Harnessing AI for Enhanced Forest Mapping in East Asia

the cloud pillars of Zhangjiajie national park in China rise through the mist.

(Credit: Unsplash.com)

by | Sep 18, 2023

This article is included in these additional categories:

Scientists at Purdue University have embarked on a groundbreaking endeavor to create a comprehensive spatial database of planted forests in East Asia. Through a fusion of on-site data collection and satellite technology, they have harnessed artificial intelligence (AI) to produce high-resolution forest maps, with a focus on common tree species.This pioneering initiative promises to revolutionize our understanding of East Asian forests, their distribution, and sustainable management.

The Significance of East Asia’s Planted Forests

The Food and Agriculture Organization of the United Nations (FAO) estimates that East Asia holds a staggering 36% of the world’s planted forests. In comparison, Africa, Europe, and the United States account for 19%, 7%, and 9% respectively. Within East Asia, the majority of planted forests — 87% — reside in China, followed by Japan (11%), South Korea (1%), and North Korea (0.9%).

Bridging the Data Gap with AI

Prior to this endeavor, maps of East Asia’s planted forests were fragmented, covering only portions of the region, and relied on inconsistent and unverified data sources. This new initiative led by Purdue University leverages AI to analyze an extensive dataset, combining information obtained from both ground-based measurements and remote sensing.

“We used an AI approach to help us understand a massive amount of data, measured both from the ground and also from remote-sensing sources,” says Jingjing Liang, associate professor of quantitative forest ecology at Purdue. “This study complements the research portfolio of the Institute for Digital Forestry with an international perspective, enriching our understanding of global forest ecosystems and their sustainable management.”

Collaborative Efforts Yield Breakthrough

The project’s success is a testament to collaborative efforts, with lead author Akane Abbasi, a PhD student in forestry and natural resources, and 15 co-authors working in tandem. These researchers are part of Science-i, a web-based platform uniting over 300 scientists worldwide, and the Global Forest Biodiversity Initiative (GFBI), which has amassed a database of 1.3 million sample plots and 55 million trees.

Nancy Harris, research director of Land and Carbon Lab and Global Forest Watch at World Resources Institute, emphasizes the practical significance of this collaboration, stating, “Our partnership with Purdue and FAO embodies our mission to deploy breakthroughs in geospatial monitoring that power solutions for sustainable landscapes.”

Unifying Disparate Data Sources

One of the significant challenges faced by the team was integrating data from various sources that differed in format, type, and spatial extent. Some datasets covered only specific countries, while others encompassed broader regions.

Abbasi elaborates: “Some data cover only China. Some data cover only Japan. Some data cover China and South Korea. They differ in spatial extent. They also differ in relatability in terms of whether it’s measured on the ground or from space.”

The Power of Ensemble Machine Learning

To overcome these challenges, the researchers employed an ensemble machine-learning approach. This method utilizes AI to train three different machine learning models, accounting for imperfections in both the data and the model.

Liang, who is also co-director of the Forest Advanced Computing and Artificial Intelligence Lab (FACAI), underscores the role of technology in their work.

“When I say I study forest ecology, people assume that I go to the forest, measure something, then come back and analyze the data,” says Abbasi. “But this is not what we’re doing here. We deal with very large spatial extents, and we study nature using cutting-edge AI and machine learning.”

A Global Impact

This comprehensive spatial database not only reinforces the FAO’s Global Forest Resource Assessment (FRA) but also offers detailed spatial locations of forest plantations at small resolutions. Javier Gamarra of FAO commends Purdue’s use of AI to merge data from satellites and on-the-ground sources, making vast amounts of forest resource information widely available. FAO’s Forestry Division is actively collaborating with Professor Liang’s FACAI Lab and the World Resources Institute to further leverage artificial intelligence in estimating global forest growth.

Lessons from China’s Reforestation Efforts

China’s significant reforestation efforts have garnered attention. Over the past few decades, China has planted more forest than any other nation, covering an area greater than the combined size of Texas and New York. Japan and South Korea also played a pivotal role by establishing forests in response to post-war demand for forest products.

As Beijing grapples with sandstorms, it has initiated a costly and labor-intensive project of planting trees on sand dunes. Liang emphasizes the importance of learning from China’s experiences to combat deforestation globally, stating, “Learning from what China did would be one thing we can do to help the world plant more trees to stem deforestation.”

In conclusion, Purdue University’s pioneering spatial database, empowered by AI and collaborative efforts, promises to enhance our understanding of East Asia’s planted forests and their sustainable management. This milestone underscores the potential of artificial intelligence in addressing complex ecological challenges and encourages global cooperation in conserving vital forest resources.

 

Additional articles you will be interested in.

Stay Informed

Get E+E Leader Articles delivered via Newsletter right to your inbox!

This field is for validation purposes and should be left unchanged.
Share This