Acoustic and Elastic wave Exploration and Monitoring using Dense Passive Arrays

Abstract

Passive recording systems have become increasingly popular and important in various marine acoustic and geophysics branches. Data from these systems provide valuable information and can be used in various applications, e.g., to characterize whales, earthquakes, or, more generally, the subsurface. These applications contain different frequency ranges, and various recording systems have been developed to capture them all. For example, strainmeters are sensitive to the lowest frequencies, like those generated by Earth’s tide, while hydrophones, geophones and broadband seismometers can pick up higher frequencies, such as those produced by whales and small-magnitude earthquakes. Hence, the choice of receiving system depends on the application. These systems contain pointsensors, therefore making it challenging to sufficiently instrument the whole world. This is mainly due to the cost of installation and maintenance, hence leaving the world sparsely instrumented. In recent years, new technologies have been developed to increase instrumentation. One example is Permanent Reservoir Monitoring (PRM) systems, which have been installed near oil and gas fields to improve recovery. Such a system typically contains 1000s of 4-component sensors per PRM system and is placed near interesting features in the subsurface. Another example is Distributed Acoustic Sensing (DAS), which has, in recent years, emerged as a new technology to help increase the density of sensors worldwide. The receiver unit, the DAS interrogator, is connected to an optical fiber within a fiber telecommunication cable. This can re-purpose the fiber to a distributed sensor, sensing acoustic signal every meter for up to 150 km, thus creating a far-reaching, easily accessible receiver unit. This thesis consists of several manuscripts investigating how dense passive receiver arrays can extract information from the water column through whale songs and the subsurface through different seismic waves. Most of the work has focused on DAS data from three datasets. The first DAS dataset was acquired in 2020 in Svalbard by re-purposing one of two telecommunication cables connecting Longyearbyen and Ny-Ålesund. The data were used to study the signals from a local earthquake. Standard preprocessing techniques were used to enhance the on-set of P- and S-waves, which were further used to locate the event. The obtained epicenter location was compared to those found using conventional receiver systems. The second DAS dataset was acquired in 2022 in Svalbard by re-purposing both telecommunication cables connecting Longyearbyen and Ny-Ålesund. We used data recorded on the cables to track up to eight whales for five hours, using two different localization methods. Additionally, seismic shots from a single air gun were used as ground truth information to assess the accuracy of the tracking methods. The third DAS dataset was acquired using a dedicated fiber in Rissa, Norway, trenched roughly 40 cm into a known quick clay area. This data were used to monitor the quick clay as a new road was built on top of the clay by investigating changes in the shear-wave velocity depth profile. Inversion of the dispersion curves of Rayleigh waves was used to estimate the shear-wave velocity profiles. Moreover, the Rayleigh waves were generated from sledge hammer shots and the ambient background noise. The final dataset used in this thesis was recorded on a PRM system South of the Oseberg C platform in the Norwegian North Sea at a water depth of 107 m. The data were used to estimate the average quality factor of a sedimentary package under the platform by applying a spectral ratio of earthquake signals recorded on a station pair, one station placed on sediments near the Oseberg C platform and one on bedrock in Bergen, Norway. Three different earthquakes were analyzed. By studying the local earthquakes in the Svalbard 2020 dataset, we showed how a simple beamforming procedure could find the direction of the earthquake, the apparent velocity and help to increase the onset of P- and S-waves. In addition, we could estimate the epicenter location of a local earthquake within 17 km of the estimates from conventional receiver systems. Using the two fiber cables in Svalbard, we could see when the various whales vocalized whilst accurately locating them. A dedicated single air gun fired at known positions close to the fibers allowed us to estimate the average localization error to be 100 m. These capabilities demonstrate the potential for near-real-time whale tracking using DAS that could be applied anywhere in the world where whales and fiber-optic cables are present. Using the Rayleigh waves recorded in the quick clay DAS data, we could estimate shearwave velocity depth profiles over a seven-month acquisition period. The obtained dispersion curves, and shear-wave velocity profiles, showed small time-lapse variation during the acquisition period, suggesting that the construction work did not alter the quick clay’s property. Nevertheless, the obtained results capture the non-repeatability effects within the acquisition period and provide reference curves for the study at undisturbed conditions. From the Oseberg PRM data, we found average Qp and Qs values for the sedimentary sequence underneath the PRM system. TheQp values were found to be more scattered and hence more uncertain than the Qs values due to the P-wave being closer to the background noise level. The quality factors for one of the studied earthquakes were found to be 92±18 for the P-wave and 84 ± 13 for the S-wave, suggesting more attenuation of S-waves than P-waves in this sedimentary layer.