Applications of remote sensing on vineyard studies: A review on two recent papers

Abstract

In two recent papers Ducati et al. (2014a,b) reported results on the application of Remote Sensing techniques to vineyard studies. Reflectance data were extracted from satellite images of vineyards in two regions in France : Loire Valley and Burgundy. A Factorial Discriminant Analysis was applied on both data sets. In Loire Valley, conventional vineyards were well separated from organic and biodynamic vineyards ; in Burgundy’s Côte d’Or, categories of quality of vineyards as Grand Cru, Premier Cru and Communales were also separated. Since the set of techniques used has been seldom reported on viticultural studies, we presently make a review of the whole problem. The basics of Remote Sensing applied to land studies is introduced ; applications focused on vineyards and precision viticulture are reviewed ; possible problems arising from pixel mixture of vine and non-vine classes are discussed, as well as shadow contributions. The physico-chemical causes of the reported separability are also discussed. Chemicals can be the reason of the observed significant reflectance differences between conventional and organic vineyards. On the other hand, discrimination between quality categories in Côte d’Or can be due to specific soil properties, as capacity of water storage and mineral composition. In Burgundy’s case, soil factors are apparently transmitted to vine canopy, as reflectance in summer images comes mostly from vine leaves. Future developments of these techniques and potential limitations and perspectives are discussed.