The Innovation Team of Agricultural Remote Sensing at the Institute of Agricultural Resources and Regional Planning (IARRP) of the Chinese Academy of Agricultural Sciences (CAAS) has achieved groundbreaking progress in the field of dynamic monitoring of farmland in the agro-pastoral ecotone of northern China. The findings have been successively published in prestigious international journals such as "GIScience & Remote Sensing," "Computers and Electronics in Agriculture," "International Journal of Applied Earth Observation and Geoinformation," "Scientific Data," and "Science of the Total Environment."

The agro-pastoral ecotone in northern China is influenced by factors such as poor soil quality, water scarcity, the "Grain for Green" policy, and labor force migration, leading to complex and varied changes in farmland. The distribution of fallow land, abandoned land, and reclaimed land is challenging to accurately capture, with traditional remote sensing methods often limited by insufficient accuracy or difficulties in data integration.

To address these challenges, the research team integrated multiple sources of satellite imagery and ground data, combined with advanced technologies, to overcome these obstacles. The team developed various innovative methods. For instance, by utilizing Landsat TM and OLI imagery along with Google Earth Engine, the team proposed a land use change trajectory method that accurately differentiated dispersed farmland from grasslands with similar spectra for the first time. They revealed that the farmland area expanded rapidly from 1990 to 2019, then gradually decreased, falling below the 1990 level by 2019. Through vegetation index analysis, they confirmed the continuous recovery of vegetation after returning farmland to forests.

Furthermore, by integrating Sentinel-1 and Sentinel-2 data, a dataset containing radar and spectral features was constructed, generating a distribution map of fallow land in 2020 with a resolution of 10 meters, achieving a high accuracy of 95.82%. In monitoring abandonment in Inner Mongolia, the team optimized the LandTrendr algorithm, introducing an analysis method based on ground samples and vegetation health indices, significantly improving monitoring accuracy under complex climatic conditions to reach 82.02%.

Moreover, the team developed a time-series segmentation method (TSARC), combined with key growth period data, to generate a 30-meter resolution farmland status map for Inner Mongolia from 2000 to 2022, with an annual accuracy ranging from 97% to 99%, clearly depicting the spatial trends of abandonment and reclamation. Building upon this, utilizing Sentinel-1/2 data, they produced a dataset covering the period of 2016-2023 in Inner Mongolia at a 10-meter resolution, distinguishing between active cultivation, unstable fallow, continuous abandonment, and reclamation.

Figure 1: Workflow for Monitoring Fallowing, Abandonment, and Reclamation of Farmland Based on High Temporal and Spatial Resolution Remote Sensing Images and the TSARC Method

The above-mentioned achievements of the research team provide reliable data support for the refined management of fallow land, abandoned land, and reclamation, while offering scientific evidence for the ecological effects of the "returning farmland to forests" policy and effectively revealing the profound impact of farmland changes on vegetation. Additionally, the release of the ARCC10-IM dataset provides valuable resources for land planning, environmental monitoring, and food security management in the arid and semi-arid regions of northern China.

Dr. Wuyundeji, a postdoctoral researcher at the Institute of Agricultural Resources and Regional Planning of the Chinese Academy of Agricultural Sciences, is the first author of the series of papers, with Dr. Liang Sun as the corresponding author. This research was supported by the National Key Laboratory for Efficient Utilization of Arid and Semi-Arid Farmland, the National Key R&D Program "Development and Application of Big Data Platform for Grain Production," and the Agricultural Science Talent Program of the Chinese Academy of Agricultural Sciences. Researchers from the Brazilian Soybean Research Center and the Food and Agriculture Organization of the United Nations also participated in the aforementioned studies.

Article Links:

1. https://doi.org/10.1038/s41597-025-04614-8

2. https://doi.org/10.1016/j.jag.2025.104399

3. https://doi.org/10.1016/j.compag.2024.109541

4. https://doi.org/10.1080/15481603.2022.2026638

5. https://doi.org/10.1016/j.scitotenv.2022.150286

Dataset Links:

1. https://doi.org/10.6084/m9.figshare.25687278.v4

2. https://doi.org/10.6084/m9.figshare.26360647.v1