Research SUMMARY
EPA Region 1 and ORD Merrimack River Projects
This project summary provides ari overview of the
work EPA Region 1 and the EPA Office of Research and
Development (ORD) have done for the Merrimack
River in the Lawrence, Massachusetts area from 2015
to 2019. EPA produced several deliverables during this
time and collected data sets for research use. Below is
a discussion about the project history and data sets
that EPA collected.
Background of EPA Work on the Merrimack River
The Merrimack River is the source of drinking water for
approximately 600,000 people in New Hampshire and
Massachusetts. The river begins in Northern New
Hampshire and flows into the Atlantic Ocean in
Newburyport, Massachusetts. EPA has been involved in
projects on the Merrimack River for several decades.
In 2015, EPA began working directly with the City of
Lawrence, Massachusetts and other partners as part of
the regional "Making a Visible Difference in
Communities" project. Lawrence is the farthest
downstream of five Massachusetts communities along
the Merrimack River which use the river as their only
source of drinking water. The other Massachusetts
communities using the Merrimack River for drinking
water are Methuen, Andover, Tewksbury, and Lowell.
The Merrimack River is a critical, but threatened
resource. In addition to providing drinking water, it
also receives the discharge of wastewater treatment
effluent, combined sewer overflow, and stormwater
discharges, many of which are from communities
upstream of Lawrence. EPA learned more about the
community's priorities by hosting stakeholder
meetings from 2015 -2017 with City of Lawrence
officials, citizens, planning agencies, non-profit
organizations, and state agencies. Priorities included
addressing water quality concerns and improving the
resiliency of the drinking water treatment plant.
In 2015, staff from ORD, based in Cincinnati, visited
Lawrence to see and learn about some of the city's
water quality and flooding issues. As part of EPA
Region l's "Making a Visible Difference" project in
Lawrence, EPA ORD staff were able to offer technical
assistance to the City of Lawrence. One outcome of
ORD's visit was to assist the community in developing a
"comprehensive water strategy" for the river. This
included conducting research to assess the issues and
possible solutions.
EPA worked closely with Lawrence water officials,
Groundwork Lawrence, the Merrimack River
Watershed Council, and the U.S. Army Corps of
Engineers as a research plan was developed. EPA's
research focused on three objectives:
•	Flooding Vulnerability
•	Water Quality
•	Environmental Justice
Lawrence's drinking water treatment plant is
vulnerable to flooding, as it is in the 100-year flood
zone. One component of the water strategy was to
develop a climate and flooding vulnerability
assessment of the Lawrence drinking water treatment
plant, situated along the Merrimack River. The water
strategy also included mapping and analyzing water
quality data to advance the community's priorities. EPA
captured local community knowledge of sensitive sites
and locations where Lawrence residents go boating,
swimming, and fishing to identify potential exposure
locations. Collaborating with the stakeholders, EPA
gathered historical water quality data and information
on point source discharges (e.g., sewer overflows) to
map against social vulnerability metrics and identified
exposure points. EPA mapped and analyzed flood
zones, using an updated analysis of precipitation data.
EPA developed a mapping tool for the watershed,
bringing together interactive data to visualize the
watersheds greatest challenges and attributes. The
Merrimack mapping tool can be found in the mapping
section of the EPA Merrimack River webpage
(www.epa.gov/merrimackriver). This tool also allows
users to add their own data, enabling those who do not
Office of Research and Development
Center for Environmental Solutions and Emergency Response
Page| 1

-------
have access to GIS tools to create maps to analyze
multiple data layers.
To support these efforts, EPA gathered additional
water quality data from two monitoring stations in the
river. The stations were funded by ORD
(https://www.epa.gov/merrimackriver/basic-
information-about-lower-merrimack-river-monitoring-
station) and operated by the EPA Regional Laboratory.
Water quality monitoring data were used to
characterize the variability of river conditions and to
develop predictive models of contamination levels
impacting its use. EPA collected real-time water
chemistry measurements as well as grab samples for
microbial analysis during wet and dry weather. These
data allowed ORD to evaluate the potential for
nowcasting water quality using real-time monitoring,
observed meteorological information, and river flow
data. Nowcasting is a short time forecasting of water
quality.
Data Collected
During the environmental monitoring and research,
EPA both produced original data sets and gathered
data from other sources to complete their analyses.
EPA deployed two real-time monitors that collected
data every 15 minutes, from December 2016 to 2019.
The preliminary data were displayed in near real-time
on EPA's public website
(https://www.epa.gov/merrimackriver).
Measurements are available on EPA's website and
included the following:
•	Temperature
•	Dissolved oxygen
•	Specific conductance (conductivity)
. pH
•	Turbidity
•	Chlorophyll
•	Phycocyanin
Additional water quality parameters were collected
every 15 minutes at each station. These parameters
were not transmitted to EPA's web page because these
data are considered preliminary due to the
experimental nature of operating this equipment in the
field. EPA scientists and water quality managers used
data from these two stations to assess and understand
water quality conditions. The experimental research
data that were collected included:
•	Total organic carbon (TOC)
•	Fluorescent dissolved organic matter (FDOM)
•	Nitrate
•	Phosphate
The above datasets have not been published or QA/QC
reviewed.
Lastly, from 2016 - 2018, EPA conducted seventeen
grab sampling events in the Lawrence area to examine
bacteria levels. EPA tested for presence of Escherichia
coli and nutrients at six sites during dry and wet
weather conditions. These data have gone through
EPA's internal review and clearance, and are available
upon request.
EPA also gathered additional secondary datasets to use
in their analyses:
•	Merrimack River water level data for the USGS
stations
•	Precipitation datasets for Lawrence and Lowell, MA
•	Lowell CSO event dates and discharge volumes
Analyses Conducted
•	Flooding Analysis: A detailed flooding analysis was
conducted for Lawrence. The results pertain to the
probability of flood levels overtopping the
protective berm of the Lawrence drinking water
plant located at the northern side of the river. Also,
the flooding risk in Spicket River was analyzed to
determine the potential impact on Lawrence water
supply and wastewater collection systems. The
results and datasets include:
o Hydraulic profiles related to the river and
water treatment plant
o River flow and stage modeling
o Areal precipitation and river stage variations in
hydrological modeling
o Flooding recurrence interval and design river
stages in hydrological analysis
o Datasets for both Lawrence and Lowell
Page|2

-------
o Reconstruction of the historical 2006 flooding
map for Lawrence
o Flooding risk and water stages for Spicket River
developed for small-probability floods
• Indicators of Pathogen Levels: River water quality
nowcasting (models) and analyses were conducted
based on datasets from the sensor monitoring
data, river flow data, CSO discharge data, and
other hydrological datasets. The results are in a
summary presentation that is available upon
request. Datasets from the nowcasting analysis
include:
o Nowcasting equations and methods for
estimating E. coli in the river water based on
real-time sensor monitoring results, CSO data
in Lowell, and area precipitation
o River water turbidity nowcasting using river
flow and water turbidity variations, and their
correlations
o Datasets for identifying flow-contaminant
events using coupled sensor stations, and river
stage data
Through modeling and preliminary engineering
analysis, EPA found that the changes in precipitation,
watershed hydrology, and aged water infrastructure
are the major factors affecting water quality and water
supply resilience. EPA presented its research to the City
of Lawrence in 2017 to help the city understand risks to
their water supply in the event of extreme flooding or
power loss.
Future Areas of Research
This work could lead to further investigation by
Merrimack River stakeholders. Below are potential
research questions and technical support needs that
could be further explored.
Nowcasting models of fecal indicator bacteria could be
used to develop a real-time notification system for
recreational activities on the Merrimack River. Such a
system would color-code water quality conditions via a
web app or flagging system to inform the public about
anticipated contamination levels that exceed
acceptable thresholds. These thresholds could be
based directly on recreational water quality criteria for
E. coli, or on pathogen infection risks predicted using
quantitative microbial risk assessment (QMRA). QMRA
would: use the fecal indicator concentrations to model
likely pathogen levels based on their respective
densities in wastewater (a worst-case assumption) and
the fate and transport of each; combine these with
reported water ingestion rates to estimate exposure
doses during various recreational activities; and then
use pathogen-specific dose-response relationships to
quantify associated infection risks for comparison
against defined acceptable rates.
While this modeling introduces additional uncertainty
through its assumptions, it has the benefit of relating
contamination events to explicit risk-based
conclusions. In doing so, the different levels of risk
associated with swimming or non-contact recreation
(e.g., boating or fishing) could be differentiated,
informing which types of activities are suitable under
current water quality conditions. However, because
water quality sensors used to develop the nowcasting
model are no longer in place, and if they cannot be
replaced, new correlations using readily accessible data
sources (e.g., precipitation levels and CSO reporting)
would need to be developed in order to implement the
notification system.
Additional monitoring would be needed to support
future research involving modeling water quality
conditions. Monitoring could be conducted to support
model development or validation. A monitoring plan
would be developed as part of the research needs.
Disclaimer
The views expressed in this document are those of the
authors and do not necessarily represent the views or
the policies of the U.S. Environmental Protection
Agency. This document has been reviewed in
accordance with U.S. Environmental Protection Agency
policy and approved for release.
Contacts
Modeling
Page|3

-------
Jeff Yang, Office of Research and Development, Water
Quality Modeling, vang.ieffffiepa.gov
Michael Jahne, Office of Research and Development,
Microbial Risk Modeling, Jahne.michael(a)epa.gov
Dan Murray, Office of Research and Development, CSO
Infrastructure Technical Support, murrav.danffiepa.gov
Monitoring
Tom Faber, Region 1, Laboratory Services and Applied
Science, faber.tomffiepa.gov
Drinking Water Program
Kira Jacobs, Region 1, Water Division, Source Water
Protection, iacobs.kiraffiepa.gov
Page|4

-------