Seismic input for chennai using adaptive kernel density estimation technique

Abstract

Seismic input at a particular site can be estimated quantitatively using probabilistic or deterministic approach. Probabilistic seismic hazard analysis (PSHA) provides a framework in which uncertainties in the size, location, rate of recurrence and effects of earthquakes are explicitly considered in the evaluation of seismic hazard. The probabilistic way of analyzing the seismic hazard was developed conventionally by introducing zones in the seismogenic regions based on regional seismotectonic and geologic setting. The seismic uniformity is assumed within these source zones. Later, many researchers found that the conventional approach has many drawbacks viz., difficulty in delineating seismic sources into various zones, difficulty in applying Gutenberg-Richter (G-R) recurrence relationship to characterize the seismic source for low seismicity regions and distributed seismicity, and the consideration of uniform seismicity within the zone is also questionable. Because of these issues, several alternative methods to hazard estimation have been proposed in the literature. In the present study, zone free approach is proposed to evaluate the spatial distribution of seismicity based on kernel density estimation technique. The kernel technique provides a spatial variation of the seismic activity rate unlike the conventional approach where it is constant for a seismic source zone. The fixed bandwidth kernel poorly evaluates the earthquake distributions since the earthquake catalogue has several areas of high activity clusters and low background seismicity. Therefore in this study, clustering based adaptive kernel technique is proposed to find the spatial activity rate and integrated with other forms of uncertainty in magnitude and distance to determine the probability of exceedance of the selected ground motion parameter. The proposed methodology of seismic hazard analysis has been used for Chennai, southern India and the seismic input is provided in the form of Peak Ground Acceleration (PGA) and Uniform Hazard Spectra (UHS) for return periods of 475 and 975 years. The UHS obtained are compared with the Cornell-McGuire approach and IS 1893: 2002.