Evaluation of a Low-Cost UAS and Phenocams for Measuring Grapevine Greenness

Unpersoned aerial systems (UAS) could provide winegrowers with the potential to monitor vineyard productivity with ultra-high-resolution imagery and low operational costs. This ability could prove particularly valuable in the challenging cool-climate viticultural areas of Appalachia. Especially in this mountainous region of increasingly variable microclimates, there could be of great value from an ability to use UAS-measured greenness to monitor wine grape phenology and predict harvest quality and quantity. In this study, I assess how UAS-measured greenness relates to three complementary measures of field-based: leaf angle measurements, phenocam measured greenness, and leaf spectral measurements of greenness. After correlating these field-based measures of greenness to UAS-measured greenness, I evaluate whether UAS-sensed greenness can predict spatial and temporal patterns in seasonal wine grape phenology. I collected imagery every other week between June and September 2020 from a DJI Mavic 2 Pro UAS platform, focusing on three blocks of Marechal Foch varietal grapes at the Christian W. Klay Winery in Chalkhill, Pennsylvania. I also collect weekly leaf angle and greenness measurements from consumer-grade phenocams. From these data, I employed multivariate regression to assess the relationship of UAS-measured greenness to the three field-based measures of greenness. Based on these tests, I concluded that UAS-measured greenness is not highly correlated across space and time with field-based measures of greenness. I hypothesize this might be due to highly vertical leaf angles present on the grapes vines limiting the amount of assessable green area from the perspective of the UAS