Use of Cone Penetration Testing in the Interpretation of Ground Shaking From Paleoliquefaction Evidence

Abstract

To evaluate seismic hazard in regions of infrequent earthquakes it is often necessary to rely on evidence left from earthquakes that occurred before seismographs were available. This evidence can consist of human observations or of evidence left in the geologic record. The New Madrid Seismic Zone is a region where geologic evidence of previous earthquakes can be found in abundance in the form of soil liquefaction features. Present day measurements of liquefaction resistance can be used to estimate the severity of the ground shaking that had to occur to cause the observed liquefaction features. The database of present-day liquefaction cases is refined to add recent data and to cull older data based on a set of criteria that allow the effect of measured fines content on liquefaction resistance to be quantified. This refined database allows for the development of liquefaction resistance relationships based on the magnitude of the earthquake-induced shear stress and on the energy dissipated in the soil during earthquake shaking. To apply the energy-based liquefaction resistance relationship in the NMSZ a relationship between the number of equivalent cycles of shaking and site characteristics is developed based on regional seismological characteristics and site conditions. Back-analyses conducted at paleoliquefaction sites in the NMSZ using these refined tools suggest that the magnitude of the 1811--1812 event that caused the paleoliquefaction features was between 7.0 and 7.9 depending on the choice of ground motion prediction relationship, and the choice of magnitude and ground motion prediction relationship for design should be consistent to avoid significantly overestimating or underestimating ground motions at a site.