Studying phenological stages of cherry (Prunus avium L.) using field observations and satellite-derived vegetation indexes

ABSTRACT In the framework of a continuous search for technological innovations that allow fruit growers to practice a more competitive and sustainable fruit production, satellite-derived vegetation indexes could be used as tools to support the decision-making processes of technicians and territory planners. The objectives of the present study were to characterize phenological stages of cherry trees (Prunus avium L.) in the Maule region of Chile using field observations and to evaluate the capacity of multispectral satellite imagery (Landsat ETM+) and the derived vegetative indexes (NDVI, SAVI), as tools for remotely identifying these phenological stages in cherry trees. Phenological transition stages in cherry trees, estimated from Landsat 7 (ETM+) data did not agree with in situ measurements of phenology made at the different sites. Landsat 7 (ETM+) derived vegetation indices (NDVI, SAVI) could discriminate only gross changes associated with an increase and then a decrease of canopy greenness in cherry trees, but could not discriminate between successive phenological states in which the proportion of foliage is only slightly different. The number of days after the first of July (Southern Hemisphere) tended to be a less variable indicator than growing degree-days (GDD) for predicting the occurrence of different phenological stages in cherry trees.