In precision agriculture, most spatio-temporal analyses involve the analysis of time series of remotely-sensed images whose dates are expressed according to the Gregorian calendar. However, in the agronomic literature, it is commonly accepted that calendars based on a thermal index allow a better representation of the actual state of development of the plant. This paper aims to demonstrate that potential new information at the block scale can be revealed by adjusting a NDVI time series extracted from Sentinel 2 images using a site-specific thermal index. This is achieved by removing part of the inter-annual differences in the time-series that is due to variations in phenology by using a temperature-based time adjustment. To do so, the paper is based on an operational case study in viticulture exploring whether a link between NDVI profiles and the consistency over years of harvest quality parameters could be hypothesised at the block scale.