摘要 : The focus of this paper is to explain a proper way of using climate model simulations in climate change impact studies. Special attention is addressed to reducing model bias, a systematic model error present in all climate integra... 展开 The focus of this paper is to explain a proper way of using climate model simulations in climate change impact studies. Special attention is addressed to reducing model bias, a systematic model error present in all climate integrations. Studies are done using climate projection results obtained by the Coupled Regional Climate Model EBU-POM, for three periods: 1961-1990 (experiment 20c3 m), 2001-2030 (A1B scenario) and 2071-2100 (A2 scenario) for the Balkan area with special analysis of vineyard regions in Serbia. A Statistical Bias Correction method, based on daily values, is applied to the model results in order to minimize bias. Corrected values for 2 m air temperature and precipitation are used to calculate several climate indicators important for grape growing. Obtained results show that projected increase in temperature may lead to an extended growing season duration, as well as an increase in Growing Degree-Days by 1000 units, by the end of the 21st century. The grapevine dormant period is likely to be shortened and affected by warmer winters with less frost days. Changes in water supply and overheating could lead to changes in vineyard locations or changes in the selection of grapevine cultivars. Presented results show that the present climate conditions in vineyard regions could be shifted to higher altitudes (1000 m) by the end of the century. 收起