Unbiased Moment Rate Spectra and Absolute Site Effects in the Kachchh Basin, India, from the Analysis of the Aftershocks of the 2001 Mw7.6 Bhuj Earthquake

Abstract

What can be learned about absolute site effects on ground motions, with no geotechnical information available, in a very poorly instrumented region? In addition, can reliable source spectra be computed at a temporary deployment? These challenges motivated our current study of aftershocks of the 2001 Mw 7.6 Bhuj earthquake, in western India, where we decouple the ambiguity between absolute source radiation and site effects by first computing robust estimates of coda-derived moment-rate spectra of about 200 aftershocks in each of two depth ranges. Crustal attenuation and spreading relationships, based on the same data used here, were determined in an an earlier study. Using our new estimates of source spectra, and our understanding of regional wave propagation, for direct S waves we isolate the absolute site terms for the stations of the temporary deployment. Absolute site terms for each station were determined in an average sense for the three components of the ground motion via an L1-norm minimization. Results for each site were averaged over wide ranges of azimuths and incidence angles. The Bhuj deployment is characterized by a variable shallow geology, mostly of soft sedimentary units. Vertical site terms in the region were observed to be almost featureless (i.e., flat), with amplifications slightly 1.0 within wide frequency ranges. As a result, the horizontal-to-vertical (H/V) spectral ratios observed at the deployment mimic the behavior of the corresponding absolute horizontal site terms, and they generally overpredict them. This differs significantly from results for sedimentary rock sites (limestone, dolomite) obtained by Malagnini et al. (2004) in northeastern Italy, where the H/V spectral ratios had little in common with the absolute horizontal site terms. Spectral ratios between the vector sum of the computed horizontal site terms for the temporary deployment with respect to the same quantity computed at the hardest rock station available, BAC1, are seriously biased by its nonflat, nonunitary site response. This indicates that, occasionally, the actual behavior of a rock outcrop may be far from that of an ideal, reference site (Steidl et al., 1996).