From fields to forests: tracking cropland abandonment from space

A sweeping new review reveals how satellites are helping scientists track a quiet but widespread shift in global agriculture: the abandonment of cropland. By analyzing decades of data and hundreds of studies, researchers have mapped out where farmland is being left idle, why it’s happening, and what it means for ecosystems and food security. Their findings offer a powerful new framework for understanding land-use change in a warming, urbanizing world.

Around the world, once-productive farmland is falling silent. Cropland abandonment—when farmers stop cultivating their land—is a growing but often invisible trend. Whether driven by economic hardship, shifting policies, or environmental degradation, this phenomenon carries huge implications: food insecurity, carbon dynamics, biodiversity, and even wildfire risk. Yet global understanding remains patchy. Satellite-based remote sensing offers the unique ability to track these shifts over time and space, providing crucial data on where, when, and why land is being abandoned. Due to these unresolved challenges, it is vital to conduct a systematic, global-scale review of remote sensing methods for cropland abandonment monitoring.

In a new study (DOI: 10.34133/remotesensing.0584) published on May 23, 2025, in Journal of Remote Sensing, a team of scientists from Tsinghua University, the Chinese Academy of Sciences, and Cornell University unveiled the most comprehensive review to date on monitoring cropland abandonment using satellite technology. By synthesizing findings from over 250 studies, the researchers examined how satellites are detecting abandonment patterns, what the underlying drivers are, and how abandoned land reshapes ecosystems. The review introduces a novel “cause–pattern–effect” framework that aims to guide future land-use policies and ecological restoration strategies.

The study shows that cropland abandonment is not random—it follows patterns, often unfolding in mountainous areas, post-conflict regions, or places with poor infrastructure. By tracking vegetation growth trends over time, likes using time series NDVI, satellites can now distinguish abandoned fields from active ones with over 89% accuracy. The review reveals that beyond visible signs of disuse, abandonment reflects deeper stories—of migration, poverty, changing diets, and climate stress. The team’s new “cause–pattern–effect” framework links satellite-detected land changes with socioeconomic and ecological data, offering policymakers a dynamic lens to view abandonment not just as a loss, but sometimes as a recovery of nature. This approach could transform how we manage land globally.

The authors reviewed global studies spanning Europe, Asia, Africa, and Latin America. Eastern Europe showed some of the highest abandonment rates—over 27% in certain post-socialist regions—while China’s southwest mountains and arid zones are modern hotspots. Remote sensing tools used include Landsat for historical tracking, Moderate Resolution Imaging Spectroradiometer (MODIS) for large-scale pattern recognition, and high-resolution Gaofen satellites (GF) and Satellite Pour l’Observation de la Terre (SPOT) imagery for detecting small, fragmented plots. Machine learning algorithms such as Support Vector Machine (SVM) and random forests enhanced detection precision, while Light Detection and Ranging (LiDAR) added vegetation structure insights. Beyond mapping, the study also explores the ecological ripple effects of Cropland Abandonment (CA): abandoned fields can absorb carbon, support biodiversity, or, conversely, fuel wildfires and harbor pests. One striking insight is how different definitions of abandonment—seasonal, passive, or active—change what satellites see. This complicates international comparisons and calls for better standardization in monitoring practices.

 “Abandonment is more than empty land—it’s a signal of deeper transitions in how we live, farm, and interact with the environment,” said Dr. Le Yu, corresponding author of the study. “By pairing satellite data with ecological and social insights, we’re beginning to understand this hidden landscape. It’s both a warning sign and an opportunity.”

The researchers performed a meta-review of 254 publications, categorized into mapping techniques, abandonment drivers, and ecological impacts. They compared the accuracy and applicability of eight major remote sensing methods, from vegetation phenology tracking to deep-learning image analysis. Data sources included multi-sensor satellite imagery (Landsat, MODIS, Sentinel, GF, LiDAR) and cloud platforms like Google Earth Engine. The review also compiled 50 abandonment drivers and scored their relevance to remote sensing evidence.

As farmland continues to vanish in parts of the world and rewild in others, understanding cropland abandonment will be critical to future planning. The authors call for more integrated monitoring that combines satellites, field surveys, and socioeconomic data. Their “cause–pattern–effect” framework could be applied to model food security risks, carbon recovery, or reforestation potential. With Asia and Africa poised to see rising abandonment, timely action could turn silent losses into sustainable opportunities for both people and the planet.

###

References

DOI

10.34133/remotesensing.0584

Original Source URL

https://doi.org/10.34133/remotesensing.0584

Funding information

This research was supported by the National Key R&D Program of China (grant number: 2022YFE0195900).

About Journal of Remote Sensing

The Journal of Remote Sensing, an online-only Open Access journal published in association with AIR-CAS, promotes the theory, science, and technology of remote sensing, as well as interdisciplinary research within earth and information science.