Probabilistic estimation of earthquake source location and magnitude using inverse analysis of regional paleoliquefaction studies

Liquefaction is one of the most significant and remarkable causes of ground failure in geotechnical earthquake engineering. The phenomenon mostly occurs in saturated cohesionless soils when subjected to seismic loading. Studies on past liquefaction evidence, also known as paleoliquefaction studies, have helped several researchers predict a particular region’s future vulnerabilities. However, it is always difficult to prepare human life for future devastation resulting from ground failures. But, prior estimation of the magnitude and likelihood of earthquakes that may strike a location shortly can create an environment involving fewer risk factors. Several methods are available to back-calculate the strength of shaking and earthquake magnitude from seismic evidence, such as paleoliquefaction. Knowing the origin of an earthquake aids in locating the fault zone. As a result of these historical investigations, information for seismic hazard analyses and ground motion forecasts for a particular region becomes possible. The present study is designed on similar grounds to carry out the investigation. A total of nine sites are selected in the Roorkee region, which is vulnerable to earthquakes. The region is also prone to liquefaction based on experimental evidence available from the past studies. The Standard Penetration Test data analysis performed on all nine sites is used for site characterization. For probabilistic earthquake source characterization, magnitudes between 3.5 and 8.5 and PGA between 0.05 and 0.5 are considered. For the interpretation of the most likely source location and its corresponding likelihood of magnitude, both site and source data are utilized in the ground motion model. The findings show that with the increase in source-to-site distance, the likelihood of source occurrence reduces, whereas the most likely magnitude increases. Eventually, this framework illustrates a probabilistic method for determining the seismic source parameters based on paleoliquefaction inverse analyses.