The establishment and description of the model is the basis of seismic data analysis and processing, and the reasonable model parameterization method has achieved good results for all aspects of seismic forward and inversion. Ray tracing plays an important role in the fields of forward modeling, tomography, and migration. The authors studied the ray tracing method, designed a model of least squares parametric method,parameterized the model of velocity and travel time respectively, and then tested it by using the velocity gradient model and error analysis. The results show that the model parameterization improves the precision of travel time calculation and the tracking ray path. The validity and applicability of the model parameterization method are proved by the example of the classical Marmousi model and the Sigsbee 2A model.

Seismic refraction exploration is one of the most widely used methods in engineering and environmental geophysics. As a simple, convenient and economical exploration method, it can provide the engineering geology with the fluctuation of the ground floor, the variation of velocity,and the variation of the diving surface. With the development of engineering geological and urban geological exploration, the data processing and interpretation of seismic refraction wave method is even more important. In this paper, we mainly introduce the development history and application conditions of the shallow seismic refraction wave method, and summarize and discuss the principle of several main refractive wave interpretation methods. On this basis, we compare and discuss the advantages and disadvantages of various methods in the application, expounds the shallow refraction wave method at home and abroad in recent years, the development status and trend, and emphatically introduce the research focus as a method of refraction tomographic imaging. The research shows that when the detection depth is relatively shallow and the boundary of the interface is sufficient to smooth the angle of the interface, the most convenient interpretation method is the intercept time method; when the depth of exploration reaches 25 m, the t₀ difference method is most applicable; the generalized interchange method should be used when the exploration target is buried deeper than 25 m. As its high precision, the use of refraction wave traveltime tomography technology can satisfy the near-surface strata with horizontal velocity changes, including the formation of large dip angle, concealed layer, interface fluctuation, etc.

Most conventional ground penetrating radar (GPR) is single polarimetric radar, which can only obtain single polarimetric data, and is difficult to identify the pipeline accurately and quickly in a complex environment. In order to identify the pipelines in a complex environment, we applied the method of full polarimetric GPR to identify the pipeline. We extracted the polarimetric attributes of a single pipeline and the polarimetric attributes of the pipeline influenced by other targets. The results show that the target can be well recognized,although its polarimetric attributes are under the influence of other targets. The method of full polarimetric GPR can get more comprehensive polarimetric information of a target,and can effectively and quickly identify a pipeline in a complex environment.

The classification of remote sensing image is an important part of extracting effective information of images. In order to explore the optimal classification methods, many machine learning algorithms are gradually applied to the classification of remote sensing images. Because of the high efficiency, speediness, and good performance generalization, extreme learning machines (ELM) have been widely used in pattern recognition. This paper aims to classify high-resolution remote sensing images, and analyze the performance of extreme learning machine algorithms for the accuracy of classification of remote sensing images, and compare ELM algorithm with support vector machines (SVM) algorithm and Maximum likelihood method. The GF-2 data from some area in Changchun City were selected to test the accuracy of all the three methods for classification with the fused image as the original data. The results show that the overall accuracy of ELM algorithm is more than 85%, and the kappa coefficient is 0.718. Compared with the other two classification methods, the ELM classification accuracy is the best, and its running time is faster than the support vector machine by 2 480 s, 1/8 of the support vector machine's. An even better performance can be obtained in a shorter training time. ELM is valuable for classification of remote sensing images.

To compare the pre-stack migration in the time domain (PSTM), the pre-stack migration in the depth domain (PSDM) can adapt to the seismic data which has the high dip angle and the large variation of the velocity in the horizon. Based on the imaging principle of Kirchhoff pre-stack depth migration method, we established the initial velocity model at first; then added the target geological body-Salt Dome to the initial velocity model,and further optimized it;under the constraints of the horizon of the salt dome, we computed the difference of two travel times through a lot of iterations so as to update the initial velocity model. By the updated initial velocity model, the Kirchhoff integral method is used to migrate the seismic data. The job can be finished if the corrected reflection phase of the common imaging point gather is at the same level;if not, the horizon of the salt dome has to be modified again and again. We can modify the horizon by use of the well logs and the geological knowledge or consult with the geological experts. Taking this new modeling approach, the latter image has higher S/N and resolution compared to the results of the PSTM imaging methods under the constraints of the horizon of the salt dome. The specific performance is the clear breakpoint, outstanding wave group characteristics; especially, the outline of the salt dome has very clear boundaries, and the bottom of the salt dome image is also much more improved than that of the image of the PSTM.

With the dramatic development of the geophysical equipment and exploration technology, the demand of rapid processing of large-scale geophysical data has been increasing. In order to solve the time-consuming problem of three dimensional gravity data inversion, a parallel implementation of preconditioned conjugate gradient algorithm is proposed to improve the computation efficiency. We adopted and tested two different preconditioned strategies with synthetic data, and evaluated their acceleration quality through observing the residual error of iterations and the running time. The results indicate that the symmetric successive over-relaxation (SSOR) preconditioner has a better performance with density results closer to the true model. Compared with the conventional conjugate gradient algorithm in serial code, our parallel preconditioned conjugated gradient algorithm can result in up to approximately 19x speedup. We used it to the real gravity data from the Vinton salt dome in USA, and the inversion results well delineated the location of the cap-rock, which verified the high efficiency and feasibility of this parallel preconditioned conjugate gradient algorithm in 3D fast gravity inversion.

In potential field data processing, high-order vertical derivative has important physical significance. It can divide superimposed anomalies generated by sources with different depth and size. With the order increase of the derivative, the resolution becomes higher, however the calculation of the higher order derivative is unstable. In this study we proposed the Tikhonov regularization iteration method for high-order vertical derivatives based on the combination of Tikhonov regularization method and iteration method,and obtained the recursive formula of Tikhonov regularization iterative method. Through analyzing the filter characteristics of the method, we can see that the method has certain stability and amplitude preservation in the calculation of high-order vertical derivative. The results of the model test and the real data show that the stability and practical value of the method are higher than that of the routine FFT method.

Borehole radar is a kind of ground penetrating radar (GPR). Conventional borehole radar can determine the depth of the geological body, but it's unable to determine the azimuth of geological target precisely. Directional borehole radar can determine not only the depth of the geological body, but also the orientation of the geologic body. In this paper, the directional borehole radar, consisting of one transmitting antenna and four receiving antennas, is applied to determine the positions of several targets by the modified residual algorithm through adding moving windows to the traditional residual algorithm. A three dimensional imaging method is discussed in this paper too. We used the model of synthetic data obtained with azimuth angle to realize 3D imaging, and 3D imaging results in the east and south direction of the well showed an obvious anomaly, which is in good agreement with the original model.

In order to compute the global-scale electromagnetic induction responses, this paper introduces a 3-D forward method of the frequency domain geomagnetic depth sounding (GDS). The method is based on the Staggered-grid finite difference method (FDM) in spherical coordinate. The difference equations are derived from the integral form of Maxwell equations. The PARDISO Solver is used to solve the discrete equations to avoid the divergence correction of iterations. To validate our code, we compared the results to the solutions of the finite element method and FDM. The relative error between the finite difference numerical result of the three-dimensional staggered grid and the analytical solution is less than 5%. The accuracy of the two hemispheres model is also high enough. The calculation results of the three-dimensional checkerboard model show that the magnetic field components have a good resolution to the size and position of an abnormal body.

On the basis of the traditional methods for edge detection of potential-field data, various field source edge enhancement techniques have been studied to improve signal-to-noise ratio (SNR) and localization accuracy. But problems still exist, such as noise interference and divergence of source edges. The authors made an improvement based on the traditional methods of edge detection using wavelet combined with traditional methods to enhance the ability of noise suppressing, and put forward power transform to make edges convergent effectively. In this paper, the principle was introduced and the method was applied to edge detection. Three commonly used numerical methods for edge detection were selected and compared with the modified methods in edge detection effect with the model data, gravity anomalies of Sichuan basin, and the magnetic anomalies in Zhurihe area. The results showed that the method of wavelet combined with power transform could effectively detect the edges of geological bodies in the study area. It plays a good role in noise suppressing, and can be used to identify the edges with convergence. The proposed method improves the accuracy of edge detection and obtains satisfactory results.

In the application of airborne electromagnetic detecting method, underground detection environment is complex, and the physical parameters of underground media are anisotropic. If we use the conventional model of isotropic medium, serious deviations will occur during the data interpretation. This paper presents three-dimensional frequency-domain airborne electromagnetic modeling in an arbitrary anisotropic medium based on the vector finite element method. By decomposing the total field into primary and secondary fields, the uniform space of the air medium is subjected to the analytic calculation of the primary field, and the vector finite element method is used to solve the double-rotation equation of the secondary electric field. The parallel solution of the large-scale sparse matrix is calculated by using the shared-memory direct solver PARDISO, so that the 3D model calculation is greatly speeded up. Then the 3D anisotropic media simulations of four typical target-models are carried out, including isotropic rock-anisotropic target (rotate around the z axis) model,isotropic rock-anisotropic target (rotate around x axis) model, isotropic rock-anisotropic target (rotate around z axis) model, and anisotropic rock-anisotropic target rotate (around x axis) model. The change characteristics of the real and imaginary components of the magnetic field under different rotating angles in different models are analyzed and compared, and the influence law and identification method of the anisotropic parameters in airborne electromagnetic response are summarized. The research can provide a reference for the accurate interpretation and inversion of the airborne electromagnetic data.

Alpha Particle X-Ray Spectrometer (APXS) is one of the payloads of Chang'e-3 Yutu Lunar Rover for China's Exploration Project. Based on the raw data processing, which acquired by APXS, the accumulated characteristic X-ray count rate spectrum were obtained. After the data filtering with wavelet filter, the influences of radioactivity fluctuation were eliminated. By using the energy calibration formula, the lunar regolith's elemental characteristic X-ray energy spectrum was obtained. Through an analysis, the elemental composition of the lunar regolith at the lunar rover landing point consists of Si, Ca, Al, Mg, K, P, S, Fe, Ti, Ni, Cr, Mo, Sr, Y, Zr, and Nb. This result provides certain favorable parameter constraints for the research on the lunar crust geological evolution.

Santanghu basin is a typical volcanic oil and gas basin in the west of China. Based on the high precision aeromagnetic data, integrated interpretation was completed, combining with the regional geology, Bouguer gravity, field reconnaissance,and magnetic susceptibility information. Firstly, the magnetic and gravity characteristics of Santanghu basin, as well as the relationship with the structure division were described in detail; and then, the volcanics distribution range was delineated, and the magnetic basement relief was revealed with the depths of the magnetic body calculated by the Euler deconvolution method; finally, the oil-gas-bearing prospecting was conducted based on the aeromagnetic data, together with the volcanic reservoir features and hydrocarbon accumulation pattern. The sub-sags with concealed volcanics distribution in the center of Santanghu basin depression are favorable for oil reservation. Two sub-sags in Hanshuiquan and one sub-sag in northern Tiaohu are key region for oil gas exploration. Shitoumei and Chahaquan uplifts are important regions, and Fangfangliang uplift and the southern area may be the pre-study regions. The exploration direction and prospective areas presented in the paper can provide reference for the oil and gas exploration in the future.

The latest high-precision aeromagnetic data is used to interpret the regional magnetic characteristics of the East Kunlun-Altyn Tagh area in the light of the different magnetic anomalies, which strik generally in NE and nearly EW, and remarkable for their obvious partitioned and segmented features. With the assistance of the Satellite gravity data and the processes of the analytic continuation and horizontal derivative, the geophysical and geological data revealed the magnetic anomaly of the fault structure, especially the deep fracture. The magnetic anomaly is mainly characterized by the boundary, gradient zone, and beaded anomalies. The faults act not only as the important tectonic boundaries in this region, but also the important tectonic belts, controlling the distribution of mineralization. Based on the magnetic interpretation and practical applications, we described the relation between the deep fracture and mineralization. In the East Kunlun-Altyn Tagh region, many known medium and pint-sized deposits are located nearby the deep fractures.

The full waveform inversion (FWI) has been used as a high resolution imaging method for cross-hole radar inversion; however, it is still trapped by local minimum. An effective way is to provide an accurate initial model for FWI. The fact is that the underground media is unknown, and the low-frequency information are usually not available in the GPR field data, which lead to difficult to achieve accurate result in FWI. The gradient formula of envelope waveform inversion (EWI) is deduced in detail by using the method of derivation of the envelope objective function. In comparison with the inversion result of the traditional FWI, the EWI has a better inversion ability for low-frequency missing data, and can effectively restore the missing low-frequency information. It can provide quantitative interpretation for subsurface media.

A conventional GPR system includes computer, network analyzer, rectangular coordinate robot, and a single antenna mode for transmission and reception, responding only to the co-polarization signal. However, we expect that more polarmetric information can be obtained. So we developed a full-polarimetric GPR system including computer, vector network analyzer, position controller, switch driver, and polarimetric multiple coverage antenna array. The arrangement of the antenna array consists of ten Vivaldi antennas oriented perpendicularly to each other that can obtain co-polarization and cross-polarization signals for each transmission. One group of experiment was performed. This group targeted to a metallic plate and a metallic dihedral. The result of this experiment shows that this GPR system with the proposed antenna array is capable of detecting the existence of buried targets, and the application of the polarimetric multiple coverage antenna array can improve the signal-to-noise ratio dramatically. We use Pauli polarimetric decomposition technique to obtain subsurface colour-coded reconstructed object images, which can be employed to interpret both the geometrical information and the scattering mechanism of the subsurface objects.

In this paper, we apply the instantaneous attribute analysis method to ground penetrating radar (GPR) antenna-target polarization to detect the Light Non-Aqueous Phase Liquids (LNAPL) contaminated soil. We set up a model of LNAPL contaminated soils using quartz and diesel in laboratory and use the GPR to detect the contaminated soils with 0° antenna and 90° antenna, separately. The shape, size, and structure of LNAPL soil are not regular. Therefore, we can analyze the antenna-target polarization characteristics through 0° antenna measurement and 90° antenna measurement, and get the underground material information such as the direction and shape of the target. Through analyzing the instantaneous attribute of the pre-processed information including antenna-target polarization characteristics, the LNAPL contamination range can be clearly identified, and the detection ability of GPR can be improved.

The development of the massive data processing, interpretation and integration of the modeling workflow software technology based on the high efficiency and high accuracy rapid moving platform joint detection technology would provide practical and feasible technical guarantee for the effective implementation of large-scale resource exploration investigation and major prospecting breakthrough targets under the complex environmental conditions. Based on the existing seismic data processing and interpretation, the combination of the advanced seismic software platform with non-seismic data analysis is improved utilizing the introduction and independent research and development, the route of the "red blue" dual-track joint, and an integrated software platform with a cross-platform data fusion, secondary development and data sharing management. It fills the gap in the joint analysis of seismic and non-seismic data in foreign software products, and provides technical support for the data analysis of deep earth exploration.

The authors deduced that the distribution of the horizontal derivative of the gravity anomaly of the fault structure can realize the recognition of the fault tendency, which indicates the tendency of the fracture on the side of the small change rate of the fracture horizontal derivative, so it can directly identify the fracture characteristics and avoid the complicated operation of the previous methods. The technique is applied to the treatment of gravity anomaly in Songliao basin, and the tectonic characteristics of Songliao basin are obtained. There are obvious differences of gravity anomaly between the east and the west of Songliao basin, and the reasons of the difference are difficult to be revealed by electric method and seismic data. The boundary inversion of Songliao basin is conducted by gravity data, and the interface features of Moho surface and below are obtained by using seismic data. It is proved that the difference of gravity anomalies in Songliao basin is caused by the boundary of 10km under Moho in Songliao basin, and the great fluctuation occurs in Songliao basin due to the subduction of the Pacific plate.

Using the gravity and magnetic data with wide area and deep exploration depth, we use enhanced local phase boundary identification and optimization Euler deconvolution to get the structural distribution of the East China Sea shelf basin. The enhanced local phase boundary identification method is the ratio of function data of the first order and the second order horizontal derivatives, which can effectively balance the different depth of the geological body boundary response, and improve the resolution of the deep structure; the optimization of Euler deconvolution method can effectively reduce the interference of background anomaly to result in more accurate inversion results. The theoretical model shows that the enhanced local phase boundary identification method can be more accurate and clear, the optimized inversion results of Euler deconvolution method can be more accurate and convergent. The comprehensive use of enhanced local phase boundary identification and optimization Euler deconvolution method can accurately divide the distribution of the East China Sea continental shelf fault structure, which has important reference value for guiding regional oil and gas exploration.

In order to evaluate the quality of observation,we interpreted the horizontally layered homogeneous medium model of Dabaishe station according to the electric sounding data. The measurement arrays at the station are the fixed Schlumberger arrays. The deep-well measurement has two set of arrays. One was horizontally placed shorter than 100 meters, and the other was vertically arranged. Under the layered model, we analyzed the sensitivity coefficient variations of deep-well apparent resistivity measurement versus the different electrode depth and the spacing of current electrodes. The results show that the horizontal deep-well measurement has strong ability on inhibiting from surface disturbances. However, its current electrodes spacing is relatively short, only 60 meters, thus the measurement data didn't effectively reflect the resistivity variations from deep mediums. The electrodes of the vertical deep-well measurement array is located within the layer where underground water level changes; therefore, its measurement data suffer a strong affection from water level variations, and it is very difficult to distinguish the resistivity variations of deep mediums from the data. In order to better detect the information associated with earthquakes, the spacing of horizontal arrays of AB/2 should be set as 250-350 m, and the bury depth should exceed 150 m. A horizontal array along NS direction is also recommended for a better result.

A set of high pressure metamorphic rocks are exposed in the Toudaoqiao area, Inner Mongolia. In this study,the investigations on petrography and mineral chemistry are carried out, and two types of metamorphic rocks are classified to blueschist and greenschist on the basis of the mineral assemblages. The mineral assemblage of the blueschist is amphibole+epidote+albite+chlorite+quartz+hematite±phengite±calcite±titanite, and the mineral assemblage of the greenschist is chlorite+albite+quartz±amphibole±epidote±phengite±hematite. The blueschists were suffered from the epidote-blueschist facies metamorphism, with the peak metamorphic conditions of T=400-600℃, p=1.2-1.4 GPa. The greenschists were suffered from the epidote-amphibole facies metamorphism. According to the previous geochronology data, the high pressure blueschist and greenschist are related to the suture zone between Eerguna and Xing'an massifs.

The Gucheng lower uplift is an important exploration area in eastern Tarim basin, and its good prospect has been proved by the success of the Gucheng Well 6, Well 8 and Well 9. Years of exploration practices in Tarim basin show that the faults play an important role on controlling the formation and accumulation of reservoirs, are of great significance for oil and gas discovery. Based on the analysis of geometric characteristics of faults, the authors established the Paleo-Mesozoic fault development pattern based on the 2D-3D seismic data of this area. Based on the Multi-window angle sweep technique, the planar characteristics of the concealed strike-slip faults were identified. The deformation mechanism of the faults in different stages was analyzed systematically. The influence of the tectonic evolution on hydrocarbon accumulation was studied in detail. The study shows that there are seven different tectonic styles in this area, including extensional normal faults, torsional negative flower structure, positive "Y" structure, inverse-"Y" structure, compresso-shear structure, near vertical strike-slip faults, torsional strike-slip positive fault. The faults have three plane development directions, NW, NE and NNE. According to the geometric elements and dynamic characteristics, the faults activity period is divided into Early Caledonian stage, Middle Caledonian, Late Caledonian-Early Hercynian stage, and Indo-Yanshan stage. The Middle Caledonian faults are important channels for the upward migration of oil, gas and the lower fluid through establishing the framework of graben block and horst block structure. The "pinnate" faults zone formed in the Indosinian -Yanshanian were caused by the re-activation of Late Caledonian-early Hercynian faults, which has a strong damaging effect on the reservoirs.

Deep seismic reflection technology is an important method to detect the deep structure of large-scale metallogenic belt and ore concentration area and reveal the background of mineralization. The authors explored the subsurface structure of the magmatic fluid passage in the vicinity of the Tibetan Niangre ore concentration area with iron, copper and other polymetallic deposits in the middle section of the Gangdese metallogenic belt, revealed the background of the mineralization and the corresponding fracture formation. Through the data processing such as tomostatics, de-noising, and predictive deconvolution of the 130 km deep seismic reflection profile crossing the Niangre ore concentration area, we obtained the reflected images of the upper crustal structure of the Niangre ore concentration area. Based on the geologic and geophysical data in this area, we studied the upper crustal structure of the ore concentration area, and found out the spatial relationship between the seven existing faults in the area and the distribution of the veins in the Gangdese metallogenic belt, and point out that some special structures such as the bright spot structure control the magma activity rising channel, and closely connected with the deposits.

In order to investigate the water environment of the Second Songhua River, based on analyzing the content of heavy metals(As、Cd、Cr、Cu、Hg、Pb、Zn) of the sediment in Erdaosonghua River and its tributary, the characteristics of heavy metals were discussed,and the potential ecological risks of heavy metals in sediment were assessed by the method of potential ecological risk index. The results show that the content of heavy metals (As 4.47×10^-6,Cd 0.158×10^-6,Cr 35.81×10^-6,Cu 10.57×10^-6,Hg 0.040×10^-6) in sediment is relatively low, with uneven distribution. The potential ecological risks of the heavy metals in sediment of Erdaosonghua River belong to "slight" level;while the risk level is even high in Jinyinbie River, with mercury as the primary risk.

Jidong region, located in the North China platform and eastern Yanshan platform fold belt, is one of the important metallogenetic gold provinces in China. There are plenty of granites in this region, such as Dushan granitic batholith, Shepantu granite batholith, Qingshankou, Gaojiadian, Jiajiashan, Yuerya, and Luowenyu granites. The LA-MC-ICP-MS zircon U-Pb ages of the Gaojiadian and Shepantu granites show that the formationtime of the Gaojiadian granite is (198.8±0.75) Ma,and the Shepantu granite is (184.7±1.0) Ma, belonging to the Early Jurassic of Yanshanian. In the study area, the rocks are of I-A transition type with weakly peraluminous characteristic and high-K calc-alkaline. Among them the Early Jurassic granitoids is much more enriched in (Na₂O+K₂O), LILE and LREE, and poor in HFSE. Besides, it has obvious negative europium anomaly with the δEu of 0.14-0.82. All these show that the diagenetic material came from the shell and mineralized after the collision.