NEON Lighting the Way for Future Collaboration and Innovation in Soil Health

My destination has no address, just a set of GPS coordinates in northeastern Colorado that take us down a dirt road ending in the middle of a field. It isn’t the most remote location I have ever visited when reporting on technological innovators, but it’s close.  

I exit the car and head to the only nearby structures: a shipping container and a thirty-foot-tall metal tower decked out with instruments and antennae. Welcome to the National Ecological Observation Network’s (NEON) Sterling NEON field site.

This site is one of 80 field sites that make up an information-gathering network that covers the entirety of the United States, including Alaska, Hawaii, and Puerto Rico. NEON field sites ingest about 5.6 billion data messages a day, and staff members collect thousands of samples every year. Data collected includes lidar, spectrometer, and high-resolution RGB camera data, meteorological information, vegetation greenness, soil information, and observational samplings of animals and plants.  

Dr. Mike SanClements, the Research Initiatives Lead at Battelle-NEON, shows a small group of researchers, local agricultural producers, and NSF Engine members around the site, explaining various monitoring stations and their functions. The tower isn’t the only monitoring location; there are also sensors throughout the area, including ground sensors extending from the tower to a distance of 150 meters.  

There is no other network like NEON in the United States, and SanClements has big plans to leverage the network’s data collection capabilities to make communities across Colorado and Wyoming more resilient.

The Vision

Back in early 2025, the NSF ASCEND Engine (the Engine) hosted an event in Boulder, Colorado, to bring together researchers, industry members, and federal labs interested in the intersecting areas of water quality and availability, soil health, and drought. At this event, SanClements was introduced to key technical leaders and core Engine partners from NVIDIA, and the National Renewable Energy Laboratory (NREL). PAGE Technologies, an Engine-supported soil health startup, was also present, and attendees discovered they had many compelling opportunities for collaboration.

As a result of this meeting, a project proposal titled the Experimental Digital Twin for Ecosystems: Water, Soil, and Drought, was recently funded by the NSF ASCEND Engine. The project will support critical real-world challenges in the agricultural sector by significantly advancing the fundamental science behind digital twinning technologies.

Digital twins have been applied to Earth systems, but previous efforts have largely focused on abiotic data. A multidisciplinary team — SanClements, Paula Mabee (Chief Scientist and Observatory Director at NEON, Battelle), Cedric Hagen (NEON Environmental Scientist), Chris Florian (NEON Terrestrial Instrument Science Lead), along with partners from NVIDIA and NREL — have pioneered a novel approach: developing environmental digital twins that also incorporate biological systems. This effort is grounded in data collected from three NEON sites representing ecosystems in Colorado and Wyoming relevant to the Engine’s wildfire, drought, and water quality use cases.

The project will create one of the first ecosystem-scale digital twin models; models that will increase our understanding of how specific ecosystems react to changes in meteorology, specifically differing levels of precipitation. In the context of the Sterling site, agriculture producers and land managers will be able to utilize the team’s tools to effectively anticipate key environmental conditions of interest such as soil moisture which in turn can inform resource optimization and management strategies.  

To realize this vision, SanClements is leveraging modeling capabilities at NVIDIA and NREL to transform the large-scale standardized data from NEON and novel soil sensing from PAGE Technologies. The combined team brings together unparalleled expertise and understanding of environmental datasets and full-stack AI technologies to develop new, cutting-edge digital twins.

According to SanClements, “The outputs from these digital twin simulations offer substantial and wide-ranging benefits for diverse stakeholders across Colorado and Wyoming, including land and water managers, farmers, city planners, conservation organizers, and insurance providers.”  

Additionally, “by creating high-fidelity digital twins of local ecosystems, we can offer agricultural producers more accurate forecasting tools,” said SanClements, noting that producers wrestle with complex tradeoffs in allocating scarce resources with thin operating margins, often in drought conditions.

Why This Matters

This project exemplifies the goal of the NSF ASCEND Engine, which is to create an innovation ecosystem that brings together the best capabilities in the region to foster regional economic growth and community resilience.  

Better intelligence to estimate complex systems such as those governing precipitation will support efficiency and resilience that will, in turn, minimize costs and operating disruptions due to changes. To date, there is no current technology or practice providing forecasts for these specific societally relevant environmental risks or that incorporate this amount and variety of data.

NEON Sites as a Testbed for New Technologies

PAGE Technologies, a Colorado University – Boulder spinout and one of the participants in this project, is also using NEON’s site as an opportunity to test and validate its low-cost printed soil sensors. According to Elliot Strand, PAGE’s Co-founder and CEO, this “is a huge validation step for us, being able to take our technology out of the laboratory and come to real sites, learn about how to install it, learn about how we can turn that data into actionable insights.”

“I think there's a lot of untapped power here,” says SanClements, referring to the NEON network of sites and data collection capabilities. “We have this widespread and diverse environmental testbed that has engineering and scientific expertise alongside power and communications capabilities, which can be a really great foundational testbed for many types of startups looking to deploy new sensors and technologies.” NEON is ready to support many types of commercial and academic needs through its Research Support Services.  

“The ASCEND Engine has been really great in helping us connect to our partners and collaborators,” says SanClements. “We need the Engine to provide that connection, so that we can share our resources with those companies and our expertise, and they can come to us and we can help them solve their problems.” He hopes that more startups in the future will leverage NEON sensing capabilities.

Lessons learned from deploying PAGE Technology's sensors at NEON will inform best practices to facilitate the growth of NEON’s infrastructure and expertise as a testbed for private industry startups across Colorado and Wyoming. Optimizing NEON’s support services (e.g., technical and data science expertise, data for validation, etc.) will provide resources for small businesses and startups that they would otherwise be unable to access.  

‍