Twinning: How NEON Data Is Bringing Ecosystems Online

What happens when you feed a supercomputer a decade’s worth of weather records, soil data, bug counts and plant measurements? You get a digital twin: a virtual ecosystem built from real-world data that evolves as conditions change. Battelle NEON’s rich biotic and abiotic data from northeastern Colorado are helping researchers build these high-tech mirror worlds and explore how environmental factors and life shape each other.

What Is a Digital Twin in Ecology?

Digital twins are detailed computer models that mirror real-world systems and update as new data come in. Engineers use them to simulate aircraft, power grids or even entire cities. In ecology, the idea is newer—and a lot more complex. Instead of modeling engines or buildings, researchers are trying to capture the dynamic relationships among air, water, soil and living organisms.

The idea started with a simple question: What if we could see the future of an ecosystem before it happened? For scientists studying drought, wildfire or soil health, that kind of foresight could be transformative. But ecological systems are notoriously hard to predict: many variables, few complete datasets. That’s where the concept of a digital twin began to take root. An ecosystem digital twin is a model that doesn’t just simulate the physical environment but connects it to the living systems that depend on it.

The effort, called the Experimental Digital Twin for Ecosystems: Water, Soil, and Drought, is led by Dr. Mike SanClements, Research Initiatives Lead at Battelle-National Ecological Observatory Network (NEON). He explains, “What’s new here is that we’re coupling the biological and ecological data that NEON collects with abiotic models that look at things like precipitation and temperature. So, we can look at those forecasts and ask: how will that ripple through biological factors like plant growth or soil nutrients?”

Dr. Paula Mabee, NEON Director and Chief Scientist, notes that the purpose isn’t just scientific curiosity. “In addition to providing fundamental insights into ecosystems, these models are being designed to give people—from farmers and land managers to city planners and conservation groups—better tools for decision-making. Instead of only knowing how much rain might fall next year, digital twins could show how that rainfall will affect crop yields, wildfire risk or the health of a watershed." In other words, they’re not just forecasting the weather; they’re forecasting ecosystem response. That makes them a powerful tool for disaster preparedness, food security, resource management and conservation.  

“There are so many applications for this model beyond pure research,” says Dr. Cedric Hagen, an Environmental Scientist at NEON. “City planners and insurance companies could use it to guide development or pricing in areas with fire or flood risk. Water managers could use it to forecast spring discharge and plan for drought. Farmers could apply it to understand soil health and potential impacts on crop yields. It’s about giving communities the ability to anticipate change instead of just reacting to it.”

Dr. Chris Florian, Terrestrial Instrument Science Lead at NEON, adds, “In addition to serving the scientific community, comprehensive data collection across ecosystems opens up new possibilities for understanding how we interact with the environment, both from a commercial perspective and in terms of personal risk.”

Building the Digital Twin

The project was made possible through an award from the NSF ASCEND Engine, an innovation network spanning Colorado and Wyoming. Funded by the U.S. National Science Foundation (NSF), the Engine brings together universities, research institutions, large facilities, national labs, startups, large companies and community partners to accelerate new technologies in environmental sensing, modeling and resilience.

The idea of an ecological digital twin emerged through ASCEND’s collaborations on regional challenges. The partnership connects the National Renewable Energy Laboratory (NREL), which develops advanced models for water flow and environmental processes; NVIDIA, whose Earth-2 computing platform provides the Artificial Intelligence (AI) and simulation power to drive real-time forecasting; and NEON, whose standardized ecological and biological data give those models life and context. By combining these strengths, the team is building a powerful framework for understanding how ecosystems may change and for helping communities across the region prepare for what’s ahead.

Ground Truth: Three Sites, Three Stories

The Experimental Digital Twin for Ecosystems is being tested at three NEON sites in Colorado, each chosen to explore a different ecological question and environmental challenge. Together, they provide a living laboratory for understanding how climate and life interact across diverse landscapes.

• At North Sterling (STER), a NEON site in northeast Colorado surrounded by agricultural land, the focus is on soil health and how changes in moisture, temperature and land use affect soil nutrients and productivity. Here the team is partnering with a Colorado startup, PAGE Technologies, to test new low-cost soil sensors that could make monitoring more accessible for farmers and land managers.
• At NEON’s Rocky Mountain National Park (RMNP) site, the models are being used to assess wildfire risk in high-elevation forests. By combining NEON’s vegetation and climate data with regional fire models, researchers hope to improve local forecasting and help communities plan for resilience as fire seasons grow longer and more intense.
• At Como Creek (COMO), NEON’s mountainous headwater site, the focus shifts to water—specifically, spring discharge and water quality. Here, digital twin models are being calibrated to explore how snowpack and precipitation changes influence downstream flow and nutrient transport, shaping ecosystems far beyond the site itself.

Together, these sites form the foundation for a new kind of environmental forecasting that blends climate modeling, hydrology and ecology to capture the full complexity of living systems. The lessons learned in Colorado and Wyoming will help refine the approach, testing how well these digital twins can predict ecosystem change and guide real-world decisions. Because NEON’s data are standardized nationwide, the framework developed here could one day be extended across the entire Observatory network, offering a powerful new lens for understanding—and managing—ecosystem resilience at scale.

Looking Ahead: A Testbed for Innovation

The digital twin initiative is using NEON infrastructure as a testbed for new environmental and agricultural technologies. Through NEON’s Research Support Services, external partners can access sites, data and expertise to develop, validate and refine their own innovations.

That’s already happening through collaborations like the one with PAGE Technologies. Their work demonstrates how real-world testing at NEON sites can accelerate product development while also improving the quality and breadth of environmental data available to science. By opening NEON’s infrastructure as a proving ground for technology and discovery, the Observatory is helping build the next generation of tools—and knowledge—that will support research management and risk calculation for decades to come.

“Partnering with Battelle NEON gives us an unparalleled opportunity to validate our printed soil sensor technology in real-world conditions,” said Dr. Elliot Strand, co-founder and CEO of PAGE Technologies. “By testing at NEON sites like North Sterling, we’re not only advancing affordable tools for farmers and land managers but also contributing to a broader understanding of soil health and ecosystem resilience. This collaboration demonstrates how technology and ecology can work hand in hand to address critical environmental challenges.”

For SanClements, these digital twin projects don’t just create new forecasting tools; they also help shape the future of NEON itself. “When you start thinking about how these models can benefit broader society, you also start to see how these key partnerships and collaborations can strengthen NEON for the future,” he says. “The digital twins we are building give us valuable insight into the power of our data products and how they can evolve to make the Program even more valuable in addressing societal challenges.”

With a standardized network of 80 field sites, a decade of continuous data collection, and a commitment to open access, NEON offers something unique in ecological research: consistent, standardized, high-quality data across both space and time. That structure makes it possible to connect site-level models, like those at the digital twin sites, to regional and even continental-scale models.

“NEON is really in a unique position to provide this wealth of data,” Hagen says. “This isn’t just another AI model. It’s grounded in almost a decade of high-quality data that was painstakingly collected and funded. Imagine what we can do over the next 20 years.”

As digital twins become more sophisticated, their potential will only grow, linking local insight to national understanding and, eventually, global perspective. The next decade will show how these sophisticated models can support something even larger: a living, digital mirror of the natural world that helps us see, understand and protect the ecosystems we all depend on.