Satellite monitoring of quantity and quality of available biomass in pastoral livestock systems

  Three key tools were developed to advance satellite-based grassland monitoring: a field survey protocol, a mobile application to record field observations (biomass and forage quality), and a web platform to visualize collected data and assist with sampling activities. These tools enabled the consolidation of a monitoring network with the active participation of over 79 researchers and technicians.
  Currently, the database includes more than 1300 matched observations combining field data and satellite imagery, covering different forage resources. These data are being used to calibrate models for predicting forage quantity and quality. Using a subset of 755 observations from various regions of Argentina, the performance of machine learning models (Random Forest) was evaluated for biomass prediction using spectral indices derived from Sentinel-1 and Sentinel-2 satellite imagery. The general model yielded an average prediction error of 1100 kg DM/ha (99%). In a subset focused on rotationally grazed pastures in Balcarce, the error was reduced to 544 kg DM/ha (42%). These results indicate that reducing environmental heterogeneity significantly improves estimation accuracy, and that expanding the database is essential to strengthen the models and enhance their applicability across different production contexts.
  With this in mind, the project has designed and is currently testing two simplification methodologies (spatial and temporal) that rapidly link field observations with satellite data, enabling system-scale biomass estimation. These tools are being implemented and validated in real-world grazing systems, while also contributing to the expansion of the database to support the development of more accurate predictive models. This represents a step forward in delivering practical solutions for grazing management and promotes the adoption of satellite technology in the agricultural sector. To date, dissemination activities (videos, workshops, talks) have reached over 1500 people.