Agricultural producers benefit when all parts of a field return optimum yield and protein content, but these are often highly variable and it is not entirely clear why. Advances in precision agriculture make it possible to monitor the factors that drive field-scale variability in yield and protein content, like vegetation greenness and carbon uptake via photosynthesis. Unmanned Aerial Vehicle (UAV) technologies and applications in remote sensing are rapidly developing, and offer agricultural producers detailed crop information at user-controlled revisit frequencies. During the 2016 growing season, we acquired UAV-based measurements of the Normalized Difference Vegetation Index (NDVI) on four dates in a winter wheat field near Sun River, Montana, providing spatial and temporal detail, in conjunction with high frequency eddy covariance flux observations of NDVI and plant carbon uptake and water use. Georeferenced winter wheat yield and protein concentration measurements were made over the entire field with a combine yield monitor and protein analyzer. We examined the relationships between UAV-based measurements of NDVI and field-derived measurements of wheat yield to better understand the predictive capabilities of the remotely sensed data. For comparison, we also examined the relationship of wheat yield to a more widely-used, coarser resolution remote sensing data source.
Results/Conclusions:
The observations from the 2016 growing season demonstrated that vegetation greenness differed across both space and time. Areas were categorized as having relatively high, medium, or low NDVI trajectories across the growing season, and these differences in NDVI were significantly associated with field-scale variability in wheat yield. These UAV-based measurements of NDVI (sub-meter scale) were more strongly related to yield than coarser satellite remote sensing observations of NDVI, demonstrating the utility of high spatial-resolution observations of NDVI, and suggesting that more frequent and better timed UAV observations could further enhance this predictive relationship. The impact of these types of detailed observations is expected to provide agricultural producers with the knowledge and tools to further develop prescriptive, variable-rate management practices. Because UAV mapping is becoming so widespread, it is essential that studies like this explore the boundaries of what is practical and necessary to yield improvements in agricultural management and sustainable production. Ongoing research targets the uncertainty around the optimal time during the growing season to measure NDVI to better understand spatial variability in yield and protein content. Can measurements be made early enough in a season to take corrective action?