SIMULTANEOUS DETERMINATION THICKNESS AND VELOCITY STRUCTURE OF LAYERED VELOCITY MODEL OF RED RIVER USING GENETIC ALGORITHM METHOD

Abstract

ABSTRACT Layered velocity model is important information for seismological study. Actually is being used the three layers model in Vietnam region but it is too simple because of owing to the lack of condition of data in the earthquake catalogues. Thus, it is necessary to build up a good-layered model for earthquake monitoring in Vietnam. Simultaneous determination of thicknesses and velocities of a layered model is a non-linear problem. Genetic algorithm is one of the methods to solve non-linear problem. In this study, we have tried to apply this method to determine a layered model for Red River. My results were obtained by separating the new 1D model into 7-8 layers. The first one presents a velocity of 4.3 km/s, instead of 5.8 km/s in the fore studies. For a deeper layer from 0.6 km to 7.6 km P wave velocity becomes larger 5.4 km/s – 5.8 km/s. Thus, I suggest that above the base rocks, there is a layer of crack rocks situating at the depth from 0.6 to 7.6 km. This ideal from new model seems agreeable with territory and tectonic of Vietnam. According to Cao et al. (2000), Nguyen et al. (2004) Conrad layer when depth changes from 17 km to 23 km and P wave velocity changes from 6.1 to 6.7 km/s. Finally when the depth of Moho surface changes from 24 to 35 km the P wave velocity also changes from 7.6 -8.1 km/s.