Spring frosts after budburst are responsible for crop losses and threaten local economies. As global warming tends to advance the phenological stages of plants, these become more and more subject to facing a long period of freezing temperatures. For a given meteorologic situation, the topography of the studied site and the nature of the soil of the plots constituting it, a temperature spatial variability expands within the same territory. Considering a radiative frost, the temperature can thus differ by several degrees, creating areas of cold and warm air that it is necessary to know to fight wisely against the frost. Nowadays, many solutions exist to fight against frost, including the wind machines that dot the Quincy vineyard in France. Weather variables, topographic parameters, and daily minimum temperatures from a network of connected sensors scattered throughout the vineyard are retrieved for the last three spring seasons of 2020,2021, and 2022. Then, thanks to a hierarchical clustering algorithm, it is possible to link the spatial variability of temperatures to the synoptic situation and the topography of the domain. The outcome is the assessment of the frost risk areas to propose a judicious implantation of wind machines in the vineyard of Quincy.