SEISMIC HAZARD ANALYSIS - AN ARTIFICIAL NEURAL-NETWORK APPROACH

An artificial neural network (ANN) approach is applied for the estimat ion of seismic hazard in a region. The seismicity rhythm is recognized by means of an ANN approach. The seismicity cycle may be divided into four stages, viz. energy accumulation, increasing release in energy, intense release and the remnant release of seismic energy. The seismic ity data from the earthquake catalogue (1790-1990) for the Arakan Yoma and Naga Thrust belt in NE India have been used. Future seismicity fo r the region is predicted up to the year 2040, The results show that t he intense energy release cycle will start somewhere in the year 2030 and will continue up to 2040. The successful operation of ANN and its application to predict seismicity cycle in the selected region shows t hat the approach may be applied to other areas also for the seismic ha zard evaluation.