P-wave Traveltime Research Articles

Imaging Complex Structures of the Los Angeles Basin via Adjoint-State Travel-Time Tomography

ABSTRACT In this study, we present high-resolution seismic velocity models for the Los Angeles basin (LAB) and its adjacent area using adjoint-state travel-time tomography, fitting an extensive database of P- and S-wave travel times accumulated from 1980 to 2021. We select 151,193 first P-wave travel times and 149,997 first S-wave travel times from local earthquakes archived in the Southern California Earthquake Data Center to determine the velocity models, with earthquake locations updated at each iteration. With seismic stations spaced more than 3.5 km apart, our dataset has limited resolution in the uppermost 1–2 km. However, starting from three different initial models, our VP models, which are optimally imaged between 3 and 15 km, show similar velocity heterogeneity and provide a better fit to the observed first travel-time data compared to the Community Velocity Model-Harvard 15.1.0 and Community Velocity Model-Southern California Earthquake Center 4.26. Our models provide a detailed delineation of the subsurface structure beneath the LAB, revealing significant velocity variations across active faults, a 10-km-thick sequence of sedimentary rocks within the basin, and a distinct basin margin marked by transitions from low to high-velocities. In addition, these models highlight basement structures with elevated VP and VS located at depths of 9 to 12 km and beyond. Specifically, beneath the northeastern part of the basin, the models demonstrate improved accuracy and reliability in reflecting the linear relationship between VP and VS in mafic rocks. The accurate delineation of the basin’s structure provided by our models could also offer robust constraints for seismic response modeling and seismic hazard assessment in the region.

Bimodality of P-wave travel time residuals found in the CTBTO seismological data (REB-bulletins)

Bimodality of P-wave travel time residuals found in the CTBTO seismological data (REB-bulletins)

Seismic structure of the Balmuccia Peridotite from a high-resolution refraction and reflection survey

SUMMARY In anticipation of a forthcoming scientific deep drilling initiative within the Western Alps near Balmuccia, Italy, a high-resolution seismic survey is performed at the proposed drill site. This site is situated within the Ivrea Verbano Zone (IVZ), characterized by lower crustal materials and fragments of upper-mantle rocks exposed adjacent to the Insubric Line. The 2-km-long seismic survey crosses an isolated km-scale outcrop of peridotite near the town of Balmuccia. Applying P-wave traveltime tomography, a substantial contrast in seismic velocities is identified, with velocities in the range of 1–8 km s−1. The peridotite displays velocities ranging from 6 to 8 km s−1. The higher velocities near 8 km s−1 are consistent with laboratory measurements on small-scale samples, while the low-velocity areas within the peridotite body reflect the influence of fractures and faults. The mean velocity derived for the peridotite body is ca. 7 km s−1. The reflection seismic analysis reveals subvertical reflectors positioned at the peridotite boundaries mapped at the surface, converging at a depth of ca. 0.175 km b.s.l. which images a lens-like structure for the peridotite body. However, the area beneath the imaged lens and the deeper Ivrea Geophysical Body (IGB) suggested by earlier studies is not well imaged, which leaves room for other interpretations regarding the relationship of these two bodies. Prior geophysical investigations provide only approximate depth estimates for the top of the IGB, spanning between 1–3 km depth b.s.l. Although the reflection data do not exhibit a series of continuous reflectors beneath the peridotite, a prominent reflection at ca. 1.3 km depth may indicate the top of the IGB.

Seismic Characterization of Subsurface Structures at Rock Valley, Nevada

ABSTRACT The Source Physics Experiment (SPE) aims at improving explosion monitoring techniques by investigating source characteristics of chemical explosions in geologic formations. One of the critical tasks in Rock Valley Direct Comparison (RV/DC), SPE phase III, is to prepare for the main experiment by characterizing the subsurface structures at the test site. Based on the seismic data acquired during an accelerated-weight-drop (AWD) seismic survey at Rock Valley, we first pick the P-wave first-arrival travel times and derive a P-wave velocity model using the adjoint-state first-arrival travel-time tomography. We then apply reverse-time migration to the processed seismic data and obtain high-resolution images of the subsurface structures along the two main survey lines. Our migration results show several reflectors corresponding to major geologic formation boundaries. We employ a multitask machine learning model to enhance the reverse-time migration images and identify faults from these images. We find that our automatically picked faults correlate well with the locations of known faults in the region in addition to many geologically undetected faults. Our subsurface characterization results refine our understanding of the geology in this region and provide valuable velocity and structural information for RV/DC geologic model building and fault identification.

Tracking Seismic Velocity Perturbations at Ridgecrest Using Ballistic Correlation Functions

Abstract We present results based on data of a dense nodal array composed of 147 stations, deployed in 2022 near the epicenter of the 2019 Mw 7.1 Ridgecrest earthquake to investigate characteristics of the seismic wavefields. Through array analyses, we identified two primary components. First, we observed far-field P waves dominating the 0.5–1.2 Hz frequency range, which are likely primarily generated by wind-driven oceanic swell activity. Second, we detected near-field body waves resulting from anthropogenic activities in the frequency range 2–8 Hz. We examined noise correlation functions derived from data of the dense deployment and regional stations to explore fault-zone seismic velocity changes using ballistic arrivals, with a focus on velocity perturbation shortly before and after the Ridgecrest earthquake sequence. Our findings exhibit distinct behavior compared to results obtained through standard coda-wave interferometry. Particularly, we observed a decrease in P-wave travel time on certain station pairs prior to the 2019 earthquake sequence. Supported by detailed investigation of the local seismic wavefields, we interpret the decreasing P-wave travel time as likely caused by a velocity increase away from the fault, possibly related to fluid migration. However, additional information is necessary to verify this hypothesis.

Hydrologic Impacts of a Strike-Slip Fault Zone: Insights from Joint 3D Body-Wave Tomography of Rock Valley

ABSTRACT The Rock Valley fault zone (RVFZ), an intraplate strike-slip fault zone in the southern Nevada National Security Site (NNSS), hosted a series of very shallow (<3 km) earthquakes in 1993. The RVFZ may also have hydrological significance within the NNSS, potentially playing a role in regional groundwater flow, but there is a lack of local hydrological data. In the Spring of 2021, we collected active-source accelerated weight drop seismic data over part of the RVFZ to better characterize the shallow subsurface. We manually picked ∼17,000 P-wave travel times and over 14,000 S-wave travel times, which were inverted for P-wave velocity (VP), S-wave velocity (VS), and VP/VS ratio in a 3D joint tomographic inversion scheme. Seismic velocities are imaged as deep as ∼700 m in areas and generally align with geologic and structural expectations. VP and VS are relatively reduced near mapped and inferred faults, with the most prominent lower VP and VS zone around the densest collection of faults. We image VP/VS ratios ranging from ∼1.5 to ∼2.4, the extremes of which occur at a depth of ∼100 m and are juxtaposed across a fault. One possible interpretation of the imaged seismic velocities is enhanced fault damage near the densest collection of faults with relatively higher porosity and/or crack density at ∼100 m depth, with patches of semiperched groundwater present in the sedimentary rock in higher VP/VS areas and drier rock in lower VP/VS areas. A relatively higher VP/VS area beneath the densest faults persists at depth, which suggests percolation of groundwater via the fault damage zone to the regionally connected lower carbonate aquifer. Potentially, the presence and movement of groundwater may have played a role in the 1993 earthquake aftershocks.

Emulation of seismic-phase traveltimes with machine learning

SUMMARY We present a machine learning (ML) method for emulating seismic-phase traveltimes that are computed using a global-scale 3-D earth model and physics-based ray tracing. Accurate traveltime predictions based on 3-D earth models are known to reduce the bias of event location estimates, increase our ability to assign phase labels to seismic detections and associate detections to events. However, practical use of 3-D models is challenged by slow computational speed and the unwieldiness of pre-computed lookup tables that are often large and have prescribed computational grids. In this work, we train a ML emulator using pre-computed traveltimes, resulting in a compact and computationally fast way to approximate traveltimes that are based on a 3-D earth model. Our model is trained using approximately 850 million P-wave traveltimes that are based on the global LLNL-G3D-JPS model, which was developed for more accurate event location. The training-set consists of traveltimes between 10 393 global seismic stations and randomly sampled event locations that provide a prescribed, distance-dependent geographic sample density for each station. Prediction accuracy is dependent on event-station distance and whether the station was included in the training set. For stations included in the training set the mean absolute deviation (MAD) of the difference between traveltimes computed using ray tracing through the 3-D model and the ML emulator for local, regional, and teleseismic distances are 0.090, 0.125 and 0.121 s, respectively. For tested station locations not included in the training set, MAD values for the three distance ranges increase to 0.173, 0.219 and 0.210 s, respectively. Empirical traveltime residuals for a global reference data are indistinguishable when ML emulation or the 3-D model is used to compute traveltimes. This result holds regardless of whether the recording station is used in ML training or not.

Synthesizing Missing Travel Time of P-Wave and S-Wave: A Two-Stage Evolutionary Modeling Approach

Acquiring sonic waves is an essential part of oil and gas exploration as they give critical information about the well’s data and lithography at each well depth progression. However, these measurements are not always accessible, making analysis challenging. As computational power has improved, machine learning methods may now be used to predict these values from other data. Nonetheless, one shortcoming of existing models is that the most of them are not transparent (i.e., black box models). As a result, although promising great performance, they do not offer much insights to petrophysicists and geologists. This research aims to generate mathematical models for predicting compression wave (P-wave) and shear wave (S-wave) readings using a multistage evolutionary modelling approach. In particular, a multistage equation modelling approach using tree-based Genetic Programming (GP) and Adaptive Differential Evolution (ADE) is proposed. The obtained best mathematical models yield R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.745 and 0.9066 for P-wave and S-wave regression on normalized data, respectively. The average performance of models are R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> = 0.90 (P-Wave) and R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> = 0.75 (S-Wave). The performance of these mathematical models are comparable with other ”black box” models but with more compact mathematical approach in regression thereby opening opportunities for interpretability and analysis. Finally, the ”white box” models presented in this paper can be fine-tuned further as needed.

Three-dimensional P-wave model of the shallow crustal structure, a complementary method for detecting a trapped hydrocarbon: a case study in the DehDasht region, SW Iran

Abstract Local P-wave tomography is an efficient method to study geologically complex areas where the seismic exploration methods are not ideal for unraveling the shallow crustal heterogeneity due to the great thickness of evaporitic deposits. Despite the complex geological features in the salt-rich DehDasht region, SW Iran, we used &amp;gt;11 000 micro-earthquake events, which have been recorded by a temporary seismic network (deployed between 18 October 2016 and 1 July 2017), to derive the three-dimensional velocity structure based on the first arrival time. We selected a subset of events (1571 micro-earthquakes) by various strict criteria for our processing, and then the 1D velocity model was calculated by the computer program VELEST. Afterward, the 3D initial model of the inversion procedure with 1.5-km horizontal and 1-km deep intervals was parametrized using the calculated 1D model. Finally, the observed data (first arrival P-wave traveltimes and events locations) was inverted with an optimum regularization parameter and iteration using the computer program SIMULPS14. Our tomographic results indicate the DehDasht Basin as a relatively low-velocity zone filled out dominantly by the Gachsaran Formation and surrounded by the high-velocity Asmari-Pabdeh-Sarvak Formations. The basin has a bowl shape that is elongated in the NW–SE direction or an oval on a horizontal view. The depth of the basin varies between 3 and 5 km and contains many folding-faulting systems, which lead to locally low-velocity patches. Moreover, some evaporate deposits, which are overlying the Gachsaran Formation, emerge as a thin low-velocity layer (e.g. Aghajari, etc.).

Along-Strike Variation in the Shallow Velocity Structure beneath the Chenghai Fault Zone, Yunnan, China, Constrained from Methane Sources and Dense Arrays

Abstract The fine structure of the fault zone and the surrounding area is the basis for understanding the process of earthquake nucleation and rupture propagation. To obtain the high-resolution structure of the Chenghai fault (CHF) and the nearby basins, we deployed two dense arrays and excited eight methane sources across the fault from October to November 2020. Based on the 611 P-wave travel times, we obtained the shallow velocity structure beneath the arrays using the simul2000 travel-time inversion program, and the results are as follows: (1) the shallow velocity structure beneath the CHF is very complex, with obvious velocity contrasts on both the sides of the regional fault; (2) low-velocity zones (LVZs) beneath the CHF show clear along-strike variations. The LVZs extend to ∼500 m in depth with widths of ∼2 km and ∼5 km beneath the Qina and Pianjiao arrays, respectively, which are consistent with the Quaternary sediments, and the velocity contrasts along the interface of the LVZ can reach 20%–50%; and (3) the distribution of shallow surface tectonic geomorphology is mainly controlled by regional fault activities that are formed under the combined action of regional near-east–west stretching and clockwise rotation of microblocks. Our results can help improve cognition and seismic hazard assessment for potential earthquakes on the CHF, as well as lay the foundation for understanding the seismic wave velocity variation mechanism in the fault zone.

Deep subduction of the Philippine Sea slab and formation of slab window beneath central Japan

The geometry of the Philippine Sea slab (PHS) subducting beneath the Japanese islands has been imaged to 400 km depth beneath the Kyushu and Chugoku regions, whereas the PHS slab geometry beneath the Hokuriku region has only been determined to ~ 140 km depth, thereby indicating a large east–west asymmetry in the slab subduction. However, geologic evidence suggests that there was symmetrical east–west seafloor spreading along the axis of the Kinan seamount chain when the Shikoku basin was an active spreading center in the PHS plate. This inconsistency suggests that the PHS slab should be present beneath the Hokuriku region. Here we perform P-wave travel-time tomography across central Japan and conduct a two-dimensional plate subduction numerical simulation that reproduces the dual subduction of the PHS and Pacific (PAC) plates to elucidate the PHS slab geometry beneath central Japan. The tomography results reveal a high-velocity anomaly at ~ 150–250 km depth that extends from Wakasa bay to Noto peninsula and a slab window beneath the Hokuriku region. The numerical simulation results suggest that the PHS slab may have torn when it collided with the PAC slab, with the once leading edge of the PHS slab now present along the upper surface of the PAC slab beneath Noto peninsula. These results indicate that the PHS slab exists at ~ 250 km depth beneath the Hokuriku region, although it has been torn owing to its collision with the PAC slab, with this tear propagating westward to form a triangular slab window beneath the Hokuriku region.Graphical

Ultra-low-velocity anomaly inside the Pacific Slab near the 410-km discontinuity

The upper boundary of the mantle transition zone, known as the “410-km discontinuity”, is attributed to the phase transformation of the mineral olivine (α) to wadsleyite (β olivine). Here we present observations of triplicated P-waves from dense seismic arrays that constrain the structure of the subducting Pacific slab near the 410-km discontinuity beneath the northern Sea of Japan. Our analysis of P-wave travel times and waveforms at periods as short as 2 s indicates the presence of an ultra-low-velocity layer within the cold slab, with a P-wave velocity that is at least ≈20% lower than in the ambient mantle and an apparent thickness of ≈20 km along the wave path. This ultra-low-velocity layer could contain unstable material (e.g., poirierite) with reduced grain size where diffusionless transformations are favored.

Testing observables for teleseismic shear-wave splitting inversions: ambiguities of intensities, parameters, and waveforms

We assess the capabilities of different observables for the inversion of core-refracted shear waves (XKS phases) to uniquely resolve the anisotropic structure of the upper mantle. For this purpose, we perform full-waveform calculations for relatively simple, canonical models of upper-mantle anisotropy. The models are characterized by two and four domains of different anisotropic properties. Specifically, we assume hexagonal symmetry with arbitrarily chosen strength of the anisotropy and orientation of the horizontal fast axis. XKS waveforms, generated from plane-wave initial conditions, traverse through anisotropic models and are recorded at the surface by a single station (in case of vertical variations) and by a dense station profile across the laterally and vertically varying structure. In addition to waveforms, we consider the effects of anisotropic variations on apparent splitting parameters and splitting intensity. The results show that, generally, it is not possible to fully resolve the anisotropic parameters of a given model, even if complete waveforms (under noisefree conditions and for the complete azimuthal range) are considered. This is because waveforms for significantly different anisotropic models can be indistinguishable. However, inversions of both waveforms and apparent splitting parameters lead to similar models that exhibit systematic variations of anisotropic parameters. These characteristics may be exploited to better constrain the inversions. The results also show that splitting intensity holds some significant drawbacks: First, even from measurements over a wide range of back-azimuth, there is no characteristic signature that would indicate depth variations of anisotropy. Secondly, identical azimuthal variations of splitting intensity for different anisotropic structures do not imply that the corresponding split waveforms are also similar. Thus, fitting of observed and calculated splitting intensities could lead to anisotropic models that are incompatible with the observed waveforms. We conclude that (bandlimited) XKS-splitting inversions and related tomographic schemes, even if based on complete waveforms, are not sufficient to fully resolve the heterogeneous anisotropic structures of the upper mantle and that combinations with alternative methods, based on e.g., receiver-function splitting, P-wave travel-time deviations, or surface waves, are required.

The accuracy of approximating P-wave traveltimes in orthotropic media using VTI equations

SUMMARY Analysis of azimuthally varying non-hyperbolic moveout of P-wave traveltimes in orthotropic media can provide valuable information for reservoir fracture characterization and seismic data processing. Moveout velocities and azimuthal orientation of the two vertical symmetry planes plus anellipticity parameters defined for the three symmetry planes of an orthotropic medium can be estimated from the inversion of P-wave traveltimes measured in long-spread, wide-azimuth seismic surveys. Several orthotropic moveout equations have been developed to this date that can be used in the traveltime data inversion. The accuracy of the inversion results is primarily determined by the complexity of the equation and the assumptions (i.e. elastic or acoustic) made to derive it. Here, we show that often simpler moveout equations derived for transverse isotropy media with vertical axis of symmetry (VTI) can be used to estimate anellipticity parameters of the orthotropic media with high accuracy. We make an accuracy comparison analysis between non-hyperbolic moveout equations that are widely used in modelling P-wave traveltimes in VTI media. Our comparison analysis is based on the accuracy of estimating anellipticity parameters defined for two vertical [η(1) and η(2)] and one horizontal [η(3)] symmetry planes of a wide range of orthotropic media. The anellipticity parameters are estimated from the inversion of numerical traveltimes that are generated for a single and horizontal orthotropic layer. We show that for the survey geometry used in this analysis, most of the moveout equations can produce reliable estimates of the anellipticity parameters η(1) and η(2). Despite the availability of traveltime data from wide-azimuths and long source–receiver offsets, none of the moveout equations can estimate η(3) as accurate as η(1) and η(2) estimations. We show that while all the equations produce mostly accurate estimates of η(1) and η(2), generalized moveout approximation (GMA) provides the greatest accuracy. We also assess the validity of using GMA to invert P-wave traveltimes for three anellipticity parameters in terms of anisotropy in the background medium, fracture weakness, source–receiver offset (group angle) and traveltime data coverage in both the vertical (dip) and horizontal (azimuthal) planes. We show that, for a reasonable offset range, GMA can be reliably used to estimate anellipticity parameters η(1) and η(2) for an orthotropic medium that is constructed by intensive fracturing of a host medium with strong VTI anisotropy.

Estimation of source, path, and site factors of S waves recorded at the S-net sites in the Japan Trench area using the spectral inversion technique

S-net is a large-scale ocean bottom (OB) network in the Japan Trench area, consisting of inline-type 150 observatories equipped with seismometers and pressure gauges. Among them, 41 observatories have been buried about one meter beneath the seafloors in the shallow water regions (water depth <1500 m). We analyzed the strong-motion data recorded at the S-net sites from earthquakes with magnitudes 3.5 < Mw ≤ 7 and focal depths < 70 km to understand the site amplification effect on the recorded motions. We used the spectral inversion technique and obtained some fundamental properties of the earthquake source spectra, path attenuation, and site factors from the horizontal-component S-wave portions of the recordings. We obtained that the source spectra followed the {omega }^{-2} source model generally well, and the estimated magnitudes were mostly within ± 0.3 magnitude units of the catalog magnitudes. The stress drops increased systematically with the focal depths, and the values for the shallow earthquakes in the Pacific Plate were higher than those for the interplate earthquakes with comparable focal depths. The path-averaged quality factors were generally frequency-dependent and were somewhat larger than those in the past studies. The peak site frequencies ranged between about 0.2 and 10 Hz, while the peak amplification factors ranged between 10 and 50. Even though the peak frequencies and amplification factors differed from site to site, the peak frequencies were mostly higher than 2 Hz at the outer trench stations, while they were lower than 2 Hz at many inner trench stations. The amplification factors at a few OB sites in the shallow water regions were comparable with the theoretical ones computed from the 1-D subsea model. The amplification factors at intermediate frequencies (~ 0.3 to 2 Hz) generally increased with P-wave travel time in the sediments estimated from the multi-channel seismic survey. At about 20% of the sites (mainly at the unburied stations), spurious site spectra were recognized at frequencies higher than about 4 Hz. If the site spectra between 4 and 10 Hz are required, using only the X-component records is recommended.Graphical

Prediction of permeability of highly heterogeneous hydrocarbon reservoir from conventional petrophysical logs using optimized data-driven algorithms

Permeability is an important parameter in the petrophysical study of a reservoir and serves as a key tool in the development of an oilfield. This is while its prediction, especially in carbonate reservoirs with their relatively lower levels of permeability compared to sandstone reservoirs, is a complicated task as it has larger contributions from heterogeneously distributed vugs and fractures. In this respect, the present research uses the data from two wells (well A for modeling and well B for assessing the generalizability of the developed models) drilled into a carbonate reservoir to estimate the permeability using composite formulations based on least square support vector machine (LSSVM) and multilayer extreme learning machine (MELM) coupled with the so-called cuckoo optimization algorithm (COA), particle swarm optimization (PSO), and genetic algorithm (GA). We further used simple forms of convolutional neural network (CNN) and LSSVM for the sake of comparison. To this end, firstly, the Tukey method was applied to identify and remove the outliers from modeling data. In the next step, the second version of the nondominated sorting genetic algorithm (NSGA-II) was applied to the training data (70% of the entire dataset, selected randomly) to select an optimal group of features that most affect the permeability. The results indicated that although including more input parameters in the modeling added to the resultant coefficient of determination (R2) while reducing the error successively, yet the slope of the latter reduction got much slow as the number of input parameters exceeded 4. In this respect, petrophysical logs of P-wave travel time, bulk density, neutron porosity, and formation resistivity were identified as the most effective parameters for estimating the permeability. Evaluation of the results of permeability modeling based on root-mean-square error (RMSE) and R2 shed light on the MELM-COA as the best-performing model in the training and testing stages, as indicated by (RMSE = 0.5600 mD, R2 = 0.9931) and (RMSE = 0.6019 mD, R2 = 0.9919), respectively. The generalizability assessment conducted on the prediction of permeability in well B confirmed the MELM-COA can provide reliable permeability predictions by achieving an RMSE of 0.9219 mD. Consequently, the mentioned methodology is strongly recommended for predicting the permeability with high accuracy in similar depth intervals at other wells in the same field should the required dataset be available.

P-, SV- and SH-wave eikonal approximations for HTI media based on perturbation theory

SUMMARY The eikonal equation for transversely isotropic media is a complex quartic partial differential equation that causes instability in traveltime inversion. The phenomenon of shear wave splitting in a transversely isotropic model with a horizontal symmetry axis (HTI) provides information for inversion. However, the involvement of shear waves makes the computation of forward modelling and inversion more time-consuming than when only using Pwaves. Most studies of traveltime approximation based on the eikonal equation are based on an acoustic assumption, which does not work for shear waves. Therefore, based on perturbation theory, we derive approximate solutions of the eikonal equation in a heterogeneous elastic HTI model for P, SV and SH waves to improve the traveltime computation efficiency. The approximate solutions of linear, quadratic and Shanks forms are based on the linearization of the traveltime function in terms of fracture weaknesses in an isotropic background, which are suitable for joint P-, SV- and SH-wave traveltime inversion and tomography for fracture parameters. When applied for a homogenous model, the proposed formulation results in a simple and accurate traveltime approximation. The first- and second-order perturbation coefficients are used to analyse the sensitivity of the traveltime to fracture weakness. To improve the accuracy of our approximation, we also expand the approximation of the P-wave traveltime with respect to the anellipticity parameter η in an elliptic background, which is evaluated by performing an error test. Based on the accuracy and stability of the perturbation, we select a set of linear equations with respect to the fracture weaknesses and a set of more accurate nonlinear equations to serve different needs.

Seismic velocity changes in the Groningen reservoir associated with distant drilling

In this study, we show that passively recorded data of nearby passing trains by a deep borehole geophone array could be linked to fluctuations of the gas-water contact in the Groningen reservoir in The Netherlands. During a period of 1.5 months, changes of inter-geophone P-wave travel times were detected by deconvolution interferometry of the recorded train signals. P-to-S converted waves, obtained by deconvolution of the horizontal component by the vertical component at individual geophones, showed simultaneous variations. The observed travel-time changes could be related to fluctuations of the gas-water contact in the observation well caused by pressure variations at a well drilling 4.5 km away. The sim 3.5 day delay between drilling in the reservoir and the seismic response yields a hydraulic diffusivity of approximately 5 m^2/s and suggests that the pressure front is effectively propagated over such a long distance. Our observations illustrate that downhole geophone arrays can be used to monitor changes in the subsurface if repeating noise sources are available, and that unexpected effects may occur due to drilling.

Adaptive boosting of random forest algorithm for automatic petrophysical interpretation of well logs

The power of Machine Learning is demonstrated for automatic interpretation of well logs and determining reservoir properties for volume of shale, porosity, and water saturation respectively for tight clastic sequences. Random Forest algorithms are reputed for their efficiency as they belong to a class of algorithms called ensemble methods, which are traditionally seen as weak learners, but can be transformed into strong performers and they promise to deliver highly accurate results. The study area is located offshore Australia in the Poseidon and Crown fields situated in the Browse Basin, which are gas fields in tight complex clastic reservoirs. There are 5 wells used in this study with one well manually interpreted which is subsequently used in developing a machine learning model which predicts the output for the other 4 wells. The basic open hole logs namely Natural gamma ray, Resistivity, Neutron Porosity, Bulk Density, P-wave and S-wave sonic travel-time, are used in interpretation. One of the wells has a missing S-wave travel-time log which was also predicted by developing a Random Forest Machine Learning model. The results indicate a very robust improvement in performance when Random Forest algorithm was combined with Adaptive Boosting when interpreting the well logs. The training accuracy using Random Forest alone was 98.21%, but testing was 77.62% which suggested over-fitting by the Random Forest model. The Adaptive Boosting of the Random Forest algorithm resulted in the overall training accuracy of 99.40% and an overall testing accuracy of 97.03%, indicating a drastic improvement in performance. S-wave travel-time log was predicted by preparing a training set consisting of Natural gamma ray, Resistivity, Neutron Porosity, Bulk Density, and P-wave travel-time logs for the 4 wells using Random Forest which gave a training accuracy of 99.79% and a testing accuracy of 98.54%. Machine learning algorithms can be successfully applied for interpreting well log data in complex sedimentary environment and their performance can be drastically improved using Adaptive Boosting.

3D transdimensional seismic tomography of the inner core

Body wave observations of the Earth's inner core show that it contains strong seismic heterogeneity, both laterally and radially. Models of inner core structure generated using body wave data are often limited by their parameterisation. Thus, it is difficult to determine whether features such as anisotropic hemispheres or an innermost inner core truly exist with their simple shapes, or result only from the chosen parameterisation and are in fact more complex features. To overcome this limitation, we conduct seismic tomography using transdimensional Markov Chain Monte Carlo on a high quality dataset of 5296 differential and 2344 absolute P-wave travel times. In a transdimensional approach, the data defines the model space parameterisation, providing us with both the mean value of each model parameter and its probability distribution, allowing us to identify well versus poorly constrained regions. We robustly recover many first order observations found in previous studies without the imposition of a priori fixed geometry including an isotropic top layer (with anisotropy less than 1%) which is between 60 and 170 km thick, and separated into hemispheres with a slow west and a faster east. Strong anisotropy (with a maximum of 7.2%) is found mainly in the west, with much weaker anisotropy in the east. We observe for the first time that the western anisotropic zone is largely confined to the northern hemisphere, a property which would not be recognised in models assuming a simple hemispherical parameterisation. We further find that the inner most inner core, in which the slowest anisotropic velocity direction is tilted relative to Earth's axis of rotation (ζ=55∘±16∘), is offset by 400 km from the centre of the inner core and is restricted to the eastern hemisphere. We propose that this anomalous anisotropy might indicate the presence of a different phase of iron (either bcc or fcc) compared to the rest of the inner core (hcp).