Seismic Hazard Analysis Using Fuzzy-Probabilistic Approach for Chennai City, South India

Introduction

Earthquakes occur as broadband vibratory ground motions caused by tectonic plate motions, volcanic activities, rock bursts and artificial explosions. Among these, tectonic plate motions are the major ones in causing the earthquakes. The Indian subcontinent experienced prominent earthquakes in the recent past. Concerning the seismicity of southern India, particularly Tamil Nadu, satellite images have revealed the fault lines running through Tamil Nadu and the neighbouring areas and fractures under the tectonic plateau (Ramasamy, 2006). These structural features have increased the seismic vulnerability of Chennai and a few significant towns on the eastern coast. Chennai has experienced moderate tremors during the aftershocks of Bhuj (M_w = 7.7, 2001), Pondicherry (M_w = 5.6, 2001), Sumatra (M_w = 9.3, 2004), Nepal (M_w = 7.3, 2015) and Burma (M_w = 6.9, 2016) earthquakes. Furthermore, the ground conditions in Chennai city vary dramatically in respect of subsurface layer thickness, soil types, soil stiffness and depth to bedrock, which can considerably modify the seismic hazard level in the city and accordingly the responses of different types of buildings and structures during earthquakes. Moreover, the high population density, buildings proliferation and low-quality construction also add to the seismic vulnerability of the built environment in the study area.

The earthquakes cause vast destructions and calamities throughout the world. The moderate and above magnitude earthquakes damage almost all constructed facilities; hence, they must be designed to resist a particular level of ground shaking, avoiding enormous damage and lives lost. The level of shaking expected at a site is generally established using Seismic Hazard Analysis (SHA). The SHA is carried out predominantly either by the deterministic or by the probabilistic approach. Deterministic Seismic Hazard Analysis (DSHA) is based on the worst earthquake scenario. It does not consider details regarding the likelihood of an earthquake event and which can be incorporated reliably in Probabilistic Seismic Hazard Analysis (PSHA). The currently available models and procedures in PSHA for characterising the uncertainty in earthquake size, location, recurrence, and site effects are depending on the data accumulated for a prolonged period; however, these data are not sufficient concerning the geological time scale. In SHA, the uncertainties may come from randomness due to natural variability of observations and oscillations of quantity in time, or it may result from ambiguity due to incomplete information. In reality, the uncertainties are related to both the randomness and fuzziness. Therefore, it is more realistic to quantify the different uncertainties associated with the hazard analysis accordingly. A wise choice is to use probabilistic representation with an additive uncertainty feature given by the fuzzy set theory for simplicity, efficiency and information content. Fuzzy sets and fuzzy logic were established by Zadeh (1965 & 2015). The fuzzy logic is an effective mechanism for vulnerability and risk assessment in various engineering fields, including medicine and economics. The concept of fuzzy logic is utilised in earthquake engineering field also. For example, in classification of earthquake-impacted structures (Sen, 2010; Mangalathu & Burton, 2019), earthquake damage evaluation (Mandas & Dritsos, 2004; Mangalathu et al., 2019) and seismic risk assessment (Karimi & Hullermeier, 2007; Mangalathu & Jeon, 2020).