Six-component seismology: Joint processing of translational and rotational ground-motion

When earthquakes occur, the ground does not only shake laterally and vertically but also undergoes rotational motions. Since centuries, seismologists have been arguing that such rotational motions could be important for the assessment of earthquake hazard and for imaging the Earth’s interior. However, until recently, an instrument sensitive enough to observe rotational ground-motion was not available. This has changed in the last decade and the first portable, commercial rotational seismometers are coming to market. The applications of rotational motion measurements are manifold, ranging from the reduction of nonuniqueness in seismic inverse problems to the characterization, separation and reconstruction of the seismic wavefield. As part of this thesis, I develop novel algorithms to acquire and process six-component (6-C) data (joint recordings of translational and rotational ground-motion). The aim is to lay the foundations for wide-spread applications of 6-C measurements in seismology and, in particular, land seismic exploration. Estimates of rotational ground-motion can be derived from closely-spaced arrays of conventional translational observations. However, receiver-specific perturbations can substantially exacerbate the accurate estimation of array-derived rotational motions. I present a new method that permits the accurate estimation of rotational ground-motion in the presence of such receiver perturbations by jointly inverting array-recordings of seismic waveforms for the rotational ground-motion and frequency-dependent perturbation-correction filters. I find that the estimates of rotational ground-motion can thereby be significantly improved. Additionally, I devise novel methods for joint processing of translational and rotational motions. I show that 6-C point measurements allow one to accurately estimate wave properties, such as the propagation direction, phase velocity and the wave type, even in the presence of interfering waves. I also demonstrate how the seismic wavefield can be separated into different wave types and/or waves arriving from specific directions in a fully automated fashion using single-station 6-C measurements. Finally, I illustrate the benefits of 6-C measurements on the example of data recorded on the lunar surface during the Apollo 17 mission. I use estimates of rotational ground-motion to identify S- waves in the complex coda of the lunar seismograms and extract the first comprehensive elastic parameter model of the shallow lunar crust. The methods that were developed as part of this thesis demonstrate that a single 6-C recording station allows one to extract information on the seismic wavefield that conventionally can only be obtained from extended, dense receiver arrays. Hence, ground rotation measurements have the potential to drastically reduce the amount of seismic receivers needed in seismic exploration, thereby cutting acquisition time and costs. Rotational motion recordings are expected to particularly benefit seismic exploration in areas with restricted access and where only a limited amount of sensors can be employed, such as urban or mountainous areas, the ocean bottom or on extraterrestrial objects. I conclude that rotational motion measurements have the potential to revolutionize the way seismic data is acquired and processed and eventually change the instrumentation standard in seismology and seismic exploration from conventional three-component inertial seismometers to full six-component (6-C) setups incorporating recordings of rotational ground-motion