Empirical-Statistical Methodology and Methods for Modeling and Forecasting of Climate Variability of Different Temporal Scales

Main problem of modern climatology is to assess the present as well as future climate change. For this aim two approaches are used: physic-mathematic modeling on the basis of GCMs and palaeoclimatic analogues. The third approach is based on the empirical-statistical methodology and is developed in this paper. This approach allows to decide two main problems: to give a real assessment of climate changes by observed data for climate monitoring and extrapolation of obtained climate tendencies to the nearest future (10–15 years) and give the empirical basis for further development of physic-mathematical models. The basic theory and methodology of empirical-statistic approach have been developed as well as a common model for description of space-time climate variations taking into account the processes of different time scales. The way of decreasing of the present and future uncertainty is suggested as the extraction of long-term climate changes components in the particular time series and spatial generalization of the same climate tendencies in the obtained homogeneous regions. Algorithm and methods for realization of empirical-statistic methodology have been developed along with methods for generalization of intraannual fluctuations, methods for extraction of homogeneous components of different time scales (interannual, decadal, century), methodology and methods for spatial generalization and modeling, methods for extrapolation on the basis of two main kinds of time models: stochastic and deterministic-stochastic. Some applications of developed methodology and methods are given for the longest time series of temperature and precipitation over the world and for spatial generalization over the European area.