Processing ambient noise using wavelet analysis tools: the Iraqi Seismological broadband Network Data at Iraqi Meteorological Organization and seismology (IMOS)

Abstract

The main objective of this research is to reduce the expected error between the original signal polluted by the implicit and estimated noise. The dominance of some levels of the surrounding earthquakes in general on the seismic energy resulting from human activities (i.e. “cultural") which some generators, cars and machines emit at rather close distances, directly affected the records recently recorded in the Iraqi Seismic Network at the Iraqi Meteorological Organization and seismology. Various seismic signal-processing techniques have been applied to obtain high-quality records. Wavelet analysis techniques have proven effective in increasing the accuracy of seismic records and greatly improving the signal-to-noise ratio. More than one method was applied to the local seismic group recorded by the Iraqi Seismic Network with different characteristics, and although there were minimal changes in the waveform of interest, encouraging results were obtained. The outcomes seem that ways compared with other denoising methods, using the Continuous Wavelet Transform has many natural translation-invariant and time-frequency properties and improve considerably the results compared to those using traditional filters. In addition, the distortion problem can be reduced using a more appropriate mother wavelet and decomposition levels. Whereas the other methods used in this field (traditional filters), Discrete Wavelet Packet Transform and 1-D Wavelet Analysis Tool can remove noise and improve seismic resolution but are less effective than Continuous Wavelet Transform.

Keywords

• Ambient noise
• continuous wavelet transform
• denoising
• IMOS
• wavelet transform