Strong Ground Motion Attenuation Relationships For Lanzhou And North China From Small Earthquake Records By China Earthquake Networks

Essencially, ground motion attenuation relationships are empirical, which arebased on a large number of observed strong motion data. In the field of earthquakeengineering, it is a hot point to establish ground motion attenuation relationships forthe countries and regions that lack of strong motion data. A method is developed in thispaper to establish attenuation relationships with small earthquake data from regionalbroadband digital earthquake networks, such as Lanzhou region and North China.For the target region of Lanzhou (N33°-39°, E100°-108°), small earthquake datafrom November10,2009to October21,2012is collected by Lanzhou region digitalseismic observation network. Totally, there are592velocity time histories, recorded by49regional digital observation stations in33small earthquakes (Mw=3.5-4.5, focaldepth≤30km).30s velocity time histories are abstracted to fast Fourier transform,velocity Fourier spectral envelopes are the target curves in micro-Genetic Algorithm(μGA) method. Five parameters (stress drop Δσ, quality factor parameters Q0and η,geometric attenuation parameters R1and R2), which describe the regional source andcrust medium characteristics, are inversed by μGA, and ground motion attenuationrelationship of Lanzhou region is established. Compared with empirical attenuationrelationships and several observed strong ground motion data, the result is lower in thecase of Mw=5.0and Mw=6.0, strong ground motion records are from soil stations isprobably one of the reasons. There is only one record in the case of Mw=7.0, we haveno further comment. Relatively speaking, the result is close to empirical attenuationrelationships, which is reasonable. Compared with empirical relationships, the meanvalue of errors in this study is larger in the case of Mw=5.0, the mean value is in themiddle position in the case of Mw=6.0, standard deviation is lower. We have too littlestrong motion data to comment in the case of Mw=7.0. Afterwards, we draw a regionalseismic zoning map by our attenuation relationships. Compared with “The ZonationMap of Earthquake Ground Motion Parameters in China”(GB18306-2001), most partsof regions are coincidence, and some parts of the region are lower. The differencemight be from the attenuation relationships, or from the adjustment in the zoningprocess, according to some demands.The relationships between Kappa value and moment magnitude, focal depth andepicenter distance are discussed, Kappa value is irrelevant with moment magnitude andfocal depth, the correlation between Kappa value and epicenter distance is not obvious,concluded by small earthquake records, however, linear relation is showed by strongmotion data. The coefficient of distance by regression approximately0.01-0.001times of the constant value. The Kappa value at R=0km (κ0) is dominant in Kappa value. fmaxfilter is replared by Kappa filter, κ0is inversed with other five regional parameter toestablish the strong ground motion relationship of North China.Small earthquake data from February,2002to September,2012in North China(N36°-43°, E113°-120°), recored by Hebei region digital seismic observation network,are collected. Totally, there are1995velocity time histories, recorded by136regionaldigital observation stations in28small earthquakes (Mw=3.5-4.5, focal depth≤30km).Compared with strong motion data, the result can pass through the strong motion datain the case of Mw=5.0and Mw=7.0, it is higher than strong motion data in the case ofMw=6.0. Compared with empirical relationships, the mean value of errors in this studyis the mean slight lower in the case of Mw=5.0and Mw=6.0; the mean value is theminimum but the standard deviation is the maximum in the case of Mw=7.0. In thewhole process to establish the ground motion attenuation relationships, strong motiondata are not adopted, only for the comparison, which shows that the method is feasible.