VICKSBURG, Miss.— Thanks to a historic agreement between the U.S. Army Engineer Research and Development Center (ERDC) and the U.S. Forest Service (USFS), a foundational dataset has been created to bolster the nation’s forests.
ERDC’s Environmental Laboratory (EL) Forest Ecosystem Dynamic (FED) team is partnering with USFS’s R&D Forest Inventory and Analysis Program to establish a collaborative framework that strengthens forest structure and composition modeling capabilities for both agencies. This one-of-a-kind agreement positions ERDC’s FED team as the only Defense R&D team with access to the necessary measured forest data, providing the U.S. Army and Joint Force a decision advantage in vegetated environments.
Successful collaboration between the agencies led to the landmark achievement.
“It has felt great,” said Nathan Beane, FED team lead and senior research forester for EL. “Our team is utilizing the incredible work the USFS has created to monitor and evaluate forests in the United States to enable global modeling at an unprecedented level of accuracy and depth of information. We work closely with USFS leaders to compare algorithms, test new capabilities, and support our shared high-performance computing needs. We’ve developed a solid working relationship, combining our capabilities and expertise with theirs to achieve what has been unachievable. The USFS Forest Inventory and Analysis Program arguably holds the most in-depth national forest inventory available, so building our global framework using this data was necessary,” said Beane.
Artificial intelligence (AI) merged with machine learning (ML) powers this dataset. ERDC FED leverages the two, along with High Performance Computing Modernization Program capabilities and plot-level data throughout North America to produce the data needed to advance capabilities and inform model training for global-scale application.
“Our goal for this project is to provide tools to geospatial analysts to predict forested and other vegetated characteristics anywhere in the world,” said Beane. “Many areas have limited or no data or are inaccessible for a host of reasons. Therefore, we want to enable the capability to predict vegetation structure and composition in places we cannot access by using data from places that we can.”
Plot-level information from more than 355,000 locations across the United States – including all 50 states and territories – will fuel the prediction power of modeling efforts. Gabe Powell, a contracted senior research geoscientist, simplified the role the technology plays and how its value is demonstrated.
“Think of it like building a high-definition, 3D map of every forest on Earth without having to visit every tree,” said Powell. “First, we start with the ‘ground truth’ from those hundreds of thousands of USFS forest inventory plots. Next, we gather terabytes of global environmental data to explain the existence of the structure and composition found at those forest plots. To ensure functionality in access-denied areas, our global explainers come from satellites, which include things like climate, terrain, soil types, and available sunlight. We’re talking about approximately 17 trillion 30m pixels of data.”
The framework is a game-changer for the U.S military, equipping the Warfighter with data to empower its mission. Civilians benefit, too, as the frame exposes forest health and enables the civilian sector to predict forest activity.
“The primary benefit for the Warfighter is a massive leap in situational understanding,” said Powell. “Current vegetation data is often limited to basic landcover maps, essentially telling us if an area is a forest. For the first time, we’re providing data on forest structure and composition at 30-meter resolution. This includes critical characteristics like average tree diameter and spacing, which directly inform mission planning for everything from communications and concealment to manned and unmanned vehicle routing.”
“The benefits are immense for civilians. This project is deeply rooted in the U.S. Forest Service’s Forest Inventory and Analysis (FIA) program,” Powell added. “Often called the ‘nation’s forest census,’ the FIA program has been tracking forest health since 1930. Combining our advanced modeling with the FIA’s extensive ecological data from over 355,000 plots provides the civilian sector with the best global data they have ever seen for predicting wildfire behavior, assessing disaster risks, and understanding carbon stocks and timber resources.”
From this framework research comes the global forest and rangeland toolkit— a product that gives commanders an actionable vegetation knowledge superiority gap that hasn’t existed since the Cold War. The toolkit is in the final development phase in support of the Mission Critical Environmental Intelligence/Intelligent Environmental Battlespace Awareness Project directed by ERDC Installations and Operational Environments.
“The toolkit functions as a set of add-ins for the standard mapping software for the Army’s geospatial analysts,” said Powell. “It contributes directly to our mission of bridging the gap between the raw data and the decision-maker. We are also providing the capability to select vegetation based on stratified above-ground biomass and tree type, by genus, and many other attributes of interest.”
While the toolkit is in production, the FED team has a phased delivery schedule in place for final deliverables and has begun delivering interim capabilities that meet the Army’s needs.