The Local Haze blog welcomes Glorianna Davenport, President, and Brian Mayton, Sensor Network Specialist, of the Living Observatory.
As President of this innovative nonprofit, Glorianna Davenport has helped build an interdisciplinary learning community that tells the long-term story of wetland restoration at Tidmarsh Farms and other retired cranberry bogs in southern Massachusetts. Under her leadership, Living Observatory has brought together technologists, scientists, artists, educators, and policymakers to not only measure ecological change but also create immersive experiences that invite the public to witness ecological restoration in real time.
In this interview, we asked Glorianna and Brian to share their insights on how collaboration across diverse disciplines can advance our understanding of wetland ecology and why making complex environmental data accessible to the public is essential to building support for conservation efforts.
Welcome, Glorianna! Your pioneering work at the MIT Media Laboratory helped shape how we interact with digital media and tell stories through technology. Now, as President of Living Observatory, you’re using sensor networks, live cameras, and environmental data to document ecological restoration.
How has your background in interactive cinema and digital storytelling informed the way you approach “documenting the arc of change” in these wetland ecosystems? What parallels do you see between creating interactive narratives and revealing ecological narratives?
Ecological narratives, at least the ones we are working on, are like peeling an onion. There are many layers that make up the arc of change. Story elements go back millions of years. Likewise, we are curious about events that transpired just seconds ago. How do we build a story so that our audience, be they a scientist, a practitioner, or a member of the public, can find a way in that resonates?
Our stories are about the earth and land use change. Marshland is widely recognized as an environmental filter, especially for groundwater, and to a lesser extent for the local atmosphere. Sensors and visualization technologies allow us to experience change at different scales. However, recording and graphing change is not enough; we need to use data to create experiences that amplify the impact in ways that have the potential to change human behavior.
I see these ecological narratives as an exciting subset of interactive narratives; working with this subset, we are learning along the way, modifying what we do, and hopefully coming closer to capturing aspects of “what took place” that can then be projected into an experience that is impactful.
How is monitoring performed?
The Living Observatory’s sensor network is focused on environmental factors that are simple to measure with low-cost hardware. In the air quality world, a multi-tier network of monitors has evolved, with very few “Reference” monitors – which are very expensive and typically provided by national governments. These “Reference” monitors provide a more complete picture of air quality with higher accuracy and precision than is cost-effective at scale. The monitoring performed is then supplemented by a medium-tier of monitors that measure fewer aspects with reasonable accuracy, and finally by an even larger-scale tier of low-cost monitors that measure only a couple of aspects at low accuracy.
Brian, a question for you as someone leading the technology development for Living Observatory: have you considered a tiered system of environmental monitors?
From the beginning, the ecological narrative of change brought on by wetland restoration was the central locus for our work. Rather than choose one sensor, Living Observatory is committed to a multiplicity of views or stories. We invite individual investigators to shape projects around their interests and expertise.
The question was from the start, and remains, what sensor classes would be most effective for capturing, interpreting, and communicating ecological change?
The multiplicity of projects has resulted in the deployment of a wide range of sensors. For my PhD in the Responsive Environments Group at the MIT Media Lab, I built and installed sensor nodes at the Tidmarsh Farms Restoration Project; these nodes measure multiple environmental variables. My colleagues and I also installed a network of microphones to capture sound spatially, as well as cameras to capture video and still frame images of change.
For “The Hydrologic Understory,” Christine Hatch, a Professor at the University of Massachusetts, Amherst, was interested in studying the processes and transitions that occur at the interfaces between streams (surface water) and underground reservoirs (groundwater aquifers). For this project, she deployed a weather station to log precipitation, drone imagery to visualize upwellings of groundwater, and Distributed Temperature Sensing: DTS (multiple fiber-optic cables) to track heat by which she could infer soil moisture at Foothills Preserve, the western portion of Tidmarsh Farms that would undergo restoration in 2020, four years after Tidmarsh.
To bring this back to your question about multi-tiered sensing and use of reference sensors, we have not focused on making any one sensor a standard. In the beginning, we did not know which sensors would be most effective. Rather than choosing one technology, we let many flowers bloom. Projects could look at the “arc of change” on retired cranberry farmland through different lenses. As the collaborative grew, we built up a diverse archive of data that included hydrology, soils, soil moisture, vegetation, wildlife, and several more expressions using both audio and data. Examples include “Bog People” by Hasley Burgund, in which a visitor moving across the property can tune into stories told by the farmer, practitioners and researchers, “Doppelmarsh” in which sensor values drive the graphical expression of the marsh, and a live audio stream that allows anyone to listen into the sound of Tidmarsh.
The existence of these early studies and experiments led to our first contract with the Massachusetts Department of Fish and Game’s Division of Ecological Restoration. Alex Hackman asked us to help develop a monitoring plan for the Division’s priority restoration projects on cranberry farmland. Our members had already collected data on hydrology, soils, vegetation, and wildlife on a number of active farms, retired farms, restored farms, as well as natural wetlands.
The plan was to specify critical indicators and protocols for collecting data that the Division could use to answer questions such as:
– Does the data show that the restored farmland is on a wetland trajectory?
– Does the data indicate ways in which restoration techniques could be modified or improved?
Our early work on the plan led us to engage eleven members of the collaborative as coauthors on a 2019 Benefit Assessment Report, a report that was partially funded by the State of Massachusetts. In 2026, this remains the most comprehensive publication about the restoration of cranberry farmland in Massachusetts. As the monitoring work for the Division progressed, the then Cranberry Bog Restoration Program manager, William Giuliano, asked us to narrow the scope of monitoring to the three critical wetland indicators: hydrology, soils, and vegetation. While we continue to collect this data on the Division’s priority projects, our membership collects a variety of other data on aspects of change across restoration sites.
Recently, Living Observatory members Chris Neill, Christine Hatch, Adrian Wiegman and others have become interested in tracking soil moisture as well as salinity to augment what we are learning about vegetation. As most cranberry bogs are low-lying and often quite near the coast, saltwater intrusion has begun to play a role in farm retirement. Commercially available salinity sensors are not sufficiently sensitive to measure early signs of salinity.
Soil moisture, an important attribute of wetlands, can vary significantly geographically across a restored site, as well as through the seasons and from year to year. We are implementing a multi-tiered approach to monitor soil moisture. Samples can be taken back to the lab for maximum accuracy. Field sampling can be conducted with a high quality portable probe, allowing us to manually make measurements at many locations during a visit. Continuous monitoring from permanently installed sensors can show how the soil moisture changes over time, both on a longer timescale in response to restoration and to transient events like rainstorms. The cost and logistics of permanently installing sensors limits the number of these that we can employ. Currently, we are testing a small number of commercial units as we develop our own.
For the future, we are excited that Ben Weiss, a new student at the Responsive Environments group at the MIT Media Lab, is working with Brian Mayton to build the next generation sensor node. We are hopeful that this node can be manufactured at a lower cost and with a complement of probes, including soil moisture and saline, specifically tailored to our needs. By developing our own sensor devices, we can ensure that they integrate directly with our online platform, thus enabling researchers to quickly deploy and commission new sensors and automatically publish live data from these sensors to their project archive.
For Living Observatory, were there any attempts to focus the monitoring and data collection on the impact on the external environment surrounding your project sites?
In proposals for the restoration of Tidmarsh Farms, we emphasized benefits that could accrue to the watershed. In 2022, Caroline Jaffe earned her MIT PhD for work on An Environmental and Economic Systems Analysis of Land Use Decisions in the Massachusetts Cranberry Industry. While we are still interested in watershed level evaluations and in the socio-economic impact of these wetland restorations, we do not have such projects underway at this time.
An area that continues to interest us is display or experience. How are these records best consumed? When Brian Mayton was developing his PhD, there were multiple projects in the Responsive Environments group focused on the output or display paradigm. Don Dereck Haddad developed Doppelmarsh as a cross-reality browser. Gershon Dublon, Ph.D. built HearThere, a headset that allowed the wearer to be hands-free while controlling how they experienced the surrounding environment’s audio.
Building Learning Collaboratives
Living Observatory brings together technologists, scientists, artists, and restoration practitioners in what you call a “learning collaborative.” This interdisciplinary model seems essential for tackling complex environmental challenges. You also face an operational challenge we see in our air quality monitoring work: maintaining trustworthy, long-term environmental data in the real world.
As we face accelerating climate change and ecosystem degradation, what advice would you offer to other environmental initiatives trying to build similar collaboratives? What have been the unexpected benefits – or challenges – of bringing together such diverse perspectives and expertise?
The growth and value of Living Observatory is gated by our mission, our vision, the people who choose to contribute, our ability to provide tools that support an open-data, interdisciplinary approach, and our funders whose support gates our capacity to grow and learn.
Formed in 2011 and incorporated as a non-profit in 2015, Living Observatory has focused on supporting a specific conservation opportunity, namely the rapid retirement of cranberry farmland in Massachusetts. If left unrestored, the surrounding uplands would likely be developed and the farmed “bogs” would be left to rewild; however, without restoration action, these wetlands would not reach their ecological potential for hundreds of years if ever. As a learning collaborative, we had the opportunity to help through monitoring and making these restorations more visible to the public. There is no template for this endeavor.
Essential to our growth is that we learn by doing. Take, for example, the web-based platform that Brian Mayton has been building. Brian’s initial framework for the platform grew out of his experience in developing a sensor network project for his PhD. At the time there was no integrated platform for sharing projects and data. Today, projects sit at the center of the platform, and are supported by tools and structures that we discover are needed or appropriate. For instance, projects often involve smaller or larger groups of people and can be multi-layered. There is now a team designation on the platform and a story can expand as a project web. Moreover, projects often incorporate learning as they evolve. If we take time to reflect, that learning can be shared benefitting a larger group or the collaborative as a whole. The recently added Insights tool is designed to help collaborators make their learning explicit.
The interdisciplinary nature of Living Observatory allows us to develop a community that is interested and growing and contributing to the initiative of protecting and restoring cranberry farmland to wetlands. We believe that this is best done when we can build a group that is dedicated to individual investigation and to sharing the progress of their investigation and their learning moments with others.
The work Glorianna and her team are doing at Living Observatory shows what’s possible when we combine careful observation with collaborative tools. Please visit the Living Observatory web site for more information.
Thank you Glorianna and Brian, for joining us today and for sharing your insights with the Local Haze community.
Additional reading and information
Living Observatory Projects
Comprehensive Ecological Restoration: renaturalizing retired cranberry farmland
Live Cams & Gallery at Mass Audubon’s Tidmarsh Wildlife Sanctuary
Balantine, Davenport et al, Learning from the Restoration of Wetlands on Cranberry Farmland: Preliminary Benefits Assessment, 2020.