Results are discussed of comparative analysis of long-term time series of seasonal mean and annual mean surface temperatures in Siberia. To analyze seasonal temperature variations, we employ seasonal mean temperature values, that allows us to eliminate influence of higher-frequency variations (cyclonic and diurnal ones). Empirical relation between amplitude of seasonal temperature variations and seasonal mean surface temperature was revealed for two climatic zones (Irkutsk and Tobol'sk). For both zones amplitude of seasonal temperature variations is linearly related to seasonal mean temperature in winter time (coefficient of determination is more than 0.9). Modeling and forecast of climate changes on the basis of empirical relations is shown to be promissing.
linear trends; year-average and season-average ground temperatures; many-year time series; oscillation amplitudes; correlations; warming; climate