ASSESSING PROBABILITY OF GREAT EARTHQUAKES IN CENTRAL HIMALAYAN SEISMIC GAP USING SEMI-MARKOV MODEL

Abstract

The present work aims at the introduction and application of Markov and semi-Markov models in estimating the waiting times and magnitudes of the great earthquakes in future. These model assumes that the successive earthquakes in same structural discontinuity are not independent events, which means that time and place of future earthquake events are related to previously occurred earthquake in the region considered, as demonstrated by elastic rebound theory. While other most commonly used models such as Poisson model assumes spatial and temporal independence of all earthquakes including great earthquakes i.e., occurrence of one earthquake does not affect the likelihood of a similar earthquake at the same location in the next unit of time. Such models may apply to regions characterized by moderate frequent earthquakes in larger areas. While Markov and semi-Markov models describes the sequences of events more adequately at small regions with great infrequent earthquakes. In this report, these probabilistic models are applied in the central Himalayan region, by considering the sequence of earthquakes to form a stationary Markov chain. The occurrences of earthquakes in these models is described by discrete time and discrete states for the earthquake magnitudes and locations. In Markov and semi-Markov models, the successive states are governed by the transition probabilities depending on the just previous state and not on the history of reaching to that state. The use of probability distributions for earthquake magnitude and inter arrival time as continuous random variables is unable to account for such a dependence. It has been discussed that how this dependence can be utilized to carry out a time dependent seismic hazard analyses. On the basis of seismotectonic characteristics, the study area is divided into four sub regions and the application of these models has been illustrated to predict the probability of the next earthquake in different sub regions as a function of time from the previous earthquake, conditioned on the magnitude and the sub region of the previous earthquake. The results obtained indicate that the next major earthquake can occur in any of the sub regions with almost equally high probability. Thus, the so called central Himalayan seismic gap is expected to be closed in near future.