009).
A citizen Board of Directors has been formed
to oversee the implementation of the state-
approved Watershed Management Plan. In
addition, there is a Niantic River Watershed
Nitrogen Work Group made up of techni-
cal and scientific representatives that meets
regularly to discuss monitoring challenges
and knowledge gaps pertaining to managing
nitrogen in the Niantic system.
We looked at N reduction/delivery from sev-
eral places in the watershed. Some examples
were placed near each other, to explore
changes in N delivery estimates based on small
but important changes in the flow path. Please
see Appendix A.
3.3 Case Study #2: The Saugatucket
River Watershed, Rl
The Saugatucket River watershed is a
Hydrologic Unit Code (HUC) 12 drain-
age basin in southern Rhode Island that
covers 17.2 square miles, or approximately
11,000 acres (Figure 10). The Saugatucket
drains into the northern end of Point Judith
Pond, a coastal salt pond with a perma-
nent breachway to Block Island Sound.
Approximately 13% of the watershed is in low-
intensity development, 8% in medium and 2%
in high intensity development. Agricultural
activities, such as pasture and hay cultivation,
are 4% of the watershed. Roughly 40% of
the watershed is forested, and just under 17%
is wetlands. Most of the watershed (83%)
lies within South Kingstown, 16% in North
Kingstown, and slivers of the watershed lie
within the towns of Narragansett and Exeter,
RI.
Water quality issues include both the river
itself and its receiving waters, Point Judith
Pond. Point Judith Pond is the site of restora-
tion projects focusing on both eelgrass and
scallops. The pond also supports oyster aqua-
culture. The Saugatucket River is on the state's
303(d) List of Impaired Waterbodies for fecal
col iform, with stormwater runoff identified
as a major contributor. Although bacteria has
been the primary pollution concern, several
of the watershed's ponds have been found to
have low dissolved oxygen levels due to high
ammonia and nitrate loadings (Rhode Island
Rivers Council, 2013). As a coastal salt pond,
Point Judith Pond is sensitive to nitrogen
loading from the Saugatucket, as well as from
development surrounding the pond.
The watershed has a nonprofit organization,
the Saugatucket River Heritage Corridor, that
-------
Watershed Sub-basins
&
Surface Water
Prtmwy Rd«i»
RlWfS
| Ponds
¦basins
1 3#tT JUDITH POND SUB-BASIN
SAUGATUCKET RIVER SUB-BASIN
SaugaludwIRIvarWatwfshad " 61?
4
M-r-
Figure 10, Saugatucket River watershed. The HUC-12 watershed explored by N-Sink consists of the
upper half (blue) of the overall watershed. From Rl CIS center.
was established in 1994 and designated by the
Rhode Island Rivers Council in 1999 as the
watershed council for the Saugatucket River.
This group works in partnership with the Salt
Ponds Coalition on Point Judith Pond.
We looked at N reduction/delivery from several
places in the watershed. Some examples were
placed near each other, to explore changes
in N delivery estimates based on small but
important changes in the flow path. Please see
Appendix B.
3.4 QA and Model Verification:
Comparing N-Sink N Reduction Tool
Results to Conventional Calculations.
All research projects making conclusions or
recommendations based on environmentally
related measurements and information and
funded by the Environmental Protection
Agency are required to comply with the
requirements of the Agency Quality Assurance
Program. This model was developed under an
approved Quality Assurance Project Plan.
In order to confirm that the formulas were
coded correctly in the N Reduction Tool, we
compared the N removal estimates along
several flow paths in the Saugatucket watershed
with the removal estimates calculated with the
formulas in Excel. Typically model valida-
tion also involves comparing model estimates
with field data. However, because N-Sink is a
decision support tool and does not attempt to
include all components of N cycling in a water-
shed, it is not possible to compare N-Sink esti-
mates with any existing field data. For exam-
ple, nitrate concentrations of samples taken at
a watershed outlet reflect the entire spectrum
-------
of N cycling over a period of time throughout
the entire watershed. It would be impossible
to tease out the effects of one stream reach or
pond without designing a field study for this
purpose. The equations used in N-Sink are
empirical and the distillation of many field
studies, reflecting our current understanding of
N removal in these landscape sinks (Kellogg
et al., 2010). However, work is currently
underway at the URI Watershed Hydrology
Lab to explore N removal in the landscape
sinks identified in N-Sink, especially streams
and ponds. This work will help us compare
N-Sink N removal estimates with field data
from studies designed for this purpose. We
are also exploring the possibility of extracting
the relevant estimates of N movement from a
calibrated and validated watershed model such
as SWAT. This would not be a direct compari-
son to field data, but has potential to enhance
our assessment of N-Sink performance.
3.4.1 Methods
The N Removal Tool in N-Sink was used to
create a flow path and calculate removal along
each reach of that flow path. A table is gener-
ated and shown in a pop-up box. In order to
save the information for each flow path gener-
ated, the table was copied using the clipboard
icon, then pasted into an Excel spreadsheet.
Each new flow path calculation clears the
information from the previous calculation
[Note: the next iteration of N-Sink should have
the ability to save previous flow paths and their
associated calculations],
A separate Excel spreadsheet was created to
do the calculations using the formulas detailed
in Section 2.3 and Kellogg et al. (2010). We
focused on the calculations for % removal
from stream reaches and ponds because those
are more complex, while the wetland is a
check on presence/absence. We gathered the
necessary GIS data to be plugged into the
equations by clicking on each sink (pond
or stream reach) and recording data such as
reach length (km), pond area (m2), cumulative
drainage area (km2), and average annual flow
rate (cms). The stream removal calculation in
the spreadsheet used an estimated velocity gen-
erated from the equation presented in Section
2.3, while the stream removal calculation in
N-Sink used the estimated velocity provided
by NHDPlus v2. [Note: When this comparison
was done, the NHDPlus velocities were not
available to us from the pop-up information
box for each reach. This information will be
added in future and another comparison will be
done using the velocity provided by NHDPlus
v2.] The original intent was not to compare
the velocity provided by NHDPlus v2 to that
estimated using the equation in Section 2.3.
However, circumstances were such that this
was how the comparison had to be made.
3.4.2 Comparison Results
Tables 4.1 to 4.4 show % N removals along
each of the four flow paths in the four exam-
ples generated for the Saugatucket watershed.
The absolute difference in the cumulative
removal along the flow path ranges from -4%
to +1% N removal. The absolute difference
along reaches ranges from -4.9% to +1.8%
N removal. Given that the velocity estimates
were derived from two different sources,
the difference is within acceptable limits.
Uncertainty is a component of the calculations
that has not been addressed to date. This is
something we will explore in future develop-
ment, asking how to incorporate uncertainty
into the calculations and whether to communi-
cate uncertainty to the user, and if so, how best
to do that.
-------
Table 4.1 Comparison of % N removal by reach along flow path shown in Saugatucket Example A.
Note that the difference in cumulative removal (%) is not the sum of the differences along each
reach because each % removal is applied to the output from the previous reach.
Example A
% N Removal
NHD Reach Code Feature Type
N-Sink
Excel Spreadsheet Difference
Null
Null
1090005000333
1090005000333
1090005000332
1090005000331
1090005000330
1090005000734
1090005000321
1090005000320
1090005000319
Cumulative
Non-hydric
Wetland
Pond
Stream
Stream
Stream
Stream
Pond
Stream
Stream
Stream
0
80
9.3
6.9
3.6
0.3
0.6
8.7
0.7
0.2
2.9
86
0
80
9.6
4.3
2.2
0.3
0.5
8.8
0.3
0.2
2.4
85
+0.3
-2.6
-1.4
0.0
-0.1
+0.1
+0.4
0.0
-0.5
1
Table 4.2 Comparison of % N removal by reach along flow path shown in Saugatucket Example B.
Note that the difference in cumulative removal (%) is not the sum of the differences along each
reach because each % removal is applied to the output from the previous reach.
Example B
% N Removal
NHD Reach Code Feature Type
N-Sink
Excel Spreadsheet Difference
Null
1090005000340
1090005000331
1090005000330
1090005000734
1090005000321
1090005000320
1090005000319
Cumulative
Non-hydric
Stream
Stream
Stream
Pond
Stream
Stream
Stream
0
15.6
0.3
0.6
8.7
0.7
0.2
2.9
26
0
10.7
0.3
0.5
8.8
0.3
0.2
2.4
22
-4.9
0.0
-0.1
+0.1
+0.4
0.0
-0.5
-------
Table 4.3 Comparison of % N removal by reach along flow path shown in Saugatucket Example C.
Note that the difference in cumulative removal (%) is not the sum of the differences along each
reach because each % removal is applied to the output from the previous reach.
Example C
% N Removal
MM) Reach Code
Feature Type
N-Sink
Excel Spreadsheet
Difference
Null
Non-hydric
0
0
Null
Non-hydric
0
0
1090005000346
Stream
16.0
13.5
-2.5
1090005000347
Pond
35.0
36.8
+1.8
1090005004979
Stream
3.5
2.6
-0.9
1090005004973
Pond
0.4
0.3
-0.1
1090005004972
Stream
1.7
1.3
-0.4
1090005000342
Stream
3.1
4.1
+1.0
1090005000320
Stream
0.2
0.2
0.0
1090005000319
Stream
2.9
2.4
-0.5
Cumulative
52
51
-1
Table 4,4 Comparison of % N removal by reach along flow path shown in Saugatucket Example D.
Note that the difference in cumulative removal (%) is not the sum of the differences along each
reach because each % removal is applied to the output from the previous reach.
Example D
% N Removal
N H D Reach Code Feature Type
N-SInk
Excel Spreadsheet Difference
Null
Null
1090005000346
1090005000347
1090005004979
1090005004973
1090005004972
1090005000342
1090005000320
1090005000319
Cumulative
Non-hydric
Wetland
Stream
Pond
Stream
Pond
Stream
Stream
Stream
Stream
0
80
16.0
35.0
3.5
0.4
1.7
3.1
0.2
2.9
90
0
80
13.5
36.8
2.6
0.3
1.3
4.1
0.2
2.4
90
0
-2.5
+1.8
-0.9
-0.1
-0.4
+1.0
0.0
-0.5
0
-------
4.0
Next Steps
The primary purpose of this work is to incor-
porate scientific understanding of landscape N
sinks (freshwater wetlands, streams and ponds/
lakes) into practices that local land-use deci-
sion makers can adopt and act upon to support
sustainable and healthy communities. Work
to this point has focused on construction of
a research based, accessible tool with simple
outputs that the project team has judged to be
useful, based on their extensive experience in
working with community decision makers.
As noted, the migration to a web tool format is
a tremendous step toward our goal of acces-
sibility, but it has left the team with more work
to do. So, our next steps fall into three basic
categories:
1. Improvements and enhancements
to the tool and its interface, as listed
in Section 2.6. Creating the loading
estimation tool is the highest priority.
2. Testing the model in the field, working
with decision makers and community
groups in the Ni antic and Pawcatuck
watersheds. This will be accomplished
by an expanded project team that
includes Cooperative Extension and
Sea Grant Extension faculty from both
UConn and URL Extension programs
across the U.S., implemented through
Land Grant and Sea Grant University
Systems, use an integrated research and
extension approach to educating land use
decision makers. These programs are
ideal vehicles for reaching this critical
audience (Arnold, 2000). Both UConn
and URL are members of the National
"NEMO" Network of projects focused
on assisting communities protect their
natural resources. NEMO, which stands
for "Nonpoint Education for Municipal
Officials," has programs in 30 states
(Rozum and Arnold 2004; Dickson
and Arnold 2009); the UConn and URL
programs are the two oldest of these
programs, dating back more than 20 years.
3. Evaluating the opportunities and
barriers to adapting N-Sink nationwide.
As noted, N-Sink was intentionally
developed using nationally available data
for its primary inputs, with an eye to
adaptability. Despite this fact, regional
climatic and hydrologic differences exist
to the point that "one size" definitely
does not fit all. In addition, the concept
for the prototype N-Sink tools was as an
ArcGIS extension rather than a web tool,
and the effects of this change on national
dissemination need to be evaluated. If, as
seems likely, it is found that N-Sink can
be adapted nationally, then the project
team will use its contacts in various
national networks to assess interest and
get user feed back. These networks
include:
• The NEMO Network, as described.
• The National Geospatial Technology
Extension Network that teaches and
facilitates the use of new and advanced
geospatial technologies;
• The Sea Grant Sustainable Coastal
Community Development Network
that shares resources and tools to help
coastal communities thrive.
• The Land Grant Water Program
Network of research and outreach proj-
ects focused on water resource protec-
tion in each of the 50 states.
25
-------
5.0
References
Alexander, R.B., R.A. Smith, and G.E.
Schwarz. 2000. Effect of stream channel
size on the delivery of nitrogen to the Gulf
of Mexico. Nature 403:758-761.
Alexander, R.B., E.W. Boyer, R.A. Smith, G.E.
Schwarz, and R.B. Moore. 2007. The role
of headwater streams in downstream water
quality. Journal of the American Water
Resources Association 43:41 -59
Armstrong, M.R, RJ. Densham, R Lolonis, and
G. Rushton. 1992. Cartographic displays
to support locational decision making.
Cartography and Geographic Information
Systems 19(3): 154-164.
Arnold, C.L. 1999. Educating municipalities:
Land use is the key to managing nonpoint
source pollution. Water Environment and
Technology ll(l):41-44.
Arnold, C. L. 2000. Land use is the issue,
but is Land Grant the answer? Journal
of Extension 38(6), December 2000.
Published online at: http://www.ioe.org/
ioe/2000decemher/comml. html
Arnold, C.L., D.L. Civco, M.R Prisloe, J.D.
Hurd, and J.W. Stocker. 2000. Remote
Sensing-Enhanced Outreach Education
as a Decision Support System for Local
Land Use Officials. Photogrammetric
Engineering and Remote Sensing 66(10):
1251-1260.
Beulac, M.N. and K.H. Reckhow. 1982.
An examination of land use - nutri-
ent export relationships. Journal of the
American Water Resources Association
18(6): 1013-1024.
Brown, I. 2006. Modeling future landscape
change on coastal floodplains using
rule-based GIS. Environmental Modeling
& Software 21(10): 1479-1490. (Abstract
at: http://www.sciencedirect.com/science/
article/pii/Sl364815205001544)
Carpenter, S.R., N.F. Caraco, D.L. Correll,
R.W. Howarth, A.N. Sharpley, and V.H.
Smith. 1998. Nonpoint pollution of surface
waters with phosphorus and nitrogen.
Issues in Ecology, Volume 3. Ecological
Society of America. 12 pp.
Carver, S.J. 1991. Integrating multi-criteria
evaluation with geographical informa-
tion systems. International Journal
of Geographical Information Systems
5(3):321-339.
Clausen, J. C., K. Guillard, C. M. Sigmund
and K. M. Dors. 2000. Water Quality
Changes from Riparian Buffer Restoration
in Connecticut. Journal of Environmental
Quality 29(6): 1751-1761
Conley, D.J, H. W. Paerl, R. W. Howarth, D. F.
Boesch, S. P. Seitzinger, K. E. Havens, C.
Lancelot, G. E. Liken. 2009. Controlling
Eutrophication: Nitrogen and Phosphorus.
Science. 323:1014-1015.
Connecticut Department of Energy and
Environmental Protection. 2009. Niantic
River Watershed Protection Plan Guided
Summary. 34 pp. http://www.ct. gov/deep/
lib deen water watershed management
wm plans niantic management plan
summarv.pdf
Cova, T. J. and R. L. Church 1997. Modeling
community evacuation vulnerability using
GIS. International Journal of Geographical
Information Science ll(8):763-784.
-------
David, M B., L.G. Wall, T V. Royer, and J.L.
Tank. 2006. Denitrification and the nitro-
gen budget of a reservoir in an agricul-
tural landscape. Ecological Applications
16(6):2177-2190.
Davis, T. J. and C. P. Keller. 1997. Modeling
uncertainty in natural resource analy-
sis using fuzzy sets and Monte Carlo
simulation: Slope stability prediction.
International Journal of Geographical
Information Science ll(5):409-434.
Dickson, D. and C. Arnold. 2009. A Catalyst
for Community Land Use Change:
National NEMO Network 2008 progress
report. Publication 030930.1 of the
University of Connecticut Center for Land
Use Education and Research. 38p.
Dickson, D.W., C.B. Chadwick and C.L.
Arnold. 2011. National LID Atlas:
A Collaborative Online Database of
Innovative Stormwater Management
Practices. Marine Technology Soc. Jnl
45(2): 59-64.
Gold, A.J., W.R. DeRagon, W.M. Sullivan,
and J.L. Lemunyon. 1990. Nitrate-nitrogen
losses to groundwater from rural and
suburban land uses. J. Soil Water Conserv.
45:305-310.
Gold, A.J., P.M. Groffman, K. Addy, D.Q.
Kellogg, M. Stolt, and A.E. Rosenblatt.
2001. Landscape attributes as controls
on ground water nitrate removal capac-
ity of riparian zones. Journal of the
American Water Resources Association
27:1457-1464.
Gold, A.J. and L. Joubert. 2007. Small
town science: Constraints and successes
in ecosystem management. Emerging
Issues Along Urban-Rural Interfaces II
Proceedings. 5pp.
Groffman, Peter M., D. J. Bain, L.E. Band,
K.T. Belt, G.S. Brush, J.M. Grove, R.V.
Pouyat, I.C. Yesilonis, and W.C. Zipperer.
2003. Down by the riverside: urban ripar-
ian ecology. Frontiers of Ecology and the
Environment 1:315-321.
Harrison, J.A., R.J. Maranger, R.B. Alexander,
A.E. Giblin, P.-A. Jacinthe, E. Mayorga,
S.P Seitzinger, D.J. Sobota, and W.M.
Wollheim. 2009. The regional and global
significance of nitrogen removal in
lakes and reservoirs. Biogeochemistry
93:143-157.
Kaushal, S.S., P.M. Groffman, P.M. Mayer,
E. Striz, and A.J. Gold. 2008. Effects of
stream restoration on denitrification in
an urbanizing watersheds. Ecological
Applications 18(3): 789-804.
Kellogg, D.Q., A.J. Gold, S. Cox, K. Addy,
and P. V. August. 2010. Ageospatial
approach for assessing denitrification sinks
within lower-order catchments. Ecological
Engineering 36:1596-1606.
Kellogg, D.Q., A.J. Gold, A. Wentz, and M.
Shimizu. 2011. N-Sink: AGIS-based
Decision Support Tool to Assist in
Prioritizing Watershed Nitrogen Source
Controls and Management of Sinks. Alpha
Version 1.0. Report to EPA.
Lant, C.L., S.E. Kraft, J. Beaulieu, D. Bennett,
T. Loftus, J. Nicklow. 2005. Using GIS-
based ecological-economic modeling to
evaluate policies affecting agricultural
watersheds. Ecological Economics
55:467-484. (available at: http://www.
sciencedirect.com/science/article/pii/
S0921800904004446)
Last, D.G. 1995. Incremental land-use deci-
sion making displayed by county zoning
committees. Journal of Soil and Water
Conservation 50:21-24.
Leopold, L.B. and T. Maddock, Jr. 1953. The
hydraulic geometry of stream channels
and some physiographic implications.
27
-------
Geological Survey Professional Paper 252.
U.S. Geological Survey, Washington, DC.
Lowrance, R., L.S. Altier, J.D. Newbold, R.R.
Schnabel, P.M. Groffman, J.M. Denver,
D.L. Correll, J.W. Gilliam, J.L. Robinson,
R.B. Brinsfield, K.W. Staver, W. Lucas, and
A.H. Todd. 1997. Water quality functions
of riparian forest buffers in Chesapeake
Bay watersheds. Environmental
Management 21(5):687-712.
Maine Department of Environmental
Protection. 2010. Total Maximum Daily
Load (TMDL) Report Prestile Stream (&
Christina Reservoir) Aroostook County,
Maine
Massachusetts Department of Environmental
Protection. 2008. Inter-municipal
Watershed Planning & TMDL
Implementation to Restore Embayment
Water Quality on Cape Cod: Three
Case Studies of Towns Sharing Coastal
Watersheds.
Mayer, P.M., S.K. Reynolds Jr., M.D.
McCutchen, and T.J. Canfield. 2007.
Meta-analysis of nitrogen removal in
riparian buffers. Journal of Environmental
Quality 36:1172-1180.
McKay, L., T. Bondelid, T. Dewald,, et al.
2012. NHDPlus Version 2: User Guide.
Available at: ftp://ftp.horizon-systems.com/
NHDPlus NHDPlus V21/Documentation/
NHDPlus V2_ User_Guide.pdf
Merry, K.L., P. Bettinger, and W.G. Hubbard.
2008a. Back to the future Part 1:
Surveying geospatial technology needs
of Georgia land use planners. Journal
of Extension [On-line], 46(3) Article
3R1B6. Available at: http://www.ioe.org/
ioe/2008iune/rb 6p. shtml
Merry, K.L., P. Bettinger, and W.G. Hubbard.
2008b. Back to the future Part 1:
Surveying geospatial technology needs of
Georgia Agriculture and Natural Resources
Extension Professionals. Journal of
Extension [On-line], 46(4) Article
4RIB1. Available at: http://www.i0e.0r2/
ioe/2008auzust/rb 1 php
Moore, R.B., C.M. Johnston, K.W. Robinson,
and J.R. Deacon. 2004. Estimation of
total nitrogen and phosphorus in New
England streams using spatially refer-
enced regression models. USGS Scientific
Investigations Report 2004-5012. U.S.
Geological Survey, Pembroke, NH.
National Research Council (NRC). 2000.
Clean coastal waters: Understanding and
reducing the effects of nutrient pollution.
National Academies Press.
New York Department of Environmental
Conservation and Connecticut Department
of Environmental Protection. 2000. A total
maximum daily load analysis to achieve
water quality standards for dissolved
oxygen in Long Island Sound. 57pp.
Newbold, J.D., J.W. Elwood, R.V. O'Neill,
W. Van Winkle. 1981. Measuring nutrient
spirallng in streams. Can. J. Fisheries Aqu.
Sci. 38:860-863.
Nixon, S. W. 1995. Coastal marine eutrophica-
tion: A definition, social causes and future
concerns. Ophelia, International Journal of
Marine Biology 41:199-219.
Nixon, S. W., B. Buckley, S. Granger,
and J. Bintz. 2001. Responses of Very
Shallow Marine Ecosystems to Nutrient
Enrichment. Human and Ecological Risk
Assessment 7:1457-1481.
North Carolina Department of Environment
and Natural Resources. 1999. Total
Maximum Daily Load for Total Nitrogen to
the NeuseRiver Estuary, NC; 1999
Nowicki, B. and A. J. Gold. 2008. Nutrient
transport in groundwater at the coastal
-------
margin. In: Desbonnet, A. Costal-Pierce,
B.A. (Eds), Science foe Ecosystem-based
Management: Narragansett Bay in the 21st
Century. Springer Series in Environmental
Management, New York NY. Pages
67-100.
NLCD 2006. Fry, J., G. Xian, S. Jin, J. Dewitz,
C. Homer, L. Yang, C. Barnes, N. Herold,
and J. Wickham, 2011. Completion of the
2006 National Land Cover Database for
the Conterminous United States, PE&RS,
Vol. 77(9):858-864.
Osmond, D.L., D.A. Crouse, S.C. Hodges,
A.M. Johnson, and K.J. May. 2008.
NCANAT: North Carolina Agricultural
Nutrient Assessment Tool. Version 2.0. NC
State University, Raleigh, NC. Available
at http://nutrients, soil, ncsu. edu/ncanat/
user manual, htm
Oviatt, C., P. Doering, B. Nowicki, L. Reed,
J. Cole, J. Frithsen. 1995. An ecosystem
level experiment on nutrient limitation
in temperate coastal marine environ-
ment. Marine Ecology Progress Series
116:171-179
Rhode Island Department of Environmental
Management. 2010. Total Maximum Daily
Load Analysis for the Pawcatuck River and
Little Narragansett Bay Waters.
Rhode Island Rivers Council website, accessed
6/20/13, http://www, ririvers. ore/index, htm
Royer, T.V., J.L. Tank, M.B. David. 2004.
Transport and fate of nitrate in headwater
agricultural streams in Illinois. J. Environ.
Qual. 33:1296-1304.
Rozum, J. and C. Arnold. 2004. Putting
Communities in Charge. Publication
041215.1 of the University of Connecticut
Center for Land Use Education and
Research. 35 pp.
Rozum, J.S., E. Wilson, and C. Arnold.
2005. Strengthening integration of
land use research and outreach through
innovative web technology. Journal of
Extension [On-line], 43(5) Article 5IAW1.
Available online at: http://www.ioe.org/
ioe/2005october/iw 1 php
Seaber, PR., F.P Kapinos and G.L. Knapp.
1994. Hydrologic Unit Maps. U.S.
Geological Survey Water-Supply Paper
#2294. 66pp.
Seitzinger, S.P, R.V. Styles, E.W. Boyer,
R.B. Alexander, G. Billen, R.W. Howarth,
B. Mayer, and N. Van Breemen. 2002.
Nitrogen retention in rivers: model devel-
opment and application to watersheds in
the northeastern U.S.A. Biogeochemistry
57/58:199-237.
Seitzinger, S., J.A. Harrison, J.K. Bohlke,
A.F. Bouwman, R. Lowrance, B. Peterson,
C. Tobias, and G. Van Drecht. 2006.
Denitrification across landscapes and
waterscapes: a synthesis. Ecological
Applications 16:2064-2090.
Short, F. T., and D. M. Burdick. 1996.
Quantifying Eelgrass Habitat Loss in
Relation to Housing Development and
Nitrogen Loading in Waquoit Bay,
Massachusetts. Estuaries 19:730-739.
Stow, C. A., M.E Borsuk, and K.H. Reckhow.
2011. Nitrogen TMDL development in
the Neuse River watershed: An impera-
tive for adaptive management. Journal
of Contemporary Water Research and
Education 122:16-26.
Thumerer T., A. P. Jones, and D. Brown. 2000.
A GIS based coastal management system
for climate change associated flood risk
assessment on the east coast of England.
International Journal of Geographical
Information Science 14:265-281.
29
-------
Tulloch, D.L., J.R. Myers, J.E. Hasse, P.J.
Parks, and R.G. Lathrop. 2003. Integrating
GIS into farmland preservation policy
decision making. Landscape and Urban
Planning 63(l):33-48.
(Abstract at: http://www.sciencedirect.com/
science/article/pii/SO169204602001810)
U.S. Environmental Protection Agency (EPA).
2002. On site wastewater treatment sys-
tems manual. Report EPA/625/R-00/008.
http://www. epa. gov/owm/septic/pubs/
septic 2002 osdm all.odf
U.S. Environmental Protection Agency
(EPA). 2009. Restoring the Long Island
Sound While Saving Money: Lessons
in Innovation and Collaboration. EPA
publication 841-F-09-002C. 6 pages.
U.S. EPA Science Advisory Board. 2011.
Reactive Nitrogen in the United States: An
Analysis of Inputs, Flows, Consequences
and Management Options. A Report of
the EPA Science Advisory Board (EPA-
SAB-11-013) http://www.epa.gov/sab
U.S. EPA Science Advisory Board. 2012.
Science Integration for Decision Making at
the U.S. Environmental Protection Agency
(EPA-SAB-12-008) http://www.epa.gov/sab
Valiela, I., J. Costa, K. Foreman, J. M. Teal, B.
Howes, and D. Aubrey. 1990. Transport
of groundwater-borne nutrients from water-
sheds and their effects on coastal waters.
Biogeochem 10:177-197.
Wilson, Emily H., Juliana Barrett and Chester
L. Arnold. 2011. Land cover change in
the riparian corridors of Connecticut.
Watershed Science Bulletin Fall 2(1):
27-33.
30
-------
Appendix A. Case Study 1
Screen captures from N-Sink v2 (beta)
Niantic River Watershed, CT
http://www.edc.uri.edu/nsinkv2/
-------
Niantic River Watershed
About
ogen in the Environment
'tw iJondqn"
CA'l Sj, bloc*
1. Hydrography and hydric soils.
-------
Niantic River Watershed
33
¦ RJws and Slrsamg
Watersheds, HUC12
¦ Poor Points
P{ Row Dirocaon
D HUC12 Boundaries
Hydrography, NHDPlua
Q RowUnw
Lend Cover
PI NLCD. 2006
Topography
m Qevaaon
2. Flow direction.
3. Land Cover (NLCD 2006).
-------
Niantic River Watershed
NItrogsn Removal
rv Laices and Ponda
T/ Rivers and Streams
Watersheds, HUC12
¦» Pour Points
9 Row Direction
a HUC12Boundarfa»
Hydrography, NHDPlua
rj Row Unos
Land Cover
m NLCD, 2006
Topography
OevV.
Hopy-xd
StJtfl PjiK
. 'Fa*
Hop)'.*'!.
Goll Ciutf
J KWltlrfl
Rvmrvqt
tV Jlerli
4a. Hydrography coded for N removal of stream reaches and features.
Percent Removal, Lakes/Ponds
Percent Removal. Rivers/Streams
HUC12 Boundaries
HUC12 Boundary
4b. Hydrography coded for N removal of stream reaches and features: detail
34
-------
Niantic River Watershed
Nitrogen Removal
Q Lakes and Poods
Watersheds, HUC12
m, Pour Points
¦ Flow Direction
a HUC12 Boundaries
Hydrography, NHDPIus
a Row Lines
r, Wa»rt>odies
V) HyOrtc Soils
Ci Stream Reaches
m CaWvnents
Land Cover
n NLCD.2006
Topography
¦I BevaSon
Sotmtt
CIihIh
I voivlon
Ftthefi
Island
Club
F'thto
lilard
5a. Nitrogen calculator box opened (top left). Clicking on blue dot allows user to then select any location in
watershed.
£
r
Calculate Nitrogen _ 0
Soufoe Nirogen Vaiw too
Start RoW, I—I
MdP Ctear Subnet
5b. Close-up view of N calculation box. For beta version, N value is set to arbitrary input of 100 units.
35
-------
Niantic River Watershed
Hopyard,
GO# CUf
Clicslci
U'jtrili
6a. Example 1: Flow path noting input value of 100 units.
Hopy.nd
SIJlo Talk
How <*4
Gail CM
1
/
Rmervwi
Ivory Ion
6b. Example 1: Flow path noting output value of 48 units.
36
-------
Land LofigBi (t NHD Reach Co Nitrogen m Nitrogen Out Psrcent N Rwni Waiorway Typo
011000030002 20.00
011000030007 17.05
IMS 01
Isles Goll
Chfilti
Sloo» R."
Millar
Raserva
Ivoivlon
Niantic River Watershed
U*«OI
I •>- Goll
tD«%
Hflpyjirt
SljtoPark
Hopyvd.
Goll CIujf
Ch<»ltt
7a. Example 2: Flow path noting input value of 100 units.
7b. Example 2: Flow path noting output value of 16 units.
37
-------
Niantic River Watershed
lake 01
fox
Hopy-wl
Gotf Ciif
Chfslei
Si oo# Ror
l.liiii.'H t
RotervJ
vmi_-od9..,pk..
8a. Example 3: Flow path noting input value of 100 units.
'-•fox
Hopyara-
GtAICki/f
Chfsifi
IvorVton
8b. Example 3: Flow path noting output value of 92 units.
38
-------
Niantic River Watershed
Output^ 14 units. (86% N removal) Output = 70 units. (30% N removal)
10. Two closely placed selection points showing very different outputs. Both scenarios had arbitrary
100 unit input.
Output = 15 units. (85% N removal) Output= 44 units. (56% N removal)
10. Two closely placed selection points, close-up view, showing different outputs. Both scenarios had
arbitrary 100 unit input.
-------
Appendix B.
Case Study 2
Screen captures from N-Sink v2 (beta)
Saugatucket River Watershed, Rl
http://www.edc.uri.edu/nsinkv2/
N-Sink
Tracking Nitrogen in the Environment
Nitrogen Removal
Lakes and Poods
¦ Rivera and Streams
Watersheds, HUC12
¦ Pour Points
n HUC12 Boundaries
Hydrography, NHDPIus
D Rm Linos
"" Catchments
Land Cover
Topography
NKVV I.OND(5j.("
yHope
Vdley
CORiVlClA
IVarnl
pis hers
Island
Saugatucket River Watershed
40
-------
Saugatucket River Watershed
N-Sink
Tracking Nitrogen in the Environment
Nitrogen Removal
"" Rivera and Streams
Watersheds, HUC12
!¦ Pour Points
¦ Row Direction
D HUC12 Boundaries
Hydrography, NHDPIus
Q RowUnes
O Watertodles
B( Hydrlc Soils
i" Stream Reaches
Land Cover
Topography
ami dVst own,
Umvenlty
Ot Rhode
^/len-4,
l\.i ki'li
-------
Saugatucket River Watershed
N-Sink
Tracking Nitrogen in the Environment
Nitrogen Removal
"¦ Lakes and Ponds
i" Rivers and Streams
Watersheds, HUC12
"¦ Pour Points
Bj HUC12 Boundaries
Hydrography, NHDPIus
Row Lines
Q WaiortxxJies
D Hydrtc Soils
~ Stream Roaches
w Catchments
Land Cover
Q NLCD, 2006
Topography
/••i" ¦
3. Land Cover (NLCD 2006).
C if Q www.edc.uri.edu/nsinkv2/ ^ S
Center for Land Us. \jj Exchange - UITS Rj google maps - Coo €> Intellicait - tvoryto Q Kuall Portal Q NSINK ft CLEAR Blog . Log In ® C-CAP Land Cover El Coogle Analytics ft CLEAR Blog » Lj Other Bookmarks
N-Sink
Tracking Nitrogen in the Environment
About
miu^Mou-n
Rivers and Streams
Watersheds, HUC12
¦ Pour Points
• Flow DirccS on
- HUC12 Boundaries
Hydrography, NHDPIus
F/ Row Lines
Land Cover
¦ ML CO. 2006
Topography
4a. Hydrography coded for N removal of stream reaches arid features.
42
-------
Saugatucket River Watershed
N-Sink
Tracking Nitrogen in the Environment
Lakes and Ponds
Percent Removal. Lakes/Ponds
| Low: <=25%
| Medum: 26% - 50%
¦ High:» 50%
HUC12 Boundaries
HUC12 Boundary
4b. Hydrography coded for N removal of stream reaches and features: detail
N'Sink ^ k/ ©
Tracking Nitrogen in the Environment _ _
U» N0>
.'I Ptiodo,
)
5a. Nitrogen calculator box opened (top left). Clicking on blue dot allows user to then select any loca-
tion in watershed.
43
-------
Saugatucket River Watershed
! Jul ii iiX
¥
)
Calculate Nitrogen rt
Source Nirogen Value too
Start Point f 1
©
m Ctear SuWI
5b. Close-up view of N calculation box. For beta version, N value is set to arbitrary input of 100 units.
N-Sink
Tracking Nitrogen in the Environment
About
Dutch
Seaside
Beach
[u riven ity
Of RhoJu
KiiiiVt"ll|
Land Lent NHD Reach Code Nitrogen h Nitrogen Out Percent N Removal Waterway Type
01090005000319 1459
6a. Example A: Flow path from upper watershed. N removal to the ultimate coastal outlet is 86%.
44
-------
Saugatucket River Watershed
N-Sink
Tracking Nitrogen in the Environment
About
01090005000734 16.13
6b. Example A: Close-up.
N-Sink e ^ O
Tracking Nitrogen in the Environment _ _
Witkler
Horlxft
lS8m
Us rw«l I Nwftpett
a.i Kin / stale
Neupon / Airport ,
54.24
'Slornm
75-85 0.70 StreamRlver
14llirsto«vn
Nrwpoil
7a. Example B: Flow path from upper watershed, close to site A. N removal is 26%.
45
-------
Saugatucket River Watershed
N-Sink
Tracking Nitrogen in the Environment
About
01090005000331 84.47
03-69
7535
0109000500032C 75.85
.Kingston*
7b. Example B: Close-up.
lment
84 24
0109000900033C 84.24
0109000S00033C 1823
10. Sites A and B compared side by side.
46
-------
Saugatucket River Watershed
N-Sink
Tracking Nitrogen in the Environment
Land Length (rr NHD Reach Co Nitrogen m Nitrogen Out Percent N Rem Waterway Type
0. 010900050049 5458 82.68 3.47 StmamRlver
,miul JTstown
Dutch
[l.urol Lane
) Country
I Club
Richmorx
LlmvewH/
CI Rhodo
itlarrf
Killgstoilj
8a. Example C: Flow path from Sl/K corner of watershed. N removal is 52%.
N-Sink
Tracking Nitrogen in the Environment
About
South
County
Hospital
0.5 ml
8b. Example C: Close-up.
47
-------
Saugatucket River Watershed
N-Sink
Tracking Nitrogen in the Environment
About
Land Len NMD Roe NlJrogen Nitrogen (Xt Percent N Removal Walemay Type
Mfkfutil
fi.irh.ir
f36m
Us Nm) I Ho-vport
9a(ioo / Sue
Mawptnl / Aire en I j
I .mir-.town
Nnvpoil
9a. Example D: Flow path from SW corner of watershed, close to site C. N removal is 90%.
N-Sink hj
Tracking Nitrogen in the Environment
KcfSo»:FO:
Wninocd-
Coirly
Ho»pitat
9b. Example D: Close-up.
-------
Saugatucket River Watershed
ronment
lment
Land Lengh (r NHO React) Code
01090005004978 1052
HooptlJl
0109000500497: 52.47
10. Sites C and D compared side by side.
49
-------
Appendix C
Screen captures showing selected
Modelbuilder © models
created for N-Sink analyses
50
-------
Find the reaches that intersect the flow path.
If the flow does not pass through the stream
network and flows directly into coastal waters,
a different process path is followed.
Generate a flow path from a user-
defined source location. Flow is
determined from the NFIDPIusv2 flow
direction and flow accumulation grids.
The flow path is converted to a vector
feature for the rest of the processing.
Characterize the flow path by
summarizing the land cover
types within a 50m buffer
Add required items to the
table in preparation for N
removal calculations
Calculate N removal for the terrestrial portion
of the flow path. N is only removed if flow
passes through vegetated wetlands
(hydric soils). The source N level is defined
by the user, with defa ult of 100 units.
Any N remaining after the terrestrial portion of flow
path becomes input for surface water calculations.
The iterator steps through all stream reaches and
ponds/lakes from source to outlet and calculates N
removal for each reach. Output from one reach
becomes input for the next.
The "Merge Branch" takes the output from whichever
leg of the model runs and builds the final table for
web display. In the table, the final "N OUT" value is
the amount that reaches the outlet.
Figure C-1. Overall layout of how the final model is constructed. Yellow boxes explain the procedure.
Blue ovals are input; Green ovals are output that pass to the next level. Mauve boxes are
individual sub-models that do most of the actual calculations; Gold boxes are tools; The
dotted blue lines are preconditions that have to be satisfied before the model can continue.
51
-------
A flow path is generated using the flow direction and
hydro corrected DEM. However, it does not line up
nicely with the NHDPlus v2 stream reaches due to
apparent smoothing by the developers.
HowQipPol
How Path Po
int
Raster to
Polyline
NHD_Rowrfi
ne_ForWeb
Rowline
Layer
Select Layer
By Attribute
Route
Evert
Table
A route is created to intersect with selected
reaches to accurately capture the proper
NHDPlus v2 stream segments in order to do
the N removal calculations.
This empty feature class is created in preparation for the
stream reaches that will be selected with the stream
selection iterator. This is where we eliminate all of the
reaches that are not part of the flow path but were
selected in previous steps because they are nearby.
Figure C-2. FlowPath sub-model. This is the first sub-model that runs, showing how the flow path is
generated using the user-defined start location; the NHD flow direction grid; and the NHD
hydro-corrected elevation grid. Any output (green oval) with a "P" is used as input further
along in the processing. Yellow boxes explain the procedure.
52
-------
This model calculates N removal between the source
and the first intersection with the stream network.
For flow through non-wetlands,
no N is removed; input = output.
For flow that passes through vegetated
wetlands, N is reduced based on length
of flow through the wetlands.
Save land portion of the flow
path to draw later if needed.
Determine value passing to
stream calculations.
Figure C-3. LandRemoval sub-model. The sub-model that determines how much N is removed before the
flow path connects with the stream network. Different calculations are run depending on if
the flow path crosses vegetated wetlands (determined using SSURGO hydric soils and NLCD
land cover).
53
-------
-------
&EPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
o
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |