To improve the accuracy of products derived from shared satellite observations of Earth (see M. A. Wulder and N. C. Coops Nature 513, 30–31; 2014), governments and research institutes also need to share calibration and validation data. Such data are measured on the ground or interpreted from high-resolution satellite imagery.

Satellite images are now available at very high resolution. Crowd-sourced calibration and validation data would vastly improve classification algorithms as well as the accuracy of land-cover products.

Enhanced accuracy would enable remote sensing to be better used for monitoring biodiversity loss and ecosystem dynamics, for example, and for other applications that depend on baseline and changing land cover.

Efforts such as the Global Observation of Forest Cover and Land Dynamics and Geo-Wiki.org are working in this direction. More initiatives are needed to unlock the incredible amount of data that remain confined within institutes and agencies.