High-resolution dataset tracks global cropland water-use efficiency

Published in Scientific Data, the dataset provides annual WUE estimates for croplands worldwide at a 1-kilometer spatial resolution. It is expected to be a valuable resource for promoting sustainable agricultural water management.

WUE refers to the amount of biomass or economic value produced per unit of water used. It reflects both biological and economic benefits, balancing inputs and outputs. Cropland WUE serves as a key indicator of the trade-off between food production and water consumption, playing an important role in guiding the sustainable use of regional agricultural water resources.

"Agriculture accounts for over 90% of the world's consumptive freshwater use," said Dr. Jiang Min, the first author of the study. "With increasing concerns about global water scarcity and food security, accurately assessing cropland WUE is more important than ever."

To build the dataset, the research team developed an integrated modeling framework that combines an improved light-use efficiency (LUE) model for estimating gross primary production (GPP) and net primary production (NPP) with the ETMonitor evapotranspiration model for estimating water consumption. This framework is driven by consistent forcing data and accounts for the coupled relationship between carbon and water exchange processes.

To improve the accuracy of WUE estimation, model parameters were optimized for each climate zone. The resulting cropland WUE product demonstrated strong performance, showing a high correlation efficiency (0.76) and a low root mean square error (0.5 g C/kg H2O/yr) compared with ground measurements at flux towers. Compared to existing satellite-derived WUE products, such as MODIS and GLASS, this new dataset exhibits superior accuracy and reliability.

In 2023, the dataset was presented at the 77th session of the United Nations General Assembly as one of seven Global Water Resources Data Products aimed at supporting the assessment of progress towards the UN Sustainable Development Goals (SDGs).

The dataset is publicly available on the CASEarth Data Sharing and Service Portal.

More information: Min Jiang et al, A 20-year dataset (2001–2020) of global cropland water-use efficiency at 1-km grid resolution, Scientific Data (2025). DOI: 10.1038/s41597-025-04904-1

Journal information: Scientific Data

Story Source: Chinese Academy of Sciences