S33C-4549:
Analysis of Spectral Characteristics of Seismic Noise Preceding Kachchh, India Earthquake of 19 June 2012 for Advance Warning

Wednesday, 17 December 2014
Indra N Gupta1, David Paul Schaff2, Bal K Rastogi3, P Mahesh3 and Robert A Wagner1, (1)Array Information Technology, Greenbelt, MD, United States, (2)Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY, United States, (3)Institute of Seismological Research, Gandhinagar, India
Abstract:
A significant fraction of the ambient short-period seismic noise at a given site may be attributed to P, S, Rayleigh and Love waves with spectral characteristics strongly dependent on the geological structure underlying the recording station. Unlike an underground nuclear explosion, an earthquake may have intense pre-event activity within the hypocentral region. This pre-earthquake activity may occur at one or more places at various depths mostly within the hypocentral region and over short and/or long periods of time, leading to significant temporal variations in the geological environment. Some of this activity, such as generation of new cracks, may create seismic waves contributing to short-period seismic noise. All these underground phenomena will have significant influence on the spectral characteristics of noise as observed at nearby three-component recording stations.

Kachchh, India Mw 5.1 Earthquake of 19 June 2012 was well recorded at several broadband stations at various epicentral distances. Analysis of limited data from two stations has indicated two distinct premonitory variations in the low-frequency (less than 0.5 Hz) spectral characteristics of noise, initiating several days before the earthquake: (1) systematic shift of peak frequencies to lower values and (2) significant changes among the three components (vertical V, radial, R and transverse, T) of ground motion, evidenced by spectral ratios such as T/R. These results for premonitory variations are similar to those observed for several earthquakes in the United States. Although these preliminary results need to be confirmed by analyzing considerably more data from several additional recording stations, they appear to suggest an entirely new methodology for obtaining advance warning of a few days or more before a large earthquake.