Predicting Time Lag between Primary and Secondary Waves for Earthquakes Using Artificial Neural Network (ANN)

ABSTRACT: This thesis work investigates the possibility of predicting the arrival time of the secondary seismic earthquake waves. Seismic waves are low frequency acoustic waves that are experienced prior to earthquake, they are basically two types: the primary wave (p-wave) and the destructive secondary wave (s-wave). These waves‟ approaches a destination at different times, with the p-wave experienced earlier since it travels at higher speed as compared to the s-wave. Knowledge of the time lag between this two waves recorded by a seismometer from previous earthquakes were used together with other parameters suspected to also influence the arrival of the secondary (destructive) earthquake waves which are; the magnitude from the propagating wave, the epicenter distance from the hypocenter, the seismic station‟s distance from the epicenter and the direction (in azimuths), were used for this prediction. The prediction model was carried out using neural network on MATLAB; the artificial neural network (ANN) design makes it possible to develop the correlation between the various parameters for the study. First, the network was trained with earthquake data of magnitude 6.0-7.0 Richter, validation and testing was carried out to measure the performance of the model. The result gave satisfactory performance, with regression values greater than 0.9, and the root mean square error (RMSE) computed were of the range of 0.1003 to 0.1148 for the most satisfactory network architecture. Secondly, the trained network was also tested with external values of magnitude range outside the values the network was earlier trained with. This network gave results that were not as good as the first case, so it was concluded that it‟s better to train the network with data from earthquake of all magnitude range. In general, from the experiment we concluded that the design and parameters considered is possible for predicting the time-lag of these two seismic waveforms using artificial neural networks. Keywords: Earthquake, Seismic waves, P-wave, S-wave, Seismometer, Atificial Neural Network, Hypocenter, Epicenter, Magnitude. …………………………………………………………………………………………………………………………