Ali Gamal Hafez

Clear P-wave arrival of weak events and automatic onset determination using wavelet filter banks

P-wave arrivals of many weak events cannot be precisely determined manually. Difference in power levels between noise and P-wave in wavelet detail of weak events enables us to determine P-wave arrival manually. Because of this power difference, automatic onset detection and picking algorithm is introduced using the same wavelet detail. Parameter settings are not needed as algorithm will work on data generated by either short or very broad band seismometers. Application of the proposed algorithm on data of three stations of Egyptian National Seismic Network (ENSN) in Cairo region shows a maximum standard deviation of 0.14 seconds of the corresponding manual picks.

Earthquake onset detection using spectro-ratio on multi-threshold time--frequency sub-band

Automatic onset detection and picking algorithm has been proposed by applying the spectro-ratio on time-frequency sub-band. The proposed algorithm does not need any parameter settings as it will work on data generated by either short or very broad band seismometers. Our algorithm is applied on local events from Cairo region recorded by three stations of the Egyptian National Seismic Network (ENSN). Maximum standard deviation is observed to be 0.113 s of the corresponding manual picks made by analysts.

Seismic noise study for accurate P-wave arrival detection via MODWT

The arrival timing of the onset of microseisms and weak events is difficult to be picked even manually. The proposed algorithm uses the maximal-overlap discrete wavelet transform (MODWT) to perform manual detection for such weak events. A seismic noise analysis was done to choose the best criteria for showing clear P-wave arrival. This algorithm is also used as an accurate automatic P-wave picking algorithm. The noise level at a seismic station does not affect the proposed picking algorithm because it adapts itself to the noise level in front of each earthquake. Local events recorded by the Egyptian National Seismic Network (ENSN) were used to test this proposed algorithm. The overall average error was found to be 0.02s.

Effect of SC on frequency content of geomagnetic data using DWT application: SC automatic detection

In this paper, a study is made to determine the effect of sudden commencement (SC) on the power spectrum of geomagnetic data using multiresolution analysis (MRA) of the discrete wavelet transform (DWT). The results of this study provides a guide to develop a new technique to automatically detect the SC because it could be an indicator of the onset of a geomagnetic storm. This new technique divides the original time series into different frequency sub-bands using the MRA of the DWT. Then it detects the change in a certain sub-band which shows a large change due to the SC. The geomagnetic records used in this study were 3-s resolution data collected from the Circum-Pan Pacific Magnetometer Network (CPMN). Using such high-resolution data enables us to minimize the detection error and the processing time to make a decision. The proposed algorithm is tested on one sample every three seconds of data sets collected from the CPMN. The maximum standard deviation of the algorithm detection times is observed to be fifty four seconds of the corresponding arrival times as determined by the National Geophysical Data Center (NGDC).

Geomagnetic Sudden Commencement Automatic Detection via MODWT

It is of great importance to develop an algorithm that autodetects sudden commencement (SC) because it could be an indicator of the onset of the geomagnetic storm. A geomagnetic storm is considered as one of the global processes in the ionosphere-thermosphere-magnetosphere system. Automatic detection of SCs is based on multiresolution analysis of a maximal overlap discrete wavelet transform using a Haar wavelet filter. The proposed algorithm is tested on 1-min-resolution data sets collected from magnetic observatories belonging to the International Real-Time Magnetic Observatory Network. Maximum standard deviation of algorithm detection times is observed to be 1 min of the corresponding arrival times published by the National Geophysical Data Center.

A Wavelet Spectral Analysis Technique for Automatic Detection of Geomagnetic Sudden Commencements

Maximal overlap discrete wavelet transform is used to perform spectral analysis of geomagnetic storm sudden commencements (SCs) (SSCs). This spectral analysis guided us in the development of an automatic SSC detection algorithm. The SC can be an indicator of the onset of a geomagnetic storm; in this case, it is called an SSC. The geomagnetic records used in this study were 3-s resolution data collected from the Circum-Pan Pacific Magnetometer Network. Using such high-resolution data enabled us to achieve a small detection error and short processing time. In addition to these technical merits, we introduce a new algorithm that automatically detects, for the first time, the SC from high-resolution data (sampled at the rate of 1 sample/3 s), unlike previous studies that focused on determining the SSC times automatically using 1-min data. Ninety-three geomagnetic storms were considered for testing the proposed algorithm; it was found that the average and maximum standard deviation of the errors in the detection times determined by the algorithm were 7 and 18 samples, respectively, of the corresponding manually determined arrival times.

Un-decimated discrete wavelet transform based algorithm for extraction of geomagnetic storm sudden commencement onset of high resolution records

The automatic detection of the onset time of the geomagnetic storm sudden commencement (SSC) is of great importance for many applications. The distribution of the power along the frequency axis during the SSC was investigated. This analysis guide us to build an SSC automatic detector, for the first time, of one sample per second data based on the un-decimated discrete wavelet transform (DWT), unlike previous studies that focused on determining the SSC times using one-minute resolution data. Using such high-resolution data enabled us to achieve a small detection error and short processing time. One hundred thirty four geomagnetic storms were considered for testing the proposed algorithm; it was found that the average and maximum standard deviation of the errors in the detection times determined by the algorithm were 35 and 44s, respectively, of the corresponding manually determined onset times. The proposed algorithm tested by using continuous period data (six months). The results show the capability of the algorithm to detect the SSCs successfully with low rate of false detections.

Accurate P-wave arrival detection via MODWT

P-wave arrivals of many weak events cannot be precisely determined manually. We use Maximal overlap discrete wavelet transform (MODWT) for manually determining clear P-wave arrivals of weak events. We introduce an automatic onset detection and picking algorithm that uses the multi resolution analysis (MRA) components of this MODWT. Parameter settings are not needed, as the algorithm will work on data generated by either short or very broad band seismometers. The application of the proposed algorithm on data of three stations extracted from the Egyptian National Seismic Network (ENSN) in Cairo region shows a maximum standard deviation of 0.17 seconds for the corresponding manual picks.

Detection of precursory signals in front of impulsive P-waves

The precursory signals, which are preceding the impulsive P-waves, are analyzed in this study. These signals cause an error in automatic P-wave detectors which motivated us to propose an algorithm to identify these signals automatically. The first detail of multi-resolution analysis (MRA) of discrete wavelet transform (DWT) is used to detect these precursory signals. The proposed algorithm is built based on the parameters of the second order autoregressive (AR) model. 34 impulsive P-wave segments with precursory signals are used to test the ability of the proposed algorithm to detect these signals. Length of each precursory signal is calculated manually and using the proposed algorithm, where the difference between the two calculations is the error of the automatic detection. The proposed algorithm used both Haar and Daubechies wavelet filters of order 2 (abbreviated as Dau(2)) for comparison. It is found that the Dau(2) wavelet filter gave better results than Haar wavelet filter as will be discussed later. The mean and standard deviation of the error, when using Dau(2) wavelet, are -1 and 4 samples, respectively.