Purdue launches new AI-powered global forest mapping project


WEST LAFAYETTE, Indiana — Purdue University’s Jingjing Liang received a two-year, $870,000 grant from the World Resources Institute to map global forest carbon stock rates.

“Accurately capturing the rate of carbon accumulation in forested ecosystems around the world has always been a difficult task, mostly because doing so requires a lot of terrestrial data, which is currently very limited to the scientific community,” said Liang. He is Associate Professor of Quantitative Forest Ecology and Co-Director of the Forest Advanced Computing and Artificial Intelligence Lab.

“This task is much more difficult than mapping carbon emissions from forest loss,” said Nancy Harris, research director for the Land & Carbon Lab at the World Resources Institute, a nonprofit research organization based in Washington, DC. When trees are cut, satellite imagery signals that forest carbon stocks are greatly reduced and relatively sharp atmospheric emission pulses occur. Through sequestration, forests accumulate carbon gradually and nonlinearly.

“Even state-of-the-art satellite sensors cannot reliably capture it on their own, especially in old forests where the signal is saturated. Forests stop growing tall long before they stop accumulating carbon.”

Forest carbon stock rates are sensitive to subtle changes in three forest growth factors: ingrowth, upgrowth and mortality. Ingrowth refers to the number of small seedlings that have reached a certain threshold size, called trees. Upgrowth is the gradual increase in diameter of a tree through the process of photosynthesis. Ground-source forest stock data measured at multiple time points is currently the only reliable source of information for accurate quantification of these three components of forest growth.

“Until now, people have not been able to estimate the ingrowth, upgrowth and mortality of individual forest stands on a global scale. This information gap leaves great uncertainty about the size, location and trends of global forest carbon sinks,” said Liang.

liang-grant2
Purdue’s new World Resources Institute project will leverage extensive land-supply forest inventory data through re-estimation collected by Science-i and the Global Forest Biodiversity Initiative. (Photo courtesy of Javier Garcia-Perez) Download image

Liang is combining information collected on billions of trees measured in the field with satellite and other geospatial data to develop artificial intelligence models that will map regional forest growth rates across global forest coverage.

“This will be the first AI-based forest growth model to be deployed globally,” he said. In addition to accurately quantifying carbon dynamics, Liang’s AI-based forest growth model also captures the dynamics of forest biodiversity and wood quality.

“We are excited to support the growth of this research collaboration,” said Harris. “The spatially disaggregated data this new project will provide will help us better understand the role the planet’s forests play in nature-based solutions for mitigating global climate change. The initiative’s inclusive, globally networked approach is at the heart of WRI’s Land & Carbon Lab mission.”

Developing these models requires massive computing power and comprehensive global data coverage. Purdue’s state-of-the-art high-performance computing clusters will provide ample computing support. Nonetheless, achieving comprehensive global coverage of ground-sourced plot data remains challenging, particularly in tropical countries.

“Data from these countries has historically been limited,” said Liang. “Through the newly established network of Science-i and its sister consortium, the Global Forest Biodiversity Initiative, we already have partnerships with many scientists around the world collecting and sharing these data.”

Liang founded Science-i, a web-based collaboration platform involving more than 300 scientists worldwide. He also co-founded the Global Forest Biodiversity Initiative, which has built a database of 1.3 million sample plots and 55 million trees. That database is the foundation of the project.

“We will be collecting much more data to fill these data gaps, especially in the Southern Hemisphere,” said Liang. “We will engage more people, especially those from marginalized groups.”

Collaborators on this project already include representatives of indigenous groups in North America, the Amazon, Africa and elsewhere. Rural communities, forestry practitioners and citizen scientists will also be project collaborators.

“We co-create knowledge according to FAIR’s principles of global collaboration: discoverable, accessible, interoperable and reusable,” said Liang.

“At Science-i, everyone works together as equal partners on every project. We share results openly through real-time, transparent discussions across the entire team. Then, at the end, we cross-evaluate and integrate the research results. This is a whole new way of doing international co-forestry research.”

The broad global partnerships and comprehensive forest database created by this project will complement Purdue’s Digital Forestry Initiative, which seeks to leverage technology and comprehensive expertise to measure, monitor and manage urban and rural forests.

Liang, along with Associate Professor of Biological Sciences Ximena Bernal, is co-director of the Biodiversity Research Community, part of Purdue’s recently launched comprehensive Institute for a Sustainable Future.

Author: Steve Copes

Media Contact: Maureen Manier, mmanier@purdue.edu

source: Jingjing Liang, jjliang@purdue.edu; Nancy Harris, nancy.harris@wri.org

About Land & Carbon Lab:

The Land & Carbon Lab (LCL) is the World Resources Lab’s premier hub for geospatial data, analysis and monitoring of the world’s lands and natural ecosystems. Data and monitoring solutions, including the Global Forest Watch platform, exist to accelerate the implementation and financing of nature-based solutions to climate change worldwide. LCL is committed to (1) innovation in open geospatial data for land and carbon monitoring, (2) spatial intelligence to support nature-based solutions policy and target tracking, and (3) tailor-made to help businesses, governments, civil society organizations and communities. tools. Communities make data-driven decisions. LCL works across WRI and its network of external partners to transform large-scale geospatial data into actions and impacts. www.landcarbonlab.org, www.globalforestwatch.org, www.wri.org

Agricultural Communications: 765-494-8415;

Maureen Manier, Department Manager, mmanier@purdue.edu

agricultural news page

Leave a Comment