CONTRIBUTOR
A team of Purdue University researchers is pioneering the creation of a metaverse for forestry research to help connect researchers from across the globe and plug critical data gaps used to measure the planet’s green canopies.
The project, led by Jingjing Liang and Rajesh Kalyanam, leverages Science-i, a collaborative web-based platform developed by Liang in 2016.
With a focus on overcoming limitations in accessing forestry data due to political, economic and proprietary factors, the enhanced metaverse integrates a data governance framework and machine learning tools.
Liang explained forest inventory data, especially the ground-sourced forest inventory data, is the most important information available needed to study the global forest ecosystem.
“This data includes everything from species type to local environmental conditions, and most of this data can only be captured in situ,” he says. “The platform aims to connect data owners and data contributors and data users.”
He says although there is an abundance of data available for certain regions (lack of data for other regions—across the global south, for example—is another challenge in forester data) researchers cannot get access to that data. Political, economic and national security considerations impose constraints on sharing forest inventory data, as military reservations often overlap with forests.
Reluctance from nations stems from concerns about the impact on international carbon market negotiations, while private landowners further contribute to restrictions, viewing forest asset details as proprietary information.
“The metaverse we’ve created closely resembles other social media platforms, where people can get together and to share their profiles and to get to know each other,” Liang says. “People can also form private groups and collaborate confidentially, share data and develop research papers.”
The ability to form closed groups helps protect data confidentiality.
“Another function of the platform helps us encourage underrepresented researchers from around the world to connect with other established researchers already in our platform,” Liang says. “If a researcher from Africa has a brilliant research idea about studying a local forest ecosystem, this metaverse connects those people to real data support and to expert teams.
Another asset is a high performance computing platform used to support the community.
“With computing power, with people’s knowledge and with data, those underrepresented researchers will have a better chance of turning their ideas into discoveries,” Liang says.
The platform went online in 2022 and after nearly a year and a half boasts more than 500 members, supporting more than two dozen research projects around the world.
Recent support from the National Science Foundation has given the initiative the ability to address computing capacity gaps, enhance access to critical research data, and provide global expert support, supporting essential pilot studies on forest carbon sequestration and climate change mitigation.
A data governance system has also been installed to protect the data confidentiality, often a requirement from data contributors.
“We cannot release that data to anyone or any project without the approval of the data owners,” Liang says.
He adds the machine learning tool gives researchers who propose research ideas the ability to analyze large quantities of data in a very efficient way.
“This is a very useful tool that was typically not available before,” Liang says.
Only with the recent advent of different AI technologies, including machine learning and deep learning tools, were forestry researchers able to move away from statistical models.
“This is a huge step forward for many of them,” he says. “Most of them are actually underrepresented researchers from the global South, graduate students and female scientists, so we are very proud of this platform and the opportunities it represents.”
