Climate spatial variability knowledge is essential in agronomy and forestry, in order to characterize production potential or to assess pest development risks. In viticulture, climate mainly governs grapevine development rate and berry ripening. The aim of the present work is to characterise climate spatial variability at mesoscale level, using several spatialization techniques at daily time step, applied to climate variables and agroclimatic indices, in order to evaluate its consequences on grapevine development and grape ripening. This study was led in the Bordeaux winegrowing region. Six variables were studied: minimum and maximum temperatures, solar radiation, reference evapotranspiration, rainfall and soil water balance. For each variable, the method providing the best results (i.e. the lowest errors) with reasonable computational and time costs was selected. The propagation of spatialization errors within agroclimatic models was assessed. The error propagation is considerable within degree-days and soil water balance models. Maps of climate variables and agroclimatic indices were used to analyse the spatial characteristics of climate within the Bordeaux winegrowing area and to quantify differences that were known only from indirect and empirical observations. This study provides useful information to enhance the choice of viticultural techniques and plant material in relation to the local environment.
