In the sprawling suburban landscapes of Zengcheng, Guangdong Province, China, a silent transformation is taking place. Croplands, once teeming with life, are being abandoned, and understanding this shift is crucial for effective land management and potential energy sector opportunities. A recent study led by Shanshan Feng from the Institute of Agricultural Economics and Information at the Guangdong Academy of Agricultural Sciences has developed a novel method to accurately extract and monitor these abandoned croplands using multisource remote sensing images.
The challenge of acquiring complete cloud-free images to track crop growth cycles has long plagued researchers. Feng and her team tackled this issue by proposing a method that utilizes the annual maximum of NDVI (Normalized Difference Vegetation Index) value from Gaofen-2 (GF-2) and Sentinel-2 imagery. “We wanted to enhance the accuracy of abandoned cropland extraction,” Feng explains, “and we found that using the annual maximum NDVI value provided a robust solution.”
The process involves generating a land use map using GF-2 images through object-based image analysis. Then, the annual maximum NDVI values from 2018 to 2022 are calculated based on Sentinel-2 data on the Google Earth Engine platform. The team determined that an object with an annual maximum NDVI value below 0.4 should be regarded as unplanted. This threshold allowed them to identify the spatial distribution of unplanted cropland within a year. Croplands that remained unplanted for two or more consecutive years were then extracted as abandonment.
The results were impressive, with an overall accuracy ranging from 0.80 to 0.85. This method not only provides a more accurate way to monitor abandoned croplands but also offers valuable insights for the energy sector. Abandoned croplands can be repurposed for bioenergy production, contributing to a more sustainable energy mix. “Understanding the spatial distribution and temporal dynamics of abandoned croplands is the first step towards their effective utilization,” Feng notes.
The study, published in the IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (translated as “IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing”), highlights the potential of remote sensing technology in land management. As Feng and her team continue to refine their methods, the implications for both agricultural and energy sectors are profound. This research could shape future developments in precision agriculture, land use planning, and renewable energy production, ultimately contributing to a more sustainable and efficient use of our planet’s resources.