An international research team led by the Aerospace Information Research Institute of the Chinese Academy of Sciences (AIRCAS) has developed the global forest disturbance type dataset at 30-meter resolution: the Global Forest Disturbance Type Dataset (GFD). The dataset reveals that approximately 1.247 billion hectares of forest—equating to 31% of the global forest cover—underwent disturbances between 2000 and 2020.

The study, recently published in Earth System Science Data, leverages 30-meter-resolution Landsat satellite imagery alongside machine learning and spatial analysis to identify 11 distinct types of forest disturbance. These include wildfires, logging, shifting cultivation, and urban expansion, among others.

Key findings highlight that human activities are the primary drivers of global forest disturbances. Plantation restoration accounts for the largest share of disturbed areas at 44%, followed by shifting cultivation (24%), forest fires (11%), and infrastructure development (21%).

Notably, the study also identifies signs of forest recovery: newly formed forests from afforestation and reforestation make up around 3% of the total disturbed areas, with concentrated regrowth in China, India, and Brazil.

To create the GFD dataset, the research team integrated multi-source satellite observations, land cover data, and 18 environmental indicators on the Google Earth Engine (GEE) platform. This methodology achieved an impressive classification accuracy of over 85%.

By pinpointing both the locations and causes of forest changes, the dataset fills a critical gap in global forest monitoring infrastructure. According to the researchers, it will enable more targeted forest management strategies, enhance the accuracy of carbon accounting, and strengthen global efforts to mitigate climate change.

The study was supported by the National Key Research and Development Program of China, the National Natural Science Foundation of China, among other sources.