Investigation of grapevine areas under climatic stress using high- resolution atmospheric modelling: case studies in South Africa and New Zealand Valérie Bonnardot 1 , Andrew Sturman 2 , Iman Soltanzadeh 2 , Peyman Zawar-Reza 2 , Jacobus J. Hunter 3 , Hervé Quénol 1 1 COSTEL Laboratory, UMR 6554 LETG of CNRS, University of Rennes 2, France 2 Centre for Atmospheric Research, University of Canterbury, New Zealand 3 Institute for Fruit, Vine and Wine of the Agricultural Research Council, Stellenbosch and Stellenbosch University, Department of Viticulture and Oenology, Private Bag XI, Matieland 7602, Stellenbosch, South Africa Email:valerie.bonnardot@uhb.fr Abstract High-resolution atmospheric simulations (500 m) were used to assess viticultural areas under climatic stress in South Africa and New Zealand. The potential areas in which high daytime temperature stress was likely to affect grapevine photosynthesis and grape composition were identified. Results indicated different diurnal temperature variations within the two areas due to synoptic and local environmental factors, often associated with the influence of terrain. 1. Introduction Identification and characterisation of viticultural environments are of importance for the expanding wine industries of ‘New World’ countries such as South Africa (Carey et al., 2008) and New Zealand (Tesic et al., 2002). Among environmental factors contributing to successful vineyard areas, climate variables, especially temperature, have an important effect on grapevine growth and on the development of grape aromas and thus wine quality (Coombe, 1987). In some regions of the world, excessive temperatures can create significant stress for the vine, resulting in sub-optimal conditions for wine production (White et al., 2006). Climate monitoring networks are frequently of insufficient spatial resolution to provide a clear picture of the temperature patterns in regions of complex terrain, so that it is not possible to identify sections of a vineyard region that are more susceptible to temperature increases that may result from possible global warming. Spatial mapping of temperature by means of mesoscale atmospheric modelling has therefore been performed over several wine-producing regions. The Regional Atmospheric Modelling System (RAMS) has been used to investigate sea breeze circulations over the vineyards of the South Western Cape in South Africa (Bonnardot et al., 2005). These numerical simulations showed that, for local circulations forced by topography and surface contrasts, the use of a high horizontal resolution (< 1 km) was of great value in characterising the climate potential of viticultural environments (Bonnardot and Cautenet, 2009). Increasing resolution is therefore considered necessary to properly identify and characterize the climate of wine- producing regions and to determine their vulnerability. The CCAM (Cubic Conformal Atmospheric Model) has been used to produce a 200 m climatology over the Stellenbosch district (Roux, 2009), while the Advanced Research Weather Research and Forecasting (ARW-WRF) model has been used to investigate climate variability over the Burgundy wine region (Bonnefoy et al., 2010). Similar modelling of wine- producing areas has been conducted in Australia (Lyons and Considine, 2007). This paper shows how advanced high-resolution (500 m) three-dimensional atmospheric