To run Nature’s Heat geothermal operations (Kwintsheul, The Netherlands) in a safe and efficient manner, and with the support of the local population, it is important to mitigate potential hazards. A well known potential hazard related to geothermal operations is the occurrence of induced seismicity due to injection and extraction of fluids in and from the subsurface. Monitoring the subsurface allows detecting the induced seismicity due to the geothermal operations. To monitor the seismic activity around the geothermal field of Nature’s Heat project, a passive seismic network was installed by Delft University of Technology, Seismoctech (Greece), and Gastreat- ment Services BV. This network has detected and located one single weak event on July 14, 2019.

Within this bachelor thesis, a beamforming technique is adopted to provide a framework that can be used to detect seismic events recorded by the passive seismic network. The goal is to separate coherent signals from noise and characterize the signal (e.g., estimate propagation direction). The objective of this bachelor thesis is to assess whether beamforming is capable of lowering the detection threshold, and determining the signal’s apparent velocity and back azimuth, using data recorded by the Kwintsheul seismic array. For applying the beam- forming technique, Kwintsheul data was implemented in a python code, originated from Remote Online Sessions for Emerging Seismologists (ROSES).

Following applying the beamforming technique, the results of the detection, and the parameter estimation of the back azimuth and apparent velocity, are given for the P-wave and S-wave, 300.96 deg. 6.43km/s and 315 deg. 3.31km/s, respectively. The results of the Fisher statistics and the f-k analysis show that beamforming is capable to discriminate coherent signals from noise.
For beamforming analysis, it is necessary to have array elements that are distant from the event. As the ar- ray design is fixed, it is, therefore, necessary to eliminate data from array elements in the proximity of the event. Hence only data from array elements in the far-field is used for beamforming. This approach decreases the num- ber of array elements, and therefore the signal to noise ratio is decreased as well. In this case, to improve the resolution of the applied beamforming technique, the array design needs to be modified, with all array elements situated in the far-field.

With a rapid increase in computational resources two dimensional full-waveform inversion is evolving into a promising tool for near-surface geophysics. However, near-surface applications suffer from local minima due to amplitude errors associated with two dimensional full-waveform inversion. By clever definition of the misfit function, the influence of the amplitude errors can be mitigated. Here, I use a recently proposed misfit function based on the instantaneous-phase coherency. The instantaneous-phase coherency misfit function uses complex trace analysis to create an amplitude unbiased misfit function. First, I compare the new misfit function to a traditional least-squares misfit function by inverting synthetic models where noise is added, a field dataset containing Rayleigh waves and a field dataset containing Love waves. Next, I perform inversions on layer cake models to investigate the accuracy of the full-waveform inversion using the new misfit function and finally, I test the robustness of the inversion by using a complex subsurface model. The inversions performed on the synthetic models show that the instantaneous-phase coherency misfit is more robust when noise is introduced to the data compared to the least-squares misfit. Furthermore the two field datasets, demonstrate the ability of the instantaneous-phase to deliver accurate near-surface results when used on field data. The results from the layer cake inversions where inconclusive, however I did demonstrate that a better selection of the bandwidth did improve the result. Finally, the results from the complex subsurface model show that the instantaneous-phase coherency is able to resolve parts of the complex subsurface model.

Three-component seismometers deployed on the sea bed have been used for many years. These so-called ocean bottom seismometers (OBS) are used to register both compressional and shear waves. The recorded seismic information is used to enhance our understanding of areas of interest located offshore, like hydrocarbon reservoirs, mid oceanic ridges, and continental margins. A lot of progress has been made on OBS applications. However, there are still some distinctive problems mainly related to clock drift, orientation, and coupling. We use and test \textit{OCloC}, a new method which seeks to improve the ability to detect and correct clock errors of OBSs. The method was successfully developed using an extensive dataset from a seismic network, which included many land- and ocean bottom-seismometers. One of the objectives is to understand under which conditions OCloC can be used when the dataset is relatively small. To do this, data from a network deployed in the Red Sea coastal area of the Kingdom of Saudi Arabia is used. This network includes two ocean bottom seismometers which are complemented by several land-stations. We collected all used data during five campaigns, both offshore and onshore. To analyse the data quality we designed and applied an extensive data quality control, where after the data is prepared for Ocloc through pre-processing. Both workflows are presented in a roadmap. This roadmap is made applicable for ocean bottom seismometers by including an orientation estimation, for which we used a technique based on Rayleigh wave polarization. Our quality control shows that the data from both ocean bottom seismometers show significant attenuation, primarily for short periods, which is possibly caused by a poor coupling to the loose sediments on the sea bed. The noise source illumination turns out to be highly anisotropic, with a high microseism noise source concentration in the Red Sea. This results in a poor signal to noise ratio for the interferometric surface wave responses. As a consequence, we find that the limited station deployment time and data quality interferes with the ability to obtain an accurate clock error. Nevertheless, the technique to estimate the OBS orientation turns out to be successful.

Landslides pose a risks to the communities settled in their vicinity. They pose a threat to human lives and significant economical damage to the affected communities. As the scientific community agrees that the climate change will further increase the risks associated with landslides, it is important to develop a reliable, cost effective and widely applicable technique to assess the stability of the potentially unstable slopes. Such a technique could allow to create an early warning systems meant to help evacuate the communities at risk before the landslide event happens. There have been many studies applying different methods of achieving this goal. This study will apply electrical resistivity tomography to investigate the subsurface and use machine learning methods in the form of supervised and unsupervised learning to interpret the measurements. The electrical resistivity tomography is a widely applied technique to investigate the stability of slopes. It can determine the water table in the subsurface, which plays the crucial role in the stability of the slopes. However, the manual interpretation of the electrical resistivity tomography profiles is a time-consuming and cumbersome task. In order to develop a reliable early warning system both the measurements and their interpretation have to be automatized. The first one has been already achieved. There are systems available that can perform electrical resistivity tomography automatically and continuously. In order to automatize the second part, the following research investigated the capability of machine learning to interpret resistivity profiles alongside with various methods of improving the accuracy of the results. The results show that machine learning is capable of performing this task. Some limitations were identified and guidelines for the preprocessing of the data were proposed. Several areas that require further investigation were identified.

Global phases are seismic waves that travel through the earth's core before emerging at the surface. Traditionally, global phases are used to obtain subsurface information up to 1 Hz. A recent study, however, found that signal present in the coda of strong, distant earthquakes contains higher frequencies. Quantitative analysis of the coda of earthquakes, of at least magnitude 6, shows that frequencies of, on average, 3 Hz can be found, with the most promising results coming from P- and PKIKP-phases. By autocorrelating the coda of these phases an estimate of the zero-offset reflection response below a seismic station can be retrieved.
The method allows to successfully delineate basins in both Argentina and the United States. Even higher frequencies, up to 8 Hz, are used at specific stations, that have been active for an extended period of time (~ 8 years). Again the method succeeds in retrieving structural information below these stations. The results of both studies are confirmed by existing literature as well as Ambient Noise Seismic Interferometry (ANSI) studies. ANSI is a method, which retrieves the reflection response below a station by applying seismic interferometry to ambient noise.
Advantages are that the method is computationally cheap and fast and that it is a passive measurement thus requiring low-effort. A disadvantage, however, is the unpredictability of earthquakes, which means that the exact duration the stations have to be active is unknown. Instead, the station will have to record over a prolonged amount of time, during which suitable earthquakes should occur. Research into the upper crustal structure, in particular, may benefit from the method. For example, the depth of a sedimentary basin can be discovered and imaged using this method. This gives important constraints, e.g. for geothermal prospecting.

Coda wave interferometry is a technique used in the field of seismology, which utilizes the later part of the signal (coda) to detect subtle changes in a medium. In recent years, the application of coda wave interferometry to concrete structure has been assessed for structural health monitoring purposes. Smart aggregate is a sensor which consists of a piezoelectric sheet which is sandwiched between two marble layers which are meant to be used for structural health monitoring purposes by embedding it into concrete. However, its implementation for coda wave interferometry applications had not been attempted previously. In this research, the application of coda wave interferometry in concrete structures is explored further. The aim of this research is to assess the possibility of implementing coda wave interferometry to monitor the hydration process and the evolution of elasticity-modulus of concrete, as well as to learn how the wavespeed changes in concrete specimens subjected to cyclic loading in compression and bending. Additionally, seismic interferometry is also attempted to retrieve virtual impulse response to be used for coda wave interferometry. All experiments in this research utilize smart aggregates as transducers. By implementing coda wave interferometry, it is found that wavespeed does increase as concrete ages. This wavespeed increase can be linked to the evolution of elasticity-modulus of concrete, which enables its value to be monitored through the utilization of coda wave interferometry. It is also found that the use of embedded smart aggregate yields excellent reciprocity and stable correlation coefficient throughout the recording, while attached smart aggregates do not perform as well as the embedded ones in terms of reciprocity and correlation coefficient. Positive linear wavespeed change vs. stress and strain relationships in compressive samples are observed in lower stresses. In higher stresses, both wavespeed change vs. stress and strain display gradient reductions. Under repeated cyclic loadings, the loading phase of the first load cycle tend to have lower initial wavespeed change vs. stress and strain gradients compared to the following load steps, and the wavespeed change vs. stress and strain paths of reloading phases tend to follow the paths of their previous unloading phases. Wavespeed change vs. strain is more representative compared to wavespeed change vs. stress in depicting the compressive specimens’ condition due to the occurrence of permanent deformation during loadings. In a 10m-long beam specimen subjected to bending and shear, coda wave interferometry of later arrivals reveal decrease in wavespeed in the first loading phase of the test, while earlier arrivals show increase in wavespeed in the same phase. Moreover, it is possible to detect major crack formations by utilizing coda wave interferometry, which sensitivity is determined by the location of the cracks relative to the source-receiver sensors’ proximity. By assessing earlier arrivals of the signals recorded by smart aggregate implanted in the compression zone, the shift from uncracked to cracked section is observed through changes in wavespeed change vs load gradients. Seismic interferometry attempt was unsuccessful due to poor repeatability of the hammer hits and insufficient illumination to create diffuse wavefield.