Wavelet decomposition in the Earth's gravity field investigation

Abstract

This paper presents the results of the application of wavelet decomposition to processing data from the GGP sites (The Global Geodynamics Project). The GGP is an international project within which the Earth's gravity field changes are recorded with high accuracy at a number of stations worldwide using superconducting gravimeters. Data with a 5-second sampling interval from Wettzell and Bad Homburg were used for the research. The wavelet transform enables the investigation of the temporal changes of the oscillation amplitudes or the decomposition of the time series for the analysis of the required frequencies. The wavelet decomposition was performed using the regular orthogonal symmetric Meyer wavelet. The research concerned data from an earthquake period recorded at various locations and a quiet period when the gravimeters worked without any disturbances. The decomposition was followed by the Fast Fourier Transform for signal frequency components and then by correlation analyses of corresponding frequency components (for periods from 10 to 60 000 seconds) for all sensor combinations, for the quiet and the earthquake periods separately. Frequency components defining long term changes for all sensor combinations, as well as combinations between two sensors at the same site for the quiet days are characterised by high correlation coefficients. For the time of the earthquake, the Wettzell site data proved strong correlation for all frequency components, while the Bad Homburg site data showed an unexpected decrease of correlation for the majority of frequency components. The authors also showed that wavelet decomposition can be a good method of data interpolation, especially from the time of earthquakes. Moreover, it is a very useful tool for filtering the data and removing the noises.

On the noise characteristics of time series recorded with nearby located GPS receivers and superconducting gravity meters

“…The non-linear trends in GPS (Bogusz 2015) and SG data (Van Camp and Francis 2006;Bogusz et al 2013) have already been recognized. The non-linearity for GPS stations may come from dynamics on plate boundaries as well as the system artefacts.…”

Audio waves and its loss of energy in puncture needles

“…This selection has as main objective to only take into account the signal segment corresponding to the needle response, avoiding influences of the impact per se or noise generated by the servomotor. Then, the influence of gravity was attenuated by trend removing using a multilevel one-dimensional discrete wavelet transformation with a Daubechies mother wavelet [12], (see Figure 3b). The resulting signal was then denoised using the same aforementioned wavelet transformation but with a higher frequency scale (see Figure 3c).…”

Gravity field modelling and gravimetry

“…It has been proven that wavelet decomposition is a suitable method for data interpolation, especially from the time of earthquakes. Moreover, it is a very useful tool for filtering the data and removing noise (Bogusz et al, 2013).…”