Considering climate change and assessing its impacts is challenging due to dealing with large data sets and uncertainties. This paper discusses an approach for the impact assessment of climate change based on synthesizing weather data sets out of several climate scenarios, in a way to generalize the assessment despite of the existence of climate uncertainties. The is based on creating one-year weather data, representing typical, extreme-warm and -cold conditions for 30-year periods, which results in decreasing the length of simulations enormously. The usefulness and accuracy of the results are discussed for energy and moisture simulations in buildings.