The reflection seismic method is the most frequently used exploration method for imaging and monitoring subsurface structures with high resolution. It has proven its qualities from the scale of regional seismology to the scale of near-surface applications that look just a few meters below the surface. The reflection method uses controlled active sources at known positions to give rise to reflections recorded at known receiver positions. The reflections’ two-wave travel time is used to extract desired information about and image the subsurface structures. When active sources are unavailable or undesired, one can retrieve body-wave reflections from application of seismic interferometry (SI) to sources of opportunity—quakes, tremors, ambient noise, or even man-made sources not connected to the exploration campaign. We show examples of imaging of subsurface structures using reflections retrieved from quakes and ambient noise. We apply SI by autocorrelation to global earthquake to image seismic and aseismic parts of the Nazca plate and the Moho at these places, SI by multidimensional deconvolution to P-wave coda from local earthquakes to image the Moho and the crust at the same places, and SI by autocorrelation to deep moonquakes to image the lunar Moho and to ambient noise to monitor CO2 sequestration. @en

Obtaining new seismic responses from existing recordings is generally referred to as seismic interferometry (SI). Conventionally, the SI responses are retrieved by simple crosscorrelation of recordings made by separate receivers: one of the receivers acts as a ‘virtual source’ whose response is retrieved at the other receivers.When SI is applied to recordings of ambient seismic noise, mostly surface waves are retrieved. The newly retrieved surface wave responses can be used to extract receiver-receiver phase velocities. These phase velocities often serve as input parameters for tomographic inverse problems. Another application of SI exploits the tempo- ral stability of the multiply scattered arrivals of the newly retrieved surface wave responses. Temporal variations in the stability and/or arrival time of these multiply scattered arrivals can often be linked to temporally varying parameters such as hydrocarbon production and precip- itation. For all applications, however, the accuracy of the retrieved responses is paramount. Correct response retrieval relies on a uniform illumination of the receivers: irregularities in the illumination pattern degrade the accuracy of the newly retrieved responses. In practice, the illumination pattern is often far from uniform. In that case, simple crosscorrelation of separate receiver recordings only yields an estimate of the actual, correct virtual-source response. Re- formulating the theory underlying SI by crosscorrelation as a multidimensional deconvolution (MDD) process, allows this estimate to be improved. SI by MDD corrects for the non-uniform illumination pattern by means of a so-called point-spread function (PSF), which captures the irregularities in the illumination pattern. Deconvolution by this PSF removes the imprint of the irregularities on the responses obtained through simple crosscorrelation. We apply SI by MDD to surface wave data recorded by theMalargue seismic array in western Argentina. The aperture of the array is approximately 60 km and it is located on a plateau just east of the Andean mountain range. The array has a T-shape: the receivers along one of the two lines act as virtual sources whose responses are recorded by the receivers along the other (perpendicular) line.We select time windows dominated by surface wave noise travelling in a favourable direction, that is, traversing the line of virtual sources before arriving at the receivers at which we aim to retrieve the virtual-source responses. These time windows are selected through a frequency-dependent slowness analysis along the two receiver lines. From the selected time windows, estimates of virtual-source responses are retrieved by means of crosscorrelations. Similarly, crosscorrelations between the positions of the virtual sources are computed to build the PSF. We use the PSF to deconvolve the effect of illumination irregularities and the source function from the virtual-source responses retrieved by crosscorrelation. The combined effect of time-window selection and MDD results in more accurate and temporally stable surface wave responses.@en

Torfajökull is the largest silicic volcanic centre in Iceland lying at the intersection of the rift zone (MidAtlantic Ridge) and the transform zone that connects to Reykjanes peninsula. It erupts infrequently,with only two eruptions in the last 1200 years, the latest of which was over 5 centuries ago. Yet, itsactive tectonic setting, persistent high and low frequency seismicity, deformation and geothermalactivity within its large caldera (18x12 km diameter) indicate the continued presence of a long-lasting magma chamber. Here we speculate on possible geometry, size and depth of the Torfajökullmagma chamber by using radar interferometry (InSAR) and seismic interferometry (SI).Using InSAR time series analysis we detect a surface subsidence pattern at rates of up to ~13 mm yr-1in the SW region of Torfajökull ́s caldera, on-going since at least 1993. The subsidence rate isconstant in time, and perhaps due to a cooling magma chamber. The data can be fit reasonably wellusing a model of a NE-SW oriented spheroidal body at ~5 km depth. As the deflating area correlatesspatially with the area of geothermal activity, deflation may also be the surface response due to anactive hydrothermal circulation.To gain more insight into the geometry of Torfajökull’s magmatic system and rock properties of thesubsurface, we apply ambient noise seismic interferometry (SI) by cross-correlation of ambient noise.With this technique we can detect velocity variations, which can correspond to the edges of dikes ormolten magma bodies. Our tomographic results give reliable results of velocity variations within adepth range of 2 km to 7.5 km. We find high velocity zones that we interpret as old dike intrusions.Low velocity anomalies (>5%), which usually indicate the presence of warmer material, are locatedon the southeast and southwest part of the volcano, outside the volcano caldera.Finally we compare both InSAR and SI results. The hypothesis of a magma chamber under thesubsidence area detected by InSAR does not seem to fit the tomographic results, as the expectededges of a magma body modelled by InSAR are not clearly identified by the SI results. If there is anestablished magma chamber within Torfajökull caldera this is likely to be bellow 7km depth.@en

An areal distribution of sensors can be used for estimating the direction of incoming waves through beamforming. Beamforming may be implemented as a phase-shifting and stacking of data recorded on the different sensors (i.e., conventional beamforming). Alternatively, beamforming can be applied to cross-correlations between the waveforms on the different sensors. We derive a kernel for beamforming cross-correlated data and call it cross-correlation beamforming (CCBF). We point out that CCBF has slightly better resolution and aliasing characteristics than conventional beamforming. When auto-correlations are added to CCBF, the array response functions are the same as for conventional beamforming. We show numerically that CCBF is more resilient to non-coherent noise. Furthermore, we illustrate that with CCBF individual receiver-pairs can be removed to improve mapping to the slowness domain. An additional flexibility of CCBF is that cross-correlations can be time-windowed prior to beamforming, e.g., to remove the directionality of a scattered wavefield. The observations on synthetic data are confirmed with field data from the SPITS array (Svalbard). Both when beamforming an earthquake arrival and when beamforming ambient noise, CCBF focuses more of the energy to a central beam. Overall, the main advantage of CCBF is noise suppression and its flexibility to remove station pairs that deteriorate the signal-related beampower.@en

Tomographic studies based on passive seismic measurements have proven to be a powerful tool to image the subsurface. This especially holds in areas like Iceland, where the microseism coverage arriving from the ocean is excellent. In this study, we apply Ambient Noise Seismic Interferometry (ANSI) to generate a tomographic image of Rayleigh-waves velocity anomalies to further invert for S-wave anomalies at two Icelandic locations. We derive a tomographic image over Reykjanes Peninsula geothermal system using 30 Broad-Band (BB) stations deployed under the IMAGE (Integrated Methods for Advanced Geothermal Exploration) project framework and operated for approximately one year and a half. In the other case study, we derive a tomographic image of Torfajökull volcano using 23 BB seismometers that recorded ambient noise for 100 days. The later data were acquired in 2005 by Cambridge University. We retrieve the surface-wave part of the Green’s functions by cross-correlation between station pairs and consecutive stacking of the cross-correlations to obtain coherent ballistic surface waves (BSW). We pick the arrival times of the BSW, which are the input for the tomographic analysis. Both datasets show remarkably high signal-to-noise ratio of surface-wave arrivals between 0.1 and 0.5 Hz, even with only 100 days of recorded ambient noise. A beamforming analysis indicates a broad azimuthal coverage with persistent ambient noise arrivals within three azimuthal quadrants - between 90 and 360 degrees. The highly coherent surface-wave retrieval and the wide azimuthal coverage of the microseisms explain the success of ANSI techniques in Iceland. For the tomographic inversion, we use a Tikhonov and a statistical regularisation to invert the ballistic surface-wave time-arrival to 3D frequency-dependent velocity variations. After further inversion to S-wave velocity variations, we detect low- and high-velocity anomalies with changes between -15% and 15% from an estimated average velocity, we interpret these anomalies as possible old dyke intrusions and heat sources.@en

We discuss Rodney Calvert's work on the Virtual Source method in the context of seismic interferometry. Moreover, we present a systematic analysis of seismic interferometry by cross-correlation versus multi-dimensional deconvolution and we discuss applications of both approaches.@en

Magmatic plumbing systems beneath active and moderately active volcanoes are often poorly constrained. A better knowledge of the shape, size and location of the magma bodies would enable us to better predict magma movements preceding an eruption. Surface displacements estimated from radar interferometry (InSAR) can be used in geophysical modelling to constrain location, geometry and pressure changes in magma systems. However, the resolution of the inferred magma chamber is typically poor. More insight into the location and geometry of magmatic systems can be gained using active-source reflection seismic surveys, which allow detection of velocity contrasts at the edges of the magma bodies. The drawback of this technique is that controlled-source surveys are expensive. As an alternative, seismic interferometry (SI) uses cross-correlation of natural signals to generate new seismic records that simulate active sources. Under the premise that both seismic and radar observations would help to narrow down the location, shape and size of a magma system, we present the first results of combined radar and seismic interferometric processing over Torfajökull volcano. Torfajökull is located in the neovolcanic zone, in the south of Iceland. The volcano is characterized by intense thermal manifestations (hot springs forming ground steaming and fumaroles) with higher activity nearby the faults of the active NE-SW fissure swarm. Torfajökull erupts infrequently, with only two eruptions in the last 1200 years, the latest of which was over 5 centuries ago. However, ongoing seismicity, deformation and geothermal activity indicate the continued presence of a long-lasting magma chamber. Although historical eruptions have been relatively small, the large caldera (18x12 km diameter) and high geothermal activity within the caldera is evidence of a massive eruption in the past, and the potential for a further eruption of similar size is unknown. We applied InSAR time series analysis to data acquired by Envisat over the Torfajökull region along 6 different tracks (3 tracks in ascending and 3 tracks in descending mode). The estimated velocity maps show subsidence beneath the SW part of the caldera. This subsidence has been on-going since at least 1993, with rates of up to ~13 mm/yr. This has been interpreted as a cooling magma chamber, shrinking at linear rates. To obtain a rough approximation of the geometry and location of the source of subsidence we ran a forward model. The model suggest an ellipsoidal magma chamber within the volcano caldera with a NE-SW orientation and at ~5km depth. For the SI processing we use two types of natural signals: microseisms and local earthquakes. We use seismic data acquired in 2005 at 30 stations sparsely distributed around the Torfajökull area. Using microseisms, we divide the noise, recorded at two stations in portions of 1h, cross-correlate the corresponding portions and then sum the correlated results. The result is a retrieved surface-wave part of the Green’s function between the two stations. This is repeated between all stations. Careful assessment of the quality of the retrieved Green’s functions for small time windows allows analysis of the microseism noise. The results show that the microseisms are dominant in the NW-SE direction and the resulting retrieved surface waves propagate at ~3 km/s in the double-frequency microseism band. The retrieved surface waves between the stations will be used in tomographic inversion that will allow derivation of the 3D S-wave velocity distribution in the subsurface. We will then use these results to better constrain our magma source model, currently constrained only by InSAR. @en

Seismic interferometry (SI) retrieves virtual seismic signals from measurements at two receivers from surrounding sources. Studies have demonstrated that SI can image subsurface reflectivity. Claerbout (1968) showed that the reflection response can be obtained by autocorrelating the transmission response assuming a 1-D acoustic medium. Migration techniques using primary and multiple reflections in the autocorrelogram effectively imaged the subsurface structure (Schuster et al., 2003; Yu et al., 2003). This research aims to contribute to the knowledge of the subsurface structure at Planchón-Peteroa Volcano Complex (PPVC) by using SI. Inspired by the theory and applications in Wapenaar (2003) and Ruigrok and Wapenaar (2012), this work applies SI to fracture seismicity originated at PPVC or in active geologic faults located nearby this volcanic complex. Autocorrelating an extracted time window for selected events, zero-offset reflection responses were retrieved for each station. This response is further used to image shallow subsurface reflectors underneath each station. This application uses seismic data recorded by stations deployed in Argentina and Chile. The Argentine data was recorded by a temporary array of six 2-Hz 3-component stations deployed in 2012 on the eastern flank of the volcano. The Chilean data is provided by OVDAS-SERNAGEOMIN (South Andes Volcanic Observatory, Chile). OVDAS has a permanent array of six 30-seconds 3-component stations. Three of these stations recorded data simultaneously with the Argentine stations and were used for this research. Very local seismicity is required for this study, therefore event selection procedure is of great importance. The two arrays were initially used independently to locate fracture events (Casas, 2014; RAV-SERNAGEOMIN, 2012). In order to obtain accurate locations, the two datasets were used together to relocate the detected events (Casas et al., 2016; Olivera Craig et al., 2016). This preprocessing step enhances the resolution of the subsurface images obtained by application of SI at PPVC since appropriate wave paths are selected for processing @en

Obtaining new seismic responses from existing recordings is generally referred to as seismic interferometry (SI). Conventionally, these seismic interferometric responses are retrieved by simple crosscorrelation of recordings made by separate receivers: a first receiver acts as a 'virtual source' whose response is retrieved at the other receivers. When surface waves are retrieved, the newly retrieved responses can be used to extract receiver-receiver phase velocities. These phase velocities often serve as input parameters for tomographic inverse problems. Another application of SI exploits the temporal stability of the multiply scattered arrivals (the coda). For all applications, however, the accuracy of the retrieved responses is paramount. In practice, this accuracy is often degraded by irregularities in the illumination pattern: correct response retrieval relies on a uniform illumination of the receivers. Reformulating the theory underlying seismic interferometry by crosscorrelation as a multidimensional deconvolution (MDD) process, allows for correction of these non-uniform illumination patterns by means of a so-called point-spread function (PSF). We apply SI by MDD to surface-wave data recorded by the Malargüe seismic array in western Argentina. The aperture of the array is approximately 60 km and it is located on a plateau just east of the Andean mountain range. The array has a T-shape, which makes it very well suited for the application of SI by MDD. We select time windows dominated by surface-wave noise traveling in a favorable direction, that is, traversing the line of virtual sources before arriving at the receivers at which we aim to retrieve the virtual-source responses. These time windows are selected based upon the slownesses along the two receiver lines. From the selected time windows, virtual-source responses are retrieved by computation of ensemble-averaged crosscorrelations. Similarly, ensemble-averaged crosscorrelations between the positions of the virtual sources are computed: the PSF. We use the PSF to deconvolve the effect of illumination irregularities and the source function from the virtual-source responses retrieved by crosscorrelation. The combined effect of time-window selection and MDD results in more accurate and temporally stable surface-wave responses.@en

Seismic interferometry (SI) studies the interference phenomenon between pairs of signals in order to obtain information from the differences between them. SI is now regularly used in exploration and global seismology with active and/or passive sources, i.e., artificial sources (dynamite, vibroseis, sledge hammer, etc.) or natural sources (earthquakes, anthropogenic noise, ocean microseisms, etc.). SI allows one to extract subsurface information from complicated or random wavefields.This research aims to contribute to the knowledge of the subsurface structure at Planchón-Peteroa Volcano Complex (PPVC) by using SI technique. Inspired by the theory and applications in Wapenaar (2003) and Ruigrok and Wapenaar (2012), this work applies SI to fracture seismicity originated at PPVC or in active geologic faults located nearby this volcanic complex. Applying autocorrelation to a selected time window at each event, zero-offset reflection responses were obtained for each station. This response can be used to determine the location of shallow subsurface reflectors underneath each station.This application uses seismic data recorded by stations deployed in Argentina and Chile. The Argentine data was recorded by an array of six 2-Hz 3-component stations on the eastern flank of the volcano, deployed during the MalARRgue project in 2012. The Chile data is provided by OVDAS-SERNAGEOMIN (South Andes Volcanic Observatory, Chile). OVDAS has six 3-component 30-seconds stations located on the western flank of the volcano; these stations overlap in the same time period as the Argentine data.Events had been identified and located independently by the arrays deployed in each of the flanks (Casas, 2014; RAV SERNAGEOMIN, 2012). In order to obtain accurate locations of the detected events, the two datasets were used together to relocate them. This result constitutes a necessity for enhancing the resolution of subsurface images obtained by application of SI at PPVC.Preliminary results of this research will be presented.@en

Generating new seismic responses from existing recordings is generally referred to as seismic interferometry (SI). Conventially, the new responses are retrieved by simple crosscorrelation of recordings made by separate receivers: a first receiver acts as `virtual source' whose response is retrieved at the other receivers. The newly retrieved responses can be used to extract receiver-receiver phase velocities, which often serve as input parameter for tomographic inverse problems, or which can be linked to temporally varying parameters such as hydrocarbon production and precipitation. For all applications, however, the accuracy of the retrieved responses is of great importance. In practice, this accuracy is often degraded by irregularities in the illumination pattern: correct response retrieval relies on a uniform illumination of the receivers. Reformulating the theory underlying seismic interferometry by crosscorrelation as a multidimensional deconvolution (MDD) process, allows for correction of these non-uniform illumination patterns by means of a so-called point-spread function (PSF). We apply SI by MDD to surface-wave data recorded by the Malargüe seismic array in western Argentina. The aperture of the array is approximately 60 km and it is located on a plateau just east of the Andean mountain range. The array has a T-shape: the receivers along one of the two lines act as virtual sources whose responses are retrieved by the receivers along the other (perpendicular) line of receivers. Because SI by MDD relies on one-way wavefields, we select time windows dominated by surface-wave noise traveling in a favorable direction, that is, traversing the line of virtual sources before arriving at the receivers at which we aim to reconstruct the virtual-source responses. These time windows are selected through a frequency-dependent slowness analysis along the two receiver lines. From the selected time windows, virtual-source responses are retrieved by computation of ensemble-averaged crosscorrelations. Similarly, ensemble-averaged crosscorrelations between virtual sources are computed: the point-spread function. We use the PSF to deconvolve the effect of illumination irregularities and the source function from the virtual-source responses. The combined effect of time-window selection and MDD results in more accurate surface-wave responses.@en

Seismic interferometry refers to the principle of generating new responses. These new responses are conventionally obtained by simple crosscorrelation of recordings made by separate receivers: a first receiver acts as ‘virtual source’ whose response is retrieved at the other receivers. The recorded wavefields may be passive (e.g. seismic noise) or active (e.g. in an industrial context). The newly retrieved responses can be used to extract receiver-receiver phase velocities, which often serve as input parameter for tomographic inverse problems. More recently, the coda of the newly retrieved responses have been found to correlate with temporally varying parameters such as hydrocarbon production and precipitation. For all applications, however, the accuracy of the retrieved responses is of great importance. Irregularities in the illumination patttern often degrade this accuracy: correct response retrieval relies on a uniform illumination of the receivers. Reformulating the theory underlying seismic interferometry by crosscorrelation as a multidimensional deconvolution (MDD) process, allows the removal of the imprint of the illumination pattern on the retrieved responses by means of a so-called point-spread function (PSF). We use a seismic array in Malargüe, Argentina, to assess the feasibility of SI by MDD on ambient seismic noise recordings. The array, which has an aperture of approximately 60 km, is located just east of the Andean mountain range. The shape of the array lends itself well for the application of SI by MDD: its T-shape allows the construction of a PSF along one of the two receiver lines. These receivers act as the virtual sources and their responses are retrieved by the receivers along the other (perpendicular) line of receivers. A frequency-dependent analysis of the slowness along both lines allows us to select time windows during which most ambient seismic surface waves propagate in a favorable direction, that is, traversing the line of virtual sources prior to arrival at the receivers at which we aim to reconstruct the responses. During these time windows the wavefield therefore fulfills the assumption underlying SI by MDD that there is only energy propagating from the virtual sources to the receivers. Both the PSF and conventional crosscorrelations between the virtual sources and the receivers are computed from the selected time windows. We find that multidimensionally deconvolving the virtual-source responses by the point-spread function improves the responses’ accuracy. @en

Seismic interferometry is an effective tool to retrieve surface waves between two receiver stations by cross-correlating ambient background noise over sufficiently long recording times. This method assumes an azimuthally uniform distribution of noise sources. Unfortunately this assumption is not always fulfilled in practice. If noise sources are located on one side of a receiver array only, surface waves can also be retrieved by multi-dimensional deconvolution of passive records. We show how this method can effectively correct for azimuthal variations in the noise source distribution. We do not take backscattering of the surface waves into account, but this can be overcome if wavefield decomposition is incorporated.@en