EPA Nonpoint Source (NPS) Baseline Analysis Workgroup Summary

Workgroup Members

Name

Affiliation

Email

Joanna Ashford

KY Division of Water

joanna.ashford@ky.gov

Katie Basiotis

KS Department of Health and Environment

Katie.Basiotis@ks.gov

Kamilah Carter

EPA Region 4

carter.kamilah@epa.gov

Chris Janssen

KS Department of Health and Environment (moved
to new position at EPA)

janssen.christopher@epa.gov

Aseem Kumar

NY Department of Environmental Conservation

aseem.kumar@dec.ny.gov

Robyn Leto

EPA Region 3

Leto.Robyn@epa.gov

Maria Lopez

EPA Region 10

Lopez.Maria@epa.gov

Cindy Osborn

Minnesota Pollution Control Agency

cynthia.osborn@state.mn.us

Pete Vincent

Ml Department of Environment Great Lakes and
Energy

vincentp@michigan.gov

NPS Equity Initiative and Workgroup Background

The Baseline Analysis Workgroup was initiated as part of the EPA Nonpoint Source (NPS) program's
efforts to assess and expand equity and inclusion across the national program. This work was initiated
by the Near-term Actions to Support Environmental Justice in the Nonpoint Source Program memo. In
the memo EPA has committed to take actions to ensure equitable and fair access to the benefits from
environmental programs including the Nonpoint Source Program.

A piece of this commitment is the "Baseline Analysis" which is a geospatial study that includes
developing a methodology to assess how Section 319 state project funds have historically been
distributed when compared to overburdened/underserved communities. Additionally, this analysis
should develop a means for state programs to assess how future work may better target NPS water
quality problems in overburdened/underserved communities.

EPA NPS HQ program staff began developing this analysis in early 2022 as the program conducted a
series of regional, state, and tribal listening sessions on equity and inclusion in the NPS program. During
this time the team received some initial feedback on data and methods from agency experts in
environmental justice and public health research. Following these listening sessions, a series of
workgroups consisting of EPA, state, territory, and tribal program staff were initiated to further explore
topics that emerged as common themes in the listening sessions.


-------
Baseline Analysis Workgroup Goals and Topics

The overarching goal of this workgroup was to receive feedback and input on the draft Baseline Analysis
data and methodology, and on how to begin in engaging with state programs in piloting the Analysis and
ground truthing findings. This workgroup met between April 4 - June 7 and consisted of four 1.5-2 hour
meetings (some split into multiple time blocks). Specific goals and topics explored by the Baseline
Analysis workgroup include:

•	Existing/ongoing state EJ analysis/mapping efforts

•	Data sources and layers to be included

•	Methods for conducting the Analysis (including use of EPA's Recovery Potential Screening Tool)

•	Engaging with state programs on the Baseline Analysis

This summary was drafted using notes taken during each of the workgroup meetings.

Workgroup Themes/Findings
Data and Methodology

1.	In determining where NPS management efforts can be focused to improve equity and inclusion
across the program, this analysis should consider both social and environmental data.
Specifically, in conducting this analysis the program should endeavor to identify and consider
where the NPS efforts will be useful/able to address water quality issues.

a.	Inclusion of water quality data related to nonpoint source

b.	Program dollars if available (by state)

c.	Public water supply data

d.	Parse out NPS specific TMDLs

2.	Issues related to the scale of the analysis/available datasets

a. The workgroup noted several challenges that can arise when conducting this analysis
using national-scale datasets (i.e., Watershed Index Online (WSIO) data library) and
attempting to set a uniform threshold for determining overburdened/underserved
communities for a national program. Points included:

i.	The socioeconomic context will vary by state. Many factors that contribute to a
community being overburdened/underserved including, urban vs rural
distribution within a state, average income, cultural context/makeup, climate
risk/vulnerability, etc. will differ across the country. A particular concern raised
was that states with majority rural/agricultural areas will not have EJ issues
adequately captured if overburdened/underserved communities are defined
only at the national level. This was of interest to the group as many NPS
programs focus significant effort on agricultural areas/NPS issues.

ii.	CWA Section 319 funding, NPS issues, and priorities also vary greatly by state.
Workgroup members highlighted differences in types of NPS their program
focuses on (i.e., groundwater/drinking water in KS, agriculture NPS issues in KY).


-------
Because a "one size fits all" approach does not fit the national NPS program,
workgroup members expressed the desire for similar flexibility in identifying
overburdened/underserved areas in their states and where to focus future
work.

iii. To address these issues in state variability, one suggestion is to approach the
Baseline Analysis with a two-part methodology. Phase one includes setting
course national baseline so the program can be evaluated at the national level.
Phase two will then allow for states to work one on one with HQ to conduct
more in-depth analysis. A draft procedure is detailed in the "Future
Analysis/Engagement Process" section,
b. Varying geographic scales of available data (watershed vs census/block group level)

i.	This analysis takes a suite of data sources into account, and the geographic
scales of these datasets range from the HUC12 watershed scale to the block
group scale. Both workgroup members and agency experts that have been
consulted on this project have indicated that the watershed scale is too broad to
understand if and how program funds have been spent in
overburdened/underserved communities. Because watersheds often cover
multiple communities/census blocks/block groups, which are the units of
measure most often used in EJ research, the team acknowledges that there may
be a diverse makeup of communities within one watershed. The HUC12
watershed scale was chosen as the primary guiding unit of measure in this
analysis be because that is the level of granularity available for both Section 319
funding data and in the WSIO data library.

ii.	The workgroup raised that the 319 Grants Reporting and Tracking System
(GRTS) allows users to more specifically geolocate project BMPs. This could be
highlighted as an option for states looking to better understand where work is
occurring in relation to census/block group level demographic data.

iii.	If there are state-level datasets that map 319 projects, these maps may also be
included in a state-specific analysis conducted between the state and EPA HQ.

iv.	Data layers from EJSCREEN are available in the Baseline Analysis map.

Watershed scale can be used to home in on areas of interest, and EJSCREEN
data can be used to understand block group makeup with the watershed.

3.	Workgroup members expressed the desire to have the ability to consider additional state and
local data sources in a state-specific analysis, as opposed to relying solely on the national-scale
data that is currently available. The data layers that are currently included in the Baseline
Analysis are hosted on a data viewer in EPA's GeoPlatform which allows the user to add and run
analysis using additional datasets map.

a. Example datasets that workgroup members expressed interest in including in an analysis
of their state: Map of ongoing state efforts to target work in vulnerable communities
(e.g., KS groundwater well mapping with goal to conduct work in low-income areas),
NPS program priority areas, funding/project areas from other programs (i.e., SRF
funding (if available), other state agency priority areas).

4.	Recovery Potential Screening (RPS) tool


-------
a.	Some workgroup members have used the RPS tool in their states as a way to assess
watershed recovery potential. Those with RPS experience had a generally positive
opinion about the tool.

b.	Workgroup suggestion: Use RPS as a potential method of identifying target watersheds
based on social and ecological factors, then bringing into GISto compare against
historical 319 funding. This method works toward the goal of identifying areas that are
potentially overburden/underserved and are experiencing water quality issues that may
be NPS related and/or may be addressed by the NPS program.

c.	RPS allows the user to consider indicators from the Watershed Index Online (WSIO) data
library. When using RPS we would primarily rely on WSIO data because it is available at
a national scale. WSIO indicators were discussed in the workgroup and may be
considered include:

Ecological

Stressor

Social

% Natural Cover (N-lndex)

% Agriculture

% Low- Income Population in WS

% Forest

Number of Septic Systems

% Minority Population in WS

% Wetlands

% Impervious Cover

%< HS Educated Population in WS

% Woody Vegetation

% Tile Drained Cropland

% Linguistically Isolated
Population in WS

Soil Resilience

Linear % Channel Through

% Vulnerable Age Group



Agriculture

Population in WS

% Stream Length Unimpaired

% Urban



Watershed Shape

Channelization



Watershed Size

Hydrologic Alteration



Bank Stability/Soils

Water Use Intensity



Bank Stability/Woody

Number of 303(d) Listed Causes



Vegetation





Corridor Slope

CSS or MS4 Areas



Natural Channel Form

Severity of Loading



Channel Slope

SPARROW Nitrogen Loading
Estimate



Sinuosity

SPARROW Phosphorus Loading
Estimate



Natural Flow Regime

Stream Miles Impaired



Median Flow Maintenance

Waterbody Acres Impaired



Low Flow Maintenance

Number of Impaired Segments



Strahler Stream Order

Specialized Agricultural Practices



Biotic Community Integrity

Impaired Waters % of Watershed



Trophic State

Nutrient Impaired Waters % of
Watershed



Stream Density

Sediment Impaired Waters % of
Watershed



Contiguity with Green





Infrastructure Corridor





Proximity to Green





Infrastructure Hub






-------
d.	Workgroup members expressed the desire to maintain flexibility to include state-
specific datasets in RPS analyses. So, in a more in-depth analysis, a state could also
choose to import state/local datasets for consideration as well.

i. Examples of state-specific data to consider that were mentioned by the
workgroup include: Mining areas, floodplains/flood risk maps, harmful algal
bloom (HAB) data, Karst area, and groundwater watersheds.

e.	RPS allows the user to weight different factors 0-1 based on relative importance.
Workgroup members have not used variable weighting in their RPS work. If weighting
was explored, it should be done in partnership between EPA and the state with
supporting data/reasoning.

Baseline Analysis Limitations

1.	Currently, the Baseline Analysis overlays socioeconomic, environmental, and 319 funding data
to directly compare these datasets within watersheds. This approach limits the ability to capture
and understand upstream work that may be benefitting communities downstream. Workgroup
members had a few suggestions for addressing this issue including:

a.	Strahler stream order in a more in-depth analysis - look at order 3 or above as target
perennial streams - recognizing 1-3 have sig impact but perhaps 3-5 order stream may
have higher impact.

b.	Floodplain targeting - Map floodplain areas to better understand the range of
land targeted that likely expands based on size of order of stream.

c.	Factor in any load reduction modeling conducted as part of state NPS projects

i. Workgroup members also expressed the desire to have access to NRCS EQIP
data

2.	Limitations in available data

a.	319 funding by HUC12: projects may span across multiple HUC12 watersheds. If this is
the case, total funding for the project will be allocated equally across watersheds.

b.	States that put 319 funds in PPGs may not have project dollars fully captured in GRTS,
leading to an underrepresentation of 319 funds spent.

c.	GRTS dollar amounts (appropriation dollars) are skewed by state allocation amounts.

d.	GRTS project locational data has had increasing quality over time. For example, the
accuracy and consistency of GRTS locational data, and their HUC12 co-location, is
substantially better over the previous few years versus 2010.

3.	Assumption that dollars in a HUC12 watershed provide benefit to a disadvantaged community
has high likelihood of being overly optimistic in this level analysis (as noted in
Data/Methodology section). Additionally, depending on the types of practices and the
pollutant(s) being addressed the work may or may not have positive impact on the
overburdened/underserved community within a watershed.

4.	Inherent Bias of 319 funds going towards HUC12 where known water quality issues exist, and
are specifically NPS related impairments. If there is a lack of assessment data in an
overburdened/underserved area, it is less likely 319 funds would have been allocated there.

5.	319 funding levels in urban, MS4 areas may be low. Program guidance prohibits using 319 funds
to meet/comply with NPDES permit requirements. 319 funds may be used in MS4 areas if the


-------
projects are unassociated with or going above and beyond permit requirements. (Include link to
guidance document on spending 319 in urban areas).

Future Analysis/Engagement Process

Due to the complexity of this type of analysis, geographic variability in demographic and water quality
issues, and the desire for state-specific input that was expressed by the group, we have developed a
two-phase analysis approach. This methodology includes setting a broad, national baseline in Phase 1 so
we may achieve a course understanding of how the program has historically been working in potentially
overburdened/underserved communities, as well as a plan to conduct more in-depth, state-specific
analyses. The goal of this Phase 2 state-specific analysis is to address issues of geographic scale
(identified previously in the report), incorporate state/local/Tribal level data in identifying DACs,
consider and understand ongoing work already occurring in this space at the state/local/Tribal level, and
incorporate both water quality and social data in identifying areas of interest for potential future work.

Phase 1 - National Analysis at HUC12 scale (completed June 2022)

Phase one: Feb - August 2022. This first phase sets a coarse, national screening threshold to identify
HUC12 watersheds that cover potentially overburdened/underserved communities. In Phase one,
HUC12 watersheds were selected if they met or exceed the 80th percentile (assessing nationally) of at
least two of the Watershed Index Online (WSIO) Social Indicators that are bolded below (those not in
bold were not included in Phase 1).

WSIO Social Indicators:

•	Low income %

•	Minority %

•	Linguistically isolated %

•	Vulnerable population %

•	Less than HS education %

•	Count of mobile home parks per watershed

After identifying watersheds that met the Phase 1 criteria, they were compared to Section 319 state
project funding data to determine:

•	Number and percent of HUC12 watersheds that meet or exceed the 80th percentile for WSIO
Social Indicators and have received Section 319 funding.

•	Sum of funds and percentage of 319 funds geospatially located in HUC12 watersheds that meet
or exceed the 80th percentile for WSIO Social Indicators

There are several caveats to be aware of in examining Phase 1 results. These topics were discussed in
with the baseline analysis workgroup: A more detailed description of Phase 1 analysis is found in the
workgroup report: (link to Baseline workgroup doc)


-------
Phase 2 - Fine Tuning and In-Depth Analysis at State-Level

Phase 2 of this analysis will include more complex state-specific water quality and demographic
variables. Beginning in FY23 with 5-10 states, EPA will work with interested states to further examine
the watersheds identified in Phase 1 using the Recovery Potential Screening (RPS) tool, which includes
both environmental and socioeconomic factors. Applicable state/local/Tribal water quality data
(including data on NPS priority areas) can also be included to sharpen the screening approach to
watersheds with identified NPS concerns.

This information will then be examined at a sub watershed scale to better understand where DACs are
located within watersheds (potentially using EJScreen or state/local/Tribal socioeconomic data). HQ will
work closely with the state to determine data sources.

The results and lessons from the Phase 2 analysis will be applied to the Phase 1 results to refine the list
of screened DAC areas and will allow for the consideration and exploration of both water quality and
socioeconomic/demographic data.

Phase 2 methodology may be adjusted based on findings from the pilot projects.

States interested in participating in the Phase 2 analysis should contact Ellie Flaherty at
Flaherty.Ellie@epa.gov


-------