[S-SS09] Earthquake prediction and forecast

[SSS09-P01] Application of correlation fractal dimension in earthquake precursory study

*Haritha Chandriyan¹, Ramakrushna Reddy, P.N.S. Roy (1.Department of Geology and Geophysics, IIT Kharagpur, West bengal, India.)

Keywords:Seismicity pattern, Earthquake precursor, Fractal dimension

Epicentral and hypocentral distribution of earthquakes exhibits fractal statistics (e.g. Kagan and Knopoff, 1978; Kagan, 1981; Ogata and Katsura, 1991; Bak et al., 2002). This concept has been implemented to study the seismicity patterns associated with strong earthquakes in southern and Baja California regions based on variations in correlation fractal dimension (Dc). The study region has witnessed four Mw > 7 earthquakes, 7.3Mw 1992 Landers, 7.1Mw 1999 Hectormine, Mw7.2 2010 El-Mayor Cucapah, and 7.1Mw 2019 Ridgecrest, within a span of three decades. Unexpected rupture of 2019 Ridgecrest has raised serious concern over the stability of the Garlock fault (GF), it is an inactive fault that has not ruptured in the last 500 years, however, is capable of producing Mw ~7.8 earthquakes. The alignment of GF is perpendicular and connected to the great transform boundary fault in the earth, i.e. San Andreas Fault (SAF), an active plate boundary between the continental North American plate and oceanic Pacific plate. The stress perturbations induced by 2019 Ridgecrest has caused increased coulomb stress in the vicinity of GF, therefore, any movement in GF can possibly affect the state of stress under SAF that can lead to catastrophic consequences for mankind. We monitored the spatio-temporal Dc variations for thirty years (1990-2020) using the correlation integral method. The low Dc values are characterized by high stressed zones (Main, 1992; Maus and Dimri, 1994; Sunmonu and Dimri, 2000; Telesca et al., 2001; Thingbaijam et al., 2008). Subsequently, the decrease in fractal dimension preceding large earthquakes can be considered as a numerical precursor (e.g. Hirata et al., 1987; Murase, 2004; Lu et al., 2005; Roy and Padhi, 2007; Mondal et al., 2019). From this study, we could successfully demonstrate the presence of numerical precursors for all the strong (Mw >7) mainshocks. We observed spatio-temporal clustering of earthquakes has a remarkable correlation with the low value of the fractal dimension. Different cluster types were observed for 2019 Ridgecrest compared to the other three larger shocks. The analysis of the 1995 and 2019 Ridgecrest sequences shows great similarity in the distribution of clusters. By understanding the historical patterns we would be able to determine the behavioral pattern of future bigger events that can help in earthquake forecasting.