On the Application of Surface Wave Surveys for Seismic Site Response Evaluation

Abstract: Popular Abstract in English The work presented here is about methods that will enable more precise estimates of soil and rock parameters related to seismic hazard mapping, in particular when resources are limited. This will allow for evaluating the seismic risk in countries where earthquakes are frequent and large budgets are not available. Particularly, the study of ground response due to the energy released by an earthquake is of great interest in sites prone to earthquakes. The reason is that when the uppermost soil layers are subjected to seismic waves produced by an earthquake, there is a potential of increasing the level of ground movement (amplification), which in turn is one of the major causes of damage to buildings. The most accurate way to quantify the actual ground response is by using measurements of the vibrations of the ground at the instant when earthquakes occur. These earthquake records can be obtained simultaneously at the surface and beneath the surface by using buried sensors that are placed at various depths, called seismic vertical arrays. These arrays, however, are very expensive because the installation normally requires a great deal of resources, for example drilling of several deep wells and equipment for data acquisition. A more economical alternative for estimating the ground response is by using a non-invasive seismic method, i.e. measure the vibrations at the surface. With this method it is possible to characterize the ground stiffness by means of the shear wave velocity variation with depth. The shear wave velocity profile is used to model the site response and to predict the actual ground response in a reasonably accurate way. When using this approach, current building codes require a velocity model to at least 30 m depth. However, estimating a velocity profile down to this depth is not easy due to practical limitations. Such limitations can be size and availability of the equipment needed, and often also the lack of a suitable source to generate the low frequency energy needed to reach that depth. This often requires increasing the complexity of the equipment and yields higher costs. In this thesis possible alternatives to maximize the measuring capabilities while keeping the instrumentation to a minimum are proposed. For example, when the installation of multiple sensors to record earthquakes is not possible, a single sensor that is installed in a single well can be used. The sensor is then successively installed at different depths, recording multiple earthquakes. This thesis shows that the resulting combined earthquake records will yield results similar to those from a fully instrumented seismic vertical array. For the non-invasive methods, the coverage of a short survey line resulting from the use of standard engineering instruments is increased by acquiring several measurements while moving the survey array away from the source. When doing this, imperfections in the ground and at the source create artifacts that potentially affect the evaluation. In this thesis this is handled by signal processing, and the result is a “virtual” survey line with adequate length. Further, to be able to penetrate deeper than 30 meters, low frequency energy is retrieved in a controlled way by locating the measurement close to a road, and use the energy produced by the traffic.