We present a review of empirical regularities recently revealed based on statistical processing of the instrumental data compiled. Among these regularities there are: trends and trajectories of the warming observed in Siberia, interseasonal variations of surface temperature, and temporal periodicities of the warming in different Siberian cities. The possibility of applying modern mathematical climate models to description of regional climatic changes are also discussed. Among the results of mathematical modeling that encourage there are the following: retrievals of regional surface temperature using joint model of general atmospheric and ocean circulation, assessments of the regional consequences of the global warming, the account for regional hydrological processes. Combination of these two methodological approaches (the empirical and mathematical modeling ones) is recognized to be promising; also, some unresolved problems arising at such a combination are shown to be necessarily addressed.