Satellite-based estimation of herbage mass: comparison with destructive measurements and UAV model’s estimation

Regular estimation of herbage mass (HM) is a prerequisite for efficient pasture management. In addition to classical estimation using rising plate meters, remote-sensing methods using unmanned aerial vehicles (UAV) or satellites are available. Pasture.io has developed a model that estimates HM based on daily satellite data, herbage growth models and herbage-ingested input data recorded by farmers combined with artificial intelligence. This study compared the accuracy of Pasture.io HM estimations with UAV estimations and destructive measurements. Pastures from three Swiss farms were assessed regularly in May, June and July 2021. It was found that Pasture.io estimates HM with an error value RMSE 342 kg dry matter (DM) ha-1 while the UAV model’s estimation showed a higher RMSE of 447 kg DM ha-1. The results suggest that even in small pasture structures (mean paddock size: 1.2 ha), it is possible to estimate HM with reasonable accuracy based on satellite data and artificial intelligence