Overview of the Preliminary Healthy Watersheds Assessments Project:
Evaluating Relative Health and Vulnerability of Conterminous US Watersheds
EPA Office of Water Healthy Watersheds Program, February 2017
1.0 Introduction
In 2009, the US Environmental Protection Agency (EPA) Office of Water created the Healthy Watersheds
Program to help bring more emphasis to protecting high quality waters under the Clean Water Act. This
program supports EPA's state, tribal and other partners by taking non-regulatory, collaborative approaches
to maintaining high quality waters through assessing and protecting watershed health. A major part of this
approach is providing technical analysis, tools and data to help identify healthy watersheds that may
represent good prospects for protection.
This document provides an overview of the Preliminary Healthy Watersheds Assessments (PHWA) project,
which was carried out during 2016-2017 to improve the availability of comprehensive, national information
about watershed health and vulnerability. The goal of this project was to help our partners better target
watershed protection efforts by systematically identifying where healthier and more vulnerable watersheds
occur. In particular, this project was designed to:
>	Ensure that states and other users have basic, statewide and ecoregionally based information on
watershed condition that can help them identify and prioritize opportunities for healthy waters
protection, communicate with protection partners, and support protection approaches;
>	Provide a foundation of nationally available watershed condition data that can be built on and
enhanced as better data become available;
>	Support state efforts to implement the protection goal of the 303(d) Program Vision1 and include
healthy watersheds protection as part of their Integrated Reporting and nonpoint source pollution
control strategies; and
>	Support the Clean Water Act goal of "maintaining" as well as restoring the integrity of waters
across the nation.
2.0 The Healthy Watersheds Assessment Framework
The PHWA's approach was based on the Healthy Watersheds Assessment Framework,2 an analytical
approach influenced by EPA's Science Advisory Board3 writings that identified essential ecological
attributes that support healthy ecosystems. The Healthy Watersheds Assessment Framework focuses on six
key attributes of watershed health: Landscape Condition, Geomorphology, Habitat, Water Quality,
Hydrology, and Biological Condition (Figure 1). This assessment approach was refined over several years
through the completion of projects in 12 states, including in-depth statewide assessments of California,
Wisconsin, Alabama and Tennessee. Whereas specific assessment objectives varied, each was generally
aimed at developing a screening tool to evaluate relative watershed condition (i.e., in comparison to
watersheds across the whole state) to help resource managers plan and target future watershed protection
efforts. Because the Healthy Watersheds Assessment Framework is not designed to make a statement on
the absolute condition of any watershed or water body, these assessments do not define a "healthy
watershed" threshold. In addition to characterizing watershed health, some included estimates of relative
1	USEPA (2013). New Vision for the CWA 303(d) Program - An Updated Framework for Implementing the CWA 303(d)
Program Responsibilities. www.ena.gov/liiidlAiew-vision-cwa-303tl-mDgrani-updated-fTamewoTk-iiirolemenliiig-cwa-303(l-
program-resDonsibililies Accessed 09 February 2017.
2	USEPA (2012). Identifying and Protecting Healthy Watersheds: Concepts, Assessments, and Management Approaches.
https://www.epa.gov/sites/productioii/files/2015-10/docum.eiits/hwi-watersheds-coiiiplete.pdf Accessed 09 February 2017. See
also www.epa. gov/hwp generally.
3	USEPA Science Advisory Board (2002). A Framework for Assessing and Reporting on Ecological Condition.
https://vosemite.epa.gov/sab/sabproduct.nsf/7700D7673673CE83852570CA0075458A/$File/epec02009.pdf.
1

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
watershed vulnerability, defined as the potential for future degradation of watershed processes and aquatic
system health.
Landscape Condition
Patterns of natural land cover, natural disturbance regimes,
lateral and longitudinal connectivity of the aquatic
environment, and continuity of landscape processes.
Habitat
Aquatic, wetland, riparian, flood plain, lake, and shoreline
habitat. Rydrologic connectivity.

Hydrology
Hydrologic regime: Quantity and timing of flow or water
level fluctuation. Highly dependent on the natural flow
(disturbance) regime and hydrologic connectivity, including
surface-ground water interactions.
Geomorphology
Stream channels with natural geomorphic dynamics.
Water Quality
Chemical and physical characteristics of water.
Biological Condition
Biological community diversity, composition,
relative abundance, trophic structure, condition,
and sensitive species.
Figure 1. Six attributes of watershed health described in Identifying and Protecting Healthy Watersheds: Concepts, Assessments,
and Management Approaches (USEPA 2012). Measurement of watershed indicators related to each attribute (i.e., "sub-index")
provides the basis for the Watershed Health Index score.
Within the framework, key ecological and stressor attributes are combined to estimate overall watershed
health and watershed vulnerability. Watershed health is characterized by the presence of natural land cover
that supports hydrologic and geomorphic processes within their natural range of variation, good water
quality, and habitats of sufficient size and connectivity to support healthy, native aquatic and riparian
biological communities. An overall Watershed Health Index is calculated by first identifying measurable
indicators closely associated with each of the six key attributes that appear in Figure 1, compiling sub-
indices for each attribute, then developing the index from all sub-indices. Watershed vulnerability is scored
by a similar process of indicator selection followed by sub-index and index calculation. Key terms are
described in Table 1.
Table 1:
	Key Terms Used in Healthy Watersheds Assessments	
METRIC - a general term for any watershed raw data, indicator, sub-index, or index value.
INDICATOR - a measurable attribute of a watershed that is relevant to a component watershed health
or vulnerability.
SUB-INDEX - a watershed score obtained by considering several indicators relevant to one primary
component of watershed health or vulnerability.
INDEX - a major, overall score per watershed obtained by combining several sub-index scores. The
Preliminary Healthy Watersheds Assessment calculates a Watershed Health and Watershed
Vulnerability index score for each HUC12 watershed.
•	WATERSHED HEALTH INDEX - an integrated measure of the current condition of an
aquatic ecosystem and its surrounding watershed. The PHWA Health Index is comprised of
landscape condition, habitat, hydrology, geomorphology, water quality, and biological condition
sub-indices.
•	WATERSHED VULNERABILITY INDEX - an integrated measure of the potential for future
degradation of watershed processes and aquatic system health. The PHWA Vulnerability Index
is comprised of land use change, water use, and wildfire risk sub-indices.
2

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
3.0	PHWA Methods Overview
3.1	Adapting the Healthy Watersheds Assessment Framework to the PHWA
EPA-assisted healthy watersheds assessments were originally designed and carried out as intensive, state-
specific or basin-specific projects. This approach was too expensive, labor-intensive and data-limited for
all but a few states given the limited resources available. Tradeoffs would be necessary for EPA to assist
all states with basic watershed health and vulnerability information at a still-useful watershed scale. The
substantial number and variety of datasets and indicators covering the lower 48 states presented an
opportunity to advance the availability of healthy watersheds information for most states in one project.
Toward these ends, the PHWA was designed and carried out to help all conterminous states obtain and use
at least a basic level of healthy watersheds assessment data. The PHWA's design remained consistent with
the Healthy Watersheds Assessment Framework while adapting the watershed scale, data sources, and
products to provide more widespread state assistance within the available budget.
Watershed Scale
Healthy watersheds assessments address a specific watershed scale. Some of the past healthy watersheds
assessments compiled scores at the catchment (1 to 2 sq mi in area on average) scale, which produced
highly detailed but labor-intensive and data-limited assessments. The PHWA used the 12-digit hydrologic
unit code4 (HUC12) scale (36 sq mi on average). This scale was selected for three primary reasons: (1)
most watershed planning efforts occur at the HUC12 level, as the area represents a geographic scope that
is small enough to manage but large enough to address water quality problems and the concerns of
stakeholders5; (2) this choice of scale made it possible to use already available data across the
conterminous US; and (3) healthy watersheds data at this scale would be compatible with the wide variety
of other data already at the same scale.
Data Sources
As a national project, use of nationally available datasets was not only efficient but also enabled the
PHWA to assess watersheds in a consistent manner across states and ecoregions. Data sources included
EPA's Watershed Index Online6 and the Recovery Potential Screening7 project, which together have
compiled a database of several hundred HUC12 environmental indicators for the lower 48 states. The
usefulness of these watershed indicators to states and others was largely demonstrated through state-
specific tools for 48 states and projects in 28 states over a ten-year span. Many of these same indicators
have been used in previous healthy watersheds assessments, and it was possible to characterize all key
Healthy Watersheds Assessment Framework attributes in the PHWA using these national data sources
and tools.
PHWA Products
Past healthy watersheds assessments (2011-2015) generated an extensive state-specific technical report,
geodatabase, and project-specific workshop for each project. PHWA outputs for each state include a mix
of generic and state-specific products. State-specific PHWA products include the geodatabase of map
layers, including HUC12 watershed health and vulnerability indicators, sub-index and index scores, plus
more user-friendly maps of high-scoring watersheds at risk and summary scores in data tables. Generic
products include this overview and PHWA overview webinars that will be available upon request.
4Hydrologic Unit Codes (HUCs) in a hierarchical series of scales make up the national Watershed Boundary Dataset or WBD.
Although the WBD undergoes continuous updates, the exact version used for the PHWA HUC12 boundaries is accessible at
http://www.epa. gov/enviroatlas/enviroatlas-data-download-step-2
5	USEPA (2008). Handbook for Developing Watershed Plans to Restore and Protect Our Waters.
https://www.epa.gov/sites/production/files/2015-09/documents/2008	04	18_nps	watershed handbook_handbook-2.pdf.
Accessed 09 February 2017.
6	USEPA (2015). The Watershed Index Online, www.epa.gov/wsio. Accessed 09 February 2017.
7	USEPA (2017). Recovery Potential Screening Tool. www.epa.gov/rps. Accessed 09 February 2017.
3

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
3.2 Statewide and Ecoregional Watershed Assessment
The PHWA is a national project, but is actually comprised of 48 statewide and 85 ecoregional assessments
at the HUC12 watershed8 scale. The PHWA scored each watershed in comparison to the gradients of scores
in both the entire Omernik Level III Ecoregion9 and the entire state in which it occurs (Figure 2). Whereas
the healthier watersheds often score well both statewide and ecoregionally, the two scoring approaches
occasionally contrast in useful ways.
The gradient of watershed health scores across a state allows for watershed comparisons across the political
unit within which many watershed protection decisions take place. Statewide scores provide useful input
to statewide planning processes that protect or avoid impacts to the state's healthier watersheds. These can
include Clean Water Act Section 319 Nonpoint Source Program Plans and, under the Section 303(d) Vision,
Statewide Prioritization Plans. However, top scoring watersheds at the statewide scale are sometimes
clustered in one small region of the state due to exceptional conditions. Although it is certainly useful to
identify and act on such areas, they can outshine the healthier watersheds in other areas when there is a
need for a state program to more equitably recognize and act on healthier watersheds in other parts of the
state as well. Further, statewide-only relative scoring misses the opportunity to compare in-state watersheds
with similar watersheds in adjacent states when prioritizing limited resources and actions.
The PFIWA dual scoring approach also assesses relative watershed health by ecoregion. There are multiple
ecoregions in nearly all states, and examining top watersheds within each in-state ecoregion can reveal a
more geographically distributed group of high scorers. Because most ecoregions cross multiple state lines,
the ecoregionally based scores may reveal whether the in-state portion of one ecoregion is exceptional. For
example, in Figure 2, the majority of Ecoregion 70"s top-scoring HUC 12s (darkest blue) occur within West
Virginia despite that state containing less than a quarter of the ecoregion "s total area. Yet, statewide scoring
alone in this West Virginia example (Figure 2, left side) failed to detect Ecoregion 70's top scorers due to
better relative scores from Ecoregion 67. Further, the ecoregional scoring alternative can help users explore
partnering approaches with neighboring states, such as protecting large interstate high-scoring patches or
restoring healthy watershed corridors across state lines.
PHWA Draft Pccifc Seorct
Statewide (WV)
Ecoregional (ER67, 69, and 70)
NotAnolyted/NoScete
Figure 2. Watershed Health Index measured on West Virginia HUC 12s statewide (left) and individually across whole, multi-
state ecoregions (three maps at right). Each individual HUC12 has two watershed health scores: one relative to all HUC 12s in its
state and the other relative to all HUC 12s in its ecoregion. Statewide and ecoregional scores are often similar but sometimes
differ markedly (e.g., when a HUC 12 scores high statewide but low ecoregionally, or vice versa). Both are provided to offer
different perspectives on identifying the healthier watersheds for protection. (Draft data for demonstration purposes only)
8	For brevity, the PHWA uses the terms watershed and HUC 12 interchangeably. Although HUCs are delineated based on breaks
between overland drainage patterns, many HUCs are not true, full watersheds.
9	USEPA (2013). Level III and IV ecoregions of the continental United States. U.S. EPA, National Health and Environmental
Effects Research Laboratory, Corvallis, Oregon, Map scale 1:3,000,000. Available online at:
https://catalog.data.gov/dataset/omerniks-level-iii-ecoregions-of-the-conterminous-united-states
4

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
3.3	Index Calculation Methods Overview
The following processing steps are repeated for each statewide and ecoregional assessment:
Step 1: Select watershed indicators for assessment area (single state or ecoregion)
A core set of watershed health and vulnerability indicators served as the starting point for all ecoregional
assessments. Indicator sets were sometimes modified at the ecoregional scale depending on ecological
setting and the relevance of core indicators in the ecoregion. Statewide indicator selection included all
indicators used in any of its component ecoregions. When a HUC12 watershed straddled more than one
state or ecoregion, it was scored only in the state/ecoregion in which its majority by area resided.
Step 2: Calculate and normalize watershed indicator values
Directionality of indicator measurement was first configured to ensure that all health indicators have 'higher
is healthier' values, and all vulnerability indicators have 'higher is more vulnerable' values. Watershed
indicator values were calculated for each HUC12 watershed, then scores for each indicator were unity-
normalized within a range of 0 to 1. Indicator values were normalized separately within each ecoregion or
state to correct for differences among the indicators in unit of measurement and to ensure that each indicator
received equal weighting later when calculating sub-index scores. Because scoring is relative to the
geographic area being assessed, ecoregional and statewide scores for the same indicator in the same
watershed can differ.
Step 3: Calculate Sub-Index scores
Watershed indicators were grouped into sub-indices according to the six attributes of watershed health or
three attributes of vulnerability. A watershed's sub-index score was calculated as the mean of its normalized
indicator values. If a watershed had no data for a given indicator, that indicator was not calculated as part
of the sub-index scoring and the sub-index value reflected the mean of the remaining indicators. Watershed
Health sub-indices included Landscape Condition, Hydrology, Geomorphology, Habitat, Water Quality,
and Biological Condition. Watershed Vulnerability Sub-Indices include Land Use Change, Water Use, and
Wildfire Risk. As with their component indicators, all sub-index scores potentially range from 0 to 1. Sub-
indices were not normalized in order that raw sub-index scores more clearly show the actual range and
distribution of scores among watersheds in the ecoregion or state.
Step 4: Calculate Index scores
An overall Watershed Health Index and Watershed Vulnerability Index score was calculated for each
HUC12 as the mean of the sub-index values for that watershed. All sub-indices were equally weighted
within each index. Index scores were presented within a potential maximum range from 0 to 1 where higher
scores denote healthier or more vulnerable watersheds. Like the sub-indices, indices were not normalized
in order that raw scores could more clearly show the actual range and distribution of scores among
watersheds in the ecoregion or state.
3.4	Watershed Indicators
In total, the PHWA utilized 20 indicators representing 6 sub-indices of watershed health and 9 indicators
representing 3 sub-indices of watershed vulnerability. Indicator usage involved a core, national set as well
as limited addition or removal of specific indicators when necessary to address ecoregional differences. For
example, for landscape condition indicators, ecoregion-specific determinations were made regarding the
classification of 'Barren Land' as natural or human use land cover.
The PHWA utilized watershed indicators measured in three different spatial zones of HUC12s: (1) the
watershed (Ws; used to identify properties measured throughout the entire HUC12); (2) the riparian zone
(RZ; delineated by placing a 100-meter buffer around the surface water feature layer from the National
Land Cover Dataset (NLCD) and flowline and waterbody features from the NHDPlus version2); and (3)
the hydrologically active zone (HAZ; defined by the riparian corridor adjacent to surface waters conflated
with areas of high topographic wetness potential that are contiguous to surface waters).
5

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
3.4.1 Watershed Health Indicators
Watershed health indicators selected for the PHWA (Figure 3) reflected the six sub-indices of watershed
health previously discussed. Indicators that comprise each sub-index are described briefly below. Indicator
choices included those with positive as well as negative effects on watershed health; metrics associated
with negative effects were inverted to be directionally consistent with positive (i.e., higher score is
healthier) metrics. Detailed descriptions and source metadata are available in the PHWA data table
spreadsheets compiled for each of the lower 48 states.
Landscape
Condition

% Natural Land
Cover (Ws)
% Natural Land
Cover (HAZ)
Population
Density (Ws)
Population
Density (RZ)
Mining Density
(Ws)
Watershed Health Index
Hydrology
% Ag. on Hydric
Soils (Ws)
Dam Storage
Ratio (Ws)
% Impervious
Cover (Ws)
% Forest
Remaining (Ws)
% Wetlands
Remaining (Ws)
Road Stream
Crossing Density
(Ws)
Geomorphology
Dam Density
(Ws)
% Ditch Drainage
(Ws)
Road Density
(RZ)
% High-Intensity
Land Cover (RZ)
Habitat
NFHP Habitat
Condition Index
Local
Watershed

yv
Biological
Condition
Mean
Probability of
Good Biological
Condition (Ws)
\	/
%
Biological
Condition at
Watershed
Outlet

Water Quality
-
Difference
Between %
Assessed HUC12
Streamlength
Supporting vs.
Impaired y
Difference
Between %
Assessed HUC12
Waterbody area
Supporting vs.
Impaired

( |= Metric score
C3
[ ]= Index score (avg. of sub-index scores)
Watershed (Ws)
~= Sub-Index score (avg. of normalized metric scores) Riparian Zone {RZ)
( 5	' Hydrologically Active Zone (HAZ)
Figure 3. PHWA Watershed Health Index and Sub-Index structure with component indicators (blue boxes). Indicators measure
watershed-wide (Ws), riparian zone (RZ) or hydrologically active zone (HAZ) attributes as marked.
Landscape Condition
Landscape condition is described by the extent and connectivity of natural land cover throughout a
watershed and within key functional zones such as floodplains, riparian areas, and wetlands. Natural land
cover supports watershed health by maintaining hydrologic and geomorphic processes and protecting
aquatic ecosystems from nonpoint sources of pollution. Further, natural land cover in and around the
riparian zone, floodplains, and wetlands filters pollutants, serves as habitat for aquatic and semi-aquatic
species and supports connectivity between habitat patches.
Landscape condition indicators for the PHWA included direct measures of natural land cover extent and
measures of anthropogenic sources of landscape alteration. The extent of natural land cover was measured
throughout the entire HUC12 watershed and within the hydrologically active zone. Indicators of
anthropogenic sources of landscape alteration measured the size of human populations in the watershed and
riparian zone, and the number of mining operations in the watershed. Large human populations reduce
natural vegetative cover through conversion to urban and agriculture lands, while human settlement in
riparian corridors removes the buffer between waterbodies and upland development. Coal and mineral
6

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
mining operations can alter the landscape through vegetation loss and excavation. Other common human
influences on landscape condition such as agriculture and urbanization were reflected in the percent natural
land cover metrics.
Hydrology
The flow regime refers to a stream's characteristic pattern of flow magnitude, timing, frequency, duration,
and rate of change. The flow regime plays a central role in shaping aquatic ecosystems and the health of
biological communities. Aquatic organisms have adapted to the range of physical and chemical conditions
brought about by natural flow patterns. Alteration of natural flows can reduce the quantity and quality of
aquatic habitat, degrade aquatic life, and result in the loss of ecosystem services. Yet, flow measurements
and data are generally not consistently available across the country, or even across most states, requiring
use of surrogate measurements of the factors that strongly influence flow.
Hydrology indicators in the PHWA characterized the maintenance of natural land cover types in the
watershed that support natural flow regimes, and also characterized anthropogenic watershed features that
have the potential to alter hydrologic processes on the landscape or within stream channels. Forest and
wetland cover types help to maintain key hydrological processes such as canopy interception, soil
infiltration rates, evapotranspiration, groundwater recharge, and flood storage within their natural range of
variability. Impervious cover in a watershed can increase the flashiness of streamflow, with high rates of
stormwater runoff, reduced infiltration, and reduced groundwater recharge. The presence of agriculture on
hydric soils in place of wetland cover can also increase flow magnitudes through artificial drainage of wet
soils. While the effects of dams on riverine hydrology vary with storage and release operations, alterations
can include attenuation of high flows, augmentation of low flows, or an increase in the frequency of extreme
low-flows. Road-stream crossings can alter stream velocity and depth above and below their culverts and
interrupt upstream-downstream hydrologic connectivity.
Geomorphology
Stream channel shape, sinuosity, slope, and bed substrate reflect a dynamically stable balance between the
stream's sediment supply and sediment transport capacity. This balance can be destabilized under altered
flow and/or sediment regimes, resulting in a long-term change in channel form, such as incision or
widening. While periodic changes to channel form occur naturally with disturbances such as floods, human
activities often instigate and accelerate geomorphic change that can reduce the quality, extent, and
connectivity of habitat for aquatic organisms. Like hydrology, geomorphically relevant data are not
consistently available across the US and geomorphic condition is inferred through surrogate metrics.
Geomorphology indicators in the PHWA described watershed features that have the potential to alter
geomorphic processes and stream channel form, including dams, artificial drainage ditches, near-stream
roads, and high-intensity land use in the riparian zone. Dams can alter channel geomorphology by slowing
water velocity and increasing sediment deposition above the dam and releasing sediment-deficient water
below the dam outfall. Artificial drainage ditches increase water delivery to streams, resulting in changes
to flow magnitudes and velocities that can drive geomorphic change. The presence of roads and high-
intensity land use types (cropland and medium/high-density urban) in the riparian zone can alter channel
forms through changes to flow patterns or through direct channel manipulations (i.e., channel straightening,
bank armoring, or levee construction).
Habitat
The term aquatic habitat encompasses a host of physical, chemical, and biological characteristics of aquatic
ecosystems, and the optimal set of conditions for aquatic life will vary from one species to another. The
PHWA characterized aquatic habitat using a national, multi-metric index of reach-scale fish habitat
condition developed by the National Fish Habitat Partnership (NFHP)10. The 2015 NFHP Habitat Condition
10 Crawford, S., Whelan, G., Infante, D.M., Blackhart, K., Daniel, W.M., Fuller, P.L., Birdsong, T., Wieferich, D.J., McClees-
Funinan, R., Stedman, S.M., Herreman, K., and Ruhl, P. 2016. Through a Fish's Eye: The Status of Fish Habitats in the United
States 2015. National Fish Habitat Partnership. Accessed on 09 February 2017, at http://assessment.fishhabitat.org/.
7

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
Index (HCI) scored fish habitat condition in stream reaches across the US according to the following
characteristics of the local drainage area of each stream reach: extent of urban and agricultural land cover
types, human population density, road length, number of road-stream crossings, number of dams, number
of mine operations, number of facilities with National Pollutant Discharge Elimination System (NPDES)
wastewater discharge permits, number of sites in the EPA Toxic Release Inventory program, and number
of sites on the Superfund National Priorities List. Each of these measures were selected by the NFHP
because they were determined to be meaningful for assessing fish habitat condition. NFHP Index
calculation inputs and methods were validated by comparison with states and regional fisheries datasets
across the country. The NFHP HCI was first calculated nationally in 2010 and recalculated in 2015.
Biological Condition
A stream's biological condition can be described by the abundance, diversity, and functional organization
of fish, invertebrates, and other aquatic fauna. A healthy biotic community demonstrates a balance of native
species that are integrated across trophic and functional levels and are able to adapt to short- and long-term
variation in ecosystem conditions. EPA's 2008-2009 National Rivers and Streams Assessment (NRSA)11
evaluated the ecological condition of perennial rivers and streams in the conterminous United States by
sampling 1,924 sites randomly selected based on a probability-based design. A multi-metric index based
on measures of benthic macroinvertebrate community was used to represent overall biological condition.
Biological condition at sampling sites was classified as "poor", "fair", or "good" by comparing scores to
the distribution of scores observed as least-disturbed reference sites.
Whereas nationally consistent data estimating biological condition has long been desired but unavailable,
modeled estimates are now emerging. The PHWA characterized biological condition using data from a
landscape metric-based model of NRSA biological condition, developed by EPA's Office of Research and
Development (ORD)12. ORD modeled the probability of "good" biological condition in perennial stream
reaches using local (catchment surrounding stream reach) and cumulative (total upstream contributing area)
watershed-scale landscape predictor metrics from EPA's Stream-Catchment (StreamCat) dataset.
The Mean Probability of Good Biological Condition per HUC12 offered an area-weighted average of all
NHDPlus perennial stream catchments that have a 'probability of good condition' value calculated by the
ORD model. Catchments without a score in the HUC12 do not affect this indicator, which is sensitive only
to what has been modeled. The second metric, Biological Condition Score at Watershed Outlet, reflected
the 'probability of good condition' score for the most downstream NHDPlus perennial stream catchment
receiving flow from upstream portions of the HUC12. As the outlet or 'pour point' integrates much of the
stresses and responses from upstream, it differs from and complements the first metric. Due to the reliance
on perennial streams for this sub-index, HUC12s with no perennial waters did not receive a sub-index score.
Water Quality
Under natural conditions, water chemistry in a waterbody varies within a characteristic range that is
determined by geography, geology, topography, and other characteristics. Aquatic biota adapted to such
conditions and the presence of water quality parameters in their natural range are key features of healthy
streams. The PHWA characterized water quality using assessment data generalized from EPA's Assessment
and Total Maximum Daily Load Tracking and Implementation System (ATTAINS) database13, which
contains nationwide data on assessed and impaired waters assembled from state-specific biennial
assessment reports. Like the Biological Condition sub-index, the development of Water Quality indicators
was challenging because of state-to-state differences in assessment and reporting, as well as limitations in
data availability across the conterminous U.S. For this reason, the Water Quality indicators focused only
11	USEPA (2017). National Rivers and Streams Assessment, https://www.epa.gov/national-aquatic-resource-surveys/nrsa.
Accessed February 9, 2017.
12	Hill, R.A., E.W. Fox, S.G. Leibowitz, A.R. Olsen, D.J. Thornbrugh, and M.H. Weber. Predictive mapping of the biotic
condition of conterminous-USA rivers and streams. Ecological Applications (submitted).
13	USEPA (2017). Assessment and Total Maximum Daily Load Tracking and Implementation System (ATTAINS).
https://ofmpub.epa.gov/waterslO/attains index.home Accessed July 2015.
8

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
on summarizing extent of assessed and impaired versus unimpaired waters in general rather than attempting
to represent pollutant-specific conditions.
PHWA water quality indicators were developed using ATTAINS data 011 the lengths and areas of assessed
and impaired waters within each HUC12 watershed. The first indicator measured the percent of the
Hl.'CI2's assessed stream length fully supporting Designated Uses minus the percent of assessed stream
length that was listed as impaired. By representing the extent of fully supporting stream length as a
proportion of total assessed stream length in the HUC12, this indicator controlled for variability in extent
of assessed waters among watersheds while also considering impairments and attainments. The second
indicator measured the percent of assessed waterbody area (lakes, ponds, etc.) fully supporting Designated
Uses minus the percent of assessed waterbody area that was listed as impaired. Because state-to-state
differences in assessment were substantial in some areas, this sub-index was not calculated in some of the
ecoregional assessments. It was used in every statewide assessment. As with the Biological Condition sub-
index, the Water Quality sub-index may be a good candidate for refinement using state-specific data.
3.4.2 Watershed Vulnerability Indicators
The impact of stressors on different ecosystems in different regions of the United States depends on the
vulnerability of those systems and their ability to adapt to the changes imposed on them. A wide variety of
factors can be related to watershed vulnerability, and the most relevant factors can van substantially from
region to region around the country; national data on some of these factors is also limited. As a result, a
very limited number of vulnerability factors could be assessed nationally in the PHWA, and the
Vulnerability Index should not be considered inclusive of all major forms of watershed vulnerability.
Consideration of individual vulnerability sub-indices may be more appropriate than the overall index in
many areas. The PHWA's approach characterized vulnerability to future degradation based on direction
and magnitude of recent change in combination with selected traits relevant to likelihood of future change.
The PHWA included three vulnerability sub-indices: Land Use Change, Water Use, and Wildfire Risk
(Figure 4).
Watershed Vulnerability Index
Land Use
Change
Water Use
Wildfire
% Human Use
Change (Ws)
(2001-2011)
Agricultural
Water Use (Ws)
Mean Wildfire
Risk(Ws)
Domestic Water
Use (Ws)
% Human Use
Change(RZ)
(2001-2011)
% High or Very
High Wildfire
Risk(Ws)
Industrial Water
Use (Ws)
' Projected >
Change in
Impervious
Cover (Ws)
V (2010-2050) y
% Protected
Lands (Ws)
f ]= Metric score	Watershed (Ws)
]= Sub-Index score (avg. of normalized metric scores) Riparian Zone (RZ)
1 >	B	' Hydrologically Active Zone (HAZ)
| = Index score (avg. of sub-index scores)
Figure 4. PHWA Watershed Vulnerability Index and Sub-Index structure with component indicators (blue boxes).
Indicators measure watershed-wide (Ws) or riparian zone (RZ) attributes as marked.
9

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
Indicator choices included those with positive as well as negative effects on watershed vulnerability; some
metrics (such as percent protected land) were inverted to be directionally consistent with "higher score is
more vulnerable" sub-indices and indices. Detailed descriptions and source metadata are available in the
PHWA data tables spreadsheets compiled for each of the lower 48 states.
Land Use Change
Although a watershed may presently contain large patches of natural vegetative cover, future land cover
change may occur with the expansion of urban and agricultural lands. PHWA land use change indicators
described the vulnerability of watersheds to future land use change. Indicators included measures of recent
increases in urban and agricultural cover in the watershed and in the riparian zone, projected future
impervious cover in the watershed, and the extent of lands with protections in place to moderate impacts
from future land uses.
Water Use
Surface and groundwater withdrawals can greatly alter a watershed's natural hydrologic regime. In many
areas, future water demand will increase beyond current levels as a result of population growth and
expansion of agriculture, industry, and mining. Water use indicators in the PHWA were drawn from EPA's
EnviroAtlas14 and characterized current agricultural, domestic, and industrial water use in the watershed as
an available surrogate measure for future water demand.
Wildfire
As a result of vegetation loss, wildfires can cause changes in hydrologic processes such as overland flow,
interception, evapotranspiration, and as a result can increase runoff, erosion, and sediment delivery.
Characterization of wildfire vulnerability used a measure of wildfire risk from the USDA Forest Service
Wildfire Hazard Potential dataset.15 Areas mapped with higher values represent fuels with a higher
probability of experiencing torching, crowning, and other forms of extreme fire behavior under conducive
weather conditions, based primarily on 2010 landscape conditions.
4.0 PHWA Products
Although the PHWA was national in scope, products were compiled per individual state and include both
statewide and in-state ecoregional results. Each PHWA state package includes this overview document, a
state-specific watershed data file, and a file geodatabase containing PHWA map layers.
Watershed Data File
This Microsoft Excel file was formatted to include user-friendly products such as maps (Figure 5) and data
summaries (Figure 6), as well as all the PHWA data tables and metadata, in a single file format that would
be more accessible to a broader user audience than the geodatabase. These PHWA products can be used to
answer some critical questions:
•	Where are the top-scoring watersheds?
•	Among the healthiest watersheds, which are most vulnerable?
•	Statewide, are watersheds generally high-scoring, low-scoring, or varied?
•	Are there spatial patterns (e.g. basin clustering, high scoring corridors, one high-scoring ecoregion)
across the state?
•	Where might in-state protection efforts be focused? How about interstate collaborative efforts?
•	How similar or different are the sets of healthiest watersheds from within-state compared to the
best from within their respective ecoregions?
Each state-specific PHWA watershed data file includes several tabbed worksheets as described in Table 2.
14	USEPA (2017). EnviroAtlas website. https://www.epa.gov/eiwiroatlas Accessed February 9,2017.
15	https://www.landfire.gov/ Accessed July 2016.
10

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
File Geodatabase
A state-specific ArcGIS file geodatabase was also produced for each state in order to facilitate easier
integration of PHWA results with other state datasets and, if desired, further modification of state-specific
index calculation and data sources. Each state geodatabase includes the HUC12, state and ecoregional
boundaries, with values from all PHWA indicators, sub-indices, and indices as attributes of the HUC12s.
Statewide Watershed Health Index (Percentile)
Water Use Sub-Index
Sub-tmtex/lndex (Perceniite>
2] Iowa Boundary
Ecoregron Boundaries
40 Ecoregron Number
~	state
~	Ecoregion Boundaries
40 Ecoregion Number
Figure 5. Example interpretive maps of Iowa watersheds from a state PHWA Watershed Data File show statewide Watershed
Health Index scores for all HUC12 s (above), and three different sub-indices of vulnerability for only the top 2 5% based on statewide
watershed health (below). Percentile cutoffs represent relatively higher-scoring watersheds and do not imply an absolute healthy
or vulnerable threshold.
11

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
Table 2:
Worksheets and their content in each state-specific PHWA watershed data file.
Introduction
Key to contents of the file and several common terms.
Index Summary by
Watershed
A user-friendly summary of each HUC12 watershed's overall health and vulnerability index
scores. Highlights watersheds that are top-scorers and high-scoring watersheds that are also
vulnerable. Searchable by watershed name.
Health and Vulnerability
Map Series
A set of seven statewide maps that provide a visual summary of PHWA results. Includes
ecoregions, statewide health index (raw and by percentile), top 25% state health index, top 25%
ecoregional health index, top 10% and 25% by both state and ecoregional health index, and four
vulnerability maps for watersheds in the top 25% state health index.
State Health Index and
Sub-Index Scores
Full dataset of HUC12 statewide watershed health index and six component sub-index scores.
Each HUC 12's score is relative to the gradient of scores statewide (max range 0.0 to 1.00). State
border HUC12s are limited to those with >50% of total area in-state.
State Vulnerability Index
and Sub-Index Scores
Full dataset of HUC12 statewide watershed vulnerability index and three component sub-index
scores. Each HUC12's score is relative to the gradient of scores statewide (max range 0.0 to 1.00).
State border E[UC12s are limited to those with >50% of total area in-state.
ER Health Index and
Sub-Index Scores
Full dataset of HUC12 ecoregional watershed health index and six component sub-index scores.
Each HUC 12's score is relative to the gradient of scores ecoregion-wide (max range 0.0 to 1.00),
although only FtUC 12s within this state appear in this table. Ecoregional border HOC 12s are
assigned only to the ecoregion containing the majority of the HUC 12's total area.
ER Vulnerability Index
and Sub-Index Scores
Full dataset of FIUC12 ecoresional watershed vulnerability index and three component sub-index
scores. Each HUC 12's score is relative to the gradient of scores ecoregion-wide (max range 0.0 to
1.00), although only HUC12s within this state appear in this table. Ecoregional border HUC12s
are assigned only to the ecoregion containing the majority of the HUC 12's total area.
Raw Indicator Scores
Raw indicator values per HUC 12 for all indicators used in sub-index and index calculation. These
data are provided in order to support any future effort to revise and recalculate sub-indices or
indices using new indicators, different combinations, or different weights.
Metadata and Definitions
Definitions and sources of indicators, and their usage per ecoregion and state. Also includes basic
field definitions and a glossary of frequently used terms about the data sources.
Colorado's Preliminary Healthy Watersheds Assessment (PHWA) evaluated the relative watershed health
and vulnerability of Colorado's 2,988 12-digit hydrologic unit code (HUC 12) watersheds Watersheds were
assessed at both the statewide and ecoregional scale, resulting in paired Watershed Health and Watershed
Vulnerability scores per HUC12 watershed (i.e., one set of statewide scores and one set of ecoregional scores
per watershed). Together, these scores provide insights on a watershed's condition relative to others within
the state, as well as those watersheds sharing similar ecological characteristics across the ecoregion.
Statewide and ecoregional index scores are presented below as both raw scores ("Score", between 0 and 1)
and percentiles (0 to 100%). The "Top 10%" and 'Top 25%" columns denote watersheds scoring in the top
percentiles of watershed health, both within the state and their ecoregion.
Blue-hiahliahted watershed names indicate those scorino in the Tod 25% of watershed health both within the
state and their ecoreaion. Amona these Tod 25% "healthiest" watersheds, vellow-tiiahliahted watershed
names indicate those that also have an elevated (> 75th percentile) statewide vulnerability score. This
information helps distinguish between healthy watersheds and healthy watersheds most at risk to degradation.
Please note that the full PHWA dataset, including indicator and sub-index scores that comprise each overall
index, is available in other worksheets of this file.

Jienvvr
Pi*Wo
~ 1
¦	Wyoming Basin (18)
¦	Colorado Plateaus (20)
¦	Southern Rockies (21)
Arizona/New Mexico Plateau (22)
_ High Plains (25)
¦	Southwestern Tablelands (26)
~ Colorado Boundary
Figure: Colorado includes
parts of six Omemik Level III
ecoregions
PHWA Watershed Index Summary
WATERSHED HEALTH INDEX
WATERSHED VULNERABILITY INDEX
STATEWIDE
ECOREGIONAL
Top Scoring
Watersheds
STATEWIDE
ECOREGIONAL
Watershed Name |»T
HUC12 ~
ECOREGION "
STATE -
Score »
Percentile^ ¦»
Score »
Percentile

Top 10% ~
Top 25% ~
Score ~
Percentile ~
Score •»
Percentile ~
Bronco Canyon-Purgatorie River
110200101604
26
CO
0.66
17.7
0.82
63.8
No
No
0.13
49.0
0.13
30.4
Browns Canyon
110200010708
21
CO
0.76
41.0
0.74
52.3
No
No
0.11
35.1
0.12
34.8
Browns Creek
110200010704
21
CO
0.81
59.2
0.75
55.5
No
No
0.13
49.6
0.14
47.3
Browns Draw
140500020605
20
CO
0.90
90.7
0.90
84.0
No
Yes
0.35
96.7
0.32
90.0
Brum ley Va I ley-Disappointment Creek
140300020506
20
CO
0.91
94.4
0.91
87.7
No
Yes
0.25
86.0
0.23
69.9
Brunker Creek
102500020101
25
CO
0.81
58.4
0.78
45.1
No
No
0.19
75.1
0.19
76.5
Brush Creek
110200011001
21
CO
0.79
52.6
0.73
48.2
No
No
0.06
6.8
0.07
13.9
Brush Creek
140100050904
20
CO
0.86
76.6
0.83
54.1
No
No
0.23
83.4
0.25
75.4
Brush Creek
140100051106
21
CO
0.77
45.9
0.75
57.2
No
No
0.15
59.4
0.17
54.6
Brush Creek
140200010202
21
CO
0.89
88.5
0.88
96.0
No
Yes
0.05
4.1
0.05
6.4
Brush Creek-Cedar Creek
101900120807
25
CO
0.94
98.0
0.91
89.1
No
Yes
0.11
37.9
0.10
11.5
Brush Creek-Roaring Fork River
140100040602
21
CO
0.66
16.5
0.47
1.7
No
No
0.17
67.0
0.25
75.9
Brush Hoilow Creek-Arkansas River
110200020408
26
CO
0.64
14.3
0.59
5.2
No
No
0.16
62.7
0.20
66.6
Buck Canyon-Two Butte Creek
110200130106
26
CO
0.89
87.3
0.85
77.6
No
Yes
0.16
66.6
0.17
54.2
Buck Creek
101900130404
25
CO
0.71
27.5
0.70
25.3
No
No
0.12
40.6
0.11
25.9
Buck Creek-Hermosa Creek
140801040407
21
CO
0.87
81.5
0.83
84.3
No
Yes
0.20
75.8
0.23
73.4
Bucktaii Creeks-San Miguel River
140300030702
20
CO
0.88
83.9
0.84
58.6
No
No
0.28
90.6
0.26
77.1
Buffalo Creek
101900020303
21
CO
0.74
36.4
0.79
70.8
No
No
0.16
65.7
0.18
60.2
Buffalo Gulch
101900010302
21
CO
0.80
55.2
0.73
47.9
No
No
0.07
9.7
0.07
14.3
Figure 6. Example PHWA Watershed Index Summary from Colorado offers watershed health and vulnerability index scores that
are searchable by watershed name. Higher-scoring watersheds in health alone and health plus vulnerability are highlighted.
12

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
5.0 Using PHWA Results
The PHWA was developed to help EPA and state partners protect high quality waters and maintain their
benefits to society and the environment, by identifying healthier and more vulnerable watersheds and
making this information available to others involved in watershed protection. Potential uses include:
Support state actions to prioritize, protect and maintain high quality waters.
The PHWA was supported by EPA's CWA Section 303(d) Listing and Impaired Waters Program and the
Section 319 Nonpoint Source Control Program, because both programs have encouraged states to more
actively protect high quality waters in recent years. Under the Section 303(d) Vision initiative, states have
systematically identified priority waters and watersheds, including healthy watersheds priorities for
protection over the next several years. Although healthy watersheds assessments exist in some states, the
PHWA supports the Vision's protection goal by nationally improving the basic level of information about
watershed health and the location of high quality watersheds in all the conterminous states. Similarly, the
nonpoint source control programs in states may be able to utilize PHWA information in developing
statewide protection strategies and targeting protection efforts. The PHWA may also prove useful to other
protection-oriented programs within and beyond the Clean Water Act. The PHWA can complement case-
by-case approaches to protecting high quality waters with a more comprehensive, statewide characterization
of watershed health and vulnerability.
Raise awareness of where the healthiest watersheds occur.
One of the most important objectives of the PHWA was to identify the healthier watersheds throughout the
country, and it remains equally important to communicate those findings effectively. In addition to
generating and delivering geospatial data in commonly-used formats including Excel data tables and a file
geodatabase, the PHWA results were packaged in more user-friendly forms. A map series was developed
for each state including statewide maps of watershed health and vulnerability scores, the highest-scoring
watersheds, and the relative vulnerability of those watersheds. The Index Summary sheet of the PHWA
Watershed Data File includes health and vulnerability index scores, and is searchable by individual
watershed name. Overall state summary statistics were also compiled. These PHWA materials can be useful
in raising public awareness of healthy watersheds and their beneficial effects.
Raise awareness that healthy watersheds are sometimes highly vulnerable.
It may come as a surprise that even some of the healthiest watersheds are vulnerable to degradation and the
loss of their beneficial qualities. The PHWA provides insights into which healthy watersheds may be
vulnerable to specific threats and pressures, including land use change, water use, and wildfire. Better
understanding these specific vulnerabilities can help identify good HUC12 candidates for management
efforts, like conservation easement programs, water conservation efforts, and fuel reduction treatments.
This information can also help educate watershed managers, protection partners, and the general public
about the importance of watershed health and resilience to disturbances, such as fires, floods, and droughts.
Broader understanding of vulnerability can help avoid the loss of benefits that healthy watersheds provide
to communities and the environment.
Improve communication and coordination among states, EPA and other partners by providing
nationally-consistent measures of watershed health and vulnerability.
As partners in implementing the Clean Water Act, EPA headquarters, regional offices and states all interact
and communicate about their respective geographic areas of responsibility. Having nationally consistent
sources of watershed condition information can help EPA and the states better understand others' protection
options, plans, and priorities.
Provide a basis for EPA, states and others to promote high quality waters protection in cross-program
interactions and partnering with other landscape management efforts.
Large-scale planning activities such as utilities or transportation routing often seek information on
environmentally sensitive or high quality areas to avoid or minimize impacts. A common concern is
13

-------
PHWA Overview - https://www.epa.gov/hwp/healthy-watersheds-integrated-assessment-reports
avoiding impacts to waterways. A watershed-based dataset on health and vulnerability may be more useful
to such efforts than waterbody-based data because it addresses the whole landscape rather than just the
waters themselves. Further, the PHWA's orientation toward protection of high quality waters makes it
potentially complementary to more terrestrial-oriented landscape protection efforts such as federal, state
and private wildlife management plans. Because all watersheds are scored in the PHWA, it provides a basis
to find locations of mutual interest with other programs.
Provide an initial dataset upon which others can build better watershed condition information.
To some extent, developing the PHWA needed to sacrifice some data richness and variety to establish a
nationally consistent dataset. Thus, the PHWA should be considered a limited, initial dataset on watershed
health and vulnerability that has great potential to be improved and enhanced with more data at statewide
or other scales. For example, an individual state may have extensive bioassessment data that could improve
health index and sub-indices well beyond the more basic data available nationally. A state may choose to
identify thresholds of health or vulnerability score that are appropriate to their watersheds and priorities.
Further, whereas the PHWA used impairments in general to score the Water Quality Sub-index, a single
state may be better-suited to use data on impairment from several specific pollutants to score this attribute
in a more informative manner. Vulnerability assessment as well was able to address just three sub-indices
due to national data limitations, but potential exists to develop and score additional vulnerability factors
where state or other data can support the analysis. As the PHWA assigned equal weights to all indicators
and sub-indices, weighting could also be changed on a statewide or ecoregional basis to match
environmental conditions or program priorities. The full PHWA dataset for each state is provided to
facilitate whatever improvements might be considered over time. As resources permit, EPA's Healthy
Watersheds Program may be able to help states' and others' efforts to build on and improve PHWA data.
6.0 Contact Information and Disclaimer
The Preliminary Healthy Watersheds Assessments project (PHWA) was carried out during 2016 and early
2017 by the US Environmental Protection Agency Office of Water, Healthy Watersheds Program, with
contractual assistance from The Cadmus Group, Inc. Questions, comments and requests pertaining to the
PHWA should be routed through the online "contact us" form available at
https://www.epa.gov/liwp/forms/contact-us-about-healthv-watersheds-protection .
Disclaimer
The information compiled in the Preliminary Healthy Watersheds Assessments project (PHWA) and
presented in this document and related data files is intended to support screening-level assessments to
inform potential watershed protection priorities and is based on modeled and/or aggregated data that may
have been collected or generated for other purposes. Results should be considered in that context and do
not supplant site-specific evidence of watershed health or vulnerability. Scores represent relative gradients
from lowest to highest with reference to watersheds statewide and ecoregion-wide, and no absolute
threshold values of health or vulnerability are implied.
At times, this document refers to statutory and regulatory provisions, which contain legally binding
requirements. This document does not substitute for those provisions or regulations, nor is it a regulation
itself. Thus, it does not impose legally binding requirements on EPA, states, authorized tribes, or the public
and may not apply to a particular situation based on the circumstances.
Reference herein to any specific commercial products, process, or service by trade name, trademark,
manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or
favoring by the U.S. Government. The views and opinions of authors herein do not necessarily state or
reflect those of the U.S. Government and shall not be used for advertising or product endorsement purposes.
14

-------