【作者】 廉玉广；

【导师】 韩立国； 吕庆田；

【作者基本信息】 吉林大学 ， 地球探测与信息技术， 2011， 博士

【摘要】 传统的金属矿勘查主要依赖重、磁、电等方法,一般适应于中浅层的矿产勘查,当探测目标埋深逐渐增大时,由于方法本身的限制,异常强度和分辨率急剧下降,难于适应深部矿产勘查的需求。地震方法具有探测深度大,分辨率高和探测结果准确可靠等特点,成为深部矿产勘探最有前景的技术。为了验证金属矿区开展地震勘探的可行性,在庐枞地区罗河、泥河和大包庄矿区开展了金属矿区的地震试验,包括地震的数据采集和处理解释以及地质-地球物理模型的建立和地震波场正演模拟等研究。与石油地震勘探相比,金属矿地震勘探有其自身的技术难点：比如金属矿区往往地质构造非常复杂,地层连续性差,探测目标小,岩石蚀变严重,地下介质不均匀性非常强,矿区通常背景干扰严重。采集的地震资料信噪比较低,缺乏标志层和测井资料,金属矿区的地震解释十分困难。金属矿地震勘探一般需要在地质构造复杂的火山岩和变质岩有关的山区或丘陵地带进行数据采集。野外数据采集时应根据具体地质条件和参数试验合理布设激发炮点和选择采集参数,并保证检波点与地面的良好耦合,确保采集质量。金属矿区地震数据处理时,经过对资料信噪比、频率、干扰波、静校正、动校正和速度分析等问题的具体分析及应用针对性处理技术措施后,资料品质得到明显提高,构造特征也较为清晰。合理的应用叠前和叠后去噪,提高了资料的信噪比；应用“无射线层析反演静校正”技术,较好地解决了研究区因高差和低降速带厚度变化等而引起的静校正问题,同相轴连续性有所提高；高精度交互速度分析及分频剩余静校正的应用,更加直观也更加准确的拾取叠加速度,较好地解决了低信噪比区域的速度拾取及短波长静校正问题；应用“无拉伸NMO”技术较好地解决了因拉伸畸变切除影响浅层成像的问题,浅层资料得到较大改善。由于研究区有较多钻孔信息和岩石物性测量数据,有利于开展矿区模型的地震正演模拟。选取地震测线附近的钻孔,将钻孔投影到地震测线上,并根据这些钻孔信息,向两侧外推岩性分布,绘制出地质模型。在矿区进行物性参数测量,并统计出各种岩性的密度和速度转换关系,建立合理的物性参数库,将物性参数赋予地质模型的各个地质体,使之转换为复杂的地球物理模型；为研究地震方法对矿体和围岩的分辨能力,将复杂模型进行适当简化变为简化的地球物理模型。然而,地下介质往往并不是均匀的,尤其是在构造活动剧烈,岩石蚀变严重的金属矿区,这种不均匀性更为显著,通常建立的均匀介质模型并不符合实际地下情况。随机介质就是非均匀介质中较为典型的一种,它主要考虑了小尺度条件下的介质非均匀性。自相关函数、尺度、方差(或者标准差)、粗糙系数是随机算法的四个要素,另外为了描述随机介质的方向性,加入了倾角参数。粗糙系数从小到大,局部由粗糙变得逐渐平滑；方差越大,介质速度扰动越大；横、纵尺度的对比关系体现了介质整体的分布特征,当横纵尺度比为几倍时,介质扰动呈现断续的似层状分布特征,当横向尺度远远大于纵向尺度时,介质呈现层状分布特征；不同倾角的随机介质能更准确的体现地下介质扰动的方向特性。本文通过建立矿区的多种随机介质模型,讨论了随机参数对正演模拟地震波场的影响。地震波场正演模拟技术,因其直观、形象地反映地下介质中波场传播特征和规律,对人们理解矿体波场特征、解释实际地震资料,表征地下介质结构与岩性特征等,均具有重要的理论和实际意义。金属矿区地质构造复杂,矿体形态多变,地震资料的信噪比一般较低,波组特征不太明显,地震解释十分困难,对比研究正演模拟结果和实际资料,有助于认识和识别各种地震波场特征。本文用波动方程交错网格有限差分法对各种理论模型进行正演模拟,并对模拟结果做纵波叠后、叠前时间偏移和转换波叠前时间偏移处理。结果发现：复杂模型中很多同相轴出现了一些间断,但整体形态能看到矿体和主要目标层位的反射信息；简化模型同相轴比较连续并且更清晰,突出了矿体和主要目标层位的反射信息,易于识别和分辨矿体和围岩的反射；随机介质模型波场复杂,产生了很多小的反射,形态上和实际资料更相似,偏移剖面仍能看到一些波阻抗差较大层位的反射,能够看到矿体反射。CMP叠加剖面能量在倾斜层位和矿体上没有正确归位,绕射现象明显,需要进行偏移处理；叠后克希霍夫时间偏移使同相轴归位,能够大体确定矿体的分布；叠前克希霍夫时间偏移使能量聚焦效果更好,同相轴更清晰；转换波叠前时间偏移分辨率明显提高,同时矿体边界反射更连续,矿体下边界及陡倾角层位的反射更清晰,在纵波能量不足的地方转换波能量相对更强；处理时应结合叠前克希霍夫时间偏移和转换波叠前时间偏移共同进行解释。通过改变地震子波主频和随机介质模型中的各种随机参数,讨论频率和随机参数对地震波场的影响。发现：频率越高,地震分辨率越高,地震子波主频大于70Hz时能分辨一些复杂矿体内部的反射：随机方差大于5%时,火山岩层的反射多数淹没于随机介质产生的杂乱波场中,但由于矿体和围岩有较大波阻抗差,依然能见到矿体反射；横向相关长度较大时,产生断续的似层状地震波场；几组不同相关长度随机介质构成的组合模型产生的波场包含了这几种相关长度的组合特征；粗糙系数越大,随机介质越平滑,地震波场杂乱的现象有所减弱。另外通过建立立陡状矿体随机介质模型,并进行正演模拟,认为当矿体的随机方差远大于围岩介质随机方差时,能够看到矿体区域内部的反射和散射特征,因此随机方差相差很大时用地震方法可以探测到陡倾矿体。根据区域地质情况和测线附近的钻孔信息,结合正演模拟和实际采集资料的地震波场特征进行地质解释,发现了沉积红盆的“箕状”结构和内部分层结构及火山岩层的3层结构。对偏移剖面进行属性分析,发现不同性质的岩石其地震动力学特征存在明显差异。从频率上看,沉积岩低频成分丰富,有第四纪覆盖的火山岩区也含低频成分,而第四纪覆盖较少的火山岩区则缺少低频信息,火山岩区高频信息相对丰富。沉积岩反射同相轴连续性较好,能量强,火山岩层和矿体反射连续性较差,多见一些凌乱的强能量团。小波时频分析的分频剖面能更好的观测每个频段各个不同反射区域的能量特征。总体看来,矿体规模较大,产状相对平缓时,在地震剖面上可能直接观测到矿体反射信息。另外火山构造对成矿具有重要的控制作用,如果能够对控矿构造进行有效探测,则可以间接地为找矿提供有用信息。由于金属矿区地下反射的复杂性,直接探测矿体还存在一定问题,核心是如何区分矿体反射与其它反射,因此,如果作为一种勘探方法在未知区应用时,更可信的是追踪和探测已知控矿地质体,或控矿构造的深部延伸。更多还原

【Abstract】 Traditional metal mineral exploration depends on gravity, magnetic, electric and other methods, generally adapting in the shallow. When the probe target depth increases, abnormal strength and resolution sharp decline because of the limitation of the method itself. So it is difficult to adapt to the needs of deep mineral exploration. Seismic exploration method has great depth, high resolution, and is accurate and reliable, is the most promising technology in deep mineral exploration.To verify the feasibility of seismic exploration in metal mineral region, we carried out seismic experiment on Luohe, Nihe and Dabaozhuang deposits in Luzong area, including seismic data acquisition, processing, interpretation, geological geophysical model construction, seismic wave field modeling and other research.Compared with the oil seismic exploration, the seismic exploration in metal mining area has its own technical difficulties:Such as geological structure in metal mining is often very complex, stratigraphic continuity is poor, the detection targets is small, rock alteration is serious, subsurface heterogeneity is very strong, and mine has usually a serious background interference. Seismic interpretation of metal mining is very difficult for the low signal to noise ratio of seismic data, the lack of marker beds and log data.Seismic data acquisition generally occurs in mountainous or hilly areas where volcanic metamorphic rocks have complex geological structure. Field data acquisition should select reasonable shot position and acquisition parameters based on circumstances and parameters test, and ensure receivers coupled well with the ground, ensure acquisition quality.Metal mining seismic data processing, through the specific analysis of SNR, frequency, interference wave, static correction. NMO correction, velocity et al., and the application of specific measures, while processing seismic data of metal mining, data quality has been significantly improved, structural features is also more clear. Reasonable application of prestack and post-stack noise removal, the SNR of seismic data is improved; application of "no-ray tomography inversion static correction" technology, static correction problem due to elevation changes and low-velocity layer thickness is better resolved in the study area, and the continuity of phase axis is improved; The application of high-precision interactive velocity analysis and residual static correction of frequency splitting, more intuitive and more accurate stacking velocity picking, can solve the speed picked up and short wavelength static problem in the low SNR region; Application of "no stretch NMO" technology can solve the shallow imaging problems affected due to removal of the stretch distortion, shallow data improved greatly.Since the study area have more information of boreholes and petrophysical measurement data, mining model is conducive to carry out seismic modeling. Select boreholes nearby seismic lines, and project onto the seismic, and extrapolate to both sides of the rock distribution based on these drilling information, then draw a geological model. Take the petrophysical parameters measurement, and establish reasonable petrophysical parameters library using the statistics results of the relationship of density and velocity, and then convert the geological model to the complex geophysical model by put the petrophysical parameters to the geological bodies; To research the resolution of Seismic methods to the ore body and surrounding rock, complex model will be transformed to a simplified geophysical model.However, the underground media are often not uniform, especially in the metal mining with intense tectonic activity and serious alteration of rock, this heterogeneity is more significant. The homogeneous medium mode established is often not meet with the actual ground conditions. Random media is a typical heterogeneous medium, it is mainly considered under the conditions of small-scale. Autocorrelation function, scale, variance (or standard deviation), roughness coefficient are the four elements for the random algorithm. Roughness coefficient from small to large, local media change from rough to smoothing; Variance more larger, the disturbance of velocity is more larger; The relation of horizontal and vertical scale reflects the overall distribution of the media, when the ratio of horizontal and vertical scale is several times, the media shows like layered distribution with intermittent disturbances, when the horizontal scale is much larger than the vertical scale, the media presentation layer distribution; Different angle of the random medium can more accurately reflects the direction of subsurface media disturbance characteristics. Discuss the influence of random parameters to the forward modeling of seismic wave field by establishing a variety of mining random medium models.Seismic wave field modeling technique have important theoretical and practical significance, for its visual image to reflect the characteristics and the law of subsurface wave field propagation, people understand the wave field characteristics of the ore bodies, to explain the seismic data, characterization of subsurface structure and lithology. Geological structures are complex in metal mining, ore body shape is changeable, SNR is generally low, characteristics of composite waves is less obvious, seismic interpretation is very difficult. Compared to the results of forward modeling and practical information to help understand and identify characteristics of various seismic wave field.The various theoretical models are modeled using wave equation with staggered-grid finite difference method, and then do wave post-stack, the pre-stack time migration of p-wave and pre-stack time migration of converted wave. The results showed that:phase axis is intermittent partly in complex model, but the reflection information of overall shape of ore bodies and the main objective layers can be seen; The phase axis is more continuous and clearly in simplified model, highlight the reflection information of ore bodies and the main objective layers; The wave field is very complex in random medium model, resulting in many small reflections, is more similar to practical information in form, Some reflection of ore bodies and the layers with larger impedance difference can be still seen in the migration profile. Energy of CMP stack section in an tilted layer and ore bodies does not have the correct homing, diffraction phenomenon is obvious, so needed for migration; Kirchhoff post-stack time migration can generally replace phase axis and determine the distribution of ore bodies; Kirchhoff pre-stack time migration is better to focus the energy, has clearer phase axis; Converted wave pre-stack time migration significantly improve the resolution, while the reflection of ore boundary is more continuous, the reflections of the lower boundary of ore bodies and the steep angle layers are more clearly, converter Wave energy is relatively stronger where in the lack of p-wave reflection energy place; So p-wave Kirchhoff pre-stack time migration and converted wave pre-stack time migration should be combined to do interpretation.Discuss the influence of frequencies and random parameters to seismic wave field, through changing the seismic wave dominant frequencies and random parameters. Found:the higher the frequency, the seismic resolution is better, when the seismic wave frequency is higher than 70Hz, some internal reflection of the orebody can be distinguished; When random variance is greater than 5%, the majority reflection of volcanic layer submerge in the random medium wave field, but still able to see the ore body reflection for the large impedance difference between ore body and surrounding; When the horizontal correlation length is large, generate intermittent liked layered seismic wave field; when roughness coefficient is greater, the random medium on smooth, the messy phenomenon of seismic wave field has been declined. Through the establishment of steep-shaped ore bodies random medium model, and the forward modeling, the reflection and scattering characteristics within the ore bodies can be seen when the random variance of ore body is much larger than the surrounding rock. So when the random variance is large difference, steeply dipping ore bodies can be detected using seismic methods.Carry out geological interpretation based on regional geology and drillhole information near the seismic line, combined with forward simulation and actual data characteristics of the seismic wave field, and found that the deposition red basin’s "Kei-like" structure and the internal hierarchical structure and volcanic 3 layers structure. Attribute analysis of migration profiles can found that the different nature of the rock significantly different dynamic characteristics. From the frequency, the sedimentary rock is rich in low frequency, there are also low-frequency components in the volcanic region covered with quaternary, and volcanic region without quaternary coverage is lack of low frequency, high frequency information is relatively rich in volcanic areas. Sedimentary areas is better continuity of phase axis, the energy intensity, volcanic areas is poor continuity, and has some messy strong energy group. Wavelet time-frequency analysis can better observe the energy characteristics of different reflection region of the each frequency band profile.Generally speaking, the large and relatively flat ore bodies may generate effective information in the seismic reflection profile. And volcanic structures is very important to control the mineralization. If it can be effectively detected, then will provide useful information for prospecting. There are still some problems of direct detection of ore bodies, because of the complexity of the subsurface reflection, the core of the problem is how to distinguish between ore bodies reflection and other reflections. So when the seismic exploration method is used in unknown area, the tracking and detection of known ore-controlling geologic body or structures in the deep extension is more credible.更多还原