Extraction of a dynamic component from variations in geophysical fields using the convergence of a sample average

Observations of a complex of geophysical fields sensitive to deformation processes in the Earth’s crust point to the presence of components of different origins in their time variations. The variations contain not only regular components (seasonal, tidal, anthropogenic, and others) but also random and chaotic components. The chaotic component can be related to variations in the stress-strain state, as was shown in Smirnov et al., 2005]. Some main properties of the behavior of the chaotic component can be determined on the basis of ideas of nonlinear dynamics. A method of extracting such a “dynamic” component in the process studied requires that its inherent properties be taken into account. This paper develops a method for separating components of a dynamic system from the superimposed component of an external process using invariant properties of the system. The applicability and effectiveness of the method is demonstrated by comparing model calculations with the use of a smoothing moving window of a variable width and real geophysical data.