Search

Microseismic technology has been widely used in geological hazard monitoring. However, the effective identification and classification of microseismic events in mines have always been a challenge in hazard monitoring. In this paper, a new method combining the empirical mode decomposition (EMD) algorithm and artificial neural network is proposed for the classification and identification of microseismic waveforms. This is the first study to explore such a model in the microseismic monitoring of coal mines. The data for this study were collected from a coal mine in northern China. A total of 2768 microseismic events and 2435 non-microseismic events were manually selected from a large amount of data. Firstly, the microseismic waveform reconstruction was performed using the 4th to 8th Intrinsic Mode Function (IMF) obtained by the EMD algorithm. Secondly, the peak factor, clearance factor, impulse factor, kurtosis, skewness, and waveform factor of each reconstructed microseismic waveform were used as features. Subsequently, the convolutional neural network (CNN) was used to classify microseismic events. Among them, data from 3600 microseismic events were used as the training set, and data from 1603 microseismic events were used as the test set. Finally, the original waveform data were used as input and compared with the classification results processed with the EMD algorithm. The classification methods of the BP neural network and Radial Basis Function (RBF) neural network were used for verification. The results of the processed features showed a significant increase in accuracy. The classification accuracy of CNN, BP neural network, and RBF neural network are 97.37%, 91.99%, and 94.23%, respectively. The results show that the utilization of the feature processing technique and CNN algorithm in this study demonstrates superior efficacy in microseismic event classification, which can be used in practice. • The EMD algorithm and ANN was constructed to classify microseismic waveforms. • Six features of the reconstructed waveform were selected as the classification basis. • The EMD algorithm was used to make the six waveform characteristics more significant. [ABSTRACT FROM AUTHOR]

With the increasing depth of mining, the disasters of water inrush above confined aquifers tends to increase annually, which seriously threatens the safe mining of coal seam, especially the extra-thick coal seam floor in deep mining. This research is to study the characteristics of deformation and failure of extra-thick coal seam floor in deep mining, the field measurement is carried out by the distributed fiber optic sensing technology (BOTDR) and the 3D resistivity method of studying coal field in China. The axial strain curves and resistivity distribution characteristics of the mining face at different depth under the coal seam floor are obtained by extending the working face. Through the comprehensive comparison and analysis of monitoring data in the monitoring boreholes, the characteristics of deformation and failure of extra-thick coal seam floor in deep mining can be ascertained. The results indicate that the comprehensive monitoring technology can effectively distinguish the characteristics of stress excess, distortion and failure of rock during the process of the working face mining. The influence range of mining advance stress is about 36 m, the failure depth of coal seam floor is about 12 m and the depth of mining disturbance deformation is about 26 m. The formation of the deformation and failure of the coal seam floor have obvious stratification, which are mainly influenced by the formation structure and the lithology of the rock strata. The comprehensive monitoring technology can effectively capture the fine features of the deformation and failure of the coal seam floor, which is valuable for extra-thick coal seam floor in deep mining. • In this paper, the distributed fiber optic sensing technology is used for the first time to monitor the deformation and failure of the coal seam floor. • The coal seam studied in the paper is an extra-thick coal seam and the buried depth of the coal seam is large, belonging to the extra-thick coal seam in deep mining. • The high frequency of data acquisition and the long duration of acquisition, so as to realize the dynamic monitoring research on the deformation and failure of extra-thick coal seam floor in deep mining. [ABSTRACT FROM AUTHOR]

The Laojunshan ore cluster region is situated at the north Song Chay dome, which is the junction of the Yangtze, the Cathaysia, and the Indo-China Blocks. The north Song Chay dome characterized by the Laojunshan metamorphic core complex is an important ore-forming region for W-Sn-Pb-Zn-In elements related to the Late Cretaceous granitic magmatism. In this paper, the Bi-dimensional empirical mode decomposition (BEMD) is applied to separate gravity components from the original gravity data, combined with the known geological and mineralized information, to explore the relationships between the gravity components and various ore-controlling factors at different depths. Four intrinsic mode functions (IMFs) and a residue are obtained. They may imply the spatial distribution of the geological bodies with featured gravity components at different depths. The results are shown as follows: (1) The low-pass filtering gravity component (BIMF4) implies the deeply buried geological units in the study region. The Cretaceous Laojunshan granitic complex and the Nanwenhe gneissic granite display negative gravity components and weak positive gravity components, respectively. The former is closely related to Sn-Zn-In mineralization, the latter is closely related to W polymetallic deposits. Therefore, the negative gravity components showing the Laojunshan granitic complex are the prospecting targets for deeply buried Sn-Zn-In deposits. The weak positive gravity components displaying the Nanwenhe gneissic granite are the prospecting targets for deeply buried W polymetallic deposits. (2) The band-pass filtering gravity component (BIMF3) implies the geological units at middle depth. The skarn-type Sn-W polymetallic deposits are observed in the transition zones from the positive to the negative gravity components, which coincide with the contact metasomatic zones between the Cretaceous granite and the Cambrian-Devonian carbonate rocks. Some small quartz vein W deposits occur within the Cretaceous Laojunshan granite with negative gravity components, and some layered W deposits occur in the Silurian gneissic granite with a weak positive gravity component. (3) High-pass filtering gravity component (BIMF2) implies geological units at the shallow crust, which shows that most of the deposits (occurrences) are located within the positive gravity component zones. These positive gravity component zones should be ore-finding targets for further prospecting for the undiscovered Sn and W polymetallic deposits. (4) The gravity components separated by BEMD coincide with the spatial distribution of the granites associated with various types of mineralization and alteration zones at different depths, which will give insights for further prospecting for concealed ore bodies in this study region. [Display omitted] • The mineralized characteristics of both the Sn-Zn-In and the W polymetallic deposits are recognized. • The gravity components implying ore-controlling factors at different depths are separated using BEMD. • The ore-finding targets of both the Sn-Zn-In and the W polymetallic deposits are identified using the gravity components. [ABSTRACT FROM AUTHOR]

Abandoned coal tunnels (ACTs) pose risks to coal mine safety by accumulating gas and water, causing surface subsidence, and endangering infrastructure. Accurate detection of ACTs is vital for advanced prediction and geotechnical hazard warnings. Seismic methods, particularly the surface-wave method, have demonstrated reliability in detecting shallow buried tunnels within complex geological environments, but improvements in accuracy and resolution are still needed. In this paper, we suggest utilizing the multiscale window analysis of surface waves (MWASW) method for ACT surveys. We conduct a synthetic test and a real-world case study to evaluate the feasibility and validity of this approach. The synthetic test on a 2D ACT model reveals that MWASW can reconstruct anomalous structures and identify the presence of an ACT, while also highlighting some limitations of the method. In a real-world scenario, we applied MWASW to a multi-survey line dataset from a coal mine in Shanxi Province, China. The resulting meter-scale resolution pseudo-3D S-wave velocity structure uncovers a primary ACT and two sub-ACTs buried at approximately 60 m, offering valuable insights into their morphology and orientation. To our knowledge, this is the first application of the MWASW method in such a case study, indicating its potential as an emerging alternative for characterizing complex near-surface media, such as abandoned coal tunnels or mined-out areas in coal fields. • Detect coal tunnels using spatial multiscale dispersion analysis method. • Directly reveal 2D S-wave velocity profiles with high resolution and accuracy. • Reconstruct a pseudo-3D structure model in field case. [ABSTRACT FROM AUTHOR]

The Passive Multichannel Analysis of Surface Waves (MASW) method is a non-intrusive and cost effective technology adopted in civil and earthquake engineering for site geological and geotechnical survey. However, the application of passive MASW method under urban environments meets some challenges. One of interesting problems is associated with underground railways, which can be treated theoretically as buried sources. Since Rayleigh wave energy is dominant among that of surface waves, it is critical to design a proper geophone configuration during field tests at a site close to the metro line to ensure significant Rayleigh wave energy and therefore reliable dispersion images. This paper studied the propagation of Rayleigh waves generated from a buried source and proposed an empirical formula to estimate the nearest offset, which was then verified by numerical simulations. A case study was implemented in Shanghai, China at a site close to a subway line. With a linear spread geophone configuration, the influence of different parameters, including nearest offset, spread length and geophone interval, was investigated regarding their influences on dispersion images. Field tests were conducted by following the geophone configuration designed based on the proposed estimation method. The one dimensional (1D) shear velocity structure, being inverted from field vibration measurements, matches relatively well with the geologic feature obtained from borehole testing. Study results are expected to help facilitate the application of the passive MASW method for cases in which Rayleigh waves are generated by the buried source like underground traffics in cities. • Near-field effects due to subway traffic were studied for surface wave method application. • A method was proposed for the nearest offset estimation in MASW test with buried sources. • A case study was performed at a site close to underground railway for validation. [ABSTRACT FROM AUTHOR]

Active faults in urban have a potential damage to citizens, because they can induce not only earthquakes, but also damage pavements, utilities, homes, businesses, factories and other manmade structures because of the slow, secular and differential slippage. Consequently, the researches of detecting active faults are of great significance. This paper proposes a set of geophysical methods to detect active faults by an example in Beijing, including gravity, controlled source audio-frequency magnetotellurics(CSAMT), seismic reflection, DC resistivity and paleomagnetism (Natural remanent magnetization in rocks) to locate faults and discuss their activities. In proposed methods, gravity interpretation helps us obtain the distribution and characteristics of buried faults beneath the plain, the results of CSAMT, seismic reflection and DC resistivity reveal features and characteristics of faults from the deep to shallow part; paleomagnetism associated with radiocarbon dating help us analyze the fault slip rate; 3D seismic reflection interpretation shows the structure of two faults in the three-dimensional subsurface and the interaction of each other. Also, a few acquisition parameters, data processing methods and significant suggestions are mentioned. [ABSTRACT FROM AUTHOR]

Accurate characterization of the low-velocity anomalies such as utility tunnels in the near-surface by effective geophysical imaging methods has been a valuable but challenging task for urban hazard assessment, underground space development and infrastructure construction. As a current research hotspot, the full waveform inversion (FWI) has been proven to be a promising technique for high-precision imaging of substructures, and has been applied in related practical engineering fields. However, most previous studies have focused on single-parameter (S-wave velocity) or double-parameter (S-wave and P-wave velocities) inversion. So, a complete PSV/Rayleigh-wave multi-parameter (S-wave and P-wave velocities and density) FWI method is developed in this paper to achieve accurate detection of underground utility tunnels. In this approach, the multi-parameter of the model are updated simultaneously during the inversion process to alleviate the negative influence on the S-wave velocity inversion accuracy due to the far deviation of the other two parameters from their true values. We tested the performance of the approach with a synthetic test and a real experiment. In the synthetic test, this method well characterized the position, shape and size of the embedded tunnel in the model, proving its effectiveness. Subsequently, the approach was applied to a field case investigated in Changsha city, China, and the inversion images revealed a tunnel with a burial depth of 4.5 m, height and width of 2 m and 4 m, respectively, which was identified as the detection target. The available borehole data verified the conclusion. The above results demonstrate the validity and time-effectiveness of the approach. It is expected to provide a reference for practical engineering applications at the same scale. • Detect utility tunnels using 2D PSV/Rayleigh-wave full waveform inversion. • Illustrated the validity of the method via a synthetic test and a field case. • Provide a complete workflow for handling field data. • Verify the reliability of the results by the borehole data. [ABSTRACT FROM AUTHOR]

• A noise evaluation method based on an envelope algorithm was proposed in this paper. • The method estimates the upper bound of the noise rather than statistical average. • Not only main frequencies, but also the harmonics can be evaluated by this method. • This noise evaluation method can be applied to any periodic signal. Controlled-source electromagnetic (CSEM) is affected by noise during exploration. Moreover, different frequencies in the data are affected differently by noise. However, past evaluation methods, such as the relative root mean square error (RRMSE) method, are not efficient or accurate, especially when there are many frequencies of interest. Therefore, a noise evaluation method based on a mixed spectrum envelope algorithm containing wavelet decomposition, Hilbert transform and peak envelope was proposed for the first time. This method estimates the upper bound of the noise effect at the frequencies of interest by the amplitude of the adjacent frequencies rather than by estimating its exact value or statistical average. Based on this method, the maximum possible value of noise effect can be obtained, which is then divided by its raw spectrum to obtain a newly defined noise evaluation parameter, "noise ratio". Using this new parameter, the degree of the noise effect at the frequencies of interest can be evaluated more accurately. Furthermore, using an appropriate threshold, a large number of candidate frequencies including harmonics can be filtered out. The performance of the method is tested on synthetic signals and real data from Shandong Province, China. The test results show that the evaluation method based on the envelope algorithm is more stable and effective than the conventional RMSE method. Based on the features of the method, it can be applied to all periodic signals in other stationary scenarios. [ABSTRACT FROM AUTHOR]

One of the key and difficult points in seismic data processing is seismic data denoising. Under the influence of the acquisition environment, the collected seismic data usually have low signal-to-noise ratio (SNR) and low resolution. In desert region of western China, different from other regions, the random noise there has complex characteristics of non-Gaussian, non-stationary and non-linearity. Its main frequency is quite low and its spectral overlap with those of effective signals. Moreover, large amount of data and intelligent requirements make traditional denoising methods encounter difficulties. In addition, the existing deep learning denoising methods also reveal two problems: first, they can only effectively suppress simple seismic noise such as Gaussian noise. But for the case of desert noise, they usually mistakenly judge noise as effective signals and destroy signal characteristic structure. Second, what is more important, in supervised-based methods paired noisy and pure data are usually used as training set. However, there is no noise-free data in actual desert seismic data, which severely limits the processing performance of supervised-based networks. Thus, a new parameter-shared variational auto-encoding adversarial network (PS-VAAN) is proposed for desert seismic data denoising in this paper. This new method includes two encoders, two generators and two discriminators. It can realize domain conversion from noisy data domain to pure data domain. We construct relatively complete unpaired pure and noisy data training sets for training, and design reconstruction loss function and adversarial loss function to optimize network parameters. Moreover, cycle-consistency loss is introduced to make sure two domains mapping into the same space. Compared with the state-of-art denoising methods on synthetic and actual seismic records, the proposed method not only has superior denoising ability, but also has strong feature extraction and fitting ability to recover effective signals with almost no energy loss. • Desert seismic noise has the characteristics of low frequency and non-stationary. • This paper designs a parameter-shared variational auto-encoding adversarial network. • The network use unpaired noisy and pure training sets to optimize parameters. • The proposed method has a strong ability to recover effective signals from noise. [ABSTRACT FROM AUTHOR]

The Sponge City construction currently undergoing in China aims to enhance the moisture capacity of urban pedosphere, and subsurface moisture content (SMC) is proposed to evaluate the effect of improvement. Different ground-penetrating radar (GPR) techniques have been used in previous studies to estimate moisture content, however, very few has assessed the potential of combing 2D common offset (CO) GPR with 3D multichannel GPR to resolve SMC estimates at the complex filed scale. In this paper, modified velocity analysis of in-situ 3D GPR common-midpoint (CMP) datasets was used together with 2D GPR laboratory calibration procedure to investigate spatial moisture variation and the impact of surface pavement and soil heterogeneity on subsurface water dynamics. Precise and reliable moisture information under two field sites (i.e., sponge pavement site and asphalt pavement site) was obtained, which further allowed us to quantitative characterize SMC distribution and its associated dynamics within the shallow vadose zone, and most importantly to investigate the coupling mechanism between subsurface water fluctuations with surface pavement conditions and subsoil moisture capacity in urban near surface environment. Additionally, through contrastive analysis, the superior performance of novel sponge material to detain and transit rainwater was highlighted. Data and results presented in this research provide pioneer reference for Sponge City construction and other engineering applications. • Captures precise and reliable moisture information at field scale by integrative use of 2D GPR and 3D GPR; • Reveals the coupling between subsurface water dynamics with surface pavement conditions and subsoil moisture capacity; • Provides pioneer reference for water content estimation on dense materials of different nature. [ABSTRACT FROM AUTHOR]

Deep mining commonly causes redistribution of the stress-strain field for coal seam floor strata. The study of this distribution law was of great significance to analyzing the characteristics of floor deformation and failure and predicting the risk of floor water inrush. In this paper, the damage of a coal seam floor with strong water pressure in the Junggar Basin in western China was examined, and the stress analytical formula for every point of floor confined aquifer was constructed using a stress model in the direction of coal seam strike. Combined with Mohr-Coulomb failure criterion, the maximum depth of floor failure was 17.5 m. According to geological and stratigraphic data of the stope floor, a 3D numerical model of fluid-structure coupling was built. Simulation results showed that under a 1.5 MPa water pressure, the maximum floor failure depth was 18 m. The internal stress of the coal-rock interface and soft rock increases significantly, and the weak coal seam intensifies the transfer of stress and strain at depth to some extent. Water pressure had a certain influence on floor deformation and failure. The joint simulation results of stress-seepage-fracture were closer to the real stope measurements. In the boreholes, two types of sensors, distributed fiber and electrodes were arranged. When the coal floor broke, the compressive strain increased with a peak of 4152 με. Meanwhile, the apparent resistivity of the rock seam reached 500–700 Ω•m. The comprehensive analysis showed that: the floor failure depth in the monitoring area of the 61,303 working face was −18 m, and the floor failure disturbance depth was −35 m. The results presented in this contribution can provide reference for future examinations of coal and rock interbedded floor failure over deep confined water. • Study on mining failure of deep interbedded coal and rock floor. • The analytical formula of stress at any point of floor is derived. • The influence of weak coal seam on stress transfer is studied. • Combined application of borehole strain and apparent resistivity imaging. [ABSTRACT FROM AUTHOR]

With the continuous exploitation of deep coal resources, the long-term stability control of surrounding rock of deep high stress soft rock roadway has become an urgent support problem. Because the traditional surrounding rock deformation zoning theory does not consider the creep effect of rock, it is not suitable for the support design of deep soft rock roadway, and can not ensure the long-term safe production of coal mines. This work proposes a novel four-stage strain-softening model based on rock creep characteristics, reveals the deformation zoning mechanism of surrounding rock in deep high stress soft rock roadway, and proposes the determination method of actual deep roadway surrounding rock zoning type and boundary stress. Moreover, taking Pan Yidong Coal Mine in China as an example, the feasibility and accuracy of the above theoretical method are verified by test, field monitoring and ABAQUS numerical simulation, and the influencing factors of the deformation zoning type of surrounding rock of the actual roadway are analyzed. The results show that the actual surrounding rock zoning type and boundary stress value determined by the deformation zoning mechanism of deep roadway surrounding rock in this paper are more accurate than the calculation results of the traditional deformation zoning theory, and the variations in the basic mechanical properties of surrounding rock and the strength of the in-situ stress field determined the number of actual deformation zones and the boundary stress values of each zone in the roadway surrounding rock. The research results can potentially aid the support designing and stability evaluation of deep roadway surrounding rock. • Actual deformation zones in deep roadway surrounding rock were analyzed. • Novel four-stage strain-softening model determines zoning type and boundary stress. • Factors influencing the type and number of deformation zones were investigated. • Actual deformation zones rely on mechanical properties and in-situ stress of rock. • Model accuracy was verified using field tests and numerical simulations. [ABSTRACT FROM AUTHOR]

Cross-hole seismic tomography is a high-precision method that can obtain the velocity structure between boreholes. At present, single-component geophones are mainly used for P-wave velocity estimation. However, S waves have a shorter wavelength than P waves, incurring a high-resolution tomography image. In this paper, we apply three-component cross-hole seismic tomography to shallow geological survey. First, three-component geophone is used to collect cross-hole seismic data. Secondly, the apparent velocity polarization method is used to separate the P and S wavefields from the recorded three-component waveform data. Then, a damped least-squares traveltime tomography is used to calculate the cross-hole P- and S-wave velocity structures. Finally, the distribution of Poisson's ratio between the boreholes is derived from the inverted P- and S-wave velocities. Numerical model experiments show that this method can solve some typical shallow subsurface problems, such as the detection of karst features and boulders. The proposed method is used to detect an underground air-raid shelter located in Hangzhou, Zhejiang Province, China, and the P- and S-wave velocities and the Poisson's ratio are obtained with high precision. The tomography image of the air-raid shelter agrees well with its known location. and the calculated Poisson's ratio distribution shows that the air-raid shelter may be partially filled with mucky clay. Three-component cross-hole seismic tomography achieves a high data acquisition efficiency and can simultaneously obtain shallow subsurface stratum velocities and Poisson's ratio with high precision. Thus, this technique can be used for shallow-subsurface surveys. • Apply three-component cross-hole seismic tomography to shallow geological survey. • Acquire P and S wave velocity parameters and poisson's ratio at the same time. • Experiments have proved its high-precision imaging ability in shallow surface. [ABSTRACT FROM AUTHOR]

When the traditional sea surface towed multi-channel seismic technology is applied to the deep-water stratum exploration, the Fresnel radius of the target stratum is large due to the large-depth water bodies, greatly reducing the horizontal resolution. Hence, it is difficult to achieve high-resolution exploration with the traditional technology and cannot meet the requirements for high-precision exploration and orebody description of such deep-water seabed resources as natural gas hydrate (NGH). The near-seabed deep-towed seismic exploration technology was proposed to solve this problem. This paper introduces a set of high-resolution deep-towed seismic system being developed in China, Kuiyang ST2000. The remarkable difference between the equipment and foreign counterpart is that this equipment uses a deep-towed sparker instead of a resonator source. We systematically describes the system composition of the deep-towed system, and primarily discusses the technical design of the deep-towed sparker which can still have relatively large stratum penetration depth in deep-water and high-static-pressure environment. In addition, we also introduces the unique continuous acquisition technology possessed by the system. At present, the experiment of 2000 m deep-water depth has been carried out for this system, and the experimental and related data processing results show that this equipment has had fundamental functions. Several deep-water experiments will be carried out in the next step. • A deep-towed seismic exploration technology was proposed. • A deep-towed sparker was developed. • A deep-towed multi-channel streamer was developed. • A unique continuous acquisition technology was used. • The experiment at 2000 m deep-water depth has been carried out. [ABSTRACT FROM AUTHOR]

In this paper, we introduce a technique to obtain smoother C -responses in geomagnetic depth sounding. Instead of frequency by frequency, C -responses in entire frequency range are estimated simultaneously. In order to do so, after obtaining the spectrum of H y and H z , the equations relating C -response with H y and H z are combined for a batch frequency to form a linear equation set, from which, incorporating the requirement of smooth constraint on C -responses, we could define a regularized least-squared inversion problem. And to minimize influence of noise, data squared coherencies between Z and H signals are used to weight prediction error in order to suppress unreliable data. Synthetic data tests indicate that using the optimal regularization parameter determined by V-curve method other than L-curve method, regularization method could produce rather smooth C -responses from noise interfered signals. C -responses in Northeast China are thus obtained and then 1-D inversion is applied, from the obtained electrical and combining the Global and regional tomographic results, it is speculated that the stagnant plate in MTZ leads the high conductivity anomaly in Northeast China. • A new method for has been applied for geomagnetic data processing. • More C -responses curves increase the density of the available data for inversion. • The stagnant plate may lie subhorizontal in the MTZ. [ABSTRACT FROM AUTHOR]

In this paper two different geophysical electro-magnetic techniques, long-offset time domain EM and the magnetotelluric method, are jointly inverted using a variation of Bayesian inversion called Hybrid Monte-Carlo. Long-offset time domain EM is a method that records Earth response to a controlled source square wave signal, while the magnetotelluric method records Earth response to naturally occurring electro-magnetic signals. We obtain better understanding of subsurface resistivity structures because the information they provide is complementary and joint inversion reduces the set of permissible inversion models. In most situations where long-offset time domain EM is recorded magnetotelluric data is also collected, since receiver instrumentation is shared. Hybrid Monte Carlo uses the gradient of the forward model to enable larger jumps through the solution space than are permissible using the standard Metropolis-Hastings algorithm, avoiding the pitfall of random walk behavior and giving faster convergence. The method is successfully applied to a survey area near the edge of Wuhan China. We obtain inversion results in a highly urbanized area, where the magnetotelluric method is typically compromised by background electro-magnetic noise. • HMC reduces random walks and gives faster convergence. • The joint-inversion of MT and LOTEM combines complimentary EM responses. • Problems with difficult combinations of conductors and resistor would benefit. • This is the first joint inversion of EM data using HMC. [ABSTRACT FROM AUTHOR]

The reservoir rocks in the volcanic strata of the Jilin oil field are characterized by great complexity and diversity in composition and structure of lithology. To enhance the rate of lithology identification in subsurface is very laborious. However, lithology identification is often ignored in quantitative studies, though it is the basis for reservoir characterization. In this paper, an ensemble learning algorithm named gradient boosting decision tree (GBDT) was used to establish the classification model for the volcanic lithology identification of the Lower Cretaceous Yingcheng Formation in the Songliao Basin, NE China. At the same time, support vector machine (SVM), logistic regression (LR) and decision tree (DT) classification models were also adopted in contrast with the classification accuracy of GBDT model. Subsequently, the optimal key parameters for each model were determined by employing validation curves and GridSearchCv. These results indicate that the GBDT model is superior to the single classifier and can accurately distinguish the lithologic interface of breccia tuff and rhyolite. Moreover, it also has better recognition ability for thin layer. It was concluded that the ensemble learning algorithm GBDT has significantly enhanced the accuracy of lithology identification and can be used as a lithologic identification technology. • The GBDT model is applied well in volcanic lithology identification. • Parameter optimization is conducted to optimize the classifier. • The GBDT model is superior to the single classifier. • The GBDT model and has better recognition ability for thin layer. [ABSTRACT FROM AUTHOR]

Abstract: To reduce the acquisition footprint in the geometric attribute analysis of marine seismic data, a 3D stationary wavelet transform (SWT) is presented in this paper. We use the 3D SWT to decompose the seismic volume into multiple levels of cubes in three directions, and we determine the different filter parameters, which vary with depth. Then, we implement the suppression of the acquisition footprint by adaptively reconstructing the weighted filtered wavelet components. The field data examples indicate that the method effectively removes the acquisition footprint and suppresses other types of noise while at the same time, preserving and highlighting geological features. The method can improve the accuracy and reliability of geometric attribute analysis (such as eigenstructure-based coherence and 3D volumetric curvature) in characterizing the internal structure of reservoirs and in structural interpretation. The method can be expected to be applicable to the removal of other kinds of noise on marine seismic datasets. [Copyright &y& Elsevier]

Abstract: The ground magnetic response of deep ore bodies in the Daye iron-ore deposit is relatively weak, and sometimes concealed by the strong magnetic background of shallower sources. Apart from the low-quality ground magnetic data, another critical problem for reconstructing the deep skarn-type ore bodies is developing a versatile inversion scheme that can simultaneously resolve 3D sources with arbitrary shapes. In this case, we resort to interactive 3D forward modeling solution with the joint use of two data sets-total field surface and three-component borehole magnetic data. Joint inversion of the two data sets is expected to help resolve the ambiguity associated with either data set and greatly reduces the nonuniqueness of the magnetic inversion. Such nonuniqueness is especially severe when a 3-D distribution of magnetic susceptibility, instead of a simple body, is sought from the inversion. In this paper, we calculate the magnetic field on the surface and in the borehole caused by 3D arbitrarily-shaped bodies with the triple integral method. The complex 3D magnetic sources having arbitrary shapes are constructed with cross-sections, termination points and facets in our visualization technology. We specify, interactively and in a user-friendly environment, the outline of the sources in terms of geometric elements and their magnetic parameters. The method automatically fits the observations within a prescribed precision. If dissatisfied, the user can redefine the model parameters and proceed to a new inversion. The method''s ability to interpret a complicated 3D geologic environment is demonstrated on synthetic models and real data profiles in the Daye iron-ore deposit in central China. The interactive forward modeling results in all tests demonstrate a good correlation of estimated magnetic sources with corresponding known geologic features. [Copyright &y& Elsevier]

The frequent occurrence of groundwater-related damage in operational high-speed railway tunnels in China calls for swift and accurate diagnosis. Thus, this paper describes a way to detect water-bearing defects at 5–10 m using the train-borne transient electromagnetic method in operational tunnels. Due to the short detection distance, the early electromagnetic response of the receiving coil is crucial for determining the size and location of defects. First, the turn-off current in the transmitting coil is elaborated by the equivalent electronic circuit of the transmitting loop. Then the impact of changes in the relative position between the transmitting coil and the receiving coil on the early electromagnetic response are analyzed by a numerical simulation, which shows that a square transmitting coil is suitable for ultra-close detection due to the stability of the electromagnetic response at early times in the receiving coil. Finally a laboratory experiment of close-range TEM detection is carried out to clarify the necessity and feasibility of taking the early electromagnetic response of the receiving coil into consideration at this ultra-close range (5–10 m). This research provides an important theoretical basis for the development and application of train-borne transient electromagnetic method for automatic detection of tunnel defects. • The early EM characters in the transmitting coil plane are affected by loop shape. • The early EM characters in receiver coils are affected by loop relative position. • The square loop has the most stable EM response in the receiving coil at early times. • The test shows early EM response in the coil is essential for close-range detection. [ABSTRACT FROM AUTHOR]

Noise suppression is an important step in seismic data processing. Various geographical environments make random noise in the acquired seismic records different in characteristics. Random noise in desert areas has non-Gaussian, non-linearity, non-stationary and heterogeneity. What is more, its frequency range is generally below 15 Hz, showing distinct low-frequency characteristics. Therefore, it brings much difficulty to the suppression of desert noise. Most of the existing noise suppression methods perform well on the condition of high-frequency noise, but perform worse for low-frequency noise because the denoising process of some methods depends on the waveform of noisy signal. And in some transformed methods, signal and noise are mixed with each other in the transformed domain. Thus, to resolve this problem, this paper presents a novel multiband spectrum separation (MSS) method based upon sparse nonnegative matrix factorization (SNMF) algorithm. It firstly decomposes noisy signal into some sub-signals of different narrow frequencies bands. Then, signal and noise components of these sub-signals are further separated by spectral unmixing in time-frequency domain. Several groups of experiments were carried out on synthetic records and real seismic data in Tarim oilfield in the northwest of China. Moreover, we compare its denoising effect with time-frequency peak filtering (TFPF), empirical mode decomposing (EMD) and Radon transform methods. The experimental and comparison results demonstrate that the proposed SNMF-MSS method performs better than the other three methods in the aspect of low-frequency suppression and weak signal recovery. • Random noise from desert areas has complicated and low-frequency (below 15 Hz) characteristics. • SNMF-MSS realizes signal and noise separation by time-frequency spectrum unmixing of some different frequencies sub-signals. • Experimental results demonstrate the better performance of SNMF-MSS in the aspect of low-frequency random noise suppression. [ABSTRACT FROM AUTHOR]