UAS remote sensing missions for rangeland applications

Abstract

Rangelands cover about 50% of the earth's land surface, are in remote areas and have low population densities, all of which provide an ideal opportunity for remote sensing applications from unmanned aircraft systems (UAS). In this article, we describe a proven workflow for UAS-based remote sensing, and discuss geometric errors of image mosaics and classification accuracies at different levels of detail. We report on several UAS missions over rangelands in Idaho and New Mexico, USA, where we acquired 6–8 cm resolution aerial photography and concurrent field measurements. The geometric accuracies of the image mosaics were in the 1–2 m range, and overall classification accuracies for vegetation maps ranged from 78–92%. Despite current FAA regulations that restrict UAS operations to distances within line-of-sight of the UAS, our results show that UAS are a viable platform for obtaining very high-resolution remote sensing products for applied vegetation mapping of rangelands.