But Can I Swim In It?

Communicating Water Quality Successfully

F. Ryan O’Donnell

2026-03-10

Overview

• Introduction to me
• Water quality communication challenges
• Case study: Creating Clearer Charts
• Case study: Revamping Is It Clean

Who Am I?

• B.S. Geology (2011) & M.S. Sustainability Sciences (2013) from UMass Amherst
• 2012: Volunteered in Connecticut River Conservancy’s (CRC) water quality lab testing E. coli samples
• 2013: Vermont Agency of Natural Resources Intern supporting Southeastern Vermont Watershed Alliance (SeVWA) and Ottauquechee River Group monitoring programs
• 2014+: Continued coordinating SeVWA program and supporting CRC lab and other various projects
• 2017: Started Deerfield River Watershed Association (DRWA) monitoring program covering both MA & VT portion of watershed
• 2018-2025: Became CRC’s water quality program manager bringing together CRC’s MA & CT monitoring with SeVWA and DRWA programs

Water Quality Communication Challenges

Why is it so hard?

• As scientists, we want to communicate precisely
• Audiences have varying skill levels of interpreting data
• Water quality data is complex and noisy
• Data can be both overabundant and sparse
• Water quality professionals are used to looking at this stuff!

What am I?

• As scientists, we want to communicate accurately and precisely
  • We don’t want to be misinterpreted so we attempt to communicate as precisely as we understand the data
  • I ran into this issue a lot when sending copy to communications, they would attempt to simplify the language to make it easier to understand and drastically change the meaning of what I was trying to say
  • It’s a difficult balance to avoid jargon while maintaining clear meanings
• Audiences have varying skill levels of interpreting data - We are trying to communicate to people from a variety of backgrounds and experiences
  • Some people want simple information (ok but, can I swim in it?) and others want to see all the numbers and make their own conclusions
  • Difficult to balance the needs of both ends of this spectrum
  • I.E. data can make more sense using a logirithmic scale but most people are unfamiliar with how to read them
• Water quality data is complex and noisy
  • There are so many external factors like land use, weather, flow, time of year, discharges like combined sewer overflows that affect the numbers
  • Rivers are constantly flowing and so are constantly in flux but grab samples are snapshots in time * Does this spot suddenly have high bacteria because something changed or because a goose flew by and pooped in the water right before I sampled and most of the time it’s still fine? Have to wait for another sample to confirm a trend but audiences sometimes what immediate action
• Data can be both overabundant and sparse
  • Continuous monitoring can produce data points in the thousands
  • Volunteer grab sampling can be restricted by schedule, extreme weather events, and seasonality
• Water quality professionals are used to looking a this stuff!
  • When we’re in this kind of data day in and day out, what may seem simple and common knowledge to us may actually be very niche * look at this charthere
  • some of you may know what it is immediately without being able tosee the axes and legend clearly and many of you probably do not

Start from the basics

I start every presentation with this image

• My favorite diagram of a watershed
• Explains what a watershed is clearly, demonstrates different land uses and subwatersheds
• Has enough fun things to look at for people who might be bored relearning the basics
• I wish I knew where it came from…

Case Study: Creating Clearer Charts

The Problem

• How do we present five years of E. coli data clearly?
• Samples collected by volunteers over five years in three states
• CT & MA sampled on different days of the week, VT sampled every other week
• Weather conditions may be different across such a large area
• Include in a report for colleagues, volunteers, and the public to review on their own

The source of the problem

• Through Is It clean (to be discussed later), results from CRC monitoring are available as soon as possible so reports could focus on long-term trends
• I wanted to aggregate results from a very wide geographic area but
• sites were not all sampled simultaneously
• Even if they had been sampled all that same time, weather can be different

The Data

• Here is just one year of data
• What results are good? Bad?
• Where are the sites?

• I put this on a line chart but it would be just as overwhelming on a bar chart
• Compared to the range of the test, “good results” and “bad results” are right next to each other
• CRC’s site ids provide a lot of information (site, river, and river mile) but it looks like gibberish to most people
• There are some conclusions that can be made using this chart but not a lot
• zoom in

The Data

The Data

Adding more data doesn’t make it better…

Telling a story with data

• Important information:
  • Bacteria levels are different in wet and dry weather
  • Where each site is located relative to each other
  • Some colloquial information about sites
  • What are the standards, when/where are they being met, and what does that mean for the river user?
• Tools
  • Clear, familiar color coding systems
  • Thresholds clearly marked on graph
  • Pair a simplified map with a graph of the results
  • Bonus: Interactivity!

My Favorite Graph

See the rest of the report.

• stoplight colors (also colorblind friendly!)
• The simplified line of the CT River gives someone more familiar with the geography more information without inhibiting those who don’t
• Clear demarkation of the standards for primary and secondary contact recreation simplified to swimming/boating for the layperson that also explains the color coding without adding too much information to the legend
• Use short common names for sites instead of Site IDs
• Even when I don’t have a map, I try to always order sites by upriver to downriver if possible
• If I feel a particular method of displaying the data doesn’t convey the full picture easily, I present multiple formats (ie line graphs of averages AND box and whisker plots of all results)

Case Study: Revamping Is It Clean

Is It Clean?

• Website launched in 2012
• Part of a partnership with the Pioneer Valley Planning Commission
• Showcase E. coli data being collected throughout the Connecticut River Watershed
• Bilingual
• Had a side-mission of serving as CRC’s lab’s water quality database
• Built in Drupal with lots of custom code

• The website was in its final stages of development and had newly gone live when I was a lab volunteer in 2012
• PVPC website was supposed to showcase recreation throughout the Connecticut River valley and Is It Clean was one part of the site
• Initial contributers included CRC, PVPC, SeVWA, Millers River Watershed Council, and Farmington River Watershed Association
• Planned to be available in English and Spanish from the start
• CRC lacked a water quality database so the thinking was for this website to also serve that role
• Was somewhat high-tech at its development using Drupal but had a lot of custom code because there weren’t abundant geographic data tools available yet

Through the ages

• In the beginning
  • 2013 Home Page
  • 2016 Site Page
• Rebranding and Incremental Change
  • 2022 Home Page
  • 2023 Site Page

• Challenges along the way
  • Funding for ongoing maintenance and updates
  • Web developers didn’t understand water quality
    • Lat/Long was backward the ENTIRE time even though we constantly asked for it to be fixed
  • Switched web developers and they didn’t understand website and couldn’t make any changes

• Challenges
  • Funding disappeared after the initial development so the rest of the website was never finished and further Is It Clean development was sparse
  • We had a hard time communicating with our web developers what we wanted and how we wanted it presented
  • Then when we switched web developers, the new firm didn’t understand all the custom code and couldn’t make updates to the site

From the ground up

• Worked with The Commons
• Identified available tools that would work for CRC goals
  • Moved all CRC water quality data out of Excel into AirTable
  • Separated Is It Clean partner data out of CRC data
  • Developed ideal vision for the new website and created RFP
• Fedo.Studio developed Is It Clean
  • Weekly meetings from wireframes through to custom code development
  • Took the time to understand water quality data and what we wanted to communicate

Read The Commons’ case study of the Is It Clean Development here

Meet the new website

Is It Clean

Questions?