Senator George J. Mitchell Center for Sustainability Solutions

Session J – Integrating Complex Water Quality Data Streams & Identifying Critical Messages for the Public

Afternoon Session

This session is approved for 2 Training Contact Hours through the Maine CDC Drinking Water Program. Sign-up sheets will be available in the session room.

Session Co-chairs:
Danielle Wain, Lake Science Director, 7 Lakes Alliance, djwain@colby.edu
Tahiya Chowdhury, Assistant Professor of Chemistry, Colby College, tchowdhu@colby.edu
Jeremy Deeds, Aquatic Ecologist, Maine Dept. of Environmental Protection, jeremy.deeds@maine.gov
Whitney King, Miselis Professor of Chemistry, Colby College, dwking@colby.edu
Savannah Stresser, Aquatic Business Development Manager, Yokogawa Fluid Imaging Technologies, Savannah.Stresser@fluidimaging.com

Maine is fortunate to have strong networks of environmental nonprofits and other research groups dedicated to the collection, management, and dissemination of lake environmental data. Maine’s extensive data sets depend on a large array of water quality measurements, techniques, and people. However, such a complex matrix of data sources and end users requires new analysis tools. It also demands innovative communication tools to convey critical messages to the public, including identifying critical concepts that will prepare students as the future stewards of Maine’s water resources. This session invites talks that highlight ways to collect and manage complex data streams to support water quality monitoring and modeling, and innovative ways to maximize the impact of the data through changes in behaviors.

Session Schedule

Afternoon Session

• 1:30PM-1:55PM – Sebago Lake Subwatershed Assessment: A Scorecard for Lake Health to Inspire Lake Protection | Presenters: Amanda Pratt, Nicole Merosola
• 2:00PM-2:25PM – How Is My Lake Doing? An Interactive Scorecard for Maine Lakes | Presenter: Jeremy Deeds
• 2:30PM-3:00PM – Afternoon Break
• 3:00PM-3:25PM – Forecasting Lake Water Quality from Datasets with Missing Data towards Data-driven Conservation Action | Presenter: Tahiya Chowdhury
• 3:30PM-3:55PM – Maine Stream Explorers: A Treasure Hunt to Find Healthy Streams in Maine | Presenter: Phil Keefe

Presenters are indicated in bold font.

1:30PM – 1:55PM
Sebago Lake Subwatershed Assessment: a scorecard for lake health to inspire lake protection 

Amanda Pratt, Portland Water District, apratt@pwd.org
Nicole Merosola (student), University of Southern Maine, nicole.merosola@maine.edu
Paul Hunt, Portland Water District, phunt@pwd.org
Firooza Pavri, University of Southern Maine, firooza.pavri@maine.edu

Sebago Lake is the water supply for 200,000 Mainers and is surrounded by a 450 square mile watershed that includes many other lakes and ponds, each of which is surrounded by a smaller watershed – a subwatershed – and is important to someone. Inspiring residents and visitors to care about and help protect Sebago Lake, in some cases very far downstream, will be more effective if that protection is linked directly to protection of their local lake. 

Portland Water District worked with partners to create “Lake Scorecards” for lake subwatersheds for which there was sufficient information – 25 in all.  These scorecards were first produced in 2013. Since then, the “subwatershed assessment” has been updated twice, the most recent just completed using 2024 data. This recent update scores each subwatershed for both current condition and trend for three metrics: water quality, land cover, and land conservation. This presentation will highlight the method we used to do this without gathering any new data – using existing water monitoring, satellite, and GIS data – and how this information has been and will be communicated with stakeholders in the subwatersheds.

2:00PM – 2:25PM
How is my lake doing? An Interactive Scorecard for Maine Lakes 

Jeremy Deeds, Maine Department of Environmental Protection, jeremy.deeds@maine.gov

“How is my lake doing?” is a common question for any lake, but it’s one that can be difficult to answer concisely and accurately. There are a variety of interrelated factors that can influence lake condition and communicating this information well can often be challenging. To address this need in data communication, we constructed a scorecard for Maine lakes that distills the concept of lake condition into four major categories: Water Quality, Invasive Species, Shoreland and Habitat, and Lake Vulnerability. We applied a simple color-coded three-tier scoring system (Good, Fair, Poor) to provide a basic assessment of each. The scores of these four major categories are informed by multiple subcategories, which were scored with the same three-tier system and are also represented in the scorecard. The resulting product is an interactive web-based image, supported with direct links to score calculation methods as well as results and data visualizations specific to individual lakes. The data used to inform the scorecard were collected by professional and citizen scientists. The scorecards were created in R Markdown, an open-source software program that combines the statistical and data visualization strengths of R with html display capabilities. This presentation will discuss how various types of data were used to create the scorecard as well as the development of the web-based reports.

3:00PM – 3:25PM
Forecasting Lake Water Quality from Datasets with Missing Data towards Data-driven Conservation Action

Gent Maksutaj (student), Colby College 
Adrian Gellert (student), Colby College (student)
Danielle Wain, 7 Lakes Alliance
D. Whitney King, Colby College
Tahiya Chowdhury, Colby College, tahiya.chowdhury@colby.edu

Lakes are central to life in Maine — they provide homes for wildlife, places for recreation, and resources that support local economies. Maintaining water quality is essential for healthy lake ecosystems, and declining lake health can negatively hurt both nature and communities that rely on them. While prior works have predicted water quality in a few well-studied lakes, many others lack sufficient data. To address this gap, we utilized decades of historical lake monitoring measurements from over 700 lakes collected by the Lake Stewards of Maine. By applying data imputation techniques, we filled in missing information and predicted Secchi depth, a key indicator of water quality, using chemical, physical, and morphological lake variables. Our initial analysis, conducted on ten lakes, demonstrated that our approach can predict future Secchi depth with an error less than 15%. Interestingly, working from imputed data the simpler statistical methods, such as linear regression and random forest, consistently outperformed more advanced time series forecasting techniques in predicting Secchi depth. This work highlights the power of advanced data analytics in informing sustainable lake management practices. We hope that by bridging gaps in existing databases, this line of work will foster collaborations and empower decision-makers with tools to protect Maine’s lakes and water resources for future generations.

3:30PM – 3:55PM
Maine Stream Explorers: A Treasure Hunt to Find Healthy Streams in Maine

Phil Keefe, Maine Audubon, pkeefe@maineaudubon.org
Sally Stockwell, Maine Audubon, sstockwell@maineaudubon.org
Hannah Young, Maine Audubon, hyoung@maineaudubon.org
Tom Danielson, Maine DEP, thomas.j.danielson@maine.gov
Doug Suitor, Maine DEP, Douglas.Suitor@maine.gov

Stream Explorers enlists the help of volunteer community scientists to survey macroinvertebrates in streams all across Maine.  Different species are considered “sensitive,” “moderately sensitive,” or “tolerant” to poor water quality, and the collection of species found at a site can be used as a good indicator of that stream’s water quality.  Volunteers are trained in survey techniques and identification through on-line webinars and in-person field workshops, and record their data on either paper forms or on line through Survey 1-2-3.  A Basic and Advanced guide and data forms are available on the Maine Audubon website for downloading and printing , and survey kits comprised of guides, nets, plastic tubs, ice cube trays, spoons, droppers etc are available to check out from a variety of locations. The program was launched in 2021 in the Sebago Lake watershed, and has expanded since then to many other areas of the state. Preschoolers, elementary and middle school students, home school groups, families, neighbors, retired folks, and teachers have all participated in the program.  To date, 150 people conducted 107 surveys on 58 streams.  Of those streams, 77 had mostly sensitive species, 13 had both sensitive and tolerant and 17 had mostly tolerant species. Community scientists spent 143 hours sorting and identifying their catch, had fun while learning about the connection between land and water, and now know a lot more about what lives out of sight under water in these streams. Maine DEP is using the data from volunteers as a screening tool to determine if and where they need to do more extensive follow-up biomonitoring surveys.