A new study introduces an artificial intelligence (AI)-driven remote sensing framework designed to map the potential for forage cultivation across northern China's drylands, with a particular focus on the middle reaches of the Yellow River. The study, recently published in Water Research, identifies optimal forage-growing belts at the kilometer scale, delivering data and decision-ready tools to underpin ecological protection, sustainable agricultural practices, and national feed and food security.

Led by Prof. WANG Shudong from the Aerospace Information Research Institute (AIR) of the Chinese Academy of Sciences, the study was a collaborative effort involving the Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters and the Department of Earth and Environmental Science at the University of Pennsylvania.

Northern China's drylands grapple with two critical challenges: scarce water resources and the need to secure stable supplies of both feed and food. To address these concerns, the team developed a cross-scale, multi-source fusion framework that integrates satellite observations, ecohydrological model outputs, and on-site field measurements, reducing the reliance on dense in-situ sampling.

By leveraging multi-source satellite data and mechanistic models of water balance and crop growth, the researchers generated high-quality training samples. They then applied ensemble learning and transfer learning techniques to retrieve key production factors, including irrigation water usage, vegetation net primary productivity (NPP), and soil organic carbon (SOC). Notably, the retrieval accuracy for these factors exceeded 90%. Additionally, distribution alignment and quantile mapping methods reduced regional biases by 43%, enabling the identification of optimal forage belts with a positional accuracy of over 85%.

Unlike traditional single-metric assessments, the framework frames forage planting as a spatial optimization problem, which simultaneously balances water consumption, soil carbon sequestration benefits, and forage production capacity. By quantifying ecological gains, economic returns, and water costs on a unified scale, the tool pinpoints priority planting areas and optimal input-output ratios, facilitating the efficient allocation of labor, resources, and funding.

Characterized by its replicability and cost-effectiveness, the approach offers support for ecosystem restoration and high-quality agricultural development in regions with strict water constraints, the researchers noted.