Non-physical arrivals produced by seismic interferometry, the process whereby Green’s functions are synthesized between two points by cross-correlation, crossconvolution or deconvolution, are often considered to provide little information about the Earth’s subsurface. Their contributions are usually suppressed in interferometric Green’s function estimates to suit existing methods of seismic velocity estimation which favour the more familiar physical arrivals. In this thesis we show that the non-physical arrivals retrieved in exploration-type settings are useful for determining the long-wavelength seismic velocity structure and can be used to obtain improved Green’s function estimates. First, we estimate the seismic velocity and layer thickness by measuring the signal coherency along traveltime curves between two receivers in a collection of traces consisting of cross-correlated wavefields, known as the correlation gather. The traveltime curves represent the traveltime differences between wavefields recorded at the two receivers. When the procedure is used to find the velocity and thickness of the uppermost layer, the traveltime curves implicitly incorporate the physical and non-physical wavefields in the Green’s function estimates. When the procedure is applied to a model with more than one layer, the traveltime curves correspond to non-physical wavefields only in the Green’s function estimates. Instead of suppressing multiple reflections as in conventional methods, the procedure incorporates the traveltimes of multiple reflections to constrain velocity and thickness estimates. The procedure above is most suitable for recovering the first-layer seismic velocity. We propose a simpler method to estimate the seismic velocities corresponding to deeper layers. We find that the Green’s functions contain very weak reflections, but are dominated by non-physical refractions if retrieved using a limited source aperture. The seismic velocities are easily identifiable as repeating bright spots after transforming the refraction-dominated Green’s functions to the − p domain. We show that non-physical reflections can be used constructively to provide physical reflections, and therefore improved Green’s function estimates, by using a cross-convolution operation in a new variant of seismic interferometry, called source-receiver interferometry. We also show that non-physical reflections associated with the cross-correlation of reflections from different interfaces allow for the direct estimation of interval velocities and layer thicknesses. This method removes the necessity to first find the root-mean-square velocities and two-way traveltimes required to compute the interval velocities by Dix inversion. Overall, this thesis significantly improves our understanding of how nonphysical energy in seismic interferometry both provides useful information about the Earth’s subsurface and contributes to physical energy in particular interferometric methods.

Seismic or wavefield interferometry refers to a set of methods that synthesize wavefields between pairs of receivers, pairs of sources, or a source and a receiver, using wavefields propagating from and to surrounding boundaries of sources and/or receivers. Starting from cross-correlations of ambient seismic noise recordings, which provide the signal between two receivers as if one of them had been an active source, interferometric methods developed rapidly within the last decade, revolutionizing the way in which seismic, acoustic, elastic, or electromagnetic waves are used to image and monitor the interior of a medium. Only recently, an explicit link was found between the methods of source-receiver interferometry (SRI) and seismic imaging, a technique widely used in seismic exploration to map diffractors and reflectors in the subsurface, but also in more academic studies investigating, for example, deep crustal processes. This link is particularly interesting because SRI, in contrast to classical imaging schemes, does not rely on the single-scattering assumption but accounts for all multiple-scattering effects in the medium. While first non-linear imaging schemes based on SRI have been proposed, the full potential of the method remains to be explored and a number of open questions concerning, for example, the role of non-physical energy in interferometric wavefield estimates, require further investigation. The aim of this thesis is to gain more insight into the method of source-receiver interferometry in the context of wavefield construction and analysis in multiply scattering media, especially when theoretical requirements of the method (such as complete boundaries of sources and receivers, surrounding the medium of interest) are not met. First I analyse the single diffractor case using partial surface boundaries only. I find that only two out of eight terms of the SRI equation are required to construct a robust estimate of the scattered wavefield, and that one of these two terms is also used in seismic imaging. The other term provides a pseudo-physical estimate of the scattered wave; this is a new type of non-physical energy that emulates the kinematics of a physically scattered wave. I then proceed to a multiple scattering scenario, using the pseudo-physical term to predict the travel times and exact scattering paths of multiply diffracted waves. The presented algorithm is purely data-driven and fully automated and, as a by-product, provides a new tool to isolate primary diffracted waves from a complex multiply diffracted wavefield. Finally, the concept is expanded to multiply reflecting media. In reflection seismic data, multiply reflected waves should be removed prior to migration in order to avoid artefacts in the seismic image. I demonstrate how internal multiples can be estimated and attenuated using pseudo-physical energy constructed from SRI. Moreover, an explicit link is derived between the internal-multiple equation based on SRI and the internal-multiple equation derived from the inverse-scattering series (ISS), currently the most capable algorithm for internal-multiple attenuation. Using the insight provided by the SRI approach, I suggest an alternative equation that estimates internal multiples more effciently compared to the current method. Overall, this thesis improves our understanding of how physical, non-physical, and pseudo-physical wavefields are constructed in SRI, how new information about multiply scattered wavefields can be inferred, and how SRI relates to other methods of wavefield analysis, in particular seismic imaging and the ISS.

Virtual source gathers were generated using the principles of seismic interferometry from 135 hours of ambient noise recorded during a controlled-source survey across the Salton Sea Geothermal Field in southern California. The non-uniform nature of the noise sources violated a primary assumption of the method and generated artifacts in the data. The artifacts generated by the high-energy impulsive sources (e.g. earthquakes, shots) were removable using traditional methods of amplitude normalization prior to cross-correlation. The continuous source artifacts generated by the geothermal wells and highways required an unconventional approach of utilizing only normalized impulsive sources to successfully reduce the artifacts. Virtual source gathers were produced successfully that contained strong surface waves at 0.4-2.5 Hz, an order of magnitude below the corner frequency of the geophones, and modest body waves at 22-30 Hz, which are generally more difficult to obtain due to the need for many large, well-distributed subsurface sources. The virtual source gathers compare well to nearby explosive shots and are more densely spaced, but have a much lower signal-to-noise ratio. Analysis of the surface waves was complicated by strong higher-order modes. Spectral analysis of virtual source gathers required utilization of the geothermal plant energy, which produced usable signal at offsets required for mode separation. The virtual source dispersion curve compared well to a dispersion curve from a nearby explosive shot. P-waves were observed on the virtual source gathers. Creation of a low-quality multichannel reflection stack revealed two weak reflectors in the upper 2 km. / Master of Science

This research is focused on the application of both new and established seismic interferometry techniques to a single area: the Altiplano in the Andes region. This area has already been widely studied in terms of its geological evolution. Nevertheless, a single accepted theory has not yet been developed to explain why the topography of the Andes incorporates such a large area of low relief at this altitude. The Altiplano is therefore an interesting zone to study. This research introduces and analyses new concepts and methodologies, such as retrieving surface and body waves between earthquakes by using interferometry. Nevertheless, several factors, such as the quality of recordings, the separation between sources, and the velocity gradient of the medium, had to be taken into account for body and surface wave retrieval. This research also analysed the retrieval of body waves by means of seismic interferometry applied to coda wave arrivals. Results show that due to the attenuation of S waves produced by the zone of partial molten material, when using S coda waves, seismic interferometry does not achieve the objective of wave retrieval. On the other hand, P coda waves gave good results. Also, the combined methodology of interferometry by cross-correlation and convolution was shown to account for the behaviour of the retrieved waves and provided an indication of how the distribution of sources affects the Green’s functions estimates for body waves in this area. Another point covered by this research was the analysis of passive recordings in order to retrieve surface and body waves. Results indicate that surface and body waves could be retrieved. However, in order to retrieve body waves, special circumstances are required, such as lateral continuity of the Moho, a relative strong Moho impedance contrast, and simplicity of the geologic structure because these factors will contribute to a strong signal like that obtained in critical reflections making interferometry results more successful.

The ability to synthesize recordings from surface data as if they had come from subsurface sources has allowed geophysicists to estimate subsurface properties. Either in the form of classical seismic migration which creates structural maps of the subsurface, to the more recent seismic interferometry which turns seismic sources into receivers and vice-versa, this ability has provided a rich trove of methods with which to probe the Earth's interior. While powerful, both of these techniques suffer from well-known issues. Standard migration requires data without multiply-scattered waves (multiples). Seismic interferometry, on the other hand, can be applied to full recorded data (containing multiples and other wave types), but requires sources (receivers) to be physically placed at the location from (to) one wishes to estimate responses. The Marchenko method, developed recently for the seismic setting, circumvents both of these restrictions: it creates responses from virtual subsurface sources as if measured at the surface. It requires only single-sided surface data, and a smooth estimate of the subsurface velocities. Initially developed for acoustic media, this thesis contributes the first elastic formulation of the Marchenko method, providing a more suitable setting for applications for the solid Earth. In another development, this thesis shows how the obtained virtual recordings may be used for migration. With these two contributions, this thesis shows that for elastic surface seismic data, the main drawbacks of migration and interferometry can be overcome using the Marchenko method: multiples do not harm migrated images, and sources (receivers) need not be physically placed in the medium for their responses to be accessible. In addition to the above methods, generating images devoid of multiple-related artifacts can be achieved in several other different ways. Two approaches to this are the use of a post-imaging filter, and attenuation of internal multiples in the data itself. This thesis contributes one new method using each of these approaches. First, a form of Marchenko imaging is known to create spurious reflectors, as also occurs in standard reverse-time migration (RTM). However, these artifacts usually appear at different locations in RTM and this form of Marchenko imaging. Using this insight, this thesis presents a way to combine pairs of seismic images in such a way that their differences (e.g. artifacts) are attenuated, while similarities (e.g. true reflectors) are preserved. Applying this to RTM and Marchenko-derived images markedly improves image quality. Second, this thesis presents a method to estimate multiples in the data. Multiples can either be migrated on their own to aid in interpretation, or be adaptatively removed from the data to improve image quality. However, because of the nature of adaptive subtraction, this second method may harm primary energy. To avoid this problem, this thesis develops a final method to directly image using only primary energy in the recorded data using only a small number of virtual points. This method bypasses the need for multiple removal and the estimation of subsurface responses at every depth location. In addition, primaries from particular reflectors may be particularly selected such that they can be imaged individually. Overall this thesis provides several new ways to use surface seismic data in such a way that multiples do not hamper the end product of seismic data processing: the seismic image. It demonstrates this use on synthetic and real data, proving their effectiveness.

This thesis describes new interferometric imaging methods for migration and waveform inversion. The key idea is to use reflection events from a known reference reflector to ”naturally redatum” the receivers and sources to the reference reflector. Here, ”natural redatuming” is a data-driven process where the redatuming Green’s functions are obtained from the data. Interferometric imaging eliminates the statics associated with the noisy overburden above the reference reflector. To mitigate the defocussing caused by overburden errors I first propose the use of interferometric least-squares migration (ILSM) to estimate the migration image. Here, a known reflector is used as the reference interface for ILSM, and the data are naturally redatumed to this reference interface before imaging. Numerical results on synthetic and field data show that ILSM can significantly reduce the defocussing artifacts in the migration image. Next, I develop a waveform tomography approach for inverting the velocity model by mitigating the velocity errors in the overburden. Unresolved velocity errors in the overburden velocity model can cause conventional full-waveform inversion to get stuck in a local minimum. To resolve this problem, I present interferometric full-waveform inversion (IFWI), where conventional waveform tomography is reformulated so a velocity model is found that minimizes the objective function with an interferometric crosscorrelogram misfit. Numerical examples show that IFWI, compared to FWI, computes a significantly more accurate velocity model in the presence of a nearsurface with unknown velocity anomalies. I use IFWI and ILSM for 4D imaging where seismic data are recorded at different times over the same reservoir. To eliminate the time-varying effects of the near surface both data sets are virtually redatumed to a common reference interface before migration. This largely eliminates the overburden-induced statics errors in both data sets. Results with synthetic and field data show that ILSM and IFWI can suppress the artifacts caused by non-repeatability in time-lapse surveys. This can lead to a much more accurate characterization of the movement of fluids over time. In turn, this information can be used to optimize the extraction of resources in enhanced oil recovery (EOR) operations.

Ambient seismic noise from generation to its application for determination of sub-surface velocity structures is analyzed using continuous data recordings from the Swedish National Seismic Network (SNSN). The fundamental aim of the thesis is to investigate the applicability of precise velocity measurements from ambient noise data. In the ambient noise method, a form of interferometry, the seismic signal is constructed from long-term cross correlation of a random noise field. Anisotropy of the source distribution causes apparent time shifts (velocity bias) in the interferometric signals. The velocity bias can be important for the study area (Sweden) which has relatively small velocity variations. This work explores the entire data path, from investigating the noise-source distribution to a tomographic study of southern Sweden. A new method to invert for the azimuthal source distribution from cross-correlation envelopes is introduced. The method provides quantitative estimates of the azimuthal source distribution which can be used for detailed studies of source generation processes. An advantage of the method is that it uses few stations to constrain azimuthal source distributions. The results show that the source distribution is inhomogeneous, with sources concentrated along the western coast of Norway. This leads to an anisotropic noise field, especially for the secondary microseisms. The primary microseismic energy comes mainly from the northeast. The deduced azimuthal source distributions are used to study the level of expected bias invelocity estimates within the SNSN. The results indicate that the phase-velocity bias is less than 1% for most station pairs but can be larger for small values of the ratio of inter-station distance over wavelength. In addition, the nature of velocity bias due to a heterogeneous source field is investigated in terms of high and finite-frequency regimes. Graphical software for phase-velocity dispersion measurements based on new algorithms is presented and validated with synthetic data and by comparisons to other methods. The software is used for phase-velocity measurements, and deduced azimuthal source distributions are used for velocity-bias correction. Derived phase-velocity dispersion curves are used to construct two-dimensional velocity maps of southern Sweden at different periods based on travel-time tomography. The effect of the bias correction is investigated, and velocity maps are interpreted in comparison to previous geological and geophysical information.

This thesis concerns the study of time reversal and plane-wave decomposition in various geophysical applications. Time reversal is a key step in seismic interferometry, reverse time migration and full waveform inversion. The plane-wave transform, also known as the tau-p transform or slant-stack, can separate waves based on their ray parameters or their emergence angles at the surface. I propose a new approach to retrieve virtual full-wave seismic responses from crosscorrelating recorded seismic data in the plane-wave domain. Unlike a traditional approach where the correlogram is obtained from crosscorrelating recorded data, which contains the full range of ray parameters, this method directly chooses common ray parameters to cancel overlapping ray paths. Thus, it can sometime avoid spurious arrivals when the acquisition requirement of seismic interferometry is not strictly met. I demonstrate the method with synthetic examples and an ocean bottom seismometer data example. I show a multi-scale application of plane-wave based full waveform inversion (FWI) with the aid of frequency domain forward modeling. FWI uses the two-way wave-equation to produce high-resolution velocity models for seismic imaging. This technique is implemented by an adjoint-state approach, which viii involves a time-reversal propagation of the residual wavefield at receivers, similar to seismic interferometry. With a plane-wave transformed gather, we can decompose the data by ray parameters and iteratively update the velocity model with selected ray parameters. This encoding approach can significantly reduce the number of shots and receivers required in gradient and Hessian calculations. Borrowing the idea of minimizing different data residual norms in FWI, I study the effect of different scaling methods to the receiver wavefield in the reverse time migration. I show that this type of scaling is able to significantly suppress outliers compared to conventional algorithms. I also show that scaling by its absolute norm generally produces better results than other approaches. I propose a robust stochastic time-lapse seismic inversion strategy with an application of monitoring Cranfield CO2 injection site. This workflow involves two steps. The first step is the baseline inversion using a hybrid starting model that combines a fractal prior and the low-frequency prior from well log data. The second step is to use a double-difference inversion scheme to focus on the local areas where time-lapse changes have occurred. Synthetic data and field data show the effectiveness of this method. / text

No description available.

In this study, seismic interferometry is used to analyze dynamic processes in the Earth’s shallow subsurface caused by environmental processes and ground shaking. In the first part of the thesis, the feasibility of a passive monitoring with ambient seismic noise at the pilot site for CO2 injection in Ketzin is investigated. Monitoring the expansion of the CO2 plume is essential for the characterization of the reservoir as well as the detection of potential leakage. From June 2008 until August 2013, more than 67000 tons of CO2 were injected into a saline aquifer at a depth of about 650 m. Passive seismic data recorded at a seismic network around the injection site was cross-correlated in a frequency range of 0.5-4.5 Hz over a period of 4 years. The frequency band of 0.5-0.9 Hz, in which surface waves exhibit a high sensitivity at the depth of the reservoir, is not suitable for monitoring purposes as it is only weakly excited. In a frequency range of 1.5-3 Hz, periodic velocity variations with a period of approximately one year are found that cannot be caused by the CO2 injection. The prominent propagation direction of the noise wave field indicates a wind farm as the dominant source providing the temporally stable noise field. This spacial stability excludes variations of the noise source distribution as a spurious cause of velocity variations. Based on an amplitude decrease associated with time windows towards later parts of the coda, the variations must be generated in the shallow subsurface. A comparison to groundwater level data reveals a direct correlation between depth of the groundwater level and the seismic velocity. The influence of ground frost on the seismic velocities is documented by a sharp increase of velocity when the maximum daily temperature stays below 0 C. Although the observed periodic changes and the changes due to ground frost affect only the shallow subsurface, they mask potential signals of material changes from the reservoir depths. To investigate temporal seismic velocity changes due to earthquake-related processes and environmental forcing in northern Chile, 8 years of ambient seismic noise recorded by the Integrated Plate Boundary Observatory Chile (IPOC) are analyzed. By autocorrelating the ambient seismic noise field, approximations of the Green’s functions are retrieved and velocity changes are measured with Coda Wave Interferometry. At station PATCX, seasonal changes of seismic velocity caused by thermal stress as well as transient velocity reductions are observed in the frequency range of 4-6 Hz. Sudden velocity drops occur at times of mostly earthquake-induced ground describing the seismic velocity variations based on continuous observations of the local ground acceleration. The model assumes that not only the shaking of large earthquakes causes velocity drops, but any small vibrations continuously induce minor velocity variations that are immediately compensated by healing in the steady state. The shaking effect is accumulated over time and best described by the integrated envelope of the ground acceleration over one day, which is the temporal resolution of the velocity measurements. In the model, the amplitude of the velocity reduction as well as the recovery time are proportional to the strength of the excitation. The increase of coseismic velocity change and recovery time with increasing excitation is confirmed by laboratory tests with ultrasound. Despite having only two free scaling parameters, the model fits the data of the shaking-induced velocity variation in remarkable detail. Additionally, a linear trend is observed that might be related to a recovery process from one or more earthquakes before the measurement period. A clear relationship between ground shaking and induced velocity reductions is not visible at other stations. The outstanding sensitivity of PATCX to ground shaking and thermal stress can be attributed to the special geological setting of the station, where the subsurface material consists of a relatively loose conglomerate with high pore volume leading to stronger nonlinearity compared to the other IPOC stations. / In dieser Studie werden mit Hilfe von seismischer Interferometrie kleinste dynamische Prozesse in der Erdkruste beobachtet, welche beispielsweise durch umweltbedingte oder anthropogene Einflüsse sowie Bodenerschütterungen hervorgerufen werden können. Im ersten Teil der Arbeit werden Änderungen in der seismischen Geschwindigkeit am Pilotstandort für CO2-Speicherung in Ketzin untersucht. In einer Tiefe von 650m wurden dort zwischen Juni 2008 und August 2013 über 67000 Tonnen CO2 eingelagert. In einem Frequenzbereich vom 0,05-4,5 Hz wurden Kreuzkorrelationen des seismischen Hintergrundrauschens an einem kleinräumigen Netzwerk über einen Zeitraum von 4 Jahren berechnet. Der Frequenzbereich zwischen 0,5 und 0,9 Hz weist eine hohe Sensitivität von Oberflächenwellen in der Tiefe des Reservoirs auf, ist aber nur sehr schwach angeregt und eignet sich deswegen nicht für die Analyse. In einem Frequenzbereich von 1,5-3 Hz zeigen sich periodische Geschwindigkeitsänderungen mit einer Periode von einem Jahr, welche nicht durch die Einlagerung von CO2 erzeugt werden können. Eine Analyse des seismischen Hintergrundrauschens zeigt, dass dieses über den gesamten Zeitraum hinweg hauptsächlich aus der Richtung eines Windparks kommt. Durch die Stabilität des Wellenfeldes können Änderungen in den Quellpositionen, welche sich in scheinbaren Geschwindigkeitsänderungen zeigen können, ausgeschlossen werden. Eine Amplitudenabnahme der Geschwindigkeitsänderungen hin zu späteren Zeitfenstern in der Coda lässt auf oberflächennahe Prozesse als Ursache schließen. Ein Vergleich zwischen den jährlichen Geschwindigkeitsänderungen mit Schwankungen im Grundwasserspiegel zeigt eine direkte Korrelation. Ein sprunghafter Anstieg in der Geschwindigkeit zeigt sich im Winter, wenn die Tageshöchsttemperaturen unter den Gefrierpunkt sinken und der Boden zufriert. Obwohl Bodenfrost und Änderungen im Grundwasserspiegel nur einen sehr oberflächennahen Bereich betreffen, so überdecken sie dennoch mögliche Signale durch die Einlagerung von CO2. Im zweiten Teil der Arbeit werden Geschwindigkeitsänderungen in Nordchile untersucht, welche durch erdbebeninduzierte Prozesse und umweltbedingte Einflüsse hervorgerufen werden. Dazu wurden über einen Zeitraum von 8 Jahren Autokorrelationen des seismischen Hintergrundrauschens des IPOC Netzwerkes (Integrated Plate Boundary Observatory Chile) berechnet und mit seismischer Interferometrie ausgewertet. An der Station PATCX können in einem Frequenzbereich von 4-6 Hz periodische Geschwindigkeitsänderungen beobachet werden, welche durch thermisch induzierte Dehnung hervorgerufen werden. Außerdem treten transiente Geschwindigkeitsabnamen nach Bodenerschütterungen auf, welche hauptsächlich von Erdbeben verursacht werden. Die seismische Geschwindigkeit kehrt daraufhin langsam wieder auf ihr vorheriges Niveau zurück. Für die Geschwindigkeitsänderungen wurde ein empirisches Modell entwickelt, welches auf Messungen der lokalen Bodenerschütterung basiert. Dabei wird angenommen, dass nicht nur große erdbebeninduzierte, sondern auch kleinste Bodenerschütterungen einen Abfall der Geschwindigkeit erzeugen, welche wiederum innerhalb kürzester Zeit durch Heilung in den Gleichgewichtszustand zurückkehrt. Dabei summieren sich die Effekte durch die Bodenerschütterungen mit der Zeit auf und werden am besten mit dem Integral der lokalen Bodenbeschleunigung über die Messwerte eines Tages beschrieben. Die Diskretisierung von einem Tag entspricht der zeitlichen Auflösung in der Messung der Geschwindigkeitsänderungen. Sowohl die Amplitude der Geschwindigkeitsabnahme als auch die Zeit bis der Gleichgewichtszustand wieder erreicht ist (Heilungszeit) werden im Modell als proportinal zur Größe der Anregung angenommen. Eine Korrelation der Heilungszeit und der Amplitude der koseismischen Geschwindigkeitsabnahme mit der Größe der Anregung konnte mit Hilfe von Laboruntersuchungen mit Ultraschall bestätigt werden. Mit nur zwei Parametern beschreibt das Modell die transienten Geschwindigkeitsänderungen in bemerkenswerter Genauigkeit. Desweiteren beinhaltet das Modell einen linearen Verlauf in den Geschwindigkeitsänderungen, welcher vermutlich durch einen Heilungsprozess hervorgerufen wird, der auf ein oder mehrere Erdbeben vor dem Messzeitraum folgte. Eine Beziehung zwischen Bodenerschütterung und Geschwindigkeitsänderung ist an anderen Stationen des IPOC Netzwerkes nicht erkennbar. Die herausragende Sensitivität von PATCX im Hinblick auf Bodenerschütterung und thermische Dehnung kann den speziellen geologischen Gegebenheiten an der Station zugeschrieben werden. Bei dem dort vorliegenden Material handelt es sich um ein relativ loses Konglomerat mit großem Porenvolumen, welches ein starkes nichtlineares Verhalten aufweist, was an anderen IPOC Stationen nicht zu erwarten ist.