Temporal climatic variabilities of global atmospheric, oceanic, and land surface parameters



Changes in the states of two groups of major earth climatic subsystems are investigated: (a) "thick" subsystems (atmosphere, and ocean), and (b), "thin" subsystems (vegetation, snow, and sea ice). These two groups are distinguished with respect to temporal variations of their spatial extent which is dynamic for the "thin" subsystems and stable for "thick" subsystems. Stochastic models governing monthly and longer-term anomalies of remotely sensed and in situ, globally averaged parameters of these layers (tropospheric temperature, sea surface temperature vegetation, snow and sea ice cover areas) are evaluated according to a new modification of the maximum entropy method with frequency truncation of normalized spectra. The analysis shows that global temporal variations of global thick subsystems are governed by the Bernoulli-Wiener type stochastic processes, while variabilities of the global "thin" subsystems are described by the first order Markov processes with intermediate values of coefficients. It is hypothesized that stochastic variations in global sea ice, snow, and vegetation covers depend on only a small number of independent regional variations and therefore still show the Markov characteristics of the latter. On the basis of the analysis of observational data and stochastic modeling, the authors propose a concept of mechanisms of climatic variations which takes into consideration different climatic subsystems as well as different temporal and spatial scales. The proposed hypothesis explains the stationarity of changes in the area of global "thin" covers, and the non-stationary, random walk (i.e. without deterministic trends) character of global temperature changes.

Full Text: