oEPA
United States
Environmental Protection
Agency
EPA/600/R-18/167 | September 2018 | www.epa.gov/research
pump1
Ecosystem Goods and Services Case
Studies and Models Support Community
Decision Making using the EnviroAtlas
and Eco-Health Relationship Browser
-------
EPA/600/R-18/167
September 2018
Ecosystem Goods and Services Case Studies and Models
Support Community Decision Making using the EnviroAtlas
and the Eco-Health Relationship Browser
By
Bolgrien, David W.1, Theodore R. Angradi1, Justin Bousquin2, Timothy J. Canfield3,
Theodore H. DeWitt4, Richard S. Fulford2, Matthew C. Harwell2, Joel C. Hoffman1,
Thomas P. Hollenhorst1, John M. Johnston5, Jonathon J. Launspach6, John Lovette7,
Robert B. McKane4, Tammy A. Newcomer-Johnson8, Marc J. Russell2, Leah S. Sharpe2,
Arik Tashie9, Kathleen Williams1, and Susan H. Yee2
Office of Research and Development
U.S. Environmental Protection Agency
1. Mid-Continent Ecology Division, National Health and Environmental Effects Research Laboratory,
Duluth, MN
2. Gulf Ecology Division, National Health and Environmental Effects Research Laboratory, Gulf
Breeze, FL
3. Ground Water and Ecosystems Restoration Division, National Risk Management Research
Laboratory, Ada, OK
4. Western Ecology Division, National Health and Environmental Effects Research Laboratory,
Newport, OR
5. Computational Exposure Division, National Exposure Research Laboratory, Athens, GA
6. General Dynamics Information Technology (contractor to the U.S. EPA), Duluth, MN
7. OREU contractor, Environmental and Public Health Division, National Health and Environmental
Effects Research Laboratory, Research Triangle Park, NC
8. Systems Exposure Division, National Exposure Research Laboratory, Cincinnati, OH
9. ORISE participant, Environmental and Public Health Division, National Health and Environmental
Effects Research Laboratory, Research Triangle Park, NC
-------
Notice/Disclaimer Statement
The U.S. Environmental Protection Agency (EPA) through its Office of Research and Development
(ORD) funded and collaborated in the research described herein. This document has been subjected to
the Agency's peer and administrative review and has been approved for publication as an EPA
document. Any mention of trade names, products, or services does not imply an endorsement or
recommendation for use.
This is a contribution to the EPA's ORD Sustainable and Healthy Communities Research Program.
The citation for this report is:
Bolgrien, D.W., Angradi, T.R., Bousquin, J., Canfield, T.J., DeWitt, T., Fulford, R.S.,
Harwell, M.C., Hoffman, J.C., Hollenhorst, T.P., Johnston, J.M., Launspach, J.J., Lovette,
J., McKane, R.B., Newcomer-Johnson, T.A., Russell, M.J., Sharpe, L.S., Tashie, A.,
Williams, K., and S.H. Yee. 2018. Ecosystem Goods and Services Case Studies and
Models Support Community Decision Making using the EnviroAtlas and the Eco-Health
Relationship Browser. U.S. Environmental Protection Agency. EPA/600/R-18/167.
Acknowledgments
We greatly appreciate the efforts of reviewers who took the time to read the report: Will Bartsch, Anne
Kuhn, Andy McGuire, Valerie Seidel, and Barbara Sheedy. Chloe Jackson provided editorial and
formatting assistance. Liem Tran assisted in creating the indicator selection and evaluation framework in
Chapter 2.
Cover photo credits:
St. Louis River: from Messenger, Alex. (2017). The St. Louis River. In: Open Rivers: Rethinking Water,
Place & Community, no. 7. http://editions.lib.umn.edu/openrivers/article/the-st-louis-river/ and
http://messengerphotography.com.
Milwaukee Harbor District: photo credit: Eddee Daniel for OnMilwaukee;
https://OnMilvvaukee.com/buzz/articles/harbor-district-harbor-fest.html.
-------
Abstract
This report presents multiple lines of inquiry on how data in the U.S. Environmental Protection
Agency's (EPA) EnviroAtlas and Eco-Health Relationship Browser and the concepts and tools of
ecosystem goods and services (EGS) can be used together to improve a community's ability to address
environmental, social, and economic problems. The EnviroAtlas and Eco-Health Relationship Brower
are examples of data and communication platforms needed for translational science research.
Translational research emphasizes the use of multilateral communication to connect scientific data and
the information needs of communities for decision making. When coupled with data in the EnviroAtlas,
EGS tools can be applied at various spatial scales and for diverse decision contexts. Case studies show
that EGS indicators complement stakeholder engagement processes, such as public meetings and
surveys. The Eco-Health Relationship Browser can help stakeholders use EGS to connect changes in
infrastructure policies (e.g., transportation and parks) to social fairness and human well-being. The
purpose of translational EGS tools and models is to make scientific information and approaches
practical, relevant, and accessible so that more and increasingly diverse stakeholders can make better
decisions. Diverse tools, aided by the flexibility of the EnviroAtlas, allow stakeholders to explore, and in
some cases, quantify, changes in human well-being. The report provides information on how EGS
models and EnviroAtlas data can be translated and adapted for use in new places and for novel contexts.
The report summarizes current strategies for using EGS to support community decision making using
the EnviroAtlas and Eco-Health Relationship Browser. Most importantly, the report presents paths
forward for translational EGS research and applications at EPA.
This report covers a period from October 2017 to September 2018 and work was completed as of May
2018.
iv
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the Nation's
land, air, and water resources. Under a mandate of national environmental laws, the Agency strives to
formulate and implement actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. To meet this mandate, EPA's research program is
providing data and technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how pollutants affect
our health, and prevent or reduce environmental risks in the future.
The National Health and Environmental Effects Research Laboratory (NHEERL) within the Office of
Research and Development (ORD) is the Agency's center for investigation of technological and
management approaches for preventing and reducing effects of pollution that threaten human health and
the environment. The focus of the Laboratory's research program is on methods and their cost-
effectiveness for prevention and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites, sediments and
ground water; prevention and control of indoor air pollution; and restoration of ecosystems. NHEERL
collaborates with both public and private sector partners to foster technologies that reduce the cost of
compliance and to anticipate emerging problems. NHEERL's research provides solutions to
environmental problems by: developing and promoting technologies that protect and improve the
environment; advancing scientific and engineering information to support regulatory and policy
decisions; and providing the technical support and information transfer to ensure implementation of
environmental regulations and strategies at the national, state, and community levels.
This report presents multiple lines of inquiry focused on improving the public's use of ecosystem goods
and services (EGS) concepts, data, and tools for addressing environmental, social, and economic
problems. The case studies and data tools and models in the report highlight the broad use of the
EnviroAtlas and Eco-Health Relationship Brower as data and communication platforms of EPA's
translational EGS research. The report summaries current strategies for using EGS case studies and
models to support community decision making using the EnviroAtlas and Eco-Health Relationship
Browser but it also presents paths forward for translational EGS research and applications at EPA.
Wayne Cascio, Director
National Health and Environmental Effects Research Laboratory
v
-------
Table of Contents
Acknowledgments iii
Abbreviations and Acronyms xi
Executive Summary 1
1. Introduction 3
1.1. EPA Objectives for Community Decision Support 3
1.2. Practical Strategies for Community Decision Making 4
1.3. Integration of Science into Community Decision Making 4
1.3.1. TheEnviroAtlas 5
1.3.2. Eco-Health Relationship Browser 7
1.4. Report Objectives and Organization 9
2. Assessing the EnviroAtlas and Eco-Health Relationship Browser for Supporting Decision Making in
Great Lakes Communities 10
2.1. Introduction to Great Lakes Case Studies 10
2.2. Community Indicator Selection and Evaluation Framework Using the EnviroAtlas 11
2.2.1. Introduction 11
2.2.2. The Indicator Evaluation Framework 12
2.2.3. Milwaukee Harbor Plan Case Study - Background 13
2.2.4. Milwaukee Harbor Plan Case Study - Approach 15
2.2.5. Milwaukee Harbor Plan Case Study - EnviroAtlas Indicators 18
2.2.6. Milwaukee Harbor Plan Case Study - Stage 1 23
2.2.7. Milwaukee Harbor Plan Case Study - Stage 2 24
2.2.8. Milwaukee Harbor Plan Case Study - Conclusions 27
2.3. Supporting Local Decisions with the Eco-Health Relationship Browser 27
2.3.1. Background 27
2.3.2. Imagine Duluth 2035 28
2.3.3. Qualitative Analysis of Public Comments and Input 29
2.3.4. Eco-Health Relationship Browser 30
2.3.5. Impact of Analysis 33
3. EGS-Related Tools Using EnviroAtlas and Eco-Health Relationship Browser for Community
Decision Making 35
3.1. Introduction 35
3.2. Identifying and Mapping Indicators and Metrics Using the EnviroAtlas 36
3.2.1. Mapping Direct Measures of EGS: The FEGS Framework 36
vi
-------
3.2.2. Mapping Measures of Well-being: The Human Well-being Index (HWBI) 40
3.3. Integrating EnviroAtlas Layers with External Scenario Modeling Tools 43
3.3.1. Modeling Ecosystem Services Under Alternative Scenarios: VELMA 44
3.3.2. Assessing Benefits Under Alternative Scenarios: Rapid Benefit Indicators Approach
46
3.3.3. Linking EGS to Well-being: The Services-^HWBI Framework 50
3.3.4. Making External Scenario Modeling Studies Available Through EnviroAtlas: WEDO
55
3.4. Transferring Relationships in the EnviroAtlas and Eco-Health Relationship Browser to Other
Settings 58
3.4.1. Applying EnviroAtlas Models to Calculate Community-Scale Metrics 59
3.4.2. Assessing the Applicability of Ecological Models to New Locations: Model
Transferability Assessment 67
3.4.3. Exploring EnviroAtlas Models Alongside Other Model Options: The ESML 70
3.4.4. Linking Actions to Ecosystem Services and Well-being: Network Analysis 76
3.5. Conclusions 81
4. Conclusion 82
5. References 85
6. Glossaryd 93
vii
-------
Figures
Figure 1.1. Conceptual model of Ecosystem Goods and Services (EGS) research 4
Figure 1.2. Health outcomes related to clean air (an EGS) in urban ecosystems as presented in the Eco-
Health Relationship Browser 8
Figure 2.1. Schematic diagram of the Indicator Selection and Evaluation Framework 13
Figure 2.2. Map of Milwaukee, WI Harbor District 14
Figure 2.3. Milwaukee Harbor Plan Indicator Evaluation Framework Stage 1: from Milwaukee Harbor
Plan recommendations to EnviroAtlas Indicator-Categories 16
Figure 2.4. Milwaukee Harbor Plan Indicator Evaluation Framework Stage 1: from EnviroAtlas
Indicator-Categories to EnviroAtlas Indicators 19
Figure 2.5. Screenshot from the Eco-Health Relationship Browser showing that multiple ecosystems
connect with Aesthetics and Engagement with Nature (an EGS) (green arrows) that is, in turn,
connected with multiple endpoints of health and well-being (blue arrows) 31
Figure 2.6. Screenshot from the Eco-Health Relationship Browser illustrating relationships between
mental health and ecosystem services of water and heat hazard mitigation, recreation and physical
activity, and aesthetics and engagement 32
Figure 3.1. Illustration of the three elements needed to define FEGS 36
Figure 3.2. Conceptual model showing production function linkages between intermediate EGS (IEGS),
final EGS (FEGS), and social and economic outcomes (SEO) 37
Figure 3.3. Number of EnviroAtlas metrics identified as IEGS, FEGS, or SEO for each category of
Beneficiary in the FEGS-CS across environments 39
Figure 3.4. The "Add Data" feature in EnviroAtlas can be used to access HWBI data layers from an
ArcGIS Server Web Service URL and display them alongside other EnviroAtlas data layers 41
Figure 3.5. Human Well-being Index (HWBI) composite scores for counties in Florida, and individual
domain scores for counties within the Tampa Bay area 42
Figure 3.6. HWBI Health scores for Tampa Bay area counties (left) mapped alongside two community-
scale EnviroAtlas metrics (right) to illustrate how visualizing ecosystem services could inform
health and overall well-being 43
Figure 3.7. VELMA conceptual model 44
Figure 3.8. Benefits assessed using Rapid Benefit Indicators 48
Figure 3.9. Sample comparison of calculated RBI indicators for four restoration sites in the Tampa Bay,
FL area 49
Figure 3.10. Services-^HWBI Index framework 50
Figure 3.11. Example output from the Services~>HWBI spreadsheet tool illustrating how changes in
ecosystem services scores could impact the eight domains of well-being relative to baseline
estimates 51
Figure 3.12. County maps of ecosystem services in the original HWBI (left) and example EnviroAtlas
surrogate metrics (right) 54
Figure 3.13. Watershed and Economic Data Interoperability (WEDO) workflow overview for
discovery, evaluation and integration of watershed modeling studies for reuse 55
Figure 3.14. EnviroAtlas screen shot of subwatershed and streams data layer 56
Figure 3.15. EnviroAtlas screen shot with streams with WEDO modeling studies published color-coded
red 57
Figure 3.16. WEDO Study Information screen shot 58
Figure 3.17. Map of Milwaukee and Harbor District 64
Figure 3.18. Current land cover, street configurations and park entrances compared to hypothetical
additions of 10 park entrances, 3 ha of trees, and 4 km of walkable streets in the Harbor District.. 65
VIII
-------
Figure 3.19. Comparison of current and scenario-based changes in near-road tree cover 66
Figure 3.20. Comparison of current and scenario-based changes in distances to park entrance 67
Figure 3.21. Major steps in the methodology to assess the transferability of models to new sites or
applications 68
Figure 3.22. Hypothetical example site context data drawn from EnviroAtlas 70
Figure 3.23. ESML Data Map 71
Figure 3.24. ESML data maps (labelled as variable relationship diagrams) for seven EnviroAtlas models
diagraming the relationships between predictor variables (PD: time or space varying; PC: constant
or parameter), intermediate variables (IE: in ESML; IN: not in ESML), and response variables (RC:
computed response; RM: observed response) 73
Figure 3.25. Screen shot of relationships between ecosystems (yellow arrows), aesthetics and
engagement with nature (center circle), and multiple health outcomes (blue arrows) in the Eco-
Health Relationship Browser 76
Figure 3.26. Flow diagram of the Tampa Bay Relational Browser showing expansion of Eco-Health
Relationship Browser relationships (Ecosystem-^Ecosy stem Services->Health Outcomes; light
blue arrows) to identify: a) manageable intervention points on ecosystem attributes or ecosystem
services (green arrows); b) relationships between ecosystem services, social services and economic
services and domains of human well-being (blue arrows); and c) indicators to measure services and
well-being (dashed blue arrows) 77
Figure 3.27. Illustration of emerging research to integrate the Eco-Health Relationship Browser
relationships (light blue arrows; see Figure 3.17) with ecosystem services attributes and intervention
points (green arrows), and human well-being (HWBI) relationships (blue arrows) of the Tampa Bay
Relational Browser 79
Figure 3.28. Example of the degree of connectivity, using pathway analysis, to assess sustainability of
community decisions based on indicators of network organization 80
ix
-------
T ables
Table 1.1. Screenshots of EnviroAtlas interactive map tables of content 6
Table 2.1. Milwaukee Harbor Plan recommendations, Neighborhood Model dimensions, and
Recommendation-Categories 17
Table 2.2. EnviroAtlas community data layers and indicators summarized by census block groups 20
Table 2.3. Milwaukee Harbor Plan bicycling recommendations 23
Table 2.4. Adjusted R2 from multivariate linear regression (MLR) in Stage 2 26
Table 2.5. Imagine Duluth 2035 language draft and final guiding principles for fairness and health 33
Table 3.1. Example EnviroAtlas data layers, and spatial scales available, useful for hydrologic modeling.
45
Table 3.2. Datasets used and their source in the example RBI analysis 48
Table 3.3. Metrics in the EnviroAtlas that could be used as surrogates to replace the national HWBI
metrics for each of the Ecosystem Services indicators in the Services->HWBI framework 52
Table 3.4. Input and output data, scripts, and output metrics (variable name and description) following
the EnviroAtlas' nomenclature 61
Table 3.5. Selected EnviroAtlas models and response (output) variables included in the ESML as
identified by ESML ecological model identification (EM ID) number 71
x
-------
Abbreviations and Acronyms
AOC
Area of Concern
BenMAP
Benefits Mapping and Analysis Program
BG
Block Group (U.S. Census)
BNF
Biological Nitrogen Fixation
CDC
Centers for Disease Control (and Prevention)
DASEES
Decision Analysis for Sustainable Environment, Economy, and Society
EPF
Ecological Production Function
EGS
Ecosystem Goods and Services
EM
Ecological Models
EPA
U.S. Environmental Protection Agency
ESML
EcoService Models Library
ESP
Ecosystem Services Production
FEGS
Final Ecosystem Goods and Services
FEGS-CS
Final Ecosystem Goods and Services Classification System
FIPS
Federal Information Processing Standards
GI
Green Infrastructure
GLNPO
Great Lakes National Program Office
GLWQA
Great Lakes Water Quality Agreement
HIA
Health Impact Assessment
HiAPC
Health in All Policies Coalition
HSPF
Hydrologic Simulation Program Fortran
HUC
Hydrologic Unit Codes
HWBI
Human Weil-Being Index
IEGS
Intermediate Ecosystem Goods and Services
MT
Metric Ton
MTA
Model Transferability Assessment
MULC
Meter-Scale Urban Land Cover
NHD
National Hydrologic Data
NHEERL
National Health and Environmental Effects Research Laboratory
NLCD
National Land Cover Database
NRCS
Natural Resources Conservation Service
NWI
National Wetlands Inventory
NWIS
National Water Information System
ORD
Office of Research and Development
PM
Particulate Matter (aerosol)
R2R2R
Remediation, Restoration, and Revitalization
RBI
Rapid Benefit Indicators
SDM
Structured Decision Making
SEO
Social and Economic Outcomes
STORET
STOrage and RETrieval
SWAT
Soil Water and Assessment Tool
SWM
Stormwater management
TMDL
Total Maximum Daily Loads
USD A
U.S. Department of Agriculture
USFS
U.S. Forest Service
USGS
U.S. Geological Survey
xi
-------
VELMA Visualizing Ecosystem Land Management Assessments
WEDO Watershed and Economic Data for Interoperability
WQ Water Quality
xii
-------
Executive Summary
This report provides an overview of how the EPA is developing conceptual, scientific, and practical
strategies for community decision support. The EPA is committed to promoting sustainable solutions for
protecting human health and the environment with research supporting environmental decision making
based on translational science. Practical strategies for decision support involve organizing and
translating scientific information so that information is well-aligned with the information needed for
decisions. Practical strategies developed by EPA's Office of Research and Development (ORD) involve
quantitative tools, supporting data, and transferable approaches for their use. A critical step for
developing practical strategies is the measurement of ecosystem goods and services (EGS) that are
directly relevant to communities.
Translational science is a core practical strategy in that scientific information is made useful for decision
making. A key goal of this report is the demonstration of translational science through application of
tools and approaches in real communities with a focus on the application of two tools developed by the
EPA. The EnviroAtlas is a tool for identifying and organizing spatial data on ecosystem services and
human health. It is both a source of information and a platform to combine information in useful ways
that improve translational science. This report discusses the use of the EnviroAtlas in specific tools and
case studies. The Eco-Health Relationship Browser is a visualization tool for understanding connections
between ecosystem services and human health. It is based on peer-reviewed science and allows multiple
connections to be explored at once by bringing the information closer to real-world problems that cross
over disciplinary lines. The Eco-Health Relationship Browser allows for structured approaches to
complex decisions. This report discusses use of the Eco-Health Relationship Browser in stakeholder
engagement and evaluation of decision trade-offs.
The case studies presented in Chapter 2 demonstrate how the EnviroAtlas and the Eco-Health
Relationship Browser serve as gateways between scientific data and decision makers. Successful
community problem solving depends on such gateways that facilitate effective communication among
partners and make data accessible to establish robust and mutually relevant decisions. The Milwaukee,
WI case study (Chapter 2.2) demonstrates how an expert-stakeholder partnership for Milwaukee's
Harbor District used an indicator evaluation framework to connect revitalization goals to related EGS
indicators in the EnviroAtlas. The framework included the translational models needed to connect the
community's land use and infrastructure recommendations (e.g., physical and social connectivity,
governance, and sense of place) and EnviroAtlas EGS indicators. The framework depends on
transferable science, structured decision making, and the accessibility of EGS indicators in the
EnviroAtlas, and as a result is applicable to other communities trying to make complex decisions.
Like in Milwaukee, the Duluth, MN case study (Chapter 2.3) demonstrates how expert-stakeholder
partnerships can translate public input into an EGS context for decision making. In Duluth, the Eco-
Health Relationship Browser is used to organize stakeholder input into a framework in support of public
policy on fairness and health. This step can help facilitate the selection process for EGS indicators, such
as those found in the EnviroAtlas. Qualitative analyses of data from surveys and public meetings were
needed to provide government decision makers with information on how communities equated fairness
with health. The Eco-Health Relationship Browser was used to show decision makers and the public
how fairness could first be linked to EGS and then how EGS could link to health outcomes. For
Duluth's comprehensive planning effort, the objective was to integrate the concepts of fairness and
health into public policy. The important lesson of this case study was that transparency must accompany
translation of scientific information and this is best accomplished with accessible data tools.
l
-------
The EnviroAtlas and the Eco-Health Relationship Browser can help diverse communities, including
under-represented communities, engage in decision making because the information is publicly
accessible and translatable. Clearly, EGS concepts and indicators are not relevant to every community or
decision context. But in this report the national- and community-scale EGS data in the EnviroAtlas and
the health outcomes related to EGS described in the Eco-Health Relationship Browser offer important
demonstrations of translating scientific data for use by the public.
Tools for identifying or predicting changes in EGS are important parts of practical strategies for decision
support. Like the EnviroAtlas and Eco-Health Relationship Browser, these tools can assist decision
makers with incorporating ecosystem services concepts into their decision-making process. Chapter 3
explores how the EnviroAtlas and the Eco-Health Relationship Browser can be used with other EGS
tools to improve decision support.
An important early step in decision making is to create a decision context describing the key issues,
identifying stakeholders that should be involved, and defining objectives as statements of what really
matters to stakeholders about the decision. Structured frameworks (e.g., FEGS Classification System
(FEGS-CS); Chapter 3.2.1) can provide starting points to narrow the scope of EnviroAtlas data to be
considered by first identifying key beneficiary groups and then emphasizing measures that are most
directly relevant to those beneficiaries. The flexibility of the EnviroAtlas as a mapping environment
allows decision contexts to consider many objectives, such as measuring human well-being (HWBI;
Chapter 3.2.2), In some cases, the EGS mapped in the EnviroAtlas may provide a means to achieving
broader social objectives, such as increasing social cohesion. Once the decision context is defined, the
next step is to estimate the potential consequences of the decision. Data from the EnviroAtlas or Eco-
Health Relationship Brower can help inform community deliberations and expert judgements about
potential consequences. For communities or spatial scales lacking EnviroAtlas data, communities may
develop their own using methods from the EnviroAtlas (Chapter 3.4). In other cases, decision makers
may desire more predictive modeling tools (Chapter 3.3), which can leverage input data from the
EnviroAtlas to simulate alternative scenarios and quantitatively compare outcomes. Analysts can also
use EnviroAtlas data to assess the transferability of environmental models to data-poor settings
(Chapter 3.4.2),
Various methodologies, tools, and approaches exist that can facilitate the incorporation of EGS concepts
into community decision making. Community issues and their decision processes are highly diverse,
often at different stages of implementation, and with variable levels of stakeholder experience with
ecosystem services. Practical strategies for EGS-based decision support require a wide array of
adaptable tools and approaches that can be implemented at various scales, stages of the decision process,
and levels of experience. Ultimately the use of the EnviroAtlas, Eco-Health Relationship Browser, and
companion EGS tools can be used synergistically to facilitate the integration of EGS concepts into
decision making.
The EnviroAtlas and the Eco-Health Relationship Browser represent important and valuable resources
for EPA support of state, community, and federal partners and broader goals of fostering translational
science. Decision making that is based on production and delivery of EGS is more likely to be
sustainable and promote stakeholder well-being. Tools are only useful if they are accessible, transferable
between locations and issues, and generate information in clear, decision-specific language. The EPA
practical strategies for EGS-based decision support, including tools like the EnviroAtlas and Eco-Health
Relationship Browser, are being used to inform decisions related to the protection of human-health and
the environment based on the core goal for translational science.
2
-------
1. Introduction
1.1. EPA Objectives for Community Decision Support
This report provides an overview of how the U.S. Environmental Protection Agency (EPA) is
developing conceptual, scientific, and practical strategies to help communities make decisions for
sustainable development and to solve environmental problems. EPA research is producing new
knowledge by quantifying, in space and time, the benefits of nature; those biophysical attributes of the
natural environment that humans value and that contribute to human well-being (here termed, ecosystem
goods and services, or EGS). The EPA is committed to translating scientific information for use by
communities to remediate environmental damages, restore lost or degraded EGS, and to increase human
health and well-being. Improving access to reliable scientific information is a key part of cooperation
and collaboration among federal, state, tribal, and local governments to solve environmental problems
challenging our country. This cooperative federalism (U.S. EPA 2018b) means that EPA, states, and
tribes are equal partners in decision making and are mutually accountable for environmental
remediation, restoration, and revitalization outcomes. Cooperative federalism creates distinct roles for
federal, state, and tribal governments for environmental regulation and management. The federal
government establishes standards for pollutants and the states have authority to create programs to
implement the standards (ECOS 2017).
The EPA's mission, to protect human health and the environment, is process-, not goal-, oriented
because the conditions of the built and natural environment are intertwined in complex and dynamic
relationships. The EPA has a unique mission among federal agencies to fuse science and regulation to
address critical economic, environmental, and social issues. Thus, the EPA has unique responsibility to
develop translational approaches to share the best possible scientific data and tools that contribute to
sound and sustainable decisions. Cooperative federalism is an important Agency strategic goal (U.S.
EPA 2018b). Because technical capacities vary among governmental partners, cooperative federalism
facilitates the development of accessible information, transferable analytical tools, and translational
approaches, to better enable communities (defined as those people who reside within the jurisdiction of
one or more local governments or tribal nations, or people unified by a common interest or activity) to
meaningfully engage in decision making with local, state, federal, and tribal governments.
The EPA's Office of Research and Development (ORD) has a history of conducting community-scale
research for addressing specific issues in specific places (Summers et al. 2014). To operationalize the
accumulated knowledge for wider application, it may not be adequate, nor feasible simply to "ramp up"
local or site-specific approaches. Among other efforts, EPA ORD's Sustainable and Healthy
Communities National Research Program is advancing efforts to support community-based decision
making with approaches that use EGS (Fulford et al. 2016). Many stakeholder communities and local
governments lack the capacity, knowledge, and tools with which to translate scientific data, evaluate
problems in multiple dimensions, and quantify trade-offs among potential solutions. To address this,
EPA's ORD's EGS research has focused on models, tools, and approaches for identifying and
evaluating important tradeoffs among diverse beneficiary groups.
"The Agency will also emphasize the translation of Its work products for
end user application and feedback/' - EPA Strategic Plan 2018.
3
-------
1.2. Practical Strategies for Community Decision Making
Decisions,
Alternatives
Mgmt.
Actions
Social and Economic
Services
Biophysical
State of the
Ecosystem
(includes
intermediate
EGS)
Qm>
A Final
EGS
Benefit
Functions
A Weil-Being
Information for Decision Support
Figure 1.1. Conceptual model of Ecosystem Goods and Services (EGS)
research (from Yee et al. 2017).
Community decisions, and the agencies and stakeholders that cooperate in making them, often address
relatively narrow contexts. Yet, decisions may have significant widespread economic, environmental,
and social consequences.
Structured decision analysis
provides an approach for
evaluating tradeoffs in a
way that encourages public
participation and
collaboration and allows
consideration of multiple
alternative outcomes (Liu et
al. 2010). The conceptual
framework for EPA's
ORD's Community-Based
Final Ecosystem Goods and
Services (FEGS) research
(Figure 1.1) focuses on the process of informing decision making using state-of-the-science research in
EGS, particularly FEGS that directly benefit people.
The conceptual model includes the analysis of information leading from decisions through EGS to
benefits and changes in well-being. Ecological production functions (EPFs) and benefit functions refer
to processes that link management actions (such as environmental remediation or restoration) to EGS
and well-being. Research continues to effectively quantify EGS, benefits, and well-being and to
construct production functions (e.g., Johnston et al. 2017b). The conceptual model identifies links that
help stakeholders bridge the gap between decisions and outcomes through the integration of science and
policy. Connecting decision alternatives to changes in EGS and human well-being represents the
cornerstone of a practical strategy for supporting community decision making.
Yee et al. (2017) outlined a suite of practical strategies for supporting community decision making based
on tenets of structured decision making (SDM) and EGS. The strategies emphasize evaluating relative
benefits communities enjoy from the environment, the structured collection of needed data, effective
stakeholder communication and engagement, development of decision alternatives, and an evaluation of
alternatives in terms of potential changes in well-being. All strategies include collaborative tasks for
establishing decision contexts with requisite goals and objectives, performance and monitoring
measures, and approaches for evaluating alternative trade-offs. Of particular interest is the structured
collection and evaluation of the data and analytical models needed to inform decisions, including for
developing decision alternatives and comparing tradeoffs. Data collection guidance means more than
simply making data accessible. Structured decision making and FEGS approaches require reliable tools
and mutual understanding among the parties for translating important data into the language of the
decision makers. This is the backbone of translational science and involves then integration of models
and other data sources with decision support tools that make data useful.
1.3. Integration of Science into Community Decision Making
A goal of EPA research is to translate science into knowledge with which communities can remediate
legacy environmental damages, restore lost or degraded EGS, and initiate sustainable community
revitalization that promotes social justice, health, and economic prosperity (U.S. EPA 2016). There is a
suite of EPA's ORD models, tools, and datasets that integrate EGS science into decision support tools
4
-------
(e.g., Fulford et al. 2016; Yee et al. 2017). For decision contexts involving spatial data and maps, EPA's
EnviroAtlas (Pickard et al. 2015) can be a valuable vehicle for data analysis, stakeholder engagement,
and information sharing. In addition, EPA's Eco-Health Relationship Browser (Jackson et al. 2013) can
inform those using SDM of links between environmental conditions, EGS, and health outcomes. People
with access to data and information on the distribution of EGS in their communities and how EGS can
determine physical, mental, and economic health, are empowered to make decisions supportive of the
long-term public good. This report outlines how the EnviroAtlas, the Eco-Health Relationship Browser,
and other EGS tools can be used to translate EGS and related scientific data for community decision
making.
1.3.1. The EnviroAtlas
The EPA has made significant investments in interactive tools and databases to translate EGS and
related scientific information for decision making. The most important of these is the EPA's EnviroAtlas
(EnviroAtlas website at http://www.epa.gov/EnviroAtlas; accessed 8/7/2018; Pickard et al. 2015). The
EnviroAtlas, launched in 2014, is a publicly-accessible web-based data platform designed to map the
spatial relationships between EGS and landscapes, and the functional relationships between EGS and
human health. It is an example of an operational, combined data delivery system, modeling
environment, and decision support tool specifically developed for public use (Olander et al. 2017). It
contains more than 400 open access geospatial indicators synthesized from multiple sources and
organized in multiple themes according to spatial scales. Data span the continental U.S. with national
coverage of 212 indicators accessible via interactive maps (Table 1.1). Additionally, the EnviroAtlas
has a Dynamic Data Matrix (EnviroAtlas Dynamic Data Matrix website at
https://www.epa.gov/enviroatlas/enviroatlas-dvnamic-data-matrix; accessed 8/1/2018) with a search
feature to allow the user to learn more about data layers and indicator metrics, including links to data
fact sheets and metadata. An additional suite of 97 indicators has been developed at higher spatial
resolution for approximately 50 communities. Most national data layers are summarized at 12-digit
hydrologic unit codes (HUC; or sub-watershed basins) and are based on land cover data with the spatial
resolution of 30 m of the National Land Cover Database (NLCD website at
https://vvvvvv.mrlc,gov/index.php; accessed 8/7/2018; Homer et al. 2015). Most community data are
available at the census block level. Likely the greatest usefulness of the EnviroAtlas derives from the
large volume and variety of data that are translated for diverse user applications and are available by
download, web services, or through the online interactive map.
The EnviroAtlas is translational by design with multiple access points to information that is organized
by EGS and national and community-scale geospatial layers. It is available to technical and non-
technical users, including community stakeholders. The EnviroAtlas represents EPA's commitment to
develop EGS indicators and to make available supporting data, tools, and information that have direct
relevancy for the needs and interests of policy makers, educators, researchers, and the public. Users can
apply EnviroAtlas data to create new knowledge for decisions in multiple contexts, including sustainable
development (Pickard et al. 2015). Besides the creation and delivery of data and tools, the EnviroAtlas is
a rich source of data for exploring EGS and sustainability indicators and adapting EGS concepts for use
in diverse applications (Summers et al. 2014). The value of the EnviroAtlas can be constrained by the
diversity in EGS indicators that cannot be readily integrated into science-supported decision making
without transdisciplinary and translational tools (Gret-Regamey et al. 2017; Rosenthal et al. 2015).
Stakeholders and decision makers still face the problem of extracting manageable sets of relevant
indicators from the EnviroAtlas that address their data needs. Tools are needed to allow stakeholders to
interactively select the indicators relevant to the problems they hope to address (Tran 2016).
5
-------
Table 1.1. Screenshots of EnviroAtlas interactive map tables of content. Of the 309 EGS and biodiversity
maps, 212 and 97 are at national and community scales, respectively (see EnviroAtlas Interactive Map website
at https://enviroatlas.epa.gov/enviroatlas/interactivemap/; accessed 8/14/18).
Indicators are exclusively at community scale.
Indicators are exclusively at national scale.
EGS and Biodiversity layers
Envir©Atlas Interactive Map
31 T & ©
I • I I I ¦
Ecosystem Services and Biodiversity
» Filter EnviroAtlas Data by Geography | Topic
National EnviroAtlas Communities i)
( ) ) Carbon Storage
|*D Crop Productivity
*D Ecosystem Markets Q"~)
#D Energy Potential
Engagement with
fcD
*D
CD
CD
CD
CD
*D
Outdoors ItD
Health and Economic^t^Z)
Outcomes
Impaired Waters
Land Cover: Near-
Water
Land Cover: Type
Landscape Pattern
Near-Road
Environments
ITD
#D
CD
CD
CD
#D
Pollutant Reduction:
Air
Pollutant Reduction:
Water
Pollutants: Nutrients
Pollutants: Other
Protected Lands
Species: At-Risk and
Priority
Species: Other
Water Supply, Runoff,
and Flow
Water Use
Weather and Climate
Wetlands and
Lowlands
I Search All Layers
0 of 309 Maps
I Hide Icons
People and built spaces layers
EnvirilAtlas
Interactive Map
T ® <9
People and Built Spaces
~ Filter People and Built Spaces Data
( ) ) Community Demographics
QD Commuting and Walkability
CD Employment
C ) ) Housing and Facilities
QZ) National Demographics
QZ) Quality of Life
0 of 69 Maps
Boundaries and natural feature layers
Envir©Atlas Interactive Map
M ? % ©
Boundaries and Natural Features
Ecologic Boundaries
HI GAP Ecological Systems
i)
USEPA Ecoregions (Omemik)
Hydrologic Features
~ Hydrologic Unit Code (HUC) Boundaries and labels i)
I] NHDPIus V2 features
Waterscape - Hydrologically connected zone i;
Waterscape - Riparian zone
HI Waterscape - Surface water _i)
Political Boundaries
Z] Congressional District boundaries and labels
^ EnviroAtlas Community Boundaries
~ Landscape Conservation Cooperatives ^
States, County, and Census Block Group boundaries
USEPA Regions ij
Climate model layers
EmirQAtlas Interactive Map
X T %9
Times Series Layers
~ Climate Scenarios
Show Scenario
Clear Map
Timeline: Years (1950-2099)
Select Scenario or Historical
V
Select Variable
V
Select Season or Annual
V
i)
6
-------
1.3.2. Eco-Health Relationship Browser
The Eco-Health Relationship (Eco-Health Relationship Browser website at
https://www.epa.gov/enviroatlas/enviroatlas-eco-health-relationship-browser; accessed 8/7/2018)
provides users with visual depictions of scientific evidence for relationships between ecosystems, EGS,
and human health and well-being. (Jackson et al. 2013). The Eco-Health Relationship Browser is an
interactive, publicly-accessible web-based visualization tool based on peer-reviewed scientific literature
related to 36 health outcomes or social determinants of health grouped by six EGS categories
(Millennium Ecosystem Assessment 2005; Boyd and Banzhaf 2007; aesthetics and engagement with
nature; air quality; heat hazard mitigation; recreation and physical activity; water hazard mitigation; and
water quality) in four ecosystem types (urban; agro-ecosystems; forests; and wetlands). Figure 1.2
shows the value of the visualization tool using an example of health outcomes associated with clean air
in urban ecosystems. The Eco-Health Browser uses a weight-of-evidence approach to illustrate the
multiple relationships and pathways between EGS and health. While not exhaustive, the information
provided in the Eco-Health Relationship Browser is based on extensive literature review and highlights
statistically significant, plausible associations. Results are intended to complement the EnviroAtlas and
provide a visual resource for EnviroAtlas users to organize their data search. One of the key objectives
of this report is to explore how resources in the EnviroAtlas and Eco-Health Relationship Browser can
be combined with decision support tools and integrated into practical strategies for community decision
support (Yee et al. 2017).
The Eco-Health Browser uses a weight-of-evidence approach to illustrate
the multiple relationships and pathways between EGS and health.
7
-------
Aesthetics
Engagement I
with Nature /
Recreation
& Physical
activity
Urban
Ecosystems
Clean Water
Water
Hazard
Mitigation
Heat Hazard
Mitigation
ADHD
Aggression
•Anxiety
¦ Arthritis
' Asthma
• Birth outcomes
' Bronchitis
' Cancer
'Cardiovascular diseases
'Cognitive function
Confusion
1 COPD
Depression
Fatigue
Gastrointestinal illness
k Happiness
Healing
Heat stroke
High blood pressure
'Hospital admissions
L Kidney damage
Longevity
Low birth weight
Mental health
Migraine
Miscarriage
•Mortality
lObesity
Preterm birth
PTSD
1 Respiratory symptoms
Self-esteem
Social relations
Stress
Thyroid dysfunction
Figure 1.2. Health outcomes related to clean air (an EGS) in urban ecosystems as presented in the Eco-
Health Relationship Browser. Clean air may have directly positive health outcomes (e.g., longevity) or may
ameliorate negative outcomes (e.g., post-traumatic stress disorder (PTSD), chronic obstructive pulmonary
diseases (COPD), attention-deficit/hyperactivity (ADHD) disorder).
8
-------
1.4. Report Objectives and Organization
The objective of this report is to highlight the data and tools that EPA is providing to federal, state,
tribal, and local partners to incorporate EGS concepts, data, and tools into public policy decisions and
environmental problem solving. This report presents "deep dives" into using EGS and the EnviroAtlas
and the Eco-Health Relationship Browse for translational research as introduced by Fulford et al. (2016)
and Yee et al. (2017). The applications of the EnviroAtlas and the Eco-Health Relationship Browser
described here illustrate ways that scientific knowledge can be translated and put to work for community
decision making, supporting EPA's strategic goal to increase transparency and public participation in
environmental problem solving (U.S. EPA 2018b)
"EPA will strengthen Its community-driven approach, which emphasizes
public participation to better partner with states, tribes, and communities
and to maximize the support and resources of the entire Agency to create
tangible environmental results."
- EPA Strategic Plan 2018
This report is divided into two chapters. Chapter 2 presents case studies where communities used
EnviroAtlas and the Eco-Health Relationship Browser to address problems of land use and social policy.
They show how to translate public discourse into EGS indicators and how to translate EGS indicators
for use in public discourse. Evaluating the relevancy of EGS indicators for decision making depended on
constructive interactions between experts, managers, and stakeholders to define decision contexts and
identify data needs. Chapter 3 presents models, tools, and datasets that integrate EGS, the EnviroAtlas,
and Eco-Health Relationship Browser in concept and practice. These include tools that use the
EnviroAtlas to identify and map EGS indicators, datasets that are complementary to the EnviroAtlas,
and models that describe how EGS indicators can be transferred to novel locations or applications. The
objective is to highlight intersections among EGS concepts, the EnviroAtlas, and the Eco-Health
Relationship Browser that support community decision making by improving accessibility of EGS-
based spatial data and the tools.
9
-------
2. Assessing the EnviroAtlas and Eco-
Health Relationship Browser for Supporting
Decision Making in Great Lakes
Communities
2.1. Introduction to Great Lakes Case Studies
Some coastal communities of the Great Lakes are challenged to meet the objectives of the Great Lakes
Water Quality Agreement (GLWQA) because environmental conditions have led to multiple beneficial
use impairments. The EPA is actively working with federal and state agencies, stakeholders, and
communities to apply practical strategies for translating community-relevant needs for contaminant
remediation, EGS restoration, community revitalization into EGS indicators, analytical tools, and
models. Translating and communicating these strategies is helped by discussing them collectively as
"R2R2R" (or Remediation, Restoration, and Revitalization) processes. Information and indicators from
the EnviroAtlas and the Eco-Health Relationship Browser are useful for addressing the needs of R2R2R
in two important ways. First, they are platforms from which relevant EGS indicators can be selected for
evaluation in decision contexts, and second, they are useful for organizing stakeholder engagement on
complex R2R2R issues.
Research into how EGS concepts may be applied to R2R2R requires consideration of the relationships
between experts and stakeholders, between state and federal resource management agencies, and
between local governments and the public. These relationships determine decision contexts and drive
needs for data and analysis. The case studies presented in Chapter 2 show how the EnviroAtlas
(Chapter 2.2) and Eco-Health Relationship Browser (Chapter 2.3) can be used to translate EGS
information to address diverse community objectives. These case studies demonstrate how EGS research
can create a new understanding and lead to better health and well-being outcomes. The EnviroAtlas
indicator evaluation framework developed for Milwaukee, WI demonstrates how to translate a
community's vision for R2R2R into EnviroAtlas indicators. The second case study, focused on Duluth,
MN, describes how the Eco-Health Relationship Browser can be used to translate a community's vision
health and fairness into public policy. These case studies represent two linked pieces for achieving the
overall goal of engaged and collaborative expert-stakeholder decision making based on EGS. In
addition, they show how the EnviroAtlas and the Eco-Health Relationship Browser can contribute to
public discourse on land use, infrastructure, and social policy that help communities use available data
to improve health and well-being and to evaluate potential impacts of alternate decisions.
The U.S. EPA's Great Lakes National Program Office uses the term
"Remediation to Restoration to Revitalization (R2R2R)" to characterize the
process of remediating environmental contamination and restoring
ecosystem services to help revitalize coastal communities.
10
-------
2.2. Community Indicator Selection and Evaluation Framework Using the EnviroAtlas
2.2.1. Introduction
The Brundtland Report (1987) set off a 30-year "explosion" (Tanguary et al. 2010; Turcu 2013) in the
number, diversity, and complexity of attempts to apply indicators of EGS for community decision
making, public policy formulation, and public enlightenment (Lehtonen et al. 2016). Indicator
development facilitated trans-disciplinary, trans-generational, and trans-cultural conversations on
humans in the environment and created knowledge for both experts and the public. It has advanced the
principles of translational ecology of collaboration, engagement, and communication (Enquist et al.
2017). The use of EGS indicators can help communities conceive and implement the diverse projects
needed for R2R2R.
The number and diversity of EGS indicators is impressive. Uhlmann et al. (2014) reviewed 14 studies
involving approximately 400 indicators. Costanza et al. (2016) mapped relationships between
sustainability targets and over 300 indicators. Singh et al. (2009) listed 41 indices based on 550
indicators. However, Tanguary et al. (2010) found no optimal number or type of indicators and Turcu
(2013) found it surprisingly difficult to select a set of essential indicators from a surfeit of choices.
Importantly, indicators that lack adequate conceptual foundations or are applied in undefined decision
contexts may, at best, be ambiguous or irrelevant to community decision making. At worst, they can
lead to unreliable assessments of environmental conditions that erode the credibility of sustainability and
EGS in concept and practice (Mitchell 1996; Reed et al. 2006; Singh et al. 2009; Rosales 2011; Wu and
Wu 2012; Hak et al. 2016; Dowling et al. 2017; Missimer et al. 2017a).
Selecting community relevant EGS indicators is conceptually and practically challenging. Of paramount
importance creating a decision context defining decision and data needs. The decision context guides the
indicator translation and evaluation process (Enquist et al. 2017). The use of EGS indicators is decidedly
contextual and their selection must be intentional and responsive to a well-defined problem (Valentin
and Spangenberg 2000; Reed et al. 2006).
Approaches for establishing decision contexts and subsequently selecting and evaluating EGS indicators
should be transparent and collaborative and reflect community values (Valentin and Spangenberg 2000;
McCool and Stankey 2004; Rosales 2011; Tran 2016). Top-down (or expert-led) approaches may be
viewed as more efficient than bottom-up (or stakeholder-led) approaches because less effort may be
needed to reach consensus and experts may be able to adapt existing indicators for new applications.
However, top-down approaches used for selecting indicators may be biased if they were used for
developing indicators. Participatory bottom-up decision making is responsive to, and reflective of,
community priorities and values. This is important as more communities initiate, develop, and
implement strategies for sustainable development and R2R2R.
Ecosystem goods and services indicators must be derived from accessible data, directionally
unambiguous, and sensitive to anticipated state changes. Top-down approaches may imply rigor or
objectivity of indicators that can be misunderstood as being scientifically unassailable by stakeholders
(McCool and Sankey 2004; Reed et al. 2006). Bottom-up approaches may favor indicators for their ease
of translation and interpretation, accessibility, and potential to promote collaboration. In practical
applications, there are tradeoffs in the choice of using a top-down expert-lead filtering of indicators
versus bottom-up approaches involving stakeholders in the indicator selection process. Tensions can be
exacerbated by the disparate language and methodologies among environmental, social, and economic
disciplines to develop and apply indicators (Missimer et al. 2017a and 2017b). If, for example,
stakeholders were confronted with a priori indicator selection decisions on indicators without initial
11
-------
engagement or buy-in, overall stakeholder engagement could be jeopardized. Overall, processes for
selecting indicators are often ad hoc compromises between objectivity and ease of translation, and
therefore, top-down versus bottom-up approaches. Successful processes employ expert-stakeholder
collaborations across spatial scales, domains and disciplines (Uhlmann et al. 2014). Regardless of
approach, maintaining the integrity of indicator selection and evaluation processes requires feedback or
learning loops and transparency. Stakeholders must be able to iteratively reconsider selection decisions
based on new information or the refinement of the decision context.
Establishing decision contexts and selecting and translating EGS indicators can be done using
qualitative methods (Hsich and Shannon 2005). However, quantitative methods have been used to
evaluate the information content of indicator datasets. The premise is that stakeholders need EGS
indicators to address the needs of the decision context (e.g., assess current conditions, evaluate possible
outcomes of alternate decisions, or measure progress towards goals). In theory, these needs could be met
by a single perfect indicator or an infinite number of indicators. In reality, stakeholders need to cull a
subset of relevant and practical indicators from the universe of EGS indicators. They need methods for
creating such a set. Munier (2011) applied linear programming and concepts of entropy to maximize the
information gleaned from available data (i.e., explain the most variance) with the smallest practical
number of indicators. Similarly, Recatala and Sacristan (2014) used principle components analysis to
evaluate various sets of indicators to explain data variance and reduce data collection costs. The
statistical clustering and multivariate linear regression method of Tran (2016) was adapted for the case
study presented here to optimize the number of EnviroAtlas community indicators (Pickard et al. 2015)
needed to address specific community-derived decision contexts.
2.2.2. The Indicator Evaluation Framework
For EGS indicators to be used for decision making they must be translated into language, units, and
formats readily understood by stakeholders. The Milwaukee case study demonstrated a two-stage
indicator evaluation framework for selecting and evaluating EGS indicators for use in community
decision making. The framework used data from the EnviroAtlas, but it could incorporate EGS
indicators from other sources. In Stage 1 (Figure 2.1), a community-based vision of social and
environmental conditions was abstracted using qualitative methods (blue boxes). This information was
used to create a decision context. In parallel, a universe of EGS indicators was assembled and explored
to create an initial set of context-informed indicators (green boxes). Further qualitative processing
(including human-coded document analysis) was done in a translational model to categorize common
elements of both the decision context and available indicators. The translational model was used to
distill recommendation-categories from the decision context and indicator-categories from the available
indicators. The qualitative processes in Stage 1 winnowed the universe of available indicators down to a
manageable, context-relevant set of candidate indicators. Stage 2 of the framework (Figure 2.1; gray
boxes) used quantitative statistical methods (Tran 2016) to evaluate how much variance of the entire set
of indicators could be explained by various subsets of indicators. The goal was to create a final subset of
indicators that explained a large amount of variance of all indicators that addressed the decision context.
12
-------
Community Vision
Statement
(Harbor District Plan)
Feedback:
Decision-maker, Public,
stakeholder
communication
Decision Context
(District-wide recommendations)
Feedback:
Refine decision context
J
Translational Model of Common Elements
• Aesthetics
Physical environment
• Anchor
Research
institutions
Resilience
• Connections
• Safety
Economics
• Transit
• Equity
• Air & water quality
• Governance
Land cover
Housing/where
Health costs
people live
Parks
Identity
Participation
Infrastructure
0
Recommendation
Categories
Candidate indicators
0
Indicator
Categories
1
Initial selection of context-
informed indicators
Feedback:
Redefine available indicators
I
Universe of EGS Indicators
(EnviroAtlas)
Other sources of EGS
indicators
I
I
I
I
A.
Quantitative
Indicator Evaluation
«-+
Final Indicators
Figure 2.1. Schematic diagram of the Indicator Selection and Evaluation Framework. Stage 1 consists of
decision context pathway (blue boxes) and initial indicator selection pathway (dark green boxes). Stage 2 consists
of quantitative evaluation of candidate indicators (gray boxes, red lines). Feedback learning pathways are noted
for candidate indicators in Stage 1 (dashed blue lines) or as part of Stage 2 (solid blue lines).
An important feature of the framework was that it provided multiple opportunities for experts and
stakeholders to refine the decision context and EGS indicator selection criteria. This included the
possibility of including indicators from multiple sources and re-evaluating common elements used in the
translational model. The framework helped translate stakeholder-inspired sustainability goals into
functional EGS indicators for use by decision makers, stakeholders, and the public to create equitable
social and environmental policies. However, the framework (in Stage 2) only evaluated the statistical
strength of subsets of indicators, not their acceptability to users or practical usefulness for achieving the
community's vision. Further, it did not evaluate the practicality of collecting the data needed to generate
or monitor the indicators. The expert-stakeholder collaboration remains fundamentally responsible for
assessing the value of indicators in fulfilling the community's vision.
2.2.3. Milwaukee Harbor Plan Case Study - Background
The EGS indicator selection and evaluation framework was demonstrated using community-derived
sustainability goals for industrial areas and neighborhoods adjoining the harbor in Milwaukee, WI
(Figure 2.2). The Harbor District Comprehensive Water and Land Use Plan (2017; herein called the
Milwaukee Harbor Plan) was produced through extensive expert-stakeholder collaboration and
presented recommendations for the sustainable and equitable social and economic development in the
area.
13
-------
Milwaukee Wisconsin
Harbor District
Village ol
K«*asku"<
Village ot
Belgium
E Lincoln Ave'
2 Kilometers
/lllag®
tail)
Village1
Blown D
Village
Village ol
Hatlana
Wauwatosa
iFwmingtcn
Waukesha
Village ol
Dousman
Village
ol Sulliv an
Green"61 &
MAW**'
Legend |\j
| | Harbor District boundary 0 5 10 20 Kilometers A
1—1—1—1—1—1—1—1—1 A
I I EnviroAtlas Milwaukee Community boundary ' \
Service Layer Credits: Sources: Esri, HERE. DeLorme. USGS, Inlermap. increment P Corp., NRCAN, Esri Japan. METI, Esri China (Hong Kong), Esri (Thailand). Mapmylndia, © OpenStreetMap contributors, and the GIS User Community
Figure 2.2. Map of Milwaukee, WI Harbor District.
Decision contexts, derived from content analysis of the Milwaukee Harbor Plan, guided the selection of
community-scale EGS indicators from the EnviroAtlas (Stage 1). A quantitative indicator evaluation
process (Stage 2) was initially developed for EnviroAtlas indicators for Durham, NC by Tran (2016) and
was adapted in this case study.
14
-------
The decision contexts in the Milwaukee Harbor Plan were embedded in multiple other physical,
political, and social contexts and scales. At large scales, the development of the Milwaukee Harbor Plan
reflected a tradition of collaborative natural resource management in the Great Lakes region (Jetoo et al.
2015). This tradition is evident in the major environmental protection legislation for the Great Lakes
such as the Boundary Waters Treaty of 1909, the (GLWQA, and the Great Lakes Restoration Initiative
(GLRI) (Linton and Hall 2012). Vallentyne and Beeton (1998) argued that environmental, civic, social,
and institutional collaborations make cities of the Great Lakes region laboratories of innovation for
developing and applying approaches for ecosystem sustainability and sustainable development.
The Milwaukee Harbor Plan also reflected the history of transdisciplinary innovation in urban planning
that began with The Plan of Chicago (Burnham and Bennett 1909). The Plan of Chicago is credited with
beginning a movement that recognized importance of co-functioning natural and urban landscapes and
the value of public engagement in land-use decision making. The Plan of Chicago advocated a proto-
ecosystem approach to restore and conserve EGS for the public good.
At intermediate scales, the Milwaukee Harbor Plan reflected R2R2R goals for the Milwaukee Estuary
Area of Concern (AOC). The AOCs were designated under the GLWQA to restore beneficial uses lost
to human activities. The AOCs are required to use expert-stakeholder collaborations for shared problem-
solving (MacKenzie 1997). The Harbor District comprised a relatively small area of the Milwaukee
AOC but was a priority for potential EGS restoration and neighborhood revitalization. The Milwaukee
Harbor Plan was an example of how AOC and urban development objectives can be integrated. The
R2R2R goals for the Harbor District are important to the sustainability goals for the AOC, which in turn
are important to sustainability goals for the City.
At local scales, the community vision in the Milwaukee Harbor Plan reflected the past and present
economic, demographic, and environmental conditions in the Harbor District (Figure 2.2). The land use
and infrastructure recommendations focused on: 1) economic growth consistent with past industrial
strengths and supportive of future commercial growth; 2) linking environmental remediation and
restoration to revitalization; and 3) creating social and economic equity to benefit people living and
working in the district. The 155 district-wide recommendations from the Milwaukee Harbor Plan were
analyzed and categorized into a decision context that was then translated into indicators available from
the EnviroAtlas (Figure 2.1).
2.2.4. Milwaukee Harbor Plan Case Study - Approach
The Neighborhood Model (Johnston et al. 2017b) used conventional content analysis (Hsieh and
Shannon 2005) to characterize the Milwaukee Harbor Plan recommendations and EnviroAtlas indicators
for building a translational model to connect recommendation and indicators conceptually and
functionally. The analysis entailed identifying and categorizing how the data and indicators and the
plan's recommendations reflected features of the built, social, and natural environments that contribute
to community and personal well-being. The goal was to "normalize" each dataset so they could relate to
each other. The analysis was applied independently to the Milwaukee Harbor Plan's 155 district-wide
recommendations (Milwaukee Harbor Plan 2017) and 361 EnviroAtlas EGS indicators for Milwaukee.
Results were then harmonized to create a concise set of common elements in the Translational Model of
the framework (Figure 2.1).
In the framework, the Neighborhood Model organized the recommendations through a three-step
process. First, the recommendations were categorized into four dimensions: the built environment (or
assets that facilitate public engagement with a space, such as parks, housing, transit); human-
environment relationships (e.g., sustainability, aesthetics); personal attachment to a space (e.g., safety,
15
-------
identity, self-determination, participation, social cohesion); and structural-statistical attributes (e.g.,
physical infrastructure, natural features, economy, governance) (Table 2.1). Recommendations were
then coded to define to which sub-dimension they contributed (i.e., aesthetics, equity, resiliency).
Finally, the recommendations were analyzed to provide more descriptive results about the contributions
(i.e., recommendation-categories in Table 2.1) A total of 253 recommendation-category combinations
were found that addressed the decision context (Figure 2.3). This step-wise coding ensured that data
maintained logical and hierarchically-consistent connections backward to the Milwaukee Harbor Plan
text and forward (through the translator) to EnviroAtlas indicators.
Harbor Plan Recommendations
Harbor District Recommendation Categories
Built Environment
Human-Env.
Relations.
Personal
Attachment
Structural or Statistical
Harbor Plan Section
# recs
Economic development
Environmental clean-up
Subsection
Commercial
I ndustry attraction
Industry retention & expansion
Job creation
Recreation and tourism
Small business support
Waterfront
Equity
Habitat & ecology
Equity
Housing
Job availability & access
Parks & public space
Education
Land
Water
Land use policies
Commercial
General land use & buildingform
Industrial
Parks & open space
Residential
Stormwater &WQ
Sustainable consumption
Steam energy
Water use
Transportation & utilities
Bicycles
Marine/water transportation
Personal vehicles & parking
Public transportation
Rail
Streets & sidewalks
Utilities
EnviroAtlas Indicator
Categories
Air quality improvement
Demographics
Distance to park
Health cost & disease avoidance
H istorical places
Places to play
Presence of green space
Presence of impervious
Presence of land cover
Presence of park
Presence of trees
Presence of vegetation
Presence of water resource
Presence of wetlands
Roads and sidewalks
Water quality improvement
Where people live
o- 33
n f
5f O
Figure 2.3. Milwaukee Harbor Plan Indicator Evaluation Framework Stage 1: from Milwaukee Harbor
Plan recommendations to EnviroAtlas Indicator-Categories.
16
-------
Table 2.1. Milwaukee Harbor Plan recommendations, Neighborhood Model dimensions, and
Recommendation-Categories.
Neighborhood Model Dimensions
Human-
Harbor Plan Recommendations
# Recom-
Built
Environment
Relationship
Personal
Structural or
mendations
Environment
Attachment
Statistical
Anchor inst.
Demography
Connections
Economy
Recommendation-Categories
(Figure 2.3)
Housing
Parks
Transit
Aesthetics
Equity
Resiliency
Identity
Participation
Safety
Governance
Infrastructure
Local business.
Physical
environment
Research
Economic Development
Industry attraction
4
4
Industry retention and expansion
1
1
1
Commercial development
3
2
1
1
Job creation
4
1
4
Recreation and tourism
6
3
1
5
Small business support
2
1
1
2
Waterfront
3
3
Environmental Cleanup
9
2
9
Equity and Affordability
1
Housing
8
6
4
1
4
Job availability and access
8
3
8
Parks and public spaces
10
7
4
2
8
Habitat and Ecology
1
•s
1
Education
J
Land
8
5
8
Water
4
1
2
3
Land Use Policies and Strategies
Commercial
3
1
3
General land use and built form
11
3
10
2
Industrial
5
1
3
2
1
Parks and open space
10
9
7
7
5
Residential
6
1
3
1
5
Public Art
8
1
6
8
3
SWM and WQ management and
19
11
1
16
sustainability Lake water cooling
1
1
1
District steam energy
1
1
1
Transportation and utilities
Personal vehicles and parking
1
1
Bicycles
3
2
3
Marine/water transportation
2
2
Utilities
2
2
1
2
Rail
1
1
Public transit
4
3
1
1
2
Street and sidewalks
6
2
3
1
3
Totals 155 41 71 34 107=253
17
-------
2.2.5. Milwaukee Harbor Plan Case Study - EnviroAtlas Indicators
The Neighborhood Model also was used to code the EnviroAtlas community indicators for Milwaukee,
WI. The data included 361 census block group summarized indicators distributed in 13 data layers
(EnviroAtlas Dynamic Data Matrix website at https://www.epa.gov/enviroatlas/enviroatlas-dynamic-
data-matrix; accessed 7/22/2018). Based on review of EnviroAtlas metadata, 304 indicators were
eliminated from consideration as being irrelevant to the decision context or were statistically redundant
(correlated at R>0.98; Tran 2016). Further, indicators expressed as proportions were selected over
indicators using actual values. Mean and median values were used rather than max or min values.
Incidence of health outcomes (#/yr) were included rather than their value ($). The resulting set of 57
indicators (Table 2.2) comprised the initial selection of context-informed indicators shown in Figure
2.1. The Neighborhood Model was used to assign EnviroAtlas indicators to 17 categories according to
their ecological function (e.g., air or water quality improvements), social function (e.g., health cost and
disease avoidance, presence of historic places, parks, places to play, and where people live), and the
presence of land cover types (Figure 2.4). This yielded 145 indicator-category combinations.
Analyses of Milwaukee Harbor Plan recommendations and the EnviroAtlas indicators yielded common
elements for the translator model (Figure 2.1). The former emphasized equity, stormwater management,
and the built environment. The latter emphasized land cover, demographics, air and water quality, and
where people live. A significant disparity was the availability of health indicators in the EnviroAtlas but
the absence of health recommendations in the Milwaukee Harbor Plan.
In the framework it is assumed that stakeholders were more knowledgeable about the recommendations
of the Milwaukee Harbor Plan than about EGS indicators. Therefore, stakeholders were expected to
translate their recommendations into EnviroAtlas indicators. An analogy is that language translations are
done by adapting broad knowledge of a native language rather than attempting to adapt limited
knowledge of a target language. Therefore, stakeholders were responsible to adapt Milwaukee Harbor
Plan recommendation-categories for translation into EnviroAtlas indicator-categories, not the reverse.
EnviroAtlas" indicators have been selected for their ability to describe
provision, benefits and beneficiaries, and drivers of change
of ecosystem services.
18
-------
EnviroAtlas Indicator Categories
Avgtempreductionday
BODmed
Buff Pet
BWDP Pet
Runoff
Day_Count
Day Low
FPl_lmp_P
FP1 Land P
FPl_Pop_P
Green P
Green PC
IBuff_Pop
lmp_P
lmp_PC
IWDP Pet
K12 Count
K12 Low
Lane PctIB
Lane PctSB
MFor P
MFor PC
MTCSTOR
N02 Asthma Exacerbation 1
N02_Emergency Room VisitsJ
N02AQYr
NonWt Pet
03_Acute Respiratory Symptoms J
03_Emergency Room VisitsJ
03_Mortality_l
03_School Loss Days J,
03AQYr
over_70pct
PlOAQYr
PM25 Asthma Exacerbation 1
PM25 Chronic Bronchitis 1
PM25_Emergency Room VisitsJ
PM25_Hospital Admissions CardiovascularJ
PM25_Hospital Admissions RespiratoryJ
PM25_Work Loss Days J
P25AQYr
PLx2 Pet
RB15 ForP
RB15JmpP
RB15 LABGP
RB15_VegP
SBuff_Pop
S02 Asthma Exacerbation 1
S02AQYr
SUM HOUSIN
Total his count
TSSmed
under_13pc
under_lpct
Wet P
WVT Pet
WVW_Pct
Figure 2.4. Milwaukee Harbor Plan Indicator Evaluation Framework Stage 1: from EnviroAtlas Indicator-
Categories to EnviroAtlas Indicators. Indicators are described on Table 2.2.
19
-------
Table 2.2. EnviroAtlas community data layers and indicators summarized by census block groups. Indicators comprise set of 57 initial
selection of context informed indicators of Figure 2.1. Meta-data and factsheets may be found at the EnviroAtlas Dynamic Data Matrix website
at https://www.epa.gov/enviroatlas/enviroatlas-dynamic-data-matrix (Accessed 8/1/2018). * = Candidate indicators for Milwaukee Harbor
Plan biking recommendations (n = 23 indicators).
BenMAP
Health and pollution changes by block group (based on EPA's BenMap; BenMap website
at
https://vvvvvv.epa.gov/benmap; accessed 8/14/18)
N02_Emergency Room Visits_I
N02_Asthma ExacerbationI
S02_Asthma ExacerbationI
PM25_Work Loss DaysI
PM25_Asthma ExacerbationI
PM25_Chronic BronchitisI
PM25_Emergency Room Visits_I
PM25_Hospital Admissions Cardiovascular l
PM25_Hospital Admissions Respiratory l
03 Acute Respiratory Symptomsl
03_Mortality_I
03_School Loss Days I,
03_Emergency Room VisitsI
Incidence (yr"1) of emergency room visits avoided by NO2 reductions
Incidence (yr"1) of asthma exacerbation avoided by NO2 reductions
Incidence (yr"1) of asthma exacerbation cases avoided by SO2 reductions
Incidence (yr"1) of work days lost avoided by reductions of particulate matter
<2.5 microns.
Incidence (yr"1) of asthma exacerbation cases avoided by reductions of
particulate matter <2.5 microns.
Incidence (yr"1) of chronic bronchitis cases avoided by reductions of
particulate matter <2.5 microns.
Incidence (yr"1) of emergency room visits avoided by reductions of particulate
matter <2.5 microns.
Incidence (yr"1) of hospital admissions for cardiovascular cases avoided by
reductions of particulate matter <2.5 microns
Incidence (yr"1) of hospital admissions for respiratory cases avoided by
reductions of particulate matter <2.5 microns
Incidence (yr"1) of asthma exacerbation avoided by O3 reductions
Incidence (yr"1) of mortality avoided by O3 reductions
Incidence (yr"1) of school days lost avoided by O3 reductions
Incidence (yr"1) of emergency room visits avoided by O3 reductions
BG Population demographics in census block groups. All indicators are referenced by block group (bgrp).
bgrp Census block group identifier; a concatenation of 2010 Census state FIPS, county FIPS code, Census tract
code, and block group number
BG Pop Population demographics in block groups (or demographic data)
SUM HOUSIN* # households
20
-------
underlpct
under_13pc*
over_70pct
NonWt_Pct*
PLx2 Pet*
EduLowGS
% population under 1 year of age
% population <13 years of age
% population over 70 years of age
% population that is not "White Alone" (including white Hispanics)
% population with income under two times the poverty level
Number of K-12 schools in block group with low < 25% greenspace within 100 m
K12_Count*
DayCount*
K12_Low*
DayLow*
Floodplain
FPlLandP
FPlImpP
FPlPopP
# of K-12 schools
# of daycares
# of K-12 schools with < 25% green space within 100 m
# Daycares with < 25% green space within 100 m
Population and land cover by block group in floodplains
% land in 1% annual chance floodplain hazard
% impervious surface in 1% annual chance floodplain hazard
% population in 1% annual chance floodplain hazard
Historical places Number of historic places by block group
Total his count* # historical features
iTREE Ecosystem Services by block group calculated by U.S. Forest Service (USFS) i-Tree (i-Tree website at
https://www, itreetools.org/; accessed 7/25/2018)
Avgtempreductionday Average day time temp reduction (°C)
BODmed
Runoff
MTCSTOR
N02AQYr
03AQYr
PlOAQYr
P25AQYr
S02AQYr
TSSmed
Reduction in median mass of biochemical oxygen demand (kg/year)
Reduction in runoff of the block group (m3/year)
Carbon stored (MT)
Annual mean of air quality improvement by NO2 reduction 1
Annual mean of air quality improvement by O3 reduction
Annual mean of air quality improvement by PM 10 reduction 1
PM 10 = particulate matter between 2.5 and 10 microns
Annual mean of air quality improvement by PM 2.5 reduction 1
PM 2.5 = particulate matter smaller than 2.5 microns
Annual mean of air quality improvement by SO2 reduction 1
Reduction in median mass of total suspended solids (kg/year)
21
-------
LCSum
Land Cover summarized by block group
MForP*
ImpP *
GreenP*
WetP
MForPC*
ImpPC*
Green PC*
% area as tree cover
% area as impervious area
% area as green space
% area as wetlands
Tree cover per capita (m2/person)
Impervious area per capita (m2/person)
Green space per capita (m2/person)
NrRd Pop Roads and population potentially affected by traffic pollution, summarized by block group.
Lane_PctIB* % of busy roadway bordered by < 25 % tree buffer
Lane_PctSB* % of busy roadway bordered by > 25 % tree buffer
SBuff_Pop* # population within 300 m of a busy roadway with > 25 % tree buffer.
IBuff_Pop* # population within 300 m of a busy roadway with < 25 % tree buffer
BuffPct* % population within 300 m of a busy roadway
Park Pop
IWDPPct*
BWDP Pet*
Population near parks summarized by block group.
% population within 500 m of a park entrance
% population not within 500 m of a park entrance
RB LC Land Cover Characteristics of the Riparian Buffers within block group
RB15 LABGP % land area in 15 m buffer
RB 15_ImpP % impervious area in 15 m buffer
RB 15_ForP % tree cover in 15 m buffer
RB 15_VegP % vegetated cover in 15 m buffer
TreeWY
WVT Pet*
WaterWV
Population in block group with an average of <5% forested area within 50 m of the dasvmetric population pixel
% population with minimal potential views of forest
Population in block group with waterbodv of >300 m2 within 50 m of a residence.
WVW Pet
% population with potential views of water
22
-------
2.2.6. Milwaukee Harbor Plan Case Study - Stage 1
Figure 2.3 illustrates the built-out framework for Milwaukee Harbor Plan recommendations and
EnviroAtlas indicators. A stakeholder "enters" the framework via a recommendation
(representing a decision context), establishes links to recommendation-categories, and then
moves down the column to the translator where each recommendation-category is associated
with at least one EnviroAtlas indicator-category (Figure 2.3). From there, the stakeholder is
directed to individual EnviroAtlas indicators that are relevant for those indicator-categories
(Figure 2.4; see Table 2.2 for indicator descriptions). This process yielded candidate
EnviroAtlas indicators (Figure 2.1). A decision context based on all the Milwaukee Harbor
Plan's district-wide recommendations and all 57 EnviroAtlas indicators represented the least
refined demonstration of the framework.
To better demonstrate the workflow of the framework, a scenario was considered where
stakeholders sought EnviroAtlas indicators relevant to the three biking recommendations listed
under Transportation and Utilities in Table 2.1. The Neighborhood Model was used to abstract
the language of these recommendations (Table 2.3) into two dimensions (Built Environment and
Structural/statistical) and four recommendation-categories; one recommendation each for
economics, governance, and infrastructure, and two recommendations for connectivity. In the
translation model, recommendation-categories created 37 links to 13 indicator-categories that
were then linked to 56 EnviroAtlas indicators. The selection process excluded only a single
indicator, proportion of wetland area in the census block group (Wet_P), from consideration.
This lack of specificity in selecting candidate indicators relevant to the biking recommendation
resulted from the many links between indicator-categories and indicators related to trees and tree
canopy. It is reasonable, however, to associate improved biking experiences with increased tree
cover that provide shade and buffer traffic. The EnviroAtlas uses tree cover data to calculate
numerous indicators from the i-Tree (i-Tree website at
http://www.itreetools.org/eco/overview.php; accessed 7/22/2018) and BenMap (Benefits
Mapping and Analysis Program; BenMap website at http://www.epa.gov/benmap; accessed
7/22/2018) models (Pickard et al. 2015) associated with air and water quality and related health
outcomes.
Table 2.3. Milwaukee Harbor Plan bicycling recommendations.
Neighborhood Model
Dimensions /
Recommendation-Categories
Milwaukee Harbor Plan Recommendations
Build Environment /
Connectivity
Locate civic and institutional uses along main corridors or at
prominent intersections to make them easily accessible on foot, by
car, bicycle, bus, or other means of transportation.
Built Environment /
Connectivity
Structural / Governance
Structural / Infrastructure
Complete the Kinnickinnic River Trail with the goal of creating an
uninterrupted, dedicated, and protected bicycle route connecting Bay
View to the Hank Aaron State Trail and Oak Leaf Trail. See the
Improved Access and Mobility catalytic project for more details.
Structural / Economics
Work with Bublr Bikes to ensure that any location within the Harbor
District (excluding lones Island) is within 1/2 mile of a bike share
station.
23
-------
Stakeholders could evaluate the performance of the 56 candidate indicators associated with the
biking recommendations using the quantitative methods in Stage 2 of the framework. This
iterative process could yield a smaller set of indicators explaining most of the variance in
candidate indicators (Figure 2.1; red arrows). However, it seemed likely that stakeholders would
refine the indicator selection process to reduce the number of candidate indicators before
proceeding to Stage 2. The framework provides multiple opportunities to refine the decision
context and the candidate indicators (Figure 2.1; broken lines). For example, stakeholders may
have no reason to consider indicators linking biking recommendations to EnviroAtlas indicator-
categories such as air or water quality improvements, or the presence of water resources or
general land covers. While there might be logical associations expected between biking and air
quality (Johansson et al. 2017), air quality was not referenced in the Milwaukee Harbor Plan
recommendations and therefore cannot be assumed to be relevant in the decision context as
established by the community. Further, stakeholders could exclude indicators involving people
over 70 years old and under one year old (not typical cyclists), floodplains or riparian areas
where biking is not allowed, air quality-related health outcomes, water or air quality, and carbon
storage as being irrelevant to the biking decision context. These additional hypothetical
exclusion criteria and reconsiderations of recommendation-categories and indicator-categories
through the translational model resulted in a reduced set of 23 candidate indicators for Stage 2
evaluation (Table 2.2). This demonstrated how stakeholder feedback could be incorporated at
multiple steps in Stage 1. It is highly desirable that feedback loops involve sustained stakeholder
and public engagement and communication. All rules and criteria for indicator selection must be
documented to ensure transparency and reversibility.
2.2.7. Milwaukee Harbor Plan Case Study - Stage 2
Quantitative methods in Stage 2 of the framework were used to evaluate the set of 23 candidate
indicators relevant to the biking recommendations. The premise of Stage 2 was that all candidate
indicators are relevant to the decision context, but there was yet some sub-set of candidate
indicators that adequately accounted for the variance of the full set of candidate indicators. The
objective of Stage 2 was to determine the amount of variance of the candidate indicator set
explained by various sub-sets of indicators. The initial set of EnviroAtlas community indicators
was reduced from 361 indicators to 57 indicators by expert review (Table 2.2). An additional 34
indicators were excluded in response to the decision context represented by the biking
recommendations. The set of final indicators informed the cost to collect or monitor indicators
and the need to translate and communicate indicators to the public.
The quantitative indicator evaluation process of Stage 2 involved statistical clustering and
multivariate linear regression (SAS Proc Cluster and Proc VARREG, respectively; SAS Institute
2011). Typically, missing data are not a concern for the EnviroAtlas because census blocks
groups are used to define study site boundaries. In this case study, data from all 1,156 census
block groups in the Milwaukee data set were used. While the Harbor District comprised only 15
census block groups, the biking recommendation-categories: connectivity, economics,
governance, and infrastructure, were considered sufficiently generic that they could apply to
neighborhood and other development plans throughout the metro area. Increasing the availability
of bikes and improving trail-to-trail and trail-to-neighborhood connections could be readily
adopted and benefit from the same indicators selected here.
24
-------
Some sources of EGS indicators may have missing data or non-numerical information. As in
Stage 1, it is important that rules for imputing or omitting missing data, redundancy checks, and
normalization routines be reached by consensus and are well documented. Stakeholders, the
public, and experts must understand how rules may affect the selection of final indicators and
how well those indicators address the decision context.
In Stage 2, candidate indicators were clustered to create statistically dissimilar groups containing
statistically similar indicators. The indicator that was most representative of each cluster was
defined as having a minimum 1-R2 ratio (defined as (l-R20Wn cluster) / (l-R2ciosestcluster)). Cluster
representatives were more closely correlated to indicators in the cluster (i.e., l-R20Wn cluster
approach 0) and uncorrected to indicators in the closest cluster (i.e., l-R2ciosest cluster approach 1)
(Tran 2016). Cluster representatives formed a relatively small (n = number of clusters) subset of
candidate indicators based on an objective statistical rationale. However, the cluster
representative indicators may not satisfy stakeholders' needs relative to the decision context.
Stakeholders may therefore include additional indicators from across clusters. Strategies for
explaining more statistical variance by including more indicators must consider the additional
costs or relevancy of those additional indicators to the decision context. For example, Tran
(2016) presented two circumstances where indicators were selected in addition to the cluster
representatives. First, indicators with the next lowest 1-R2 ratio values were selected to increase
the proportion of variance explained. Second, indicators were selected regardless of 1-R2 ratio
value to better distribute indicators across the EnviroAtlas EGS categories. For the biking
recommendation demonstration, the 23 indicators were distributed into nine clusters (Table 2.4).
Selected indicators were regressed using multivariate linear regression (MLR). The MLR
approach served only as an informative tool to determine how much variance of the whole
candidate indicator set was explained by the selected indicators. Total variance explained was a
surrogate for information content of the selected candidate indicators. All the variance of a
selected indicator was considered accounted for in the regression (=1). Each unselected
candidate indicator in a cluster was used as the dependent variable in the MLR. Explanatory
variables were the selected indicator(s) in the cluster. The MRL's adjusted R2 represented the
variance in the unselected indicator explained by the set of selected indicators (Table 2.4). The
total explained variance by the set of selected indicators was the sum of the adjusted R2 for all
indicators. Dividing the total explained variance by the number of candidate indicators yielded
the proportion of the total explained variance.
Tran (2016) proposed this method to "provide dynamic interaction with stakeholders rather than
just an analytical tool for experts." There was no accepted target or threshold value for explained
variance, indicator redundancy, the number of clusters, or the number of candidate or final
indicators. Strict adherence to such a scheme would seriously undercut stakeholders' prerogative
to consider other indicators. The statistical evaluation of candidate indicators in Stage 2 did no
more than to give stakeholders a way to understand how including or excluding particular
indicators impacted the statistical representativeness of the data. It did not give them insights into
the completeness, appropriateness, or relevancy of the indicators for the decision context or other
needs stakeholders have for indicators.
25
-------
Table 2.4. Adjusted R2 from multivariate linear regression (MLR) in Stage 2. N=23 candidate
indicators. Selected indicators are designated as adjusted R2=1.000. Indicator sets tested were: initial = 9
selected indicators based on minimum 1-R2 ratio (or cluster representative); random (= 9 indicators, 1
randomly selected from each cluster); and initial + (BUFF Pct or MFORPC, or SUM HOUSIN) to
simulate need for additional indicators from stakeholders. Indicators are defined in Table 2.2.
Cluster
Indicator
Initial
Random
Buff Pet
MFor PC
SUM HOUSIN
1
Buff Pet
0.433
0.349
1.000
0.433
0.574
Green P
0.994
1.000
0.994
0.994
0.994
Imp P
1.000
0.994
1.000
1.000
1.000
2
Lane PctSB
1.000
0.405
1.000
1.000
1.000
Lane PctIB
0.598
0.443
0.625
0.598
0.606
MFor P
0.666
1.000
0.673
0.715
0.668
SBuff Pop
0.458
0.236
0.465
0.459
0.545
3
NonWt Pet
1.000
0.659
1.000
1.000
1.000
PLx2 Pet
0.685
1.000
0.686
0.685
0.685
under 13pc
0.483
0.242
0.484
0.484
0.515
4
K12 Count
0.275
1.000
0.276
0.275
0.282
K12 Low
1.000
0.251
1.000
1.000
1.000
5
BWDP Pet
1.000
1.000
1.000
1.000
1.000
IWDP Pet
0.993
0.993
0.994
0.993
0.994
6
Green PC
1.000
0.471
1.000
1.000
1.000
Imp PC
0.410
0.249
0.409
0.416
0.413
MFor PC
0.849
0.437
0.849
1.000
0.849
WVT Pet
0.204
1.000
0.204
0.356
0.205
7
Day Count
1.000
1.000
1.000
1.000
1.000
total his count
0.091
0.082
0.092
0.092
0.094
SUM HOUSIN
0.495
0.467
0.621
0.496
1.000
8
Day Low
1.000
1.000
1.000
1.000
1.000
9
IBuff Pop
1.000
1.000
1.000
1.000
1.000
% of total variance explained
(Sum of adjusted R2/N)
72.3%
66.4%
75.5%
73.9%
75.8%
For the biking recommendation example, the 23 candidate indicators were clustered and
regressed. In Table 2.4, cluster representatives had adjusted R2 =1.00 in the "initial" findings.
This set of indicators explained 72.3% of the variance of the candidate indicators. For
comparison, randomly selecting an indicator from each cluster explained 66.4% of the variance.
Three more comparisons were made to estimate the impact of a stakeholder adding back one
additional indicator. Adding one additional indicator to the set of cluster representatives meant
ten indicators were used in the MLR. Each example slightly increased the % of total variance,
compared to the initial (cluster representative indicator only) set. Overall, there was no
guidelines for stakeholders to follow regarding the optimal number of indicators or the amount of
explained variance to assemble a set of final indicators.
26
-------
2.2.8. Milwaukee Harbor Plan Case Study - Conclusions
The EGS indicator selection and evaluation framework depended on expert-stakeholder
interactions to establish a decision context and to iteratively, deliberately, and transparently
consider EGS indicators relevant in that context. The set of final indicators was more
manageable than the entire set of available indicators (in this case from the EnviroAtlas), and
accounts for a relatively large (and, importantly, known) amount of information (e.g., total
variance explained) of the set of candidate indicators. Qualitative methods were used to extract
and classify elements of stakeholder sustainability visions from land use and economic
development plans in terms of EGS. This was necessary to translate between qualitative
stakeholder information and quantitative indicators.
The framework featured multiple opportunities for stakeholders to apply knowledge of the
decision context and purposes for indicators to refine and direct the selection and evaluation of
indicators. The ability of this framework to translate vision into indicators contributes to the
integration of science, policy-making, and stakeholders' involvement in sustainable development
and EGS restoration. The framework may be an effective means for selecting representative
indicators of sustainability and EGS in a scientific, participatory, adaptive, and dynamic fashion.
2.3. Supporting Local Decisions with the Eco-Health Relationship Browser
2.3.1. Background
Translational ecology seeks to "link ecological knowledge to decision making by integrating
science with the social dimensions that underlie today's complex environmental issues" (Wall et
al. 2017). It is part of a larger set of approaches, such as the intentional co-production of science
and policy, useable science (i.e., science developed for applied purposes), and boundary work
(i.e., interdisciplinary efforts conducted at the interfaces of different fields of knowledge), that is
meant to close the gap between the researcher who produces scientific knowledge, the
practitioner who would apply the knowledge, and the public that benefits from that knowledge
(Dilling and Lemos 2011; Meadow et al. 2015; Lemos and Morehouse 2005). Cash et al. (2003)
argued that boundary work can facilitate communication and translation when research is salient
and relevant to the needs of a community for scientific knowledge. Consistent with the goals of
EPA's strategic plan (U.S. EPA 2018b), the EPA's ORD developed various projects in Duluth,
MN that provided opportunities to identify, develop, and demonstrate research that is relevant to
the needs of the City of Duluth for policy formulation and for supporting community decision
making. The work reported here applies data from EPA's Eco-Health Relationship Browser
(Jackson et al. 2013) to help community decision makers incorporate concepts of fairness and
health into public policies.
In this context, and aided by relationships with local officials, EPA's ORD scientists used
qualitative research methods to help city planners integrate public input into a Comprehensive
City Plan. The development of the plan, Imagine Duluth 2035 (Imagine Duluth 2035 website at
http://www.imagineduluth.com/document; accessed 7/22/2018), provided an opportunity for
scientists and stakeholders to apply translational science principles. ORD helped city planners
integrate EGS concepts and outputs from the Eco-Health Relationship Browser into public
27
-------
discourse and policy making on urban planning, social equity, fairness, health, and
environmental restoration. Policy making here refers to findings in Imagine Duluth 2035 and
actions or directives of the city council or administration.
This research addressed two objectives. First, it helped city planners and officials to connect
changes in social, environmental, and economic policy to changes in community health and well-
being. This was consistent with research goals in EPA's Strategic Plan (U.S. EPA 2018b).
Second, it helped the city to understand fairness as a determinant of health and how fairness and
health goals could be included into comprehensive urban planning. This, also, was consistent
with EPA's focus on environmental justice involving, "meaningful (public) involvement,
procedural justice, distributional justice, justice of capabilities, and recognitional justice
collectively recognize the need to provide access to science-based, comprehensive decision-
support tools that engage community members and other stakeholders for building healthy, safe,
and sustainable communities" (U.S. EPA 2016).
The Eco-Health Relationship Browser illustrates linkages between
human health and EGS to provide information about several of our
nation's major ecosystems, the services they provide, and how those
services, or their degradation and loss, may affect people.
2.3.2. Imagine Duluth 2035
In 2016, the City officials of Duluth, MN implemented an extensive community and stakeholder
outreach strategy designed to support the development of a comprehensive plan. Such plans
serve as decadal-scale "road maps for development, the delivery of public service, the layout of
public infrastructure, and the preservation of natural beauty and open spaces" for cities. The
public engagement strategy focused on meeting with different communities (e.g., neighborhood
groups, advocacy agencies, professional and business organizations) at different types of venues
(e.g., community centers, retail settings, workplaces, public transit, libraries, churches, and
schools). Over 60 community engagement events were held in the four months preceding the
initial Imagine Duluth 2035 public meeting (i.e., "kick-off event). More than 4,000 people
responded to an on-line survey, and almost 300 people attended the initial public meeting. The
success of the outreach program generated a large amount of data.
Collected data focused on how Duluthians felt about four focal areas: economic development;
housing; open space; and transportation. Energy and conservation was added as a focal area later
in the process. To facilitate these discussions, the City utilized the governing principles from the
28
-------
2006 Comprehensive Land Use Plan (accessed 7/25/2018). The principles considered land use
policies for utilizing previously developed lands and preserving undeveloped lands, economic
policies supportive of both traditional and emerging economic sectors, and social policies that
promoted equity in public services and strengthened neighborhoods. Throughout the process, the
Mayor of Duluth emphasized the importance of public engagement so that "the comprehensive
plan becomes a reliable and steady guide through times of community change" (Duluth News
Tribune; accessed 7/22/2018). The Mayor also acknowledged the inequitable distribution of
resources and opportunities, justifying the need for two additional guiding principles: fairness;
and health.
Although the values of equity and fairness are important to urban planning (Godschalk and
Rouse 2015) guidance on using them in planning is only emerging. Thus, fairness and health
issues had to be defined de novo from public input to understand their policy implications and
priority in the context of Imagine Duluth 2035. The City of Duluth shared the transcribed public
comments from Imagine Duluth 2035 with ORD researchers for analysis. The ORD scientists
translated the public comments for use by elected officials to facilitate public discussion and to
create fairness and/or health policies, and by city staff to implement resulting policies. The Eco-
Health Relationship Browser was used to illustrate, translate, and communicate connections
between how the public defined fairness and health, EGS, and health outcomes.
Community decision-making processes and policy outcomes were likely to be perceived as fair
when they are free from apparent bias or injustice. To learn what the concept of fairness meant to
Duluth communities, the City deployed an on-line and in-person survey and conducted a
structured inquiry exercise at the Imagine Duluth 2035 kick-off event. The mayor's premise,
corroborated by public health literature (Cummins et al. 2007; Ellen and Turner 1997; Kramer
and Hogue 2009; Morello-Frosch and Lopez 2006), was that the fair and equitable delivery of
city services (including health care and public safety), access to public infrastructure (including
parks and transit), and economic opportunity could mitigate geographic and community-based
health disparities. The analysis considered: 1) how local communities equate fairness with
health; and 2) how fairness and health principles can be integrated into public policy.
2.3.3. Qualitative Analysis of Public Comments and Input
Researches from ORD utilized conventional content analysis (Hsieh and Shannon 2005) to sort
and identify patterns in the public input textual data relevant to fairness and health that were
collected at the Imagine Duluth 2035 kick-off event. All fairness comments were categorized as
being negative, positive, or neutral. Negative comments reflected a need or desire for "less of the
stated condition" (i.e., less housing segregation or less racial profiling). Conversely, positive
comments expressed that there should be "more of something" (i.e., more economic
opportunities, more access to mental health care, more public transportation). There are a few
neutral comments that did not specify a desired directional change in condition or policy. For
example, "fairness is elusive" is policy neutral and did not recommend or involve a decision.
Initial results indicated that housing segregation, economic barriers, and inequitable investments
in neighborhood parks (i.e., programming, maintenance, and access) were community priorities.
Health comments revealed that the community understood health to mean more than the absence
of disease. There were clear indications that health included physical and mental health
29
-------
endpoints with social and environmental determinants (Marmot 2005). Comments revealed the
perception that health and well-being depended on fair access to high quality green spaces.
Green space included both undeveloped natural areas for passive recreation and park spaces with
amenities and programming for active recreation.
The qualitative analysis of public comments led to consideration of approaches and methods for
creating data-informed policy recommendations within Imagine Duluth 2035 and in other city
policy decisions. The EGS concept, already familiar to city staff and some communities,
emerged as an important integrator and translator of environmental, health, and eventually,
fairness issues. The next phase of the research focused on using the Eco-Health Relationship
Browser to link EGS, and environmental and social determinants of health to health outcomes.
Translating and communicating the implications of these linkages to city officials and the public
was critical to provide context for subsequent policy decisions, including those in Imagine
Duluth 2035.
2.3.4. Eco-Health Relationship Browser
The EPA's Eco-Health Relationship Browser was introduced to city staff and communities as a
tool for translating EGS concepts into health and well-being as might be define in Imagine
Duluth 2035. The Eco-Health Relationship Browser previously has been used in urban planning
because it is publicly accessible, easily navigated, and documents the scientific literature on
EGS~>health linkages that are relevant across multiple ecosystems and decision contexts. For
example, it was used to assess hot weather health impacts in Cincinnati, OH and non-motorized
transportation connectivity in underserved communities in Hillsborough County, FL. In the
Duluth case study, the Eco-Health Relationship Browser was used helped city planners connect
Imagine Duluth 2035's high priority issues of access to public green (or open) spaces with
landscape aesthetics and opportunities to enjoy nature. Figure 2.5 from the Eco-Health
Relationship Browser illustrates how the EGS of "Aesthetics and Engagement with Nature"
present in wetlands, urban ecosystems, forests, and agroecosystems (green arrows), potentially
impact many dimensions of health (blue arrows), including positive impacts on health.
30
-------
Bibliography Eco-Health Relationship Browser: Public Health Linkages to Ecosystem Services Topics: Aesthetics Sl Engagement wit »
Click a topic bubble or choose a topic from the dropdown list above.
Hover over linkages (+) to view the relationship between elements.
Details
Description: Aesthetics a
Engagement with Nature
Many people around the world enjoy
recreating, relaxing, and spending time
outdoors. Scientific studies show that
exposure to nature is positively
associated with numerous aspects of
both physiological and psychological
health, as well as with good social
relations. Causal mechanisms for some
of these associations have been
demonstrated in the laboratory: faster
recovery from neurological fatigue
appears to be responsible for the
observed effects that greenness has on
mental concentration and the alleviation
of ADHD symptoms in children.
Exposure to natural scenery, even
through a window or a photograph,
slows the heart rate and calms anxiety.
Humans' innate affinity for nature may
be responsible for observations that
people are preferentially drawn to
community green space, where they are
more inclined to interact with neighbors
while relaxing or recreating. These
interactions are directly beneficial by
increasing social capital (Putnam 2000),
which in turn contributes positively to a
Citations / Sources
Louv, 2005; Putnam, 2000; Wilson, 1984
You are here: Social Relations I Aesthetics & Engagement with Mature
bqpwf or. Sfafyrtr's ffoloftl
Figure 2.5. Screenshot from the Eco-Health Relationship Browser showing that multiple ecosystems
connect with Aesthetics and Engagement with Nature (an EGS) (green arrows) that is, in turn,
connected with multiple endpoints of health and well-being (blue arrows). Scientific literature
documenting these connections are referenced in the box to the right of the bubble diagram.
City planners found the Eco-Health Relationship Browser beneficial for policy discussions
because it used EGS to connect public input with fairness and health. For example, several
respondents specifically noted that their health was related to "urban environments designed to
reduce stress" and "well planned land use policies that consider mental health and quality of life
needs." Using the Eco-Health Relationship Browser, city planners and stakeholders could see the
linkages and access the supporting scientific literature connecting the EGS "Aesthetics and
Engagement with Nature" to improvements of several health outcomes, including cardiovascular
disease, anxiety and stress, social relations, high blood pressure, and mental health. Figure 1.2
illustrates additional EGS-mediated (e.g., clean air) connections between urban ecosystems and
health benefits.
Social
Relations
Wetlands
ADHD
Low Birth
Weight
Aggression
Aesthetics &
Engagement
with Nature
Longevity
Anxiety
COPD
High Blood
Pressure
Healing
Cancer
Cardiovascular
Diseases
Fatigue
Confusion
Happiness
Cognitive
Function
Depression
31
-------
Mental health was further explored by city and communities because disparate access to mental
health services was frequently reported as an important issue of fairness. Figure 2.6 shows that
mental health is related to EGS including heat hazard mitigation, water hazard mitigation and
recreation and physical activity. The Eco-Health Relationship Browser also identified other
health determinants that are related to mental health, including social relations, social support,
and stress. Taken together, the Eco-Health Relationship Browser enabled the City and
stakeholders to identify how physical and mental health in the community could be connected to
open space preservation goals, particularly, fair access, and related ecosystem services. These
formed the basis of policy discussions in the context of Imagine Buhith 2035.
Bibliography Eco-Health Relationship Browser: Public Health Linkages to Ecosystem Services
Topics: Mental Health
•
Click a topic bubble or choose a topic from the dropdown list above.
Hover over linkages (+) to view the relationship between elements.
Aesthetics &
Engagement
with Nature
Details
Definition: Mental Health
Mental health is a state of well-being in
which an individual realizes his or her
own abilities, can cope with the normal
stresses of life, can work productively
and is able to make a contribution to his
or her community.
Organ System
Nervous
Mental Health
Heat Hazard
Mitigation
Water Hazard
Mitigation
Demographic
Mental and behavioral disorders are
estimated to account for 12% of the
global burden of disease. Mental and
behavioral disorders are common,
affecting more than 25% of all people at
some time during their lives. Around
20% of all patients seen by primary
health care professionals have one or
more mental disorders.
Known Contributing Factors
Social Relations, Social Support,
Spirituality/Religion, Physical Health,
Substance Abuse, Trauma, Stress
Recreation &
Physical
Activity
Citations/ Sou rces
WHO: Mental Health. Broadhead et al.,
1983; Seybold and Hill, 2001;
Annerstedt et al., 2012
You are here: Aesthetics & Engagement with Nature 1 Mental Health
Figure 2.6. Screenshot from the Eco-Health Relationship Browser illustrating relationships between
mental health and ecosystem services of water and heat hazard mitigation, recreation and physical
activity, and aesthetics and engagement.
32
-------
2.3.5. Impact of Analysis
The City of Duluth crafted policy recommendations regarding fairness and health that included
elements of the analysis shared by ORD researchers. Draft recommendations of the new guiding
principles (Table 2.5) reinforced the findings that fair access to city services and ecosystem
services was integral to achieving health and health equity. The draft principles were presented
to the public and policy makers for comment and debate. Open forums created space for further
use of the Eco-Health Relationship Browser to illustrate scientific evidence linking fairness in
city policies to health via EGS.
Table 2.5. Imagine Duluth 2035 language draft and final guiding principles for fairness and health.
Draft Guiding Principle
Final Guiding Principle
Developing a healthy community
Supporting health and well-being is a priority.
The City will encourage access to health
resources, quality food, recreation, social
opportunities, and a clean and secure
environment. Policy decisions should
consider health impacts.
Supporting health and well-being is a priority.
The City will actively promote access for all
to health resources, quality food, recreation,
social and economic opportunities, and a
clean and secure environment. Investments
and polices will advance and maximize health
and healthy equity in the City.
Integrate fairness into the fabric of the community
All people should have equitable access to
resources and opportunities that stabilize and
enhance their lives. The City recognizes
historic and ongoing disparities and will
empower inclusive and participatory decision
making that addresses systemic barriers to
success. Investments and policies should
advance and maximize equity in the City.
All people will have equitable access to
resources and opportunities that stabilize and
enhance their lives. The City recognizes
historical and current disparities and will
actively promote inclusive and participatory
decision making that addresses systemic
barriers to success. Investments and policies
will advance and maximize equity in the
City.
Additional data were examined from a survey of fairness and health issues in marginalized
communities conducted by a grass-roots advocacy group sponsored by the Health in All Policies
Coalition (HIAPC 2017; Baum et al. 2014). With ORD's assistance, this group analyzed and
reported survey results to city officials at a public meeting. Results indicated that overburdened
communities defined fairness and health in more functional terms than the general population (as
represented by the Imagine Duluth 2035 surveys). For example, many participants from
marginalized communities were solely focused on immediate, practical, and sometimes, dire
issues of avoiding homelessness or finding employment. They effectively had no opinion on the
higher-order value of fairness because their basic life-needs could not be consistently met. In
some cases, people were grateful for their situations (regardless of quality or constraints) because
they considered what they had better than known or unknown alternatives. Some respondents
considered commenting on fairness or health issues as irrelevant, beyond their control, or even
33
-------
threatening to their situations (e.g., leading to retaliatory eviction or denial of benefits or
services). City officials appreciated the efforts of the HIAPC to broaden the diversity and
representativeness of input to Imagine Duluth 2035.
Because of these analyses the final guiding principles of Imagine Duluth 2035 addressing health
and fairness were stronger than what was originally drafted (Table 2.5). Community engagement
resulted in the guiding principles addressing health and fairness in more imperative (i.e.,
changing the gerund "developing" to the active "develop"), inclusionary (adding the language
"for all"), and affirmative (i.e., that investments will advance and maximize) language. Authentic
and respectful collaborations between city and diverse communities, supplemented with access
and communication between community groups and between the city and stakeholders resulted
in the stronger principles. The Eco-Health Relationship Browser proved to be a useful tool for
translating the results of quantitative analyses of public comments into an EGS context for use in
public discourse. The graphical presentation of the Eco-Health Relationship Browser helped the
city and communities to connect fairness in policy with EGS to health and well-being.
The Milwaukee, Wl and Duluth, MN case studies show how the
EnviroAtlas and Eco-Health Relationship Browser can be used to
translate EGS information to address diverse community objectives
and create a new understanding and lead to better health and well-
being outcomes.
34
-------
3. EGS-Related Tools Using EnviroAtlas
and Eco-Health Relationship Browser for
Community Decision Making
3.1. Introduction
Chapter 2 presented case studies which examined how the EnviroAtlas can inform community
decision making. In other situations, decision makers may need additional tools and approaches
to guide or supplement their use of EnviroAtlas data. Variation in decision contexts necessitates
broad and flexible processes for integrating knowledge and skills from stakeholders in all stages
of decision making, including, problem formulation, information gathering, and trade-off
analyses. The ability for stakeholders and communities to participate should not be limited
simply by technical capacities. The "social side" of decision support (Rodela et al. 2017),
including translational assets associated with relationships, trust, and commitment, also is
important. In the following sections, several EGS related tools are described that vary in their
degree of integration with the EnviroAtlas or the Eco-Health Relationship Browser, their
complexity to implement, and their purpose within a decision-making process. These tools and
approaches can supplement and leverage information in the EnviroAtlas or Eco-Health
Relationship Browser to provide broader flexibility to decision makers.
Multiple tools and methods are available to help translate and integrate EGS concepts into the
decision-making process (Yee et al. 2017). These include those that help decision makers to:
• develop conceptual models causally linking environmental condition to EGS and
benefits;
• characterize what stakeholders care about, including EGS and their benefits;
• define stakeholders' objectives with unambiguous and meaningful indicators and metrics;
• develop and evaluate decision alternatives, including EGS as means to achieve
environmental and health objectives; or
• predict the potential consequences of decision alternatives on EGS or their benefits.
In many cases, these tools and approaches can be directly or indirectly used in conjunction with
the EnviroAtlas and the Eco-Health Relationship Browser. This complementariness may provide
guidance for prioritizing which EGS or health indicators are most relevant to a particular
decision context. Depending on the decision context, additional metrics or indicators from
multiple sources might need to be mapped alongside those in the EnviroAtlas. The spatial data in
the EnviroAtlas can provide a centralized source for data to be exported and used as input to
external tools. In addition to the tools, themselves, the underlying relationships defined by the
Eco-Health Relationship Browser or the functions used to model EnviroAtlas metrics can be
35
-------
adapted to decision contexts for developing conceptual models, setting objectives, developing
alternatives, and predicting outcomes. The following sections provide examples of EGS tools
and conceptual frameworks and their applications.
3.2. Identifying and Mapping Indicators and Metrics Using the EnviroAtlas
Data in the EnviroAtlas can help stakeholders and communities assess the potential
consequences of decisions relative to the quantity, quality, or distribution of various EGS.
Ideally, indicators of EGS should reflect what stakeholders care about, particularly what is at
stake to lose or gain for a given decision. The EnviroAtlas contains more than 400 metrics. The
sheer volume of data and effort required in learning to navigate it can be challenging to new
users (Hall and Khan 2003). The use of translational EGS-related tools can provide guidance for
decision makers to identify relevant metrics by focusing on the objectives of different
stakeholders (Bousquin et al. 2015; Fulford et al. 2016b). For example, structured frameworks
for indicator development, such as the FEGS Classification System (FEGS-CS; Landers and
Nahlik 2013) or the Human Weil-Being Index (HWBI; Smith et al. 2012) can provide starting
points for clarifying likely beneficiaries and stakeholder objectives and ways to measure them in
ways that reduce ambiguity and maximize relevance to communities
3.2.1. Mapping Direct Measures of EGS: The FEGS Framework
With a growing body of EGS data available for consideration, including data in the EnviroAtlas,
classification systems can provide a structured approach to organize data and make it relatable to
the interests of stakeholders as well as across specialized fields in the natural and social sciences
(Carpenter et al. 2009). The Final EGS (FEGS) approach (Chapter 1) explicitly connects EGS to
the people that benefit from them (Yee et al. 2017). Final EGS are defined as those "components
of nature, directly enjoyed, consumed, or used to yield human well-being" (Figure 3.1; Boyd
and Banzhaf 2007). Final EGS are biophysical qualities or features of the environment that need
minimal translation to be relevant to human well-being (Landers and Nahlik 2013).
Final Ecosystem Goods and Services (FEGS)
"[biophysical] components of nature,
directly enjoyed, consumed, or used to
yield human well-being" (Boyd&Banzhai2oo7>
Beneficiary
+
Environmental
Context
Final Ecosystem
Good or Service
m-4
¦spy
Recreational Birdwatchers
Mangroves
Charismatic bird species
Figure 3.1. Illustration of the three elements needed to define FEGS. See Figure 1.1 for the FEGS
conceptual model for decision support.
36
-------
Along a continuum of ecological production (Figure 3.2), FEGS are the last step before a user
actually benefits from the ecosystem. This focus on beneficiaries differentiates intermediate EGS
(IEGS) (e.g., water purification) from FEGS (e.g., potable water in a lake) and their ultimate
social and economic outcomes (SEO) (e.g., drinking water from your faucet). Social economic
outcomes depend both on natural systems and on human labor or capital (Boyd and Krupnick
2013).
Ecological Production
Function Labor and Capita)
Figure 3.2. Conceptual model showing production function linkages between intermediate EGS
(IEGS), final EGS (FEGS), and social and economic outcomes (SEO).
The FEGS Classification System (FEGS-CS; Landers and Nahlik 2013) provides a structured
framework for identifying relevant beneficiaries (Beneficiary Class) for each ecosystem type
(referred to as "Environment Class") to inform the selection of relevant biophysical measures of
FEGS. Cross-referencing all the EGS data layers in the EnviroAtlas with the FEGS-CS provides
an opportunity to:
1. identify the relevant beneficiaries and likely associate categories of FEGS;
2. collate and facilitate the exploration of EGS data; and
3. to assess the availability of national scale FEGS data.
A consideration of FEGS by stakeholders can be an important early step of a decision process to
help clarify the decision context, including what is at stake, what the impacts of potential actions
might be, and most of all, who stands to lose or gain (Yee et al. 2017). Multi-attribute decision
analysis approaches can be an alternative, and often more satisfying, means to inform
environmental decisions rather than monetary cost-benefit methods because of their ability to
compare evaluation measures of different types and scales (Failing et al. 2007; Hajkowicz 2007).
Decision analysis tools such as DASEES (Decision Analysis for Sustainable Environment,
Economy, and Society; U.S. EPA 2012) can help users identify measurable objectives and
explore tradeoffs under alternative decision scenarios.
Prioritization approaches, such as the FEGS Scoping Tool under development, can help users
identify and prioritize stakeholders, beneficiaries, and evaluation measures in a structured,
transparent, and repeatable process. The FEGS Scoping Tool applies three elements that build
upon each other: stakeholder prioritization; beneficiary profile development; and key attribute
identification. In the first step, users prioritize stakeholder groups based on their scores for a pre-
determined set of criteria, such as their likelihood of being impacted, their legal rights to be
37
-------
involved, or to enhance representation for underrepresented groups. Prioritized stakeholder
groups are then translated into the FEGS beneficiary classes they represent. Finally, the
environmental attributes (biophysical measures) most meaningful to those prioritized
beneficiaries are identified and prioritized themselves. This tool provides stakeholders with a set
of FEGS beneficiaries and environmental attributes most relevant for their decision context,
which can then be used to evaluate and compare the potential consequences of decision
alternatives alongside other, non-environmental attributes such as cost to implement alternatives
or job creation.
Combining the use of multi-attribute decision analysis tools (e.g., DASEES, FEGS Scoping
Tool) with the EnviroAtlas will allow decision makers to quickly find data layers in the
EnviroAtlas once they have identified relevant environmental attributes for key beneficiary
classes. For example, urban stakeholders deciding where best to locate or configure new
greenspace could consider a range of potential beneficiaries, environmental attributes (e.g.,
water, air, flora, views), and decision alternatives. Beneficiaries could include property owners
eager or reluctant to have greenspace located near their homes or teachers interested in using
greenspaces for educational opportunities. Once decision makers have prioritized FEGS
beneficiary classes and environmental attributes, relevant data layers from the EnviroAtlas could
be used to visualize and evaluate decision alternatives. For this example, relevant EnviroAtlas
metrics could include distance to park entrance, percent greenspace along walkable roads,
schools or daycares with greenspace nearby, or connectivity of natural spaces (see Table 2.2 for
more information on some of these EnviroAtlas indicators). The FEGS Scoping Tool allows
decision makers to find metrics most relevant to the decision, allowing the decision makers to be
more efficient and strategic in their use of the EnviroAtlas and visualization of their decision
area.
Most metrics in the EnviroAtlas are not FEGS (Figure 3.3), but instead are either dependent on
the addition of human inputs or are IEGS. However, SEO and IEGS can generally be causally
linked to FEGS. Intermediate EGS can be accounted for as ancillary stocks and processes that
contribute to the value of FEGS. For example, pollinator habitat does not directly benefit most
humans but it is an IEGS to a farmer growing pollinator-dependent crops. The actual pollination
of crops is the FEGS from the perspective of a farmer. Therefore, to a farmer, habitat can be
reasonable surrogate for a FEGS. Pollinator habitat is a FEGS for beekeepers because they
benefit directly from its presence.
Similarly, SEO can be causally linked to the FEGS necessary to generate them. The descriptions
in EnviroAtlas metadata help determine the most directly relevant beneficiary sub-categories.
For example, fruit crop yields (an SEO) are dependent on human labor and capital, but also
require the FEGS of pollinators and depredators, the quantity and quality of water for irrigation,
and soil. An additional consideration is that there may be a spatial mismatch between IEGS and
FEGS production. For example, the FEGS of potable water in a lake may be affected by the
IEGS of denitrification occurring in upstream wetlands. Documenting such ecosystem services
production (ESP) linkages can help decision makers select and evaluate additional ecosystem
services metrics that might be relevant to map and monitor.
38
-------
FEGS-CS Beneficiary
Figure 3.3. Number of EnviroAtlas metrics identified as IEGS, FEGS, or SEO for each category of
Beneficiary in the FEGS-CS across environments.
Relationships among IEGS, FEGS, and SEOs are being cataloged in a searchable database under
development. Database users are provided with a description of each EnviroAtlas metric, its
original classification according to EnviroAtlas taxonomy, the spatial resolution of the data, a
weblink to the metadata, a brief explanation for its reclassification with FEGS, along with the
code defining its linkage to IEGS, SEO, ESP, or potential FEGS. More than 14,000 linkages
among IEGS, FEGS, SEO, and ESP have been identified.
Linkages identified among IEGS, FEGS, SEO, and ESP in EnviroAtlas data illustrates the
multiple ways a user might approach operationalizing metrics for decision making. Coding these
relationships in a searchable database and leveraging the FEGS-CS to categorize them by
beneficiary and ecosystem provides a framework for users to navigate the data based on their
specific stakeholders, issues, or objectives. For instance, there are only 11 metrics representing
SEOs at the national scale that are relevant to agricultural irrigators reliant on streams and rivers.
Additionally, a goal of the FEGS framework is a national-scale set of data from which decision
makers can select indicators.
39
-------
A challenge faced when using EGS tools for translational science is the inconsistent use of terms
such as FEGS, IEGS, and SEO in diverse conceptual frameworks (Gret-Regamey et al. 2016).
This can limit the ability of communities, experts, and decision makers to identify the tools and
data they need. The development of meaningful, holistic EGS indicators depends not simply on
the proliferation of data, but on the effective integration of complementary data (Muller and
Burkard 2012). Cross-referencing the EnviroAtlas with the FEGS framework provides a
consistent typology for identifying spatial data metrics of most direct relevance to stakeholder
beneficiaries.
3.2.2. Mapping Measures of Well-being: The Human Well-being Index (HWBI)
In some decisions, the social, health, and economic benefits of EGS are more directly relevant
measures of what stakeholders care about than EGS, themselves. For example, components of
human well-being such as social cohesion, health, and living standards, often resonate with
communities under a variety of decision contexts (Fulford et al. 2017). The HWBI allows for a
multi-factor perspective by translating well-being in terms of eight domains of well-being that
could be impacted by changes in ecosystem services, social services, or economic services
(Smith et al. 2012). The EnviroAtlas provides a mapping platform where maps of HWBI can be
displayed and compared alongside ecosystem services data.
The HWBI is a composite index made up of eight domains of well-being - connection to nature;
cultural fulfillment; education; health; leisure time; living standards; safety and security; and
social cohesion - that are considered globally applicable to human well-being (Smith et al.
2013b; Summers et al. 2014) and that broadly resonate with self-described community goals
(Fulford et al. 2016). Domains are organized into 25 indicators that comprise 80 individual
metrics measuring various components of economic, social, and environmental well-being.
One important functionality of the EnviroAtlas is its ability to integrate nationally consistent
spatial data, including publicly available external datasets, through a web service. For example,
users can select the "Add Data" feature under "Mapping Tools" to add layers from any spatial
data set publicly available as a web service, such as ArcGIS Online. The HWBI and supporting
data are available through EPA's GeoPlatform (EPA's GeoPlatform website at
https://epa.maps.arcuis.com/home/item.html7idKJ6a0c029b2d04149b0dfa4844059e87f;
accessed 8/7/2018), allowing a user to add them as a URL from an ArcGIS Server Web Service.
The HWBI layers can then be viewed alongside other EnviroAtlas datasets (Figure 3.4). The
EnviroAtlas allows HWBI information to be broadly accessible to stakeholders and decision
makers, enhancing the ability to understand implications of decisions on ecosystem services,
human health, and well-being.
40
-------
Envir©Atlas
Interactive Map
IB IB000~ II
ArcGIS World Geocoding
Selected Community:Combined Communities
Mapping Tools
CANADA
Edmontonp
Calgary
Vancouver
Ottawa Montr
Tororr
oston
Detroit
hi lade I phi a
hi not on
Havana
MEXICO
Guadalajara
Port-au
pMexico City
Prince
Search
File
Type
An ArcGIS Server Web Service ~
URL
https://gispub.epa.gov/arcgis/rest'services/ORD/HumanWellBei
Layer Neme
HumanWellBeinglndex
SAMPLE URLfS)
Layer List
Operational layers
BE
HumanWellBeinglndex
State boundaries
HWBI Overall Score
42.84-48.41
48.41 -49.73
49.74 - 50.74
50.74-51.69
51.70-52.57
52.57-53.39
53.39-54.13
54.13-54.94
54.94-55.94
55.94 - 60.69
Figure 3.4. The "Add Data" feature in EnviroAtlas can be used to access HWBI data layers from an
ArcGIS Server Web Service URL and display them alongside other EnviroAtlas data layers. Color
scale ranges from red to dark blue indicating county scores from lower to higher HWBI.
As an example, land use decisions can affect the benefits that nature provides to people in both
planned and unplanned ways. Decision makers in the Tampa Bay, FL watershed explored how
alternative land use scenarios might impact sustainable delivery of ecosystem services and
human well-being (Russell and Harvey 2016). The 2000-2010 HWBI scores for counties in the
Tampa Bay area can be used for comparing communities (Figure 3.5) and understanding how
ecosystem services in different communities might be impacting well-being. Well-being in the
Tampa Bay area tended to be moderately high, compared to other parts of Florida, partly due to
higher scores for cultural fulfillment, education, leisure time, and living standards. Connection to
nature, health, and safety, however, tend to be lower than other parts of Florida.
41
-------
Figure 3.5. Human Well-being Index (IIWBI) composite scores for counties in Florida, and
individual domain scores for counties within the Tampa Bay area.
A side-by-side comparison of the well-being maps with ecosystem services maps can help
communities and decision makers explore how ecosystem services might be impacting well-
being. For example, some counties in the Tampa Bay area have relatively lower health scores
than other counties. The EnviroAtlas layers can be used to explore in a finer level of detail how
ecosystem services could be modified to improve well-being in each county, such as by planting
trees to remove air pollutants (e.g., % particulate matter removed annually by tree cover) or
creating additional greenspace (e.g., % of population within walking distance to a park) to
increase outdoor activity or improve mental health (Figure 3.6).
Domains of Well-Being
1 Connection to Nature
2 Cultural Fulfillment
3 Education
4 Health
5 Living Standard
6 Leisure Time
7 Safety & Security
8 Social Cohesion
too 200 Miles
I I I
42
-------
HWBI Domain
4 Health
EnviroAtlas Metrics
Figure 3.6. HWBI Health scores for Tampa Bay area counties (left) mapped alongside two
community-scale EnviroAtlas metrics (right) to illustrate how visualizing ecosystem services could
inform health and overall well-being.
3.3. Integrating EnviroAtlas Layers with External Scenario Modeling Tools
An important step in any decision process is to estimate the potential consequences of decision
alternatives on stakeholder objectives (Gregory et al. 2012). Depending on time, budget, or
technical considerations, information on consequences could come from group deliberations or
expert judgements. Data from the EnviroAtlas or the Eco-Health Relationship Browser, can
inform these deliberations. In other cases, decision makers may prefer to use predictive modeling
tools to quantitatively evaluate and compare decision scenarios. Additional information or data
may need to be collected, or analyses conducted, to align multiple endpoints under consideration.
Several EGS tools are available to assist researchers and decision makers to develop indicators
and apply them to model alternative decision scenarios (Bagstad et al. 2013; Tallis et al. 2013;
Johnston et al. 2017a). Predictive models are used to forecast alternative futures that reflect
decision scenarios. One example, the EPA tool for Visualizing Ecosystem Land Management
Assessments (VELMA; Abdelnour et al. 2011; 2013), utilizes biophysical models and
visualization tools to simulate effects of climate and land use scenarios on a suite of ecosystem
services. Other EGS modeling tools use structured frameworks to identify and estimate
endpoints that are linked to site-specific economic indicators, such as Rapid Benefit Indicators
(RBI) (Mazzotta et al. 2016), or to model the impacts of ecosystem services on indicators of
human well-being, such as the Services^HWBI framework (Summers et al. 2016). Some
examples of how these tools can be used synergistieally with the EnviroAtlas and the Eco-Health
Relationship Browser to estimate consequences of decisions are described below.
Percent removal PM10 (%)
^ 0.73
Greenspace per capita (m2/person)
^ 7.4-11.5
0.7-7.3
6.1 - 6.6
5.1 - 6.0
0.0-5.0
43
-------
3.3.1. Modeling Ecosystem Services Under Alternative Scenarios: VELMA
The VELMA tool is an eco-hydrological model combined with data visualization frameworks,
that predicts hydrologic flows and the fate and transfer of nutrients and contaminants across
multiple spatial (e.g., from small plots to regional river basins) and temporal (e.g., from days to
centuries) scales (Abdelnour et al. 2011 and 2013; McKane et al. 2011). For example, VELMA
quantifies how different climate and land use scenarios, including green infrastructure (GI)
strategies for mitigating stormwater quality and quantity, affect tradeoffs among a large suite of
EGS such as clean drinking water, flood control, food and fiber production, fish and wildlife
habitat provisioning, and climate regulation (Figure 3.7). Opportunities for leveraging the
capabilities of the EnviroAtlas with VELMA fall into two broad areas: 1) the potential of the
EnviroAtlas to facilitate VELMA modeling efforts; and 2) how runoff and nutrient loading
predicted by VELMA might be combined with the EnviroAtlas outputs to facilitate community
outreach and better decision making. The approaches to integrate VELMA with the EnviroAtlas
described here could extend to other EGS mapping and modelling tools (e.g., EPA H20 (Russell
et al. 2015) and In VEST (Tallis et al. 2013)).
Carbon
wtrogen
Water
Cycling
Climate & Land Use Effects Simulated
• Hydrology: streamflow, vertical & lateral
flow, evapotranspiration, soil moisture,
stream temperature
• Plants & Soils: uptake, transformation and
transport of carbon, nutrients and toxics
from terrestrial to aquatic systems
Ecosystem Services Simulated
• Water quality (nutrients, toxics...)
• Water quantity (floods, low flows)
• Food & fiber production
• Climate regulation (C02, N20, NOx)
• Fish & wildlife habitat
To
Stream
Figure 3.7. VELMA conceptual model (U.S. EPA 2014).
The EnviroAtlas' spatially and temporally-rich database can potentially facilitate and extend
VELMA ecosystem service assessments in a number of ways, including: (1) providi ng initial
watershed characterization metrics for model calibration and validation; (2) informing VELMA
scenario development; and (3) identifying and prioritizing watersheds needing high-resolution
model analyses for evaluating EGS under alternative decision scenarios. These features might be
particularly useful in designing VELMA scenarios focusing on ecosystem or EGS restoration.
The EnviroAtlas has a rich catalog of data layers to draw upon that could be useful for designing,
developing and parameterizing VELMA simulations, including past and present land cover,
gridded soils data, amount of carbon storage, reduction in annual runoff due to tree cover,
information about the amount of land enrolled in a Conservation Reserve Program, and
information about built environments and human demographic data (Table 3.1). In addition, the
44
-------
EnviroAtlas provides historic climate data and modeled future climate scenarios. By providing
ready access to well organized, spatially explicit data, the EnviroAtlas could expedite and
improve VELMA applications aimed at identifying potential impacts of changes in climate and
land use on ecosystem services in natural and built environments. The EnviroAtlas organizes and
provides data at both a national extent (e.g., 30-m resolution land cover and 12-digit HUCs and
at higher resolutions for more than 20 target communities (e.g., 1-m resolution land cover,
census data and models) with approximately 100 layers per community. Because VELMA was
designed to support ecosystem service assessments across a similarly wide range of scales, the
EnviroAtlas can potentially provide ancillary information concerning human impacts on the
capacity of ecosystems to sustainably provide essential services. Furthermore, the relationships
in the Eco-Health Relationship Browser can potentially assist with the translation of VELMA's
ecosystem service outputs to benefits for human health and well-being.
Table 3.1. Example EnviroAtlas data layers, and spatial scales available, useful for hydrologic
modeling.
Data Layer
Availability
Elevation
National, Community
Landcover
National, Community
Percent tree cover
Community
Reduction in mean load of total nitrogen, phosphorous due to tree
cover
Community
Reduction in annual runoff due to tree cover
Community
Total carbon stored by tree cover
Community
Acres of land enrolled in a Conservation Reserve Program
National
Agricultural water use
National
Soil drainage class
National
Surface runoff from agricultural land
National
Non-tile drainage system subsurface water flow from agricultural
lands
National
Candidate ecological restoration areas
National
Morphological spatial pattern analysis data layers
National
Potentially restorable wetlands
National
Average annual precipitation
National
Phosphorous application as manure
National
Synthetic nitrogen fertilizer application
National
45
-------
In addition to using the EnviroAtlas to facilitate VELMA's place-based ecosystem service
applications, VELMA outputs could be used to extend co-located EnviroAtlas applications. The
VELMA model produces ecosystem services outputs that might prove useful to supplement
EnviroAtlas community data layers. While there is some similarity in modeled ecosystem
services (Table 3.1), VELMA can provide highly integrated information describing how multi-
jurisdictional decisions impact a wide range of services and tradeoffs across complex urban-rural
landscapes managed for different environmental, economic and cultural goals. For example,
VELMA is being applied to identify GI best practices for enhancing water quality and ecosystem
service co-benefits in several cities - Seattle, WA; Duluth, MN; and Mobile, AL. Green
infrastructure involves the establishment of riparian buffers, rain gardens, green roofs, pervious
surfaces, constructed wetlands, and other measures to intercept, store and transform nutrients,
toxics and other contaminants in storm water that might otherwise reach surface and ground
waters (McKane et al. 2011). Although many of these communities are already using or planning
to invest in GI, sufficient data often do not exist to support informed decisions about where, how
much and what kinds of GI will be required to meet water quality goals at local and regional
scales. The VELMA model is designed to provide this information and to identify optimal
strategies for achieving target water quality criteria (e.g., Total Maximum Daily Loads or
TMDLs) and associated ecosystem service co-benefits, including improved fish and wildlife
habitat, carbon sequestration, and recreational opportunities (U.S. EPA 2014). There are
advantages to combining VELMA outputs with existing EnviroAtlas community data layers
whenever possible. For example, combining VELMA's GI water quality results with ancillary
EnviroAtlas hydrologic data layers, such as the i-Tree Hydro urban analysis data (i-Tree in the
EnviroAtlas), could lead to important insights and synergies. This could result in several
additional useful ecosystem services layers, or more resolved versions of existing layers (e.g.,
denitrification, carbon sequestration, stream temperature, GI effectiveness).
The full utility of the EnviroAtlas data as inputs for VELMA, or the challenges of integrating
VELMA model outputs as new spatial data in the EnviroAtlas have not been fully explored.
However, in some cases, mismatched spatial and temporal scales may preclude merging.
Fortunately, the EnviroAtlas has data submission guidelines available for data, widgets, and
tools. As the EnviroAtlas is currently hosted in the EPA's GeoPlatform Hosting Environment
with data published as web services (EnviroAtlas Web Services website at
https://www.epa.gov/enviroatlas/enviroatlas-web-services; accessed 8/7/2018), both the
GeoPlatform and Mapping Services templates can provide a common framework for sharing data
inputs, outputs and tools. Additional considerations for improving interoperability are described
in Chapter 3.3.4, To explore and better understand opportunities and challenges, pilot studies
could be conducted for a subset of locations where both tools are being applied.
3.3.2. Assessing Benefits Under Alternative Scenarios: Rapid Benefit Indicators
Approach
The RBI approach is a process for assessment based on non-monetary benefit indicators
(Mazzotta et al. 2016). It is translational because it was intended to be used in conjunction with
existing ecosystem service assessment approaches and tools, such as the EnviroAtlas, to connect
changes in the availability of EGS to where and how people benefit from those goods and
services. The RBI approach was originally developed for use with urban freshwater wetland
46
-------
restoration, but the general approach and indicator framework can be adapted to work with other
types of environmental changes or within different ecological systems. Like FEGS-CS, its focus
is to link FEGS (or indicators of FEGS) explicitly to beneficiaries.
The RBI approach involves a five-step process, starting off with identifying stakeholders and
their objectives. Based on these objectives, relevant ecosystem services and resulting benefits are
selected to compile indicators based on five questions:
1. Can people benefit from an ecosystem service, including whether there is demand,
sufficient quality and quantity, and necessary complementary inputs for the service?
2. How many people will benefit?
3. How much are people likely to benefit?
a. What is the quality of the service?
b. Are there substitutes for the service or is the service scarce?
c. What is the quality of complementary services?
d. How strong are people's preferences?
4. What are the social equity implications?
5. How reliable are benefits expected to be over time?
Indicators for five benefits (Figure 3.8) have been developed and integrated into two tools that
help users apply the RBI approach (Mazzotta et al. 2016). The RBI checklist tool can be used for
recording results of "desktop" or field analysis, and the RBI spatial analysis tools can be used to
compile indicators based on spatial datasets. The RBI approach and five-step process can be
followed for developing indicators that better fit the decision context, data availability or
ecological system of interest. Information in the EnviroAtlas can be used synergistically with an
RBI analysis to: 1) facilitate identification of restoration sites; 2) provide input data for
calculation of RBI; and 3) provide supplemental metrics for consideration alongside RBI.
The RBI approach (Figure 3.8) integrates information on how much restoration of EGS could
benefit people, such as how many people would benefit, how much of that EGS is already
available or available through alternative substitutes, EGS preferences, social equity
implications, and reliability of EGS delivery in the future.
Data sources from the EnviroAtlas can be used as input data to calculate RBI (Table 3.2), or to
supplement the existing RBI with additional relevant information. As a demonstration of how
data in the EnviroAtlas can help facilitate an RBI analysis, benefits provided by wetlands
restoration projects were assessed for the Hillborough County Environmental Lands Acquisition
and Protection Program, north of Tampa Bay, FL. Potential wetland restoration sites for Tampa
Bay watershed were generated using areas classified as having high potential for restoration in
the Potentially Restorable Wetlands on Agricultural Land (PRWA) dataset from the EnviroAtlas
(Pickard et al. 2015).
47
-------
Ecosystem Service How people benefit
Reduced Flood Risk: The risks from
Flood water regulation floods to people and structures are
reduced.
n
, . , Scenic Views: People can enjoy
Scenic landscapes
scenic views.
8ifii8
Environmental Education: People
Learning opportunities can benefit from studying nature or
from enhanced connection to nature.
Recreation: People can enjoy
Recreational opportunities
recreation
. , Bird Watching: People can watch
BlrdS . k" j-l
or hear birds.
Figure 3.8. Benefits assessed using Rapid Benefit Indicators (Mazzotta et al. 2016).
Areas with this classification had drainage and wetness characteristics to allow for wetland
restoration, and, once restored, for the production of wetland EGS. In addition, these areas
presented fewer land use conflicts because they were also classified as agricultural. Restoration
sites were further reduced to avoid double counting of existing benefits by removing portions
that overlap with existing wetlands, as defined by the National Wetlands Inventory (NW1
website at https://www.fws.gov/wetlands/; accessed 7/25/2018). Data from the EnviroAtlas were
combined with data from other sources to calculate wetland RBI (Table 3.2), The indicators
were then compared for potential restoration sites (Figure 3.9).
Table 3.2. Datasets used and their source in the example RBI analysis.
Parameter
Source
Link
Population Raster
EnviroAtlas National Dasymetric data
www. epa. gov/ enviroatlas
Flood Zone
FEMA National Flood Hazard Layer
https://msc.fema.gov
Dams/Levees
FEMA National Flood Hazard Layer
https://msc.fema.gov
Educational
Institutions
Homeland Infrastructure Foundation Public
Schools
http ://hifld-dhs-
gii.opendata.arcgis.com
Bus Stops
Open Street Maps
www. openstreetmap. org
Trails
Open Street Maps
www. openstreetmap. org
Roads
Open Street Maps
www. openstreetmap. org
Wetland Polygons
USFWS National Wetlands Inventory
www.fws.gov/wetlands
Land Use/
Greenspace
EnviroAtlas Communities Land use data
www.epa.gov/ enviroatlas
Social Vulnerability
Centers for Disease Control Social
Vulnerability Index
http s: //svi. cdc. gov/
48
-------
Benefit
Indicators
Site 1
Site 2
Site 3
Site 4
3.2 How Many Benefit?
2.5 mi downstream of site and in flood zone
3524
27
2182
3587
3.3.A Service Quality
Area of restoration site (acres)
10.45
0.89
75.84
10 23
a:
Features that increase retention volume?
Not Calculated
-o
o
o
3.3.B Scarcity
Dams and levees 2.5 mi downstream?
No
No
No
No |
Wetlands within 5 mi (number or % area)
Not Calculated
LL
3.3.C Complements
Not applicable (NA)
NA
NA
NA
NA i
3.3.D Preferences
Are people worried about flood risk?
Not Calculated
Number within 160 ft of site
0
0
0
3
3.2 How Many Benefit?
Number within 325 ft of site
0
0
5
5
%
Weighted number who benefit
0
0
1.5
3.6
o
Are there roads or trails within 325 ft of site?
Yes
No
No
Yes
3.3.A Service Quality
Aesthetic features or characteristics?
Not Calculated
C
green space within 2/3 mi of site
Not Calculated
0£
3.3.B Scarcity
green space within 1 mi of site
Not Calculated
green space within 12 mi of site
Not Calculated
3.3.C Complements
Infrastructure supporting recreational activities?
Not Calculated
3.3.D Preferences
Are there additional features on the site?
Not Calculated
U)
c
!c
3.2 How Many Benefit?
Number within 0.2 mi of site
25
0
48
35
Are there roads or trails within 0.2 mi of site?
Yes
Yes
Yes
Yes
o
¦4-J
CO
3.3.A Service Quality
Will the site support rare or unique species?
Not Calculated
3.3.B Scarcity
Not applicable (NA)
NA
NA
[ NA
NA ||
TJ
L—
3.3.C Complements
Supporting infrastructure or habitat on site?
Not Calculated
in
3.3.D Preferences
Will people be interested in birds at the site?
Not Calculated
V-
o
3.4 Social Equity
Score
6.59
0
0
0
JZ
4-*
O
3.5 Reliability
Score
Not Calculated
Figure 3.9. Sample comparison of calculated RBI indicators for four restoration sites in the Tampa
Bay, FL area (see Russell and Harvey 2016). Blue shaded squares represent sites with above average
results; red shaded are below average results; grey squares indicate no differences across sites. Some
indicators were not calculated for this demonstration.
49
-------
3.3.3. Linking EGS to Well-being: The Services^HWBI Framework
Beyond quantifying and mapping indices of human well-being (HWBI; Chapter 3.2.2), a
broader framework connecting environmental, economic, and social dimensions to human health
and well-being endpoints is presented in Figure 3.10 as the Services^HWBI framework. The
framework can be used to qualitatively and quantitatively relate the influence of EGS flows to
measurable endpoints of human well-being alongside the influence of social service flows and
economic service flows (Smith et al. 2014a; Summers et al. 2016). The Services-^HWBI
framework utilizes national quantitative indicator data to develop predictive (regression) models
describing relationships between EGS and human well-being (Smith et al. 2014b; Summers et al.
2016). The regression models can be used to forecast the potential consequences of changes in
social, economic, and ecosystem services for human well-being. The models provide links
between the EnviroAtlas and the Services-^HWBI framework for local decision support,
including: 1) using EnviroAtlas indicators to develop scenarios for applying the national and
community-scale Services^ITWBI regression models; and 2) using EnviroAtlas indicators to
develop customized Services^HWBI relationships for finer-scale analyses.
HWd=/(Se,Ss,Sec)
Social Services
• Activism
• Communication
• Community initiatives
• Education
• Emergency preparedness
• Family services
• Healthcare
• Justice
• Labor
• Public works
Ecosystem Services
Air quality
• Food, fiber, fuel
Greenspace
Water quality
Water quantity
Human Well-being
Connection to Nature Leisure Time
Economic Services
Capital Investment
• Consumption
• Employment
• Finance
• Innovation
• Production
• Redistribution
Cultural Fulfillment
Education
Health
Living Standards
m ft
Safety and Security
Social Cohesion
Figure 3.10. Services-MiWBI Index framework (simplified from Summers et al. 2016). Human well-
being domains (HWd) are a function of ecosystem (Sc ). social (Ss), and economic (Sec) services. The
EnviroAtlas can provide EGS data (Se).
Stakeholders may explore EGS data in the EnviroAtlas to determine how to modify input
variables in the national Services~>HWBI regression models (Smith et al. 2014b; Summers et al.
2016). For example, EnviroAtlas metrics can be used to map and assess how EGS, such as
metrics of air quality or greenspace, might be impacted by alternative decisions. The
Services-^HWBI regression models can then be used to explore the potential consequences of
EGS changes for domains of human well-being (Figure 3.11; Summers et al. 2016). At a
50
-------
national scale, for example, increases in water quantity tend to be positively correlated with
health, living standards, and safety and security (Summers et al. 2016).
I
15
$
-20
Connection to
Nature
Cultural
Fulfillment
t
Living
Standards
—©—
Safety &
Security
Social Cohesion
A
Human Weil-Being
Index
• Model Predictions Without Correlation Factors
-e-Base Line
Model Predictions with Service Correlation Factors
Model Predictions with Service and Domain Correlation Factors
Figure 3.11. Example output from the Services-^ HWBI spreadsheet tool illustrating how changes
in ecosystem services scores could impact the eight domains of well-being relative to baseline
estimates (from Summers et al. 2016). Regression models can also account for correlations among
services or domains.
The regression models that drive this forecasting are based on nationally available county-scale
data using the original Services-^HWBI metrics and are intended to provide screening or
scoping to inform potential consequences of decisions. The key to maintaining integrity of the
Services-^HWBI framework for different decision contexts is that the models are developed at
the level of indicators. Within the scope of an indicator, metrics are flexible and can be modified
(Smith et al. 2014a). As long as EnviroAtlas metrics are reasonable surrogates for the indicators
in the national Services-^HWBI framework (Table 3.3), the regression models can be used to
explore how changes in EGS might impact well-being. This assumes the relationships at a
national scale are reasonable representations of what may be happening at a local scale, but at a
minimum, can provide opportunities for discussion or raise awareness of where more
information may be needed. In conjunction, the EnviroAtlas can provide a source for developing
EGS scenarios that may be more meaningful to communities or certain decision contexts than the
metrics in the national Services-^HWBI. Furthermore, EnviroAtlas metrics are available at a
finer spatial resolution than the county-scale national HWBI, including census blocks for
EnviroAtlas communities (Figure 3.12), which may be more appropriate for exploring site-
specific decision scenarios rather than county-scale data.
51
-------
Table 3.3. Metrics in the EnviroAtlas that could be used as surrogates to replace the national HWBI
metrics for each of the Ecosystem Services indicators in the Services-^ HWBI framework.
Ecosystem
Service
HWBI
Indicator
Original HWBI Metrics
Surrogate EnviroAtlas Metrics
Air Quality
Usable
Air
• Percentage of days with good
or moderate air quality1
• Total or percent carbon
monoxide, nitrogen dioxide,
ozone, particulate matter, or
sulfur dioxide removed annually
by tree cover (kg/yr) *
• Percent tree cover within 26m of
a road edge*
• Tree cover per person*
• Average reduction in ambient
daytime temperature*
Food, Fiber,
and Fuel
Provisioning
Raw
Materials
• Copper reserves (MT)2
• Gold reserves (MT)2
• Lead reserves (MT)2
• Silver reserves (MT)2
• Zinc reserves (MT)2
No matching EnviroAtlas data
available
Food, Fiber,
and Fuel
Provisioning
Food and
Fiber
• Commercial fishery landings
(MT)3
• Saw-timber tree volume on
forest land4
• Total factor productivity5
• Agricultural land per capita*
' Hectares, number of types, or
annual yield of cotton, fruit,
grain, or vegetable crops^
' Richness of harvestable, game, or
fur-bearing species^
Food, Fiber,
and Fuel
Provisioning
Energy
• Recoverable coal reserves at
producing mines6
• Crude oil proved reserves6
• Natural gas proved reserves6
• Uranium reserves6
• Average annual daily potential
solar energy^
• Average annual daily potential
wind energy^
Greenspace
Recreation
and
Aesthetics
• Percent of people who did non-
consumptive activity within a
mile of their home7
• Percent of people who took a
wildlife observation trip within
their state7
• Area of blue space per person7
• Recreation demand for hunting,
bird watching, or freshwater
fishing^
• Percent of residential population
with water or trees in viewshed *
Percent of population within
walking distance to a park*
• Vertebrate or bird species
richness^
Greenspace
Natural
Areas
• National Parks acreage8
• Percent area designated rural
park or wildlife area5
Greenspace or tree cover per
capita*
• Percent green space*
' Percent natural land cover'
52
-------
Ecosystem
Service
HWBI
Indicator
Original HWBI Metrics
Surrogate EnviroAtlas Metrics
• Number of National Park
visitors within a state8
• Unclassified land use acres
(e.g., marsh, swamp, bare rock,
tundra)5
• Acres of land enrolled in
Conservation Reserve Program^
• Percent of land with protected
status^
Water
Quality
Usable
Water
• Percent of water bodies
assessed as 'Good'1
• Percent days under a beach
action (e.g., closure)1
• Tree or vegetative cover within
15-50m stream/lake buffer*
' Reduction in annual runoff due to
tree cover*
' Reduction in copper, nitrogen,
phosphorous, total suspended
solids loading due to tree cover*
• Total impaired stream length by
nutrients, nuisance species,
oxygen depletion, pathogens,
mercury, or other metals^
Water
Quantity
Available
Water
• Average monthly Palmer
Hydrological Drought Index3
• Water Sustainability Index9
• Daily domestic water use*^
• Agricultural water use'
• Thermoelectric water use'
• Water supply from reservoirs^
Data Sources: 1. U.S. EPA; 2. U.S. Geological Survey; 3. National Oceanic and Atmospheric Administration; 4.
U.S. Department of Agriculture Forest Inventory and Analysis Data Base; 5. U.S. Department of Agriculture
Economic Research Service; 6. U.S. Energy Information Administration; 7. U.S. Census Bureau; 8. National Park
Service; 9. Natural Resources Defense Council
fNational-scale availability (e.g., 12-digit HUCs)
*Available select EnviroAtlas communities (e.g., census block groups)
53
-------
Original HWBI Indicators
Surrogate EnviroAtlas metrics
Food, Fiber, Fuel
HIGHER
LOWER
Air Quality
HIGHER
LOWER
Greenspace
r
LOWER
Water Quality
Pi HIGHER
fl
LOWER
Water Quantity
HIGHER
LOWER
Greenspace per capita (m2/person)
^ 7.4-11.5
|b 6.1-6.6
5.1 -6.0
0.0-5.0
Reduction in phosphorous load (kg/yr)
0.60 -425.71
0.21 -0.60
0.06 -0.21
0.01 -0.06
[ j 0.00-0.01
AgriculturaI water use (xlO6 gallons/day)
Percent removaI PM10 (%)
-2.57
- 0.73
-0.47
-0.31
-0.20
Figure 3.12. County maps of ecosystem services in the original HWBI (left) and example
EnviroAtlas surrogate metrics (right). HWBI service indicators are on a relative scale (see Figure 3.6),
As an alternative to applying the national county-scale Services-^HWBI models, users may wish
to use the Services^HWBI framework to develop their own analytical models by substituting,
for example, metrics from the EnviroAtlas if they are more relevant to a particular decision
scenario or at a more appropriate spatial scale. The Services-^HWBI indicators and models can
be customized by replacing the nationally available county-scale data with alternative local-scale
data or locally-relevant metrics (Smith et al. 2014a; Smith et al. 2015), which may better reflect a
54
-------
given population, such as American Indian Alaska Natives (Smith et al. 2015), children (Buck et
al. 2018), or residents of Puerto Rico (Olando et al. 2017). The EnviroAtlas can provide a source
for alternative ecosystem services metrics that may lead to development of more meaningful
models (Table 3.3), and at finer spatial scales (Figure 3.12), than those in the national
Services-^HWBI models. For example, the EnviroAtlas does not provide analogous or surrogate
metrics for the resource extraction metrics currently used to calculate the national
Services->HWBI relationships (Table 3.3). Alternative energy metrics, such as potential
availability of wind and solar power, might be used instead based on local relevance and the
decision context (e.g., investment in energy sources for a community). This ability to customize
the Services-^HWBI indicators and models, supported in part by EnviroAtlas data, allows users
greater flexibility to explore the impacts of changes in EGS on human well-being.
3.3.4. Making External Scenario Modeling Studies Available Through EnviroAtlas:
WEDO
Watershed and Economic Data for Interoperability (WEDO; Parmar et al. 2016) was designed to
facilitate discovery and evaluation of modeling studies for reuse by other modelers (Figure
3.13)
Administrator
publishes
WEDO shape
file to
EnviroAtlas
Download
watershed
models from
WEDO
Desktop
Watershed
Modeling
studies to
Watershed and
Economic Data
Interoperability
(WEDO)
User
returns to \
WEDO from
EnviroAtlas after
selecting a stream
for detailed
evaluation and
integration
Figure 3.13. Watershed and Economic Data Interoperability (WEDO) workflow overview for
discovery, evaluation and integration of watershed modeling studies for reuse.
55
-------
The WEDO tool utilizes the EnviroAtlas to browse and locate studies of interest from the
subwatershed and streams data layer (Figure 3.14).
V»EPA
tates Environmental Protection Agency
Espanoi | 1 1 TiehgViet 1 SROj
Learn the Issues
Science & Technology Laws & Regulations About EPA
Q.
EnviroAtlas Contact Us Share
You are here: EPA Home » EnviroAtlas » Interactive Map
Clear Data Layer
Layers Matrix
Ecosystem Services
and Biodiversity
People and
Built Spaces
Supplemental
Maps
Analysis
Tools
Mapping
Tools
EnviroAtlas
Full Screen Basemap Navigate
Map Legend
NHDArea
PI Crther
J Area of Complex Channels
| | Sea/Ocean
Q Wash
Boundaries
HUC (12 Digit) Subwatershed
Hatit'ude:32r669:425lL'ongifudeTl83;059339]
Figure 3.14. EnviroAtlas screen shot of subwatershed and streams data layer. Stream and river
segments are blue (NHD Lines), while subwatershed boundaries are yellow (HUC-12).
Modeling studies that are published to WEDO become available to the community of modelers
who then can translate results of broader use by stakeholders. The WEDO tool is designed for
reuse by making all components of a modeling study available, including all input data (e.g.,
meteorology, soil types, land use), model control files (parameters, initial conditions), the model
executable used to perform the assessment, and all output files. The WEDO database saves all
files as a single compressed (zip) file for download to a modelers' computer.
The WEDO tool leverages the EnviroAtlas as a means of locating sub-watersheds and streams of
interest using the built-in query tools, and streams that have published modeling studies are
color-coded red (Figure 3.15),
56
-------
\
I
Figure 3.15. EnviroAtlas screen shot with streams with WEDO modeling studies published color-
coded red. Stream and river segments are blue and red (NHD Lines), while subwatershed boundaries are
yellow (HUC-12).
When a user locates their stream of interest in red, they then click on the stream itself to see what
modeling studies are available. Modeling studies currently supported are the HSPF (Hydrologic
Simulation Program Fortran) and SWAT (Soil Water and Assessment Tool) models. Support for
additional models, including VELMA, are planned. A user is then able to click on the hyperlink
provided in the pop-up to navigate to the WEDO study information page (Figure 3.16).
WEDO Data: Streamflow
ComID 24,432,603
WEDOdata Streamflow
WebLink http://www.epa.gov/
24432603
Zoom to Get Directions
Bridget
Nation^T
"Totest'
FLATTOP
UQtmTAIN
Esri, HERE, DeLorme, Intermap, USGS, NGA, USDA, EPA, MPS
Contents
® WEDO test - WEDO Data
II H Sip ft I •••
Modeled Data Available
No Modeled Data Available
@ Topographic
0 Save » ®® Share 41* Print ~ | «§> Directions ^ Measure Oil Bookmarks [ Find address or place
fa] Details j + Add -»¦
O I E
I Basemap I §2] Analysis
57
-------
Wedo
study Information
¦S> Upload Study
A Utilities
© FAQ
Study Information
I Mom* Sfcjdy Information
Search Studies
Find: - Select Field -
• Use the search bo»( above: to explore studies in the WE DO database.
• Search for published studies by:
° author email
° constituent
o publication status
° watershed hydrologic unit
• Study will be assigned one of the following status;
® Pending
° Widthdrawn
o Publishing
o Published
NHD(USGS) Hydrologic Units: HUC 2 Regions
Study Summary
Export CSV for Publishing Mark Setect Stucbe* Published I Withdraw Seiect Studie
Author Name
* Orgnization
Model Name
Model Start Dt
Model End Dt
Constituents
Status
Action
Maii H Gray
AQUA TERRA Consultants
HSPF
9/30/200912:00:00 AM
10/1/201012:00:00 AM
ROW
Publishing
Details
Mark H Gray
AQUA TERRA Consultants
HSPF
9/30/200912:00:00 AM
10/1/201012:00:00 AM
FlOW
Published
Details
Figure 3.16. WEDO Study Information screen shot.
The WEDO Study Information page (Parmar et al. 2016) provides additional information on the
modeling studies available at a location, such as the model used, start and end date of the
simulation, what water quality constituents were simulated, the published modeling study author.
The WEDO tool contains utilities for both downloading and publishing modeling studies to its
database, serving as a bridge between the modeler's computer and the secure cloud hosting
environment where published modeling studies are stored. The WEDO Publishing Utilities are
downloaded to a modeler's computer, and step-by-step instructions for downloading and
publishing modeling studies are available (Johnston et al. 2016). The WEDO tool is a next
generation, web service system of information technologies linked to the EnviroAtlas for
discovery of modeling studies nationwide. It is intended to facilitate open, reproducible science
through sharing of all data required to duplicate a modeling study as performed by the original
authors. The vision behind WEDO is translational. It extends beyond the current practice of
publishing modeling applications in journals or reports, with most of information not provided or
made available to the modeling community. The EPA's Water Quality Data (WOX website at
https://www.epa.gov/waterdata/water-quality-data-wqx; accessed 7/22/2018) and the U.S.
Geological Survey (USGS) National Water Information System ( WIS website at
http://waterdata.usgs.gov/nwis; accessed 7/22/2018) are examples of nationwide repositories for
sharing monitoring data on water resources. However, no such system exists for modeling
studies. The WEDO tool is designed to provide this valuable resource.
3.4. Transferring Relationships in the EnviroAtlas and Eco-Health Relationship
Browser to Other Settings
Developing or applying models for estimating impacts of decision alternatives on EGS
distribution or delivery and associated benefits can be a complex and highly technical process.
Data collection and analysis may be daunting investments, even when economic valuation is not
58
-------
considered (Granek et al. 2010). The EnviroAtlas and Eco-Health Relationship Browser help
address this concern by providing decision makers with information and data on EGS and their
socio-economic benefits, particularly human health. However, in some cases, decision makers
may wish to conduct their own ecosystem services assessments using alternative data sources,
such as locally collected data or at alternative spatial scales or tailored to their own decision
scenarios or objectives.
Practitioners can still explore the potential of transferring the model relationships from the
EnviroAtlas or Eco-Health Relationship Browser to other locations but use their own information
to build relationships or their own data to calculate metrics. For example, although the
EnviroAtlas contains more than twenty communities with fine-scale data (e.g., 1-m resolution),
other communities may wish to apply EnviroAtlas methodology to calculate metric values under
varying decision scenarios. Guidance on model transferability can help users decide when it is
and is not appropriate to apply EnviroAtlas data and models to new locations. The EcoService
Models Library (ESML) includes models used in the EnviroAtlas and other potentially relevant
models in a searchable database (Bruins et al. 2017). Conceptual models also included in the
ESML can help clarify the linkages between ecosystem structure (such as species composition,
population size, vegetation cover), ecosystem processes (such as primary productivity, nutrient
cycling, oxygen dynamics, etc.) and ecosystem services in the EnviroAtlas. More broadly,
network analysis (Chapter 3.4.4) can be used to extend the causal relationships in the
EnviroAtlas and the Eco-Health Relationship Browser to explore impacts of alternative decision
actions on human health and well-being. These tools and approaches facilitate transferring the
information and relationships in the EnviroAtlas and the Eco-Health Relationship Browser to
other applications or locations.
3.4.1. Applying EnviroAtlas Models to Calculate Community-Scale Metrics
Estimating the consequences of decision alternatives requires that stakeholders be able to
manipulate input data and quantify resulting changes in EGS or health determinants or outcomes.
Input data that are often proposed for modification by environment-health policy decisions,
especially in urban contexts, include land covers (e.g., trees, impervious surfaces), recreational
amenities (e.g., trails, parks), and infrastructure (e.g., roads, bike lanes). For example, the
Milwaukee case study (Chapter 2.2) described how recommendations for improving biking
infrastructure could be evaluated from EnviroAtlas metrics.
Enabling users to calculate their own EnviroAtlas metrics is important for three reasons. First,
users could create metrics suites for new communities. By 2019, the EnviroAtlas Team plans to
complete metrics for 36 communities containing hundreds of municipalities (Communities
within EnviroAtlas Boundaries website at http://www.epa.gov/sites/production/files/2016-
07/documents/cominbnd 2016 mav.pdf; accessed 7/22/2018). Community-scale data metrics
require 1-m resolution land cover data derived from the U.S. Department of Agriculture (USD A)
National Agricultural Imagery Program (NAIP; NA1P website at
http://www.fsa.usda.gov/programs-and-services/aerial-photographv/imagery-programs/naip-
imagery/: accessed 7/22/2018). The EnviroAtlas Development Team has developed procedures
that help users generate their own land cover data compatible with existing EnviroAtlas
community data. Second, users could construct metric suites over time to monitor changes in
59
-------
land use or land cover. Third, local stakeholders could quantify changes in EnviroAtlas metrics
in response to alternate policy decisions or recommendations. This could be to analyze tradeoffs,
assess development scenarios, or to promote public discussion of possible outcomes. This was
done here for the Harbor District in Milwaukee, WI that was introduced in Chapter 2.2. The
objective was to demonstrate how users outside of the EnviroAtlas computing environment could
calculate community metrics in response to proposed land cover or land use changes. Testing the
portability of required input data and associated python scripts (developed by the EnviroAtlas
Team) was the first step in eventually making user-direct metric calculations more practical. This
research must be expanded upon before metric processing is independent of the EnviroAtlas
Team's assistance.
Researchers from ORD adapted approximately 22 scripts involved in calculating EnviroAtlas
community metrics (Table 3.4). About half of the scripts involved actual metric calculations and
half of the scripts managed various processes (i.e., process controls, metadata generation,
geographic coordinate conversions, and data preparation). Metadata and other documentation for
metrics and input data are available for user download at EnviroAtlas (EnviroAtlas website at
http://www.epa.gov/EnviroAtlas; accessed 8/14/18). The EnviroAtlas community metrics
calculated here included types summarized by census block groups (i.e., tabular data), or
represented as lines, polygons, or rasters. Community metrics calculated by the U.S. Forest
Service's (USFS)_i-Tree and EPA's BenMAPjnodels were not included in this case study.
Metrics of i-Tree quantify stormwater, air quality, and other benefits associated with urban trees.
The EnviroAtlas has expanded upon i-Tree indicators by translating urban forestry benefits into
quantitative health and economic outcomes at the community-scale calculated by BenMAP.
While i-Tree and BenMAP are publicly available and supported by user communities, the scripts
used to create those metrics for the EnviroAtlas were not available.
Practitioners can explore the potential of transferring the model
relationships from the EnviroAtlas or Eco-Health Relationship Browser
to other locations using their own information to build relationships,
or their own data to calculate metrics.
60
-------
Table 3.4. Input and output data, scripts, and output metrics (variable name and description) following the EnviroAtlas' nomenclature.
Input data layers
Script
Output data layers
Variable name
Variable description
Dasymetric
population
Land cover,
BG
Floodplains
Floodplains
Floodplains (table)
FP1 Land M
FP1 Land P
FP02 Land M
FP02_Land_P
FPlImpM
FPlImpP
FP02_Imp_M
FP02 Imp P
FPlPopC
FP1 Pop P
FP02_Pop_C
FP02 Pop P
Total land area of block group in 1% annual chance flood hazard area (in2)
% land area of block group in 1% annual chance flood hazard are
Total land area of block group in 0.2% annual chance flood hazard area (in2)
% land area of block group in 0.2% annual chance flood hazard area
Total impervious area of block group in 1% annual chance flood hazard area (m2)
% impervious area of block group in 1% annual chance flood hazard area
Total impervious area in 0.2% annual chance flood hazard area (in2)
% impervious area of block group in 0.2% annual chance flood hazard area
Population (#) of block group in 1% annual chance flood hazard area
% Population of block group in 1% annual chance flood hazard area
Population (#) of block group in 0.2% annual chance flood hazard area
# Population (#) of block group in 0.2% annual chance flood hazard area
Land cover, green
space, water, borders
GreenProx
GreenProx (polygon)
GreenProx
% green space within 1/4 square kilometer of any given point
Land cover. Trees,
GSTCnWR
PctStGS (raster)
PctStGS
% street green space (raster; 10m cell size) (% of green space along walkable roads.)
green space, water,
walkable roads
PctStTC (raster)
PctStTC
% street tree cover (raster; 10m cell size) (% of tree cover along walkable roads.)
Land cover, imperv,
water shape, borders
ImpProx
ImpProx (polygon)
ImoProxP
% impervious within 1 sq km
Walkable Roads
IntDen
IntDen (polygon)
IntDen
Intersection Density
Land cover,
BG
LCSum
LCBinary
LCSum
(table)
Imp M
Ag_M
MForM
Green M
Wet_M
MFor P
ImpP
Green P
Ag P
WetP
MFor PC
lino PC
Green PC
As PC
Impervious area (m2)
Agricultural area (m2)
Tree cover area (in2)
Greenspace area (in2) (trees, grass, agriculture, woody wetlands, emergent wetlands)
Wetland area (m2)
% total BG area as tree cover
% total BG area as impervious area
% total BG area as green space
% total BG area as agricultural cover
% total BG area as wetlands
Tree cover per capita (nf/person)
Impervious area per capita (m2/person)
Green space per capita (m2/person)
Agriculture per capita (nf/person)
Land cover,
trees binary
NrRd
NrRdPfor (line)
NrRd Pop (table)
PctTree
Lane PctIB
% tree coverage within 30 m from road
% of busy roadway in BG bordered by < 25 % tree buffer
61
-------
Input data layers
Script
Output data layers
Variable name
Variable description
NavTeq_D (busy
streets),
BG, borders
Lane PctSB
SBuff Pod
IBuff Pod
Buff Pod
Buff Pet
% of busy roadway in BG bordered by > 25 % tree buffer
residential population (#) < 300 m of busy road with > 25 % tree buffer
BG population (#) within 300 m of a busy roadway with < 25 % tree buffer
BG population (#) within 300 m of a busy roadway
% of BG population within 300 m of a busy roadway
Park Entrances
Walkable Roads
BG
Dasymetric pop.
Borders
Parks
Park Pop (table)
Park Prox (polygon)
IWDP Pod
BWDP Pod
IWDP Pet
BWDPPct
ParkProxD
BG population (#) within 500 m of a park entrance
BG population (#) not within 500 m of a park entrance
% of BG population within 500 m of a park entrance
% of BG population not within 500 m of a park entrance
Distance to park entrance in meters
RB50 Larea
RB50 LABGP
Total land area in 50m riparian buffer (nf)
% land area in 50m buffer
RB50_ImpP
RB50 ForP
% impervious area in 50 m buffer
% tree cover in 15 m buffer
Land cover,
trees, veg, water
RB LC (line)
RB50_VegP
RBI5 Larea
% vegetated cover in 5m buffer
Total land area in 15m buffer (nf)
binaries
RB_v2
RB15 LABGP
% land area in 15 m buffer
NHD Waterbodies,
Flowlines & areas
RBI5 ImpP
RBI5 ForP
% impervious area in 15m buffer
% tree cover in 15m buffer
BG, borders
RB15m Rfor(line)
RB15_VegP
PFor
% vegetated cover in 15m buffer
% forest within a 15m moving window centered 7.5m from a waterbody edge.
RB15m_Vege (line)
RB5 lmRfor (line)
RB15m Vege (line)
PVege
PFor
PVege
% vegetation within a 15m moving window centered 7.5 m from a waterbody edge.
% forest within a 5 lm moving window centered 25.5 m from a waterbody edge.
% vegetation within a 5 lm moving window centered 25.5 m from a waterbody edge.
Land cover.
K12 Count
# of K-12 schools in the block group
Green space binary
HSIP Education Pts
Schools
EduLowGS (table)
Day Count
K12 Low
# of daycares in the block group
# of K-12 schools with <= 25% green space within 100 m
BG
Day Low
# Daycares with <= 25% green space within 100 m
Land cover,
trees binary
BG
Tree Views
TreeWV (table)
WVTPop
WVT Pet
BG population (#) with minimal potential views of forest
% of BG population with minimal potential views of forest
Land cover,
water,
BG
WaterViews
WaterWV (table)
WVWPop
WVW Pet
BG population (#) with potential views of water
% of BG population (#) with views of water
Historical_places
Various
Historical_places
(table)
totalhiscount
# historical places in the block group
62
-------
Table 3.4 lists the required input data, processing scripts, output data layers, and metrics included in this
demonstration of processing portability. Once operational, the scripts successfully reproduced the data
currently available for Milwaukee from the EnviroAtlas. From previous consideration of the Milwaukee
Harbor Plan (2017; Chapter 2.2; Figure 3.17) a scenario was created that modified land cover, roads,
and park entrances in the Harbor District. Unlike Chapter 2.2, this scenario did not come from
recommendations of the Milwaukee Harbor Plan. Patches of trees, segments of walkable roads, and park
entrances were added in and near industrial sites in the current Milwaukee land cover data (Figure
3.18). These changes targeted metrics in the Park Pop, Park Prox, LCSum, PctStGS, and PctStTc data
layers (Table 3.4). The metrics targeted for change involved people's proximity to parks or green space,
and tree cover near roads, riparian areas, and schools (if any). The scenario was designed to introduce
changes in land cover and amenities (i.e., walkable roads and park entrances) that might benefit
neighborhood residents and promote the beneficial reuse of industrial lands.
The computing environment was created on a virtual machine equipped with a local three terabyte hard
drive, an Intel® Xeon® CPU E5-2680 v2 at 2.80 GHz processor, and 16 GB of system memory. The
stand-alone configuration allowed processing to be isolated from other tasks. The virtual environment
promoted collaboration among researchers for troubleshooting scripts and reviewing results. Processing
the 66 metrics for the 1,175 census block groups in the Milwaukee EnviroAtlas boundary took
approximately 75 hours. Processing the 44 census block groups in the Harbor District project area took
approximately four hours. This suggests that, even with available input data and functioning scripts,
calculating changes in EnviroAtlas metrics for multiple scenarios or testing multiple alternate policy
outcomes currently requires a large investment in computing power and time.
The scenario resulted in modest changes in metrics in two census block groups. Converting 30,000 m2 of
grass and impervious areas to trees (about 3.5% of the area) increased the tree cover per capita
(MForPC) by 38 m2 per person. Future scenarios could be constructed where new residents are added
or current residents are redistributed in response to changing land covers. The new walkable roads
"built" on the railroad's right-of-way intersected with new tree cover (Figure 3.19) could follow
recommendations of Complete Streets or Smart Growth Initiative (McCann and Rynne 2010) or be
included in neighborhood redevelopment plans, such as the Milwaukee Harbor Plan.
Figure 3.20 shows the changes resulting from proximity to parks (as polygon data) based on the original
and virtual distribution of park entrances. The premise was that people would benefit from the
redevelopment of these sites by gaining access to green space or, more importantly, developed parklands
(i.e., with recreational facilities and programing). The latter is not considered in the metrics. Adding the
park entrances resulted in more residents living <500 m of an entrance rather than up to 3,000 m
distance. In one census block group, the percentage of residents living within 500 m of an entrance
(IWDP Pop) increased from 96% to 100%. These values reflect both the concentration of people in
these neighborhoods along the edges of industrial sites and the general lack of parks. While it is
unrealistic that a redevelopment scheme would put the entire area into parklands, this scenario
demonstrated how EnviroAtlas data can be used to quantify how residents related to small parks
distributed through neighborhoods.
Demonstrating the portability of calculating metrics was a necessary first step in enabling users to
develop new EnviroAtlas communities, assessing changes in EGS metrics over time, and for evaluating
alternative outcomes to policy decisions on land cover, infrastructures, and/or EGS. Next steps of this
research may lead to "tool kits" or User Guides that help users incorporate their own data and construct
novel metrics that complement the data provided in the current and future EnviroAtlas.
63
-------
40 Kilometers
Milwaukee
Harbor District
W 2
estate Si
W Kifocurn Avt
Milwaukee Wisconsin
Legend N
I 1 i i ! i I I I
Harbor District boundary 0 O S t 2 Kilometers
~ EnvtroAtlas Milwaukee Community boundary
5F Jlf<~ rtr-Troj T5'" KH Cii iifdr VlP (MCtiwi1Wgy/M{l Cv< (TtjTwKf i Opo* I tkj -:srTt7 t S-TO
Figure 3.17. Map of Milwaukee and Harbor District.
64
-------
Current Conditions
Park Scenario
Legend
• Park Entrances Trees
Walkable Roads | | Grass
| Water | | Crops
[Jj Developled ~ Wetlands
Highly Developed | | Woody Wetlands
0 0.25
1 1 1—
0.5
—I—
A
1 Kilometers
H 1
Legend
O New Park Entrances
= New Walkable Roads
I New Trees
Figure 3.18. Current land cover, street configurations and park entrances compared to
hypothetical additions of 10 park entrances, 3 ha of trees, and 4 km of walkable streets in the
Harbor District.
65
-------
Current Conditions
"Z1T
ris
TipHpraO../
1
-¦ 4 v8 A I 5a«aggMfi^^wrf iffli
CW^J •:
.K _m :?• f f .W8^ m ?wi
K *4.*7 §(
*- ••"- \
[. %in, • »
^Bs1kPV
vr —l _ n :moL _ .. ' v \ A
Park Scenario
W 1
1 aBH
K
\ivn
-N ¦ . - - _ .Jp
If
i?. 'C
SB
t
Legend
• Park Entrances 26 - 50
MWI_PctStTC 51 - 75
Percent Tree Cover — 76 -100
0-25
0.25 0.5
I 1 1—
1 Kilometers
H 1
A
Legend
O New Park Entrances
I New Trees
Service Layer Credits: Source: Esri, DigitalGlobe, GeoEye, Earthstar Geographies,
CNES/Airbus DS, USDA, USGS, AeroGRID, IGN, and the GIS User Community
Figure 3.19. Comparison of current and scenario-based changes in near-road tree cover (PctStTC; Table
3.4).
66
-------
Current Conditions
LMB
J
n 11
rjlH * ro
¦B rpi
®w. rwi
ITl I k a 1 Xli' v "•sir ' ^^F9n^p5?,i
ifl WS.VCV
Legend
• Park Entrances
Walkable Roads
MWI_Park_Prox
Proximity to park in meters
~~ N/A
0- 500
501 - 1000
1001 -3000
3001 - 5000
0 0.5 1 2 Kilometers
1 i 1 1 1 1 1 1 1
IN
A
Legend
O New Park Entrances
New Walkable Roads
New Trees
Figure 3.20. Comparison of current and scenario-based changes in distances to park entrance
(ParkProxD; Table 3.4).
3.4.2. Assessing the Applicability of Ecological Models to New Locations: Model
Transferability Assessment
Estimating the distribution, production, benefits, or value of EGS is frequently accomplished using
ecological models (e.g., In VEST, Tallis et al. 2013; VELMA, U.S. EPA 2014; see also ESML, Bruins et
al. 2017). Many of the EGS data in the EnviroAtlas were developed through application of such models.
Models are valuable for estimating changes in EGS in response to environmental remediation or
restoration projects or changes in land-use policy. Stakeholder-driven decision processes might be
required to use models to quantify EGS tradeoffs or scenario costs and benefits. In most cases, decision-
making processes seek to use existing ecological models because the costs and time to develop new
models may be prohibitive. The ESML includes models used in the EnviroAtlas and other potentially
relevant models in a searchable database (Bruins et al. 2017). It can help stakeholders identify existing
EGS models and required data needs (see Chapter 3.4.3), In some cases, data needs may be met with
existing sources or monitoring programs. In other cases, data must be generated de novo. Using existing
models is advantageous to users because new models are expensive to develop and validate and may
come with expensive new data requirements. Where and in what contexts models have been successfully
applied may be available in scientific or applied literature, as part of model databases (e.g., ESML;
Chapter 3.4.3), or as descriptive metadata in the EnviroAtlas. Frequently, however, stakeholders face
OOOOO
67
-------
the prospects of being the first to apply a model at a site or in a novel context. The partnership must then
evaluate whether such a model transfer is justifiable, practical, and likely to contribute positively to
solving problems. Model transfers may not be appropriate because of contextual differences between
where the model has been successfully applied and a new application site. Inappropriate transfers of
measurements or models can lead to inaccurate estimates of ecological or benefit endpoints and expose
decisions based on those estimates to legal or scientific challenge (McGarity and Wagner 2003). Users
of ecological and environmental models thus need a practical, objective, and transparent methodology
for determining whether a model may be appropriately and confidently applied at a novel site or in a
novel context. The challenge of model transfers may be further exacerbated by the scarcity of relevant
data at the application site.
The Model Transferability Assessment (MTA) framework was developed to evaluate whether a model
could be appropriately applied (i.e., transfer) to a new site or decision context based on the model's
previous performance at sites with similar ecological contexts (Moon et. al. 2017). The MTA
methodology describes a model's application niche to guide the model transfer process through stages of
evaluation and selection. This method permits users to synthesize information from databases, past
studies and/or past model transfers to depict a model's expected performance under various context
scenarios using analytical approaches of creating model performance-vs-context curves and heat maps.
The general framework for MTA is illustrated in Figure 3.21 and further description of the concepts and
statistical methods for MTA, with an example of its application for transferring a model to estimate
wetland ecological condition to sites across a broad geographic region, are presented in Moon et al.
(2017).
Model Transferability Assessment Framework.
Select model <*-—¦—
O
Find previous applications
o
Assess site context similarity
• Acceptable similarity?
(yIS) (NO
#
Assess model's expected performance
• Acceptable performance?
(yeS)
Select & transfer
endpt estimate
Conduct site-
specific study
T
Apply model., but verify
Figure 3.21. Major steps in the methodology to assess the transferability of models to new sites or
applications. A measurement transferability framework is the same, except that the expected accuracy of the
measurement is assessed in Step 4.
68
-------
Environmental data describing conditions where a model was previously used and/or at a potential new
site are critical to the MTA methodology. That data set, consisting of multiple environmental variables
that are relevant to the process or condition represented by the model, comprises the model's application
niche or the model's ecological context. This is analogous to the Hutchinsonian ecological niche
(Hutchinson 1957) describing the full range of biotic and abiotic factors that are necessary for the
survival, growth, reproduction and persistence of a species. The model's application niche specifies the
geographical locations where the model has performed successfully, which are compared to the
conditions at the new site of interest (i.e., the site-specific values of the same environmental variables
that were used to create the model application niche) to forecast the model's performance at the new
site. A caveat is that locations or conditions where models perform poorly are under-reported because of
possible publication bias. This is unfortunate because contexts under which a model performs poorly
inform the prediction of its performance at new locations. See Chapter 3.4.4 for an example where
there is value of describing model limitations to advancing the overall science of ecosystem services
assessment modeling.
Biotic and abiotic variables mapped within the EnviroAtlas can characterize conditions at sites in
multiple spatial contexts (i.e., community to national). This can include where models have previously
worked well and where models are being targeted for application. These data can be used to develop
application niches for ecological models and help predict performance at new sites, and therefore
informs the appropriateness of transferring them (Figure 3.22). Other geo-databases of ecological
variables are also useful for this task, but the wide range of environmental variables contained within the
EnviroAtlas, specified across the spatial extent of the continental U.S., makes it a particularly valuable
tool for MTA.
Models are valuable for estimating changes in EGS in response to
environmental remediation or restoration projects or changes in land use
policy. A Model Transferability Assessment framework can be used to
evaluate whether a model could be appropriately applied to a new site or
decision context based on the model's previous performance at sites with
similar ecological contexts.
69
-------
Dummy data
Forest 1
model development site
Context variable
soil type
| unconsolidated
Q carbonate rock
[J mixed rock
total annual
precipitation
0.1 m 1 m
mean annual
temperature
Forest 2
Forest 3
total tree
cover
10% 80?4
Figure 3.22. Hypothetical example site context data drawn from EnviroAtlas (using EnviroAtlas variables
of soil type, total annual precipitation, mean annual temperature, total tree cover) for a model to predict
carbon sequestration rates in polygons representing three forests. The model was developed for Forest 1. The
overall site context of Forest 3 is more similar to Forest 1 than is the site context for Forest 2; conceptually the
model would likely be more successfully transferred to Forest 3 than to Forest 2.
3.4.3. Exploring EnviroAtlas Models Alongside Other Model Options: The ESML
The EcoService Models Library (ESML website at https://esml.epa.gov/; accessed 7/22/2018) is a
searchable website and database for finding, examining, and comparing ecological models for estimating
the production of EGS, including many of the models in the EnviroAtlas (Bruins et al. 2017; Moon et al.
2017). The ESML complements and supplements the EnviroAtlas by distilling the objectives, input data
needs, output products, and analytical assumptions of ecosystem service models currently available in
the ecological, health, and economic literature. By aggregating and describing models through a
comparative typology, the ESML facilitates the discovery, use, and comparison of models useful for
quantifying EGS, including models from the EnviroAtlas. With the information in the library, users can
make informed decision on which models may be appropriate for their needs. It helps users compare the
model's original purpose and objectives, likely applicable environmental contexts, level of uncertainty,
and the feasibility of successful implementation.
Several models within the EnviroAtlas are detailed in the ESML (Table 3.5). The scope of the ESML is
currently limited to models that address EGS related to habitat and biodiversity maintenance, waste
regulation, climate and weather regulation, and water flow regulation, with a few models addressing
cultural services, such as recreation, or non-use services. Furthermore, most of EnviroAtlas data layers
are maps of actual data rather than model predictions, precluding their inclusion in ESML.
70
-------
Table 3.5. Selected EnviroAtlas models and response (output) variables included in the ESML as identified
by ESML ecological model identification (EM ID) number.
EM ID
EnviroAtlas Model
EnviroAtlas response (output)
variables1
Community-scale (1-m resolution)
EM-51
Natural filtration (water)
Tree-mediated annual reductions of
TSS, TP, TKN, SRP, NOx, BOD,
COD, Cu (i-Tree model)
EM-59
Air pollutant removal
Tree-mediated annual reductions of
CO, N02, 03, PM 10, PM 2.5 (i-Tree
model)
EM-142
Annual water recharge by tree cover
Tree-mediated change in runoff (i-
Tree model)
EM-493
Carbon sequestered by trees
Annual carbon sequestration by tree
cover (i-Tree model)
National-scale (12-digit HUC; sub-
watershed resolution)
EM-63
Natural biological nitrogen fixation (BNF)
Natural biological nitrogen fixation
EM-491
Crops with no pollinator habitat
Acres of pollinated crops with no
nearby pollinator habitat
EM-492
Restorable wetlands
Potentially restorable wetlands on
agricultural lands
1 TSS = Total suspended solids; TP = total phosphorus; TKN = total Kjeldahl nitrogen; SRP = soluble reactive phosphorus;
NOx = nitrogen oxides; BOD = biological oxygen demand; COD = chemical oxygen demand; PM 10 = particulate matter
between 2.5 micron and 10 microns; PM 2.5 = particulate matter smaller than 2.5 microns
The ESML does not contain modeling software or perform calculations; instead, it provides detailed
descriptions of models, including citations. The organizational scheme of the ESML centers on the data
map (Figure 3.23).
EM Identity
& Description
EM Locations,
Environments, Ecology
VariableType
+ EM Source Document
(uniquely identified
by Document ID)
r
w
Ecological Model (EM)
(uniquely identified by EM ID)
EM Variable
(uniquely identified
by Variable ID)
Variable Spatial &
Temporal Aspects
EM Modeling
Approach
EM Ecosystem Goods
and Services (EGS)
Variable Values,
Variability &
Validation
Figure 3.23. ESML Data Map. The ESML database includes three kinds of records: Source Documents;
Ecological Models (EMs); and EM Variables. Each record includes multiple categories of descriptors.
71
-------
The ESML includes three types of records: source documents; ecological models (EMs); and EM
variables. Over 50 individual descriptors - covering purpose, approach, and environmental use - are
applied to each ecological model, and 40 additional descriptors are applied to each of the model's
variables. The website was designed to enable the potential users of models to search and view
information and compare models against concepts of model appropriateness, and export model
descriptions.
In addition to providing descriptive information, the ESML data maps (also referred to as variable
relationship diagrams) describe how variables of ecological function are causally linked to measures of
ecosystem benefits (Figure 3.24). Such conceptual diagrams can help decision makers understand key
linkages among variables or identify intervention points (see Chapter 3.4.4), These diagrams help
translate model specifications enabling potential transfer for use in novel decision contexts or places (see
Chapter 3.4.2), Further, understanding the data needs of models permits users to find appropriate data
sources, such as the EnviroAtlas, to support applications. Examples of using EnviroAtlas data to support
model development are given in Chapter 3.3. These data map diagrams also provide a way to compare
the response variables (outputs) of one model against the predictor variables (inputs) of another, such
that different models could potentially be connected together to build more complex systems of models,
depending on a user's needs and objectives.
The EcoService Models Library complements and supplements the
EnviroAtlas by distilling the objectives, input data needs, output products,
and analytical assumptions of ecosystem service models currently available
in the ecological, health, and economic literature.
72
-------
Figure 3.24. ESML data maps (labelled as variable relationship diagrams) for seven EnviroAtlas models
diagraming the relationships between predictor variables (PD: time or space varying; PC: constant or
parameter), intermediate variables (IE: in ESML; IN: not in ESML), and response variables (RC:
computed response; RM: observed response).
a) Natural filtration (water)
Variable Relationship Diagram for: |EM-51/EnviroAtlas - Natural filtration (water) 1
from EnviroAtlas water recharge
model (EM-142)
PD: Reduction in annual runoff
(census block group) [mA3 yr*-!]
|PD Census block group
PD*: Pollutant
PC*: Pollutant pooled mean
event-mean concentration [mg
lA-l]
RC: Reduction in pollutant mean load (census block
Lroup) [kg yt"-1]
PC*: Pollutant pooled median
event-mean concentration [mg
IMJ
RC: Reaction In poiiutam median toad (census block
feroup) !kC V*-ll
b) Air pollutant removal
Variable Relationship Diagram for ]EM 59/EnviroAtlas - Air pollutant removal
Poll Jtant r err ova1 [«j yr*-l)
|PD: Ckmd CC -r.g jft]
): UpQQ- hi (ffij
PD Upper ai-tern pgr at wry (*Cl
PD: P"ec ? tat-or: | n;
latitude (00)
PD Stock growp 'on g twee (DO)
PD; Land area (block group} (m*2)
PD: Percent canopy ( '•)
PC Evt*gft€f tree cover (as oercertage of tota tree
CO1.e-;
PC: M}« eaf a*ea »nde* (trees)
PC Ltaf-Pli date (Julian day)
PC Leaf-off date (Julian day)
Wind ;peeci {rri hr"-l|
Tem perature (>«r) (f )
DewpO'int (*f}
A ? *rete* setting (in)
~|IN DeposironaI velocity (m s" ¦ 1)
73
-------
c) Annual water recharge by tree cover.
Variable Relationship Diagram for: | EM- 142/EnviroAtlas - Annual water recharge by tree cover)
from (Tree-Hydro model
(EM-137)
PC: Streamflow (of watershed per
tree cover/impervious cover
combination} (mA3 yrA-l]
F^C: land area (watershed) [mA2j
PD: Census block group
PC*: Tree cover [%]
PC*: Impervious cover [%]
PC*: Land area (census block
group} [m*2j
|IE Reduction in annual |
|runoff [mA3 mA-2 yrA-l] J
^IE: Percent annual reduction j
... stream fiow (96 mA-2j j
RC: Reduction in annual
runoff (census block
group)[mA3 yr*-!]
RC: Percent annual reduction
in stream flow (census block
d) Carbon sequestered by trees.
llE: Tree cover (per census block group) [mA2]
RC: Carbon sequestered (per census block group)
[mt C yrA-lJ
PD: Annual carbon sequestration rate
[kg C mA-2 tree cover yrA-l]
from i-Tree Eco model (EM-24)
PD: Tree cover (mA2]
PD: Census block group
Variable Relationship Diagram for: EM-493/EnviroAtlas - Carbon sequestered by tree;
74
-------
e) Natural biological nitrogen fixation.
jlE: Land cover fraction per
|HUC12 [%]
ITe: Actual
•Jevapotranspiration (AET) per
RC: Natural BNF per
' HUC12 (kg N ha-1 yr-lj
PD: 12-digit hydrological unit code
(HUC12)
PD: LULC
from Actual
Evapotranspiration
(AET) model
(See EM description)
PC*: Land cover parameter
from BNF model
(See EM description)
fif:: Biological nitrogen I
Jfrxation {BNF) per HUC12 [kg j
LNha_:Xaij_ |
PD: Annual average minimum daily
temperature (air) [*CJ
PD: Annual average maximum daily
temperature (air) [*C]
PD: Annual precipitation (cm)
Variable Relationship Diagram for: |EM-63/EnviroAtlas- Natural biological nitrogen fixation (BNF)
f) Crops with no pollinator habitat.
Variable Relationship Diagram for: |EM-491/EnviroAtlas - Crops with no pollinator habitat
PD: Cropland class
PD: Land use/land cover
PC: Average bee species foraging
distance from nest (kmj
PD: 12-digit hydrological unit code
(HUC12)
PCt: Crop requires/benefits
from pollination (y/n by pixel)
PCt: Pollinator habitat * tree
class (y/n by pixel)
jlN: Pollinated crops with no
j nearby pollinator habitat (by
pixel)
RC: Acres of pollinated crops
kvith no nearby pollinator
habitat (per HUC12) {acres)
g) Restorable wetlands.
Variable Relationship Diagram for: EM-492/EnviroAtlas - Restorable wetlands
PD: Land use/Land cover
PD: National elevation data
PD: 12-digit hydrological unit
code (HUC12)
PD: Soil class
PD: Soil survey polygon
PCt: Potentially suitable land
(Y/n)
IE: Percent poorly drained soil
idasses (per survey polygon) [%]
~(IE: Compound topographic index |-
( i
RC: Percent potentially restorable
wetlands (per HUC12) [%]
75
-------
3.4.4. Linking Actions to Ecosystem Services and Well-being: Network Analysis
Solutions to complex environmental problems are often explored using network-based conceptual
models. In their simplest form, network-based conceptual models aid decision making by translating and
clarifying linkages between variables in a diagrammatic way, such as ecosystem functions that connect
EGS (e.g., ESML variable relationship diagrams; Figure 3.24) to human health and well-being benefits
(e.g., Eco-Health Relationship Browser, Figure 3.25). Extending such diagrams to more holistically link
decision actions to desired outcomes can facilitate a visual comparison of alternatives, or become the
foundation for assessing complex suites of data or applying models.
Urban
Ecosystems
Vulnerable
Populations
ADHD
Social &
Community
Ties
Aggression
Respiratory
Symptoms
Anxiety
Obesity
Aesthetics &
Engagement
with Nature
COPD
Cancer
Low Birth
Weight
Longevity
Confusion
Healing
Cognitive
Function
High Blood
Pressure
Fatigue
Depression
Happiness
You are here: Urban Ecosystems I Aesthetics & Engagement with Nature
Bibliography Eco-Health Relationship Browser: Public Health Linkages to Ecosystem Services Topics: Aesthetics St Engagement wil *
Click a topic bubble or choose a topic from the dropdown list above.
Hover over linkages (+) to view the relationship between elements.
Details
Description: Aesthetics &
Engagement with Nature
Many people around the world enjoy
recreating, relaxing, and spending time
outdoors. Scientific studies show that
exposure to nature is positively
associated with numerous aspects of
both physiological and psychological
health, as well as with good social
relations. Causa! mechanisms for some
of these associations have been
demonstrated in the laboratory; faster
recovery from neurological fatigue
appears to be responsible for the
observed effects that greenness has on
mental concentration and the alleviation
of ADHD symptoms in children.
Exposure to natural scenery, even
through a window or a photograph,
slows the heart rate and calms anxiety.
Humans' innate affinity for nature may
be responsible for observations that
people are preferentially drawn to
community green space, where they are
more inclined to interact with neighbors
while relaxing or recreating. These
interactions are directly beneficial by
increasing social capital (Putnam 2000),
which in turn contributes positively to a
Citations/ Sou rces
Louv, 2005; Putnam, 2000; Wilson, 1984
Figure 3.25. Screen shot of relationships between ecosystems (yellow arrows), aesthetics and engagement
with nature (center circle), and multiple health outcomes (blue arrows) in the Eco-Health Relationship
Browser.
The Eco-Health Relationship Browser is an interactive network-based conceptual model that visualizes
links between provisioning ecosystem services and human health outcomes, based on relationships
identified through a review of scientific literature (Jackson et al. 2013). The Eco-Health Relationship
Browser helps users translate and visualize multiple pathways in iterative and interactive ways and can
help community decision makers understand complex multifaceted network relationships. For instance,
a community considering an investment in greenspace may wish to consider how changes in recreation
and physical activity may benefit human health, perhaps as part of a Health Impact Assessment (HIA)
76
-------
(Johnston et al. 2017b). The Eco-Health Relationship Browser allows users to visualize the links
between urban ecosystems, engagement with nature, and multiple health outcomes (Figure 3.25),
provides details on the impact of specific health outcomes, and most importantly, provides evidence
linking recreation to health outcomes so that users can weigh the evidence and decide whether
ecosystem->EGS-^health links are sufficiently strong to merit their inclusion in the decision or planning
process. Through visualization of multiple pathways, decision makers may also learn of other health
benefits for the community that were not previously under consideration.
The relationships in the Eco-Health Relationship Browser were transferred and expanded to develop a
network-based conceptual model for Tampa Bay, FL (Russell and Harvey 2016) to help community
decision makers and stakeholders. The objectives were to: a) more broadly consider the specific
attributes of ecosystem condition contributing to ecosystem services provisioning; b) to more broadly
consider benefits to human well-being beyond human health; and c) to identify indicators to measure
services and well-being benefits (Figure 3.26). The resulting Tampa Bay Relational Browser (Tampa
Bay Relational Browser website at https://archive.epa.gov/ged/tbes/web/html/relationalbrowser.html;
accessed 8/4/2018) uses the Services -> HWBI Index framework (Chapter 3.3.3) to describe
relationships between ecosystem services, social services, economic services and domains of human
well-being beyond health relationships, including connection to nature, cultural fulfillment, education,
leisure time, living standards, safety and security, and social cohesion (see Services->HWBI framework
Figure 3.10). Furthermore, the ecosystem services component of the Eco-Health Relationship Browser
(e.g., "aesthetics and engagement with nature") was extended to identify specific attributes of ecosystem
function contributing to that service (e.g., the quality of greenspace, including topography, uniqueness,
vegetation cover, water availability, and presence of wildlife). Certain ecosystem attribute nodes were
designated as being directly manageable (i.e., serving as intervention points), giving users the ability to
explore the linkages between management decisions and outcomes, through changes in ecological
attributes, human health, and well-being.
Ecosystem
Services (EHB)
Ecosystems (EHB)
Intervention
Points
Ecosystem Ecosystem
Attributes Service (HWBO^
Social
> Service
(HWBI)
Indicators (HWBI)
Economic
(HWBI)
Ecosystem (T) q
^ Services
(HWBI) © gy
Social ®o
Services Q
(HWBI) © ^
Economic ($) Q
ServicesQ /T\
(HWBI)
Health
Outcomes (EHB)
Indicators (HWBI)
Well-being
HWBI
*"* HWBI
Domain
Figure 3.26. Flow diagram of the Tampa Bay Relational Browser showing expansion of Eco-Health
Relationship Browser relationships (Ecosystem-^Ecosystem Services-^Health Outcomes; light blue
arrows) to identify: a) manageable intervention points on ecosystem attributes or ecosystem services (green
arrows); b) relationships between ecosystem services, social services and economic services and domains of
human well-being (blue arrows); and c) indicators to measure services and well-being (dashed blue
arrows).
77
-------
Ecosystem attribute nodes in the Tampa Bay Relational Browser were tailored to include those common
to the Tampa Bay region and of interest to partners such as the Tampa Bay Estuary Program and the
Tampa Bay Regional Planning Council. The objective was to more closely link the conceptual model to
specific high priority decisions in the Tampa Bay watershed making the tool relevant and useful for
decision making. As an increasing number of communities seek to integrate ecosystem services in
decision making (e.g., Yee et al. 2017), emerging research is being conducted to re-route and expand the
network pathways of the Tampa Bay Relational Browser to better describe causal relationships between
potential community actions, related manageable ecosystem attributes, and human health and well-being
for a broader range of communities. For example, the Eco-Health Relationship Browser and the HWBI
Framework use different typologies to categorize ecosystem services and human health outcomes. The
Tampa Bay Relational Browser maintains each of these as a distinct typology, resulting in some degree
of redundancy and disconnect as one navigates through the browser (Figure 3.26). An alternative
approach could be to better integrate the two networks (e.g., by aligning the HWBI ecosystem services
categories with those of the Eco-Health Relationship Browser) to more seamlessly navigate from
intervention points to attributes of ecosystem condition to ecosystem services to impacts on human
health and well-being (Figure 3.27). These causal chains could be cross-walked with multiple decision
contexts and ecological settings to broaden relevance to communities outside of Tampa Bay.
Furthermore, by integrating the typologies of a number of EGS tools, the conceptual network can
become the foundation for layering and accessing information, such as literature references, hyperlinks
to model variables in EPA's ESML (Chapter 3.4.3), data layers in the Enviro Atlas, and metrics
quantifying ecosystem services and human well-being (Chapter 3.3.3), In addition, network attributes
(e.g., nodes and arrows) allow for the application of principles and calculations from network theory to
explore the degree of connectivity within a system or identify maximally efficient pathways toward
achieving well-being goals (Figure 3.28). This could be highly useful for informing trade-off analyses
as the likelihood of success may be strongly related to making the decision most strongly connected to a
desired outcome and least strongly connected to undesirable ones. Such comparisons illustrate the utility
of applying network-based conceptual models to trade-off analysis in decision making.
Network-based conceptual models can help decision makers by visually
translating and clarifying linkages between variables in a graphical manner,
such as ecosystem functions that connect EGS to human health and well-
being benefits.
78
-------
CoZ'
Leisure Time
Safety and
Security
Living
Standards
V\f
Social
r„,
^ Cultural
fulfillment
Social
Relations
Low Birth
>ental Health Wei9ht
Longevity
High Blood
Pressure
tHHL
Bikhg /r\ Aesthetics &
Engagement
with Nature
RMmt INI I
Physical
Activity
Bo:t no
Recreational
Bird Watching
Recreational Recreational
Fishing Hunting
Running
Healing
Fatigue
Happiness
Depression
Anxiety
Cognitive
Function
Sight Seein'
Walking
Aggression
Scuba Diving
_ , Habitat And
Swhiming/Con R .
tact Recreation
Details
Description: Aesthetics &
Engagement with Nature
Scientific studies show that exposure to
nature and urban green space is
positively associated with numerous
aspects of both physiological and
psychological health, as well as with
good social relations. Causal
mechanisms for some of these
associations have been demonstrated in
the laboratory: faster recovery from
neurological fatigue appears to be
responsible for the observed effects that
greenness has on mental concentration
and the alleviation of ADHD symptoms
in children. Exposure to natural scenery,
even through a window or a photograph,
slows the heart rate and calms anxiety.
Humans' innate affinity for nature may
be responsible for observations that
people are preferentially drawn to
community green space, where they are
more inclined to interact with neighbors
while relaxing or recreating. These
interactions are directly beneficial by
increasing social capital (Putnam 2000),
which in turn contributes positively to a
variety of health and well-being issues.
Access to nature, including urban green
Citations/Sources
Louv, 2005; Putnam
1984
2000; Wilson,
Figure 3.27. Illustration of emerging research to integrate the Eco-Health Relationship Browser
relationships (light blue arrows; see Figure 3.17) with ecosystem services attributes and intervention points
(green arrows), and human well-being (HWBI) relationships (blue arrows) of the Tampa Bay Relational
Browser. Circles represent either ecosystem services (ES) attributes, metrics of human health (+), or domains of
human well-being (S) that are connected to the service category "Aesthetics and Engagement with Nature."
Similar links are defined for 22 service categories and define the collective linkages among the ecosystem and
human beneficiaries.
Network analysis explores the degree of connectivity among components
of a conceptual model and helps identify optimal pathways toward
achieving well-being goals. This can be useful for informing trade-off
analysis and decisions.
79
-------
* Heai&y tuuuftl ud baiilr cx/. uv«it:irtfl
A ctsvv to odncrtKn Qn*fiiy cfp-ofa*
% •
Ptewrv* ExfrtafcCf V^Kie
Ac#** r.o taiiihtAic
Ajinralhirr
Hou«t£$ opto out
Oca « urn ear O
Finyiff
Q
< C'xpiM Invcxmio"
o
Water Qmrflftv
Tlj&IlQ/Cf!}'
a Q
Acfrrisi
l Stnko
Q
Preditifoa
*r»d eiuw*
a
',Vi-|xzifnl
S¥««rVt $«nw flf ldUfil)UPf*c«
Supper! lilntvk
o
IflTO
Cttmd Fulltilnatl
[ft.-anry lif eikjia*K*i CL*ytfi1
Ciujn*> m<) }• ulb-kxd Uulaii\f» • oe (cpjlaiifn
Ailvmawc 0 "
A , ,, ®NlHI'H#diWlna eftiou
FjulhiMHatKC*
As®"* lD'Satanl Rpkswwi IxaJ erflw#
Figure 3.28. Example of the degree of connectivity, using pathway analysis, to assess sustainability of
community decisions based on indicators of network organization (Green - action categories; Yellow -
services; Red - domains of well-being; Pink - HWBI). This ongoing research re-routes and expands the
network pathways of the Tampa Bay Relational Browser to better integrate typologies and better represent the
causal chains outlined in Figure 3.18.
The intuitive visualization of complex environmental issues (i.e., network analysis) through network-
based conceptual models examining the shape and strength of multiple pathways within a conceptual
network, has great potential and demonstrated utility to community decision makers facing multiple
alternatives for addressing multiple stakeholder concerns. These models serve four basic purposes in
decision support. First, they can greatly aid in vi sualization of complex environmental issues. Second,
they can serve as a database entry tool for available information to support decisions. This is particularly
relevant to data and literature in the EnviroAtlas. Third, conceptual models can help identify quantitative
models that may inform specific parts of a complex decision path through linkages to model libraries
such as the ESML (Chapter 3.4.3). Finally, conceptual models provide the best framework for integrated
analysis of multiple decision options and fulfill the requisite need for evaluating trade-offs among these
options. Current visualization software allows users to step through decision pathways (from decision
alternatives to desired outcomes) and helps simplify complex relationships using network analysis.
Incorporating supporting literature allows for qualitative weighting of alternative pathways and adds the
ability to begin to analyze suites of potential alternative outcomes. Establishing links to available data
and models expands a model's utility for developing quantitative assessments of alternative decisions
impacts on multiple outcomes. Beyond that, network analysis can be applied to explore differences
among networks built for different communities or ecosystems, and how those differences impact the
optimization of ecosystem service outcomes of similar decisi ons.
80
-------
3.5. Conclusions
A range of EGS tools and approaches, used in conjunction with the EnviroAtlas and the Eco-Health
Relationship Browser, provides an opportunity to further integrate EGS concepts and approaches into
public policy development, community decision making, and environmental problem solving. Overall,
rallying around common data, models, and tools promotes more equitable engagement, transparency,
and shared accountability, and represents an important evolutionary stage in EPA's translational
research. The ultimate goal is to increase the intellectual, functional, institutional, and practical access to
diverse information needed to support federal, state, tribal, and community partnerships.
Various methodologies, tools, and approaches exist that can facilitate
incorporating EGS concepts into community decision making.
Practical strategies for making EGS-based decisions require a wide array of
adaptable tools and approaches that can be implemented at various scales,
stages of the decision process, and levels of experience.
81
-------
4. Conclusion
The EnviroAtlas and Eco-Health Relationship Browser are tools designed by the EPA's ORD to
facilitate translational science. By design, translational research spans the boundaries between natural
and social sciences. To be actionable and relevant for community decision making, scientific research
must be made accessible across the social, political, and technical domains of stakeholders, policy
makers, and scientists. Effective communication of EGS concepts, stakeholder engagement, and access
to relevant data and analytical tools can all help communities improve decision-making processes, which
yield meaningful outcomes that improve public health and community revitalization. In collaboration
with state, federal, and academic partners, the EPA has developed practical strategies that can facilitate
incorporating EGS concepts into community decision making. Supporting community decision making
with translational EGS research is challenging because processes (such as defining decision contexts and
engaging stakeholders) are highly diverse, projects and policies are at various stages of planning or
implementation, and knowledge of EGS concepts and information requirements vary among partners.
The case studies presented in Chapter 2, and analytical tools and models in Chapter 3, demonstrate
how the EnviroAtlas and the Eco-Health Relationship Browser can help communities use EGS concepts
for social and environmental problem-solving.
Successful community decision making depends on facilitating effective communication and trust
among partners. Decision making requires a relevant decision context that articulates community values,
defines objectives based on those values, and identifies the data and information needed to assess
attainment of those objectives. Additionally, community decision making, involving scientists,
government officials, stakeholders, and the public, depends on respecting the power-sharing governance
model of cooperative federalism. The importance of translational approaches and SDM applications
increases as EGS indicators and functional relationships between environmental and human health
became more familiar and accessible to more diverse users and, thus, are implemented by more
communities. An important early step of any decision process is to understand the key issues under
consideration, identify what stakeholders should be involved in the process, and to define objectives as
statements of what really matters to stakeholders about a decision. Like other EGS tools, the
EnviroAtlas and Eco-Health Relationship Browser can assist decision makers with incorporating
ecosystem services concepts into their decision-making process.
Case studies demonstrated that the EnviroAtlas (Chapter 2.2) and Eco-Health Relationship Brower
(Chapter 2.3) could be used by expert-stakeholder partnerships to link community revitalization goals
to EGS indicators, health outcomes, and public policy. Revitalization goals could range from
infrastructure to social fairness. Qualitative and quantitative methods were needed to interact with
translational models for indicator selection and evaluation and to incorporate public input. Having
multiple opportunities for stakeholder feedback to refine decision contexts, adapt models, and to
incorporate data from multiple sources were important for maintaining transparency in how EGS
indicators were selected, evaluated, and interpreted by officials and stakeholders. The case studies
integrated concepts of translational science, the organizational strengths of SDM, and the accessibility of
EGS indicators.
Chapter 3 presented a variety of additional data tools and analytical approaches for using the
EnviroAtlas and Eco-Health Relationship Browser to assist communities with incorporating EGS
concepts into decision-making processes. Multilateral communication and translational approaches were
important for sustaining critical stakeholder engagement. Structured frameworks (e.g., FEGS-CS;
Chapter 3.2.1) were shown to help decision makers focus their consideration of EnviroAtlas data by
first identifying key beneficiary groups and emphasizing metrics and indicators that are most directly
82
-------
relevant to those beneficiaries. The flexibility of the EnviroAtlas as a mapping platform allows
stakeholders to consider diverse EGS-related objectives, such as measuring human well-being (HWBI;
Chapter 3.2.2), In other cases, decision makers may prefer to use predictive modeling tools to
quantitatively evaluate and compare decision scenarios. Additional information or data could be
collected or analyses conducted to align multiple endpoints under consideration. In some cases, decision
makers may need more predictive modeling that can leverage input data from the EnviroAtlas to
simulate alternative scenarios and quantitatively compare outcomes (Chapter 3.3) to assist researchers
and decision makers to develop indicators and apply them to model alternative decision scenarios.
Chapter 3.3 presented a suite of other models that could interface with the EnviroAtlas or Eco-Health
Relationship Browser and provide additional ways to incorporate ecosystem services concepts into
community decision making. The VELMA tool can use EnviroAtlas spatial data to locate areas in
watersheds where GI infrastructure may be effective for controlling stormwater and associated
downstream nutrient loading. The RBI can be used in conjunction with EnviroAtlas data to connect
changes in the availability of EGS to where and how people benefit from those goods and services. The
Services to Human Well-being Index framework (Services^HWBI) can link with the EnviroAtlas to
develop scenarios for applying the national and community-scale Services~>HWBI regression models
and customized Services~>HWBI relationships for finer-scale analyses. Finally, the WEDO tool can
utilize the EnviroAtlas to browse and locate studies of interest from the subwatershed and streams data
layer.
Having well-defined and robust decision contexts and access to EGS data and analytical tools helps
stakeholders define the potential EGS, social, and economic costs and benefits of decision alternatives.
Data from the EnviroAtlas and Eco-Health Relationship Brower can help inform group deliberations or
expert judgements about potential consequences. For communities or spatial scales not yet included in
the EnviroAtlas, practitioners can transfer the methods from the EnviroAtlas using their own local data
to calculate metrics and indicators (Chapter 3.4.1) with which to build and compare scenarios and
evaluate potential outcomes from alternate decisions. Using EnviroAtlas data to assess the transferability
of environmental models to data-poor settings was also presented (Chapter 3.4.2), The ESML (Chapter
3.4.3) is a searchable database for finding, examining, and comparing ecological models for estimating
the production of EGS, including many of the models in the EnviroAtlas. Finally, network analysis was
shown to help stakeholders visualize complex environmental issues through network-based conceptual
models (Chapter 3.4.4), The approach can be used to examine information flows along multiple
pathways within a conceptual network, such as the Eco-Health Relationship Browser. Network analyses
allow users to step through decision pathways and analyze suites of potential alternative scenarios.
Additionally, network analysis can be used to explore differences between networks built for different
communities or ecosystems and examine how those differences impact the optimization of ecosystem
service outcomes of similar decisions.
The EnviroAtlas and the Eco-Health Relationship Browser contain information that are publicly
accessible and relevant to diverse audiences, relevant across the boundaries of natural and social
sciences, and comprehensive in scale and scope. Clearly, EGS concepts, indicators, and the application
of the EnviroAtlas and the Eco-Health Relationship Browser are not relevant to every decision context
posed by communities. However, the national and community scale EGS data in the EnviroAtlas, and
the health outcomes related to EGS described in the Eco-Health Relationship Browser, have proved
valuable for translating scientific data and information for use by the public.
Ultimately the EGS data and approaches made accessible by the EnviroAtlas, Eco-Health Relationship
Browser, and tools described in this Report, can help communities account for the gains or losses of
environmental, social, and economic benefits resulting from policy decisions. Translational research
83
-------
supporting EGS-based decision making continues to evolve through case studies and the development of
innovative models. As experience with EGS concepts and tools expands with stakeholders, elected
officials, and resource manages, so too will their relevancy, applicability, and transferability.
A key goal of this report is the demonstration of translational science
through application of tools and approaches in real communities with a
focus on the application of two tools developed by the EPA.
The EnviroAtlas is a tool for identifying and organizing spatial data on EGS
and human health. It is both a source of information and a platform to
combine information in useful ways that improve translational science.
The Eco-Health Relationship Browser is a visualization tool for
understanding connections between EGS and human health and can be
used to support structured approaches to complex decisions. It is based on
peer-reviewed science and allows multiple connections to be explored at
once by bringing the information closer to real-world problems that cross
over disciplinary lines.
84
-------
5. References
Abdelnour, A., M. Stieglitz, F. Pan, and R. McKane. (2011). Catchment hydrological responses to
forest harvest amount and spatial pattern. Water Resources Research 47(9).
Abdelnour, A., R. McKane, M. Stieglitz, F. Pan, and Y. Cheng. (2013). Effects of harvest on carbon
and nitrogen dynamics in a Pacific Northwest forest catchment. Water Resources Research 49(3): 1292-
1313.
Bagstad, K.J., D.J. Semmens, S. Waage, and R. Winthrop. (2013). A comparative assessment of
decision-support tools for ecosystem services quantification and valuation. Ecosystem Services 5:27-39.
Baum, F., A. Lawless, T. Delany, C. Macdougall, C. Williams, D. Broderick, D. Wildgoose, E. Harris,
D. Mcdermott, and I. Kickbush. (2014). Evaluation of Health in All Policies: concept, theory and
application. Health Promotion International 29(supplement 1): 130-142.
BenMAP. (2017). Environmental Benefits Mapping and Analysis Program - Community Edition. U.S.
Environmental Protection Agency, www.epa.gov/benmap: accessed 8/14/18.
Bousquin, J., K. Hychka, and M. Mazzotta. (2015). Benefit Indicators for Flood Regulation Services of
Wetlands: A Modeling Approach. U.S. Environmental Protection Agency, Office of Research and
Development, Atlantic Ecology Division, Narragansett, RI. EPA/600/R-15/191.
Boyd, J.W. and S. Banzhaf. (2007). What are ecosystem services? The need for standardized
environmental accounting units. Ecological Economics 63:616-626.
Boyd, J.W. and A.J. Krupnick. (2013). Using ecological production theory to define and select
environmental commodities for nonmarket valuation. Agricultural and Resource Economics Review
42:1-32.
Bruins, R.J.F., T.J. Canfield, C. Duke, L. Kapustka, A.M. Nahlik, and R.B. Schafer. (2017). Using
ecological production functions to link ecological processes to ecosystem services. Integrated
Environmental Assessment and Management 13(1): 52-61.
Brundtland, G., M. Khalid, S. Agnelli, S. Al-Athel, B. Chidzero, L. Fadika, V. Hauff, I. Lang, M.
Shijun, M. Morino de Botero, M. Singh, S. Okita, and others. (1987). Our Common Future. Oxford
Paperback Reference. Oxford University Press, USA.
Buck, K. D., J.K. Summers, L.M. Smith, and L.C. Harwell. (2018). Application of the Human Weil-
Being Index to sensitive population divisions: A children's well-being index development. Child
Indicators Research 11(4): 1249-1280.
Burnham, D. and E. Bennet. (1909). Plan of Chicago. The Commercial Club, Chicago, IL.
https://archive.org/details/planofchicagoOOburnuoft: accessed 8/7/2018.
Carpenter, S.R., H.A. Mooney, J. Agard, D. Capistrano, R.S. DeFries, S. Diaz, T. Dietz, A.K.
Duraiappah, A. Oteng-Yeboah, H.M. Pereira, C. Perrings, W.V. Reid, J. Sarukhan, R.J. Scholes, and A.
Whyte. (2009). Science for managing ecosystem services: Beyond the Millenium Ecosystem
Assessment. Proceedings of the National Academy of Sciences 106:1305-1312.
85
-------
Cash, D.W., W. Clark, F. Alcock, N.M. Dickson, N. Eckley, D.H. Guston, J. Jager, and R.B. Mitchell.
(2003). Knowledge systems for sustainable development. Proceedings of the National Academy of
Sciences 100:8086-8091.
Costanza, R., L. Daly, L. Fioramonti, E. Giovannini, I. Kubiszewski, L. Mortensen, K. Pickett, K.
Ragnarsdottir, R. Vogli, and R. Wilkinson. (2016). Modeling and measuring sustainable wellbeing in
connection with the UN sustainable development goals. Ecological Economics 130:350-355.
Cummins, S., S. Curtis, A.V. Diez-Roux, and S. Macintyre. (2007). Understanding and representing
'place' in health research: A relational approach. Social Science & Medicine 65:1825-1838.
Dilling, L. and M. Lemos. (2011). Creating usable science: Opportunities and constraints for climate
knowledge use and their implications for science policy. Global Environmental Change 21:680-689.
Dowling, C., S. Walsh, S. Purcell, W. Hynes, and M. Rhodes. (2017). Operationalizing sustainability
within smart cities: Towards an online sustainability indicator tool. International Journal of E-Planning
Research 6:1-17.
Eco-Health Relationship Browser. (2017). U.S. Environmental Protection Agency.
https://vvvvvv.epa.uov/enviroatlas/enviroatlas-eco-health-relationship-brovvser; accessed 8/14/18.
Ellen, I.G. and M.A. Turner. (1997). Does neighborhood matter? Assessing recent evidence. Housing
Policy Debate 8:833-866.
Environmental Council of the States (ECOS). (2017). Cooperative Federalism 2.0: Achieving and
Maintaining a Clean Environment and Protecting Public Health. Environmental Council of the States,
Washington, DC. https://vvvvvv.ecos.oru/vvp-content/uploads/2017/06/ECOS-Cooperative-Federalism-
2.0-June-17-FlNAL.pdf; accessed 8/7/2018.
EnviroAtlas. (2018). U.S. Environmental Protection Agency, www.epa.gov/EnviroAtlas; accessed
8/14/18.
Enquist, C., S. Jackson, G. Garfin, F. Davis, L. Gerber, J.A. Littell, J.L. Tank, A.J. Terando, T.U. Wall,
B. Halpern, J.K. Fliers, T.L. Morelli, E. McNie, N.L. Stephenson, M.A. Williamson, C.A. Woodhouse,
L. Yung, M.W. Brunson, K.R. Hall, L.M. Hallett, D.M. Lawson, M.A. Moritz, K. Nydick, A. Pairis,
A.J. Ray, C. Regan, H.D. Safford, M.W. Safford, M.W. Schwartz, and M.R. Shaw. (2017). Foundations
of translational ecology. Frontiers in Ecology and the Environment 15:541-550.
EcoService Models Library (ESML). (2018). U.S. Environmental Protection Agency.
https://esml.epa.gov/: accessed 8/14/18.
Failing, L., R. Gregory, and M. Harstone. (2007). Integrating science and local knowledge in
environmental risk management: A decision-focused approach. Ecological Economics 64:47-60.
Fulford, R., R. Bruins, T., Canfield, J., Handy, J., Johnston, P. Ringold, M. Russell, N. Seeteram, K.
Winters, and S. Yee. (2016). Lessons Learned in Applying Ecosystem Goods and Services to
Community Decision Making. U.S. Environmental Protection Agency, Gulf Breeze, FL. EPA/600/R -
16/136.
Fulford, R., M. Russell, J. Harvey, and M. Harwell. (2016). Sustainability at the Community Level:
Searching for Common Ground as a Part of a National Strategy for Decision Support. U.S.
86
-------
Environmental Protection Agency, Washington, DC. EPA/600/R-16/178.
Fulford, R.S., I. Krauss, S. Yee, and M. Russell. (2017). A keyword approach to finding common
ground in community-based definitions of human well-being. Human Ecology 45(6): 809-821.
doi: 10.1007/s 10745-017-9940-3.
Godschalk, D.R. and D.C. Rouse. (2015). Sustaining Places: Best Practices for Comprehensive
Plans (Vol. 578). Chicago: American Planning Association.
Granek, E. F., S. Polasky, C.V. Kappel, D.J. Reed, D.M. Stoms, E.W. Koch, C.J. Kennedy, L.A.
Cramer, S.D. Hacker, E.B. Barbier, S. Aswani, M. Ruckelshaus, G.M.E. Perillo, B.R. Silliman, N.
Muthiga, D. Bael, and E. Wolanski. (2010). Ecosystem services as a common language for coastal
ecosystem-based management. Conservation Biology 24:207-216.
Gregory, R., L. Failing, M. Harstone, G. Long, T. McDaniels, and D. Ohlson. (2012). Structured
Decision-making: A Practical Guide to Environmental Management Choices. West Sussex, UK: Wiley-
Blackwell.
Gret-Regamey, A., E. Siren, S.H. Brunner, and B. Weibel. (2017). Review of decision support tools to
operationalize the ecosystem services concept. Ecosystem Services 26:306-315.
Hajkowicz, S. (2007). Can we put a price tag on nature? Rethinking approaches to environmental
valuation. Australasian Journal of Environmental Management 14:22-26.
Hak, T., S. Janouskova, and B. Moldan. (2016). Sustainable development goals: A need for relevant
indicators. Ecological Indicators 60:565-573.
Hall, B.H. and B. Khan. (2003). Adoption of New Technology. UC Berkeley Working Paper,
Department of Economics, UC Berkeley.
Health in All Policies (HIAP) Campaign. (2017). Comprehensive Plan Input Results, Speak Up Speak
Out Campaign, http://www.imagineduluth.com/media/543179/hiap-survev-results.pdf: accessed
6/6/2018.
Homer, C., J. Dewitz, L. Yang, S. Jin, P. Danielson, G. Xian, J. Coulston, N.D. Herold, J.D. Wickham,
and K. Megown. (2015). Completion of the 2011 National Land Cover Database for the conterminous
United States - Representing a decade of land cover change information. Photogrammetric Engineering
and Remote Sensing 81:345-354.
Hsieh, H.F. and S.E. Shannon. (2005). Three approaches to qualitative content analysis. Qualitative
Health Research 15:1277-1288.
Hutchinson, G.E. (1957). Concluding Remarks. Population Studies: Animal Ecology and Demography.
Cold Spring Harbor Symposia on Quantitative Biology 22:415-427.
i-Tree. U.S. Department of Agriculture Forest Service, https://www.itreetools.org/about.php: accessed
8/14/18.
Jackson, L.E., J. Daniel, B. McCorkle, A. Sears, and K.F. Bush. (2013). Linking ecosystem services
and human health: The eco-health relationship browser. International Journal of Public Health
58(5):747-755.
87
-------
Jetoo, S., A. Thorn, K. Friedman, S. Gosman, and G. Krantzberg. (2015). Governance and geopolitics
as drivers of change in the Great Lakes-St. Lawrence basin. Journal of Great Lakes Research
41(1): 108-118.
Johansson, C., B. Lovenheim, P. Schantz, L. Wahlgren, P. Almstrom, A. Markstedt, M. Stromgren, B.
Forsberg, and J. Nilsson-Sommar. (2017). Impacts on air pollution and health by changing commuting
from car to bicycle. Science of the Total Environment 584-585:55-63.
Johnston, J.M., M. Galvin, R. Parmar, and C. Foley. (2016). Facilitating Discovery, Evaluation and
Integration through the Sharing of Watershed Modeling Data. U.S. Environmental Protection Agency,
Athens, GA. EPA/600/R-16/218.
Johnston, J.M., M. Barber, K. Wolfe, M. Galvin, M. Cyterski, andR. Parmar. (2017a). An integrated
ecological modeling system for assessing impacts of multiple stressors on stream and riverine
ecosystem services within river basins. Ecological Modelling 354:104-114
Johnston, J.M., R. de Jesus Crespo, M. Harwell, C. Jackson, M. Myer, N. Seeteram, K. Williams, S.
Yee, and J. Hoffman. (2017b). Valuing Community Benefits of Final Ecosystem Goods and Services:
Human Health and Ethnographic Approaches as Complements to Economic Valuation. U.S.
Environmental Protection Agency, Washington, DC. EPA/600/R-17/309.
Kramer, M.R. and C.R. Hogue. (2009). Is segregation bad for your health? Epidemiologic Reviews
31:178-194.
Landers, D. and A. Nahlik. (2013). Final Ecosystem Goods and Services Classification System (FEGS-
CS). U.S. Environmental Protection Agency, Washington, DC. EPA/600/R-13/ORD-004914.
Lehtonen, M., L. Sebastien, and T. Bauler. (2016). The multiple roles of sustainability indicators in
informational governance: Between intended use and unanticipated influence. Current Opinions in
Environmental Sustainability 18:1-9.
Lemos, M. and B. Morehouse. (2005). The co-production of science and policy in integrated climate
assessments. Global Environmental Change 15:57-68.
Linton, J. andN. Hall. (2012). The Great Lakes: A Model of Transboundary Cooperation. In: Norman,
E., A. Cohen, and K. Baker, (eds.) Water without Borders? Canada, the United States, and Shared
Waters. University of Toronto Press, Toronto, Ont.
Liu, S., R. Costanza, S. Farber, and A. Troy. (2010). Valuing ecosystem services: Theory, practice, and
the need for a transdisciplinary synthesis. Annals of the New York Academy of Sciences 1185:54-78.
MacKenzie, S.H. (1997). Toward integrated resource management: Lessons about the ecosystem
approach from the Laurentian Great Lakes. Environmental Management 21:173-183.
Marmot, M. (2005). Social determinants of health inequalities. The Lancet 365(9464): 1099-1104.
Mazzotta, M., J. Bousquin, J. Ojo, K. Hychka, C. Druschke, W. Berry, and R. McKinney. (2016).
Assessing the Benefits of Wetland Restoration: A Rapid Benefit Indicators Approach for Decision
Makers. U.S. Environmental Protection Agency, Narragansett, RI. EPA/600/R-16/084.
McCann, B. and S. Rynne (eds). (2010). Complete Streets: Best policy and implementation. Planning
88
-------
Advisory Service Report Number 559. American Planning Association. Chicago, IL.
McCool, S. and G. Stankey. (2004). Indicators of sustainability: Challenges and opportunities at the
interface of sciences and policy. Environmental Management 33:294-305.
McGarity, T.O. and W.E. Wagner. (2003). Legal aspects of the regulatory use of environmental
modeling. Environmental Law Reporter News and Analysis 33:10751-10774.
McKane, R., A.G. Abdelnour, A. Brookes, K. Djang, T.E. Jordan, B. Kwiatkowski, F. Pan, W.T.
Peteijohn, M. Stieglitz, A. Skibbe, and D.E. Weller. (2011). Riparian forest effects on nitrogen export
to an agricultural stream inferred from experimental data and a model. American Geophysical Union
Fall Meeting Abstracts.
Meadow, A.M., D.B. Ferguson, Z. Guido, A. Horangic, G. Owen, and T. Wall. (2015). Moving toward
the deliberate coproduction of climate science knowledge. Weather, Climate, and Society 7:179-191.
Millennium Ecosystem Assessment (MEA). (2005). Millennium Ecosystem Assessment. Ecosystems
and Human Well-Being: Synthesis Island Press, Washington, DC.
https://vvvvvv.sciencedirect.com/science/article/pii/S0921800907000341 #bbib23; accessed 8/14/18.
Milwaukee Harbor Plan. (2017). Harbor District Water and Use Plan. Milwaukee, WI Comprehensive
Plan, http: //harb ordi stri ct. org/pi an/; accessed 7/23/2018.
Missimer, M., K. Robert, and G. Broman. (2017a). A strategic approach to social sustainability - part 1:
Exploring the social system. Journal of Cleaner Production 140:32-41.
Missimer, M., K. Robert, and G. Broman. (2017b). A strategic approach to social sustainability - part
2: A principle-based definition. Journal of Cleaner Production 140:42-52.
Mitchell, G. (1996). Problems and fundamentals of sustainable development indicators. Sustainable
Development 4:1-11.
Moon, J., T. DeWitt, M. Errend, R. Bruins, M. Kentula, S.J. Chamberlain, M.S. Fennessy, and K.J.
Naithani. (2017). Model application niche analysis: Assessing the transferability and generalizability of
ecological models. Ecosphere 8(10):e01974. 10.1002/ecs2.1974.
Morello-Frosch, R. and R. Lopez. (2006). The riskscape and the color line: Examining the role of
segregation in environmental health disparities. Environmental Research 102:181-196.
Miiller, F. and B. Burkard. (2012). The indicator side of ecosystem services. Ecosystem Services 1:26-
30.
Munier, N. (2011). Methodology to select a set of urban sustainability indicators to measure the state of
the city, and performance assessment. Ecological Indicators 11:1020-1026.
National Agriculture Imagery Program (NAIP). U.S. Department of Agriculture Farm Service Agency.
https://www.fsa.usda.gov/programs-and-services/aerial-photographv/imagerv-programs/naip-imagery/:
accessed 8/14/18.
National Water Quality Information System. (2018). U.S. Geological Survey.
https://waterdata.usgs.gov/nwis: accessed 8/14/18.
89
-------
National Wetlands Inventory (NWI). (2018). U.S. Fish and Wildlife Service.
https://www.fws.gov/wetlands/; accessed 8/14/18.
Olando, J.L., S.H. Yee, L.C. Harwell, and L.M. Smith. (2017). Technical Guidance for Constructing a
Human Well-being Index (HWBI): A Puerto Rico Example. U.S. Environmental Protection Agency,
Gulf Breeze, FL. EPA/600/R-16/363.
Parmar, R., J.M., Johnston, M. Galvin, and C. Foley. (2016). Watershed and Economic Data
Interoperability (WEDO) System. 9th International Congress on Environmental Modeling and
Software. Toulouse, Fr. Brigham Young University Scholars Archive.
Pickard, B.R., J. Daniel, M. Mehaffey, L. Jackson, and A. Neale. (2015). EnviroAtlas: A new
geospatial tool to foster ecosystem services science and resource management. Ecosystem Services
14:45-55.
Recatala, L. and D. Sacristan. (2014). A minimum indicator set for assessing resources quality and
environmental impacts at planning level in a representative area of the European Mediterranean Region.
Ecological Indicators 45:160-170.
Reed, M., E. Fraser, and A. Dougill. (2006). An adaptive learning process for developing and applying
sustainability indicators with local communities. Ecological Economics 59:406-418.
Rodela, R., A.K. Bregt, A. Ligtenberg, M. Perez-Soba, and P. Verweij. (2017). The social side of
spatial decision support systems: Investigating knowledge integration and learning. Environmental
Science and Policy 76:177-184.
Rosales, N. (2011). Towards the modeling of sustainability into urban planning: Using indicators to
build sustainable cities. Iyrocedia Engineering 21:641 -647.
Rosenthal, A., G. Verutes, E. McKenzie, K. Arkema, L. Bhagabati, L. Bremer, N. Olwero, and A. Vogl.
(2015). Process matters: A framework for conducting decision-relevant assessments of ecosystem
services. International Journal of Biodiversity Science, Ecosystem Services, and Management 11:190-
204.
Russell, M. and J. Harvey. (2016). Tampa Bay Ecosystem Services. U.S. Environmental Protection
Agency, https://archive.epa.gov/ged/tbes/web/html/index.html: accessed 8/27/18.
Russell, M., J. Harvey, P. Ranade, and K. Murphy. (2015). EPA H20 User Manual. U.S.
Environmental Protection Agency, Gulf Ecology Division, Gulf Breeze, FL. EPA/600/R-15/090.
SAS Institute. (2011). Base SAS 9.3 Procedures Guide. SAS Institute Inc. Cary, NC.
Singh, R., H. Murty, S. Gupta, and A. Dikshit. (2009). An overview of sustainability assessment
methodologies. Ecological Indicators 9:189-212.
Smith, L.M., H.M. Smith, J.L. Case, L.C. Harwell, J.K. Summers, and C. Wade. (2012). Indicators and
Methods for Constructing a U.S. Human Well-Being Index (HWBI) for Ecosystem Services Research.
U.S. Environmental Protection Agency, Gulf Breeze, FL. EPA/600/R-12/023.
Smith, L.M., C.M. Wade, K.R. Straub, L.C. Harwell, J.L. Case, M. Harwell, and J.K. Summers.
(2014a). Indicators and Methods for Evaluating Economic, Ecosystem, and Social Services
90
-------
Provisioning. U.S. Environmental Protection Agency, Gulf Breeze, FL. EPA/600/R-14/20184.
Smith, L.M., L. Harwell, J.K. Summers, H.M. Smith, C.M. Wade, K.R. Straub, and J.L. Case. (2014b).
A U.S. Human Well-Being Index (HWBI) for Multiple Scales: Linking Service Provisioning to Human
Well-being Endpoints (2000-2010). U.S. Environmental Protection Agency, Gulf Breeze, FL.
EPA/600/R-14/223.
Smith, L.M., C.M. Wade, J.L. Case, L.C. Harwell, K.R. Straub, and J.K. Summers. (2015). Evaluating
the transferability of a U.S. Human Well-Being Index (HWBI) framework to Native American
populations. Social Indicators Research 124:157-182.
Summers, K., M. Mccullough, E. Smith, M. Gwinn, F. Kremer, M. Sjogren, A., Geller, and M. Slimak.
(2014). The Sustainable and Healthy Communities Research Program: The Environmental Protection
Agency's Research Approach to Assisting Community Decision-Making, p. 273-288. In: Etingoff, K.,
(ed). Urban Land Use: Community-Based Planning, Apple Academic Press and CRC Press.
Summers, J.K., L.C. Harwell, and L.M. Smith. (2016). A model for change: An approach for
forecasting well-being from service-based decisions. Ecological Indicators 69:295-309.
Tallis, H.T., T. Ricketts, A.D. Guerry, S.A. Wood, R. Sharp, E. Nelson, D. Ennaanay, S. Wolny, N.
Olwero, K. Vigerstol, D. Pennington, G. Mendoza, J. Aukema, J. Foster, J. Forrest, D. Cameron, K.
Arkema, E. Lonsdorf, C. Kennedy, G. Verutes, C.K. Kim, G. Guannel, M. Papenfus, J. Toft, M.
Marsik, J. Bernhardt, R. Griffin, K. Glowinski, N. Chaumont, A. Perelman, M. Lacayo, L. Mandle, R.
Griffin, P. Hamel, and R. Chaplin-Kramer. (2013). In VEST 3.0.0 User's Guide. The Natural Capital
Project, Stanford.
Tanguary, G., J. Rajaonson, J. Lefebvre, and P. Lanoie. (2010). Measuring the sustainability of cities:
An analysis of the use of local indicators. Ecological Indicators 10:407-418.
Tran, L. (2016). An interactive method to select a set of sustainable urban development indicators.
Ecological Indicators 61:418-427.
Turcu, C. (2013). Re-thinking sustainability indicators: Local perspectives of urban sustainability.
Journal of Environmental Planning and Management 56:695-719.
Uhlmann, V., W. Rifkin, J. Everingham, B. Head, and K. May. (2014). Prioritizing indicators of
cumulative socio-economic impacts to characterize rapid development of onshore gas resources. The
Extractive Industries and Society 1:189-199.
U.S. EPA. (2012). Decision Support Framework Implementation of the Web-based Environmental
Decision Analysis Application DASEES: Decision Analysis for a Sustainable Environment, Economy,
and Society. U.S. Environmental Protection Agency. EPA/600/R-12/008.
U.S. EPA. (2014). VELMA Eco-hydrological Model, Version 2.0. U.S. Environmental Protection
Agency. EPA/601-F-16-001.
U.S. EPA. (2015a). Community Component Emphasizes Human Health, Well-Being, and the Local
Environment. U.S. EPA EnviroAtlas Factsheet. www.epa.gov/sites/production/files/2015-
07/documents/enviroatlas community factsheet.pdf; accessed 6/23/2018.
U.S. EPA. (2018a). Evaluating a Proposed Policy to Promote Physical Fitness in Public Parks: Using
91
-------
the EnviroAtlas in a Health Impact Assessment in the Town 'n' Country area of Tampa Bay, FL.
https://www.epa.gov/sites/production/files/2018-
02/documents/tampa healthimpactassessment feb2018.pdf; accessed 7/23/2018.
U.S. EPA. (2018b). EPA Strategic Plan FY 2018-2022. U.S. Environmental Protection Agency.
EPA/190/R-18/003.
Valentin, A. and J. Spangenberg. (2000). A guide to community sustainability indicators.
Environmental Impact Assessment Review 20:381-392.
Wall, T.U., E. McNie, and G.M. Garfin. (2017). Use-inspired science: Making science usable by and
useful to decision makers. Frontiers in Ecology and the Environment 15:551-559.
Water Quality Data (WQX) System. (2018). U.S. Environmental Protection Agency.
https://www.epa.gov/waterdata/water-qualitv-data-wqx: accessed 8/14/18.
Wu, J. and T. Wu. (2012). Sustainability indicators and indices: An overview. In: Madu, C. and W.
Kuei, (eds.) Handbook of Sustainable Management. Imperial College Press. London, England, p. 65-86.
Yee, S., J. Bousquin, R. Bruins, T. Canfield, T. DeWitt, R. Jesus-Crespo, B. Dyson, R. Fulford, M.
Harwell, J. Hoffman, C. Littles, J. Johnston, R. McKane, L. Green, M. Russell, L. Sharpe, N. Seeteram,
A. Tashie, and K. Williams. (2017). Practical Strategies for Integrating Final Ecosystem Goods and
Services into Community Decision-Making. U.S. Environmental Protection Agency. EPA/600/R-
17/266.
92
-------
6. Glossary
Alternative future modeling: An approach that models the consequences of a range of choices that
decision makers face today, that help local stakeholders assess the strengths and weaknesses of
each of the policy choices inherent in scenarios.
Area of Concern: Geographic areas within the Great Lakes where significant impairment of beneficial
uses has occurred as a result of human activities at the local level. The term "Area of Concern" is
specific to the Great Lakes as defined in Annex 1 of the 2012 Protocol of the U.S.-Canada Great
Lakes Water Quality Agreement.
Beneficiary: In common usage, a beneficiary is "a person who receives benefits." However, because a
single person with multiple interests can benefit from ecosystems in multiple and distinct ways,
the Final Ecosystem Goods and Services Classification System (FEGS-CS) uses the term
beneficiary to refer to the person's awareness and interests, relative to the EGS, rather than the
person themselves. Therefore, the FEGS-CS defines beneficiary as "The interests of an
individual (i.e., person, group, and/or firm) that drive active or passive consumption and/or
appreciation of ecosystem services resulting in an impact (positive or negative) on their welfare."
Benefits: A good, service, or attribute of a good or service that promotes or enhances the well-being of
an individual, an organization, or a natural system.
Built environment: The manmade surroundings that provide the setting for human activity, including
but not limited to, land use (e.g., open space, green space, buildings and connectivity),
transportation systems (both motorized and non-motorized), buildings, infrastructure (e.g., water
supply and energy networks), and waste and materials management.
Community-based decision support: Scientifically sound and user-friendly assistance for decision
making at the community level.
Community engagement: Involvement of individuals, groups, and firms that have an interest (active or
passive) in the ecosystem/environment.
Conceptual model: A written description and/or visual representation of known or hypothesized
relationships among variables in a system (e.g., human or ecological entities), often representing
causes and effects, environmental stressors, and/or potential management strategies.
Decision analysis: The discipline comprising the philosophy, theory, methodology, and professional
practice necessary to address decisions in a formal manner. Decision analysis includes many
procedures, methods, and tools for identifying, clearly representing, and formally assessing
important aspects of a decision, for prescribing a recommended course of action, and providing
insight for the decision maker and other stakeholders.
Decision context: The environment in which a decision is made, and the environment that will prevail
when the effects of the decision are brought to bear, including the set of values, preferences,
constraints, policies, and regulations that will affect both the decision makers and those
identified as the ultimate beneficiaries.
Decision maker: Individual(s) or groups of people responsible for making choices or determining
policies that impact the functions, processes, and conditions of ecological systems.
Decision science: The quantitative and qualitative methods to understand and support decision-making
processes.
Decision support framework: An organizing structure to support decision making.
Decision support system: An interactive system to aid decision makers in identifying and solving
problems, and making decisions. These systems may use data from observations, output from
statistical or dynamic models, and rules based on expert knowledge.
Decision support tool: A tool that provides resources such as analysis methods, models, data sets,
maps, etc. in order to inform one or more types of decision-making processes.
93
-------
Ecological benefit functions: Usable expressions to quantify a change in human well-being that results
from an ecosystem change.
Ecological benefits: In the context of environmental policy and management, the term applies
specifically to net improvements in social welfare that result from changes in the quantity or
quality of ecosystem goods and services attributable to policy or environmental decision.
Synonymous with "ecosystem-derived benefits".
Ecological model: A structured description of an ecological system with qualitative or quantitative
components.
Ecological output: A biophysical feature, quantity, or quality that is relevant to human well-being. As
such, ecological outputs are the key metrics for evaluating ecosystem services.
Ecological production functions: Usable expressions, such as mathematical models, of the processes
by which ecosystems produce EGS, often including external influences on those processes.
Economic production function: A method to quantify the magnitude of an economic change in
response to a change in a biophysical quantity. Synonymous with "economic benefit function".
Ecosystem benefits: Impacts, positive or negative, on human well-being and social welfare often
resulting from an ecosystem change.
Ecosystem goods and services (EGS): Outputs of ecological processes that directly ("final ecosystem
service" or FEGS) or indirectly ("intermediate ecosystem service" or IEGS) contribute to social
welfare. Some outputs may be bought and sold, but most are not marketed.
Environmental justice: The fair treatment and meaningful involvement of all people regardless of race,
color, national origin, or income, with respect to the development, implementation, and
enforcement of environmental laws, regulations, and policies.
Final ecosystem goods and services (FEGS): Components of nature, directly enjoyed, consumed, or
used to yield human well-being. The FEGS is a biophysical quality or feature and needs minimal
translation for relevance to human well-being. Furthermore, a FEGS is the last step in an
ecological production function before the user interacts with the ecosystem, either by enjoying,
consuming, or using the good or service, or using it as an input in the human economy.
Framework: A logical structure for classifying and organizing complex information.
Health: A state of complete physical, mental, and social well-being and not merely the absence of
disease or infirmity.
Health Impact Assessment: A means of assessing the health impacts of policies, plans, and projects in
diverse economic sectors using quantitative, qualitative, and participatory techniques.
Human well-being: A multidimensional description of the state of people's lives, which encompasses
personal relationships, strong and inclusive communities, meeting basic human needs, good
health, financial and personal security, access to education, adequate free time, connectedness to
the natural environment, rewarding employment, and the ability to achieve personal goals.
Index: Mathematical aggregation of indicators or metrics.
Indicator: An interpretable value or category describing trends in some measurable aspect, often used
singularly or in combination to generate an index.
Intermediate ecosystem goods and services (IEGS): Attributes of ecological structure or processes
(including functions, characteristics, and interactions) that influence the quantity and/or quality
of ecosystem services but do not themselves quantify as final ecosystem goods and services
(because they are not directly enjoyed, consumed, or used by beneficiaries).
Market value: In economics, how much people would be willing to pay for a good or service that is
traded in markets.
Mental health: A state of well-being in which an individual realizes his or her own abilities can cope
with the normal stresses of life, can work productively, and is able to contribute to their
community.
Metric: A [singular] measurable, observable, or interpretable value.
94
-------
Place-based study: A research project focused on a specific geographic location or geographic type.
Policy: A principle or rule to guide decisions and achieve rational outcomes. Policy differs from law.
While law can compel or prohibit behaviors (e.g., a law requiring the payment of taxes on
income), policy merely guides actions towards those that are most likely to achieve desired
outcome.
Remediation: The removal or sequestration of environmental stressors to prevent or ameliorate harm to
human health or the environment.
Resilience: The capacity of a system to absorb disturbance and reorganize while undergoing change so
as to still retain essentially the same function, structure, identity, and feedbacks.
Restoration: Processes that make previously absent or degraded ecosystem goods and services or
beneficial uses of a resource more available.
Revitalization: The improvement of the environmental, social, and physical health of a community or
place.
Scalability: The degree to which a relationship (such as an ecological model) that applies at a given
spatial or temporal scale tends to hold at different (especially larger) dependent and/or
independent scales.
Social welfare function: A relationship for integrating the perceived quality, scarcity, and/or
sustainability of a FEGS into benefit metrics rendering them directly comparable with human
well-being or the benefit.
Stakeholder: An individual, group, or organization with an interest in, or potentially impacted by, the
outcome of a policy or management choice.
Stakeholder engagement: A process through which stakeholders influence and share control over
initiatives and the decisions and resources which affect them.
Structured decision making: An organized approach for identifying and evaluating alternatives that
focuses on engaging stakeholders, experts, and decision makers in productive decision-oriented
analysis and dialogue and that deals proactively with complexity and judgement in decision
making. It provides a framework that becomes a decision focused roadmap for integrating
activities related to planning, analysis, and consultation.
Sustainability: To create and maintain conditions under which humans and nature can exist in
productive harmony that permit fulfilling the social, economic, and other requirements of present
and future generations.
Tradeoff: An exchange of one thing in return for another, especially relinquishment of one benefit or
advantage for another. In a decision-making context, goods and services (including but not
limited to ecosystem goods and services) gained or lost as the result of a management choice.
This involves making judgments (e.g., a choice between different sets of outcomes) about how
much you would give up on one objective in order to achieve gains on another objective. There
is a distinction between tradeoff decisions and the decision method of optimizing among choices.
Transferability: The degree to which a relationship that was developed in a given set of circumstances
can validly be applied in another circumstance.
Translational research: A scientific line of inquiry that applies transdisciplinary techniques and
perspectives that increases the relevancy of scientific data relevant for public discourse, decision
making, and policy formulation.
Utility: In environmental management, the usefulness of a product, tool, or information source to its
intended user.
Valuation: The process of expressing a value for a particular good or service in a certain context (e.g.,
decision making), usually in terms of something that can be counted, often money, but also
through methods and measures from other disciplines (sociology, ecology, and so on).
Values: The things that people believe are important in the way they live and work.
Vulnerability: The stress to which a system is exposed, its sensitivity, and its adaptive capacity (e.g.,
95
-------
the ability of the component of a system to avoid, resist, or be resilient to the stressor).
96
-------
&EPA
United States
Environmental Protection
Agency
Office of Research and
Development (8101R)
Washington, DC 20460
97
------- |