三维成矿预测数据整合过程不确定性研究——以宁芜盆地钟姑矿田为例

基于多元地学大数据的三维成矿预测方法是开展深部找矿预测的新方法和新手段,也是当前成矿预测领域的研究热点之一。然而,大数据具有高维、混杂、非精确等特点,其分析处理过程面临多重不确定性。多元地学大数据整合是三维成矿预测的最终环节,其存在的不确定性将直接作用于预测结果,影响进一步的找矿应用和风险评估。本文以宁芜盆地钟姑矿田为例,从大数据思维出发,定量分析和度量预测要素和数学模型在数据整合过程中存在的不确定性及对三维成矿预测结果的影响。结果显示,断裂构造、背斜轴部等预测要素的不确定性对三维成矿预测结果的影响最为强烈;数据整合模型中,较之Logistic回归模型和证据权重模型,神经网络模型可能具有更高的不确定性程度。进一步工作可通过增强上述预测要素的可靠性和有效性、采用更多的数据整合模型进行更为全面的不确定性分析和评价,以获得更为可靠的三维成矿预测成果,从而降低成矿预测和找矿勘探风险。     

Mapping mineral prospectivity for Cu polymetallic mineralization in southwest Fujian Province,China

In this study, both the fuzzy weights of evidence (FWofE) and random forest (RF) methods were applied to map the mineral prospectivity for Cu polymetallic mineralization in southwestern Fujian Province, which is an important Cu polymetallic belt in China. Recent studies have revealed that the Zijinshan porphyry–epithermal Cu deposit is associated with Jurassic to Cretaceous (Yanshanian) intermediate to felsic intrusions and faulting tectonics. Evidence layers, which are key indicators of the formation of Zijinshan porphyry–epithermal Cu mineralization, include: (1) Jurassic to Cretaceous intermediate–felsic intrusions; (2) mineralization-related geochemical anomalies; (3) faults; and (4) Jurassic to Cretaceous volcanic rocks. These layers were determined using spatial analyses in support by GeoDAS and ArcGIS based on geological, geochemical, and geophysical data. The results demonstrated that most of the known Cu occurrences are in areas linked to high probability values. The target areas delineated by the FWofE occupy 10% of the study region and contain 60% of the total number of known Cu occurrences. In comparison with FWofE, the resulting RF areas occupy 15% of the study area, but contain 90% of the total number of known Cu occurrences. The normalized density value of 1.66 for RF is higher than the 1.15 value for FWofE, indicating that RF performs better than FWofE. Receiver operating characteristics (ROC) were used to validate the prospectivity model and check the effects of overfitting. The area under the ROC curve (AUC) was greater than 0.5, indicating that both prospectivity maps are useful in Cu polymetallic prospectivity mapping in southwestern Fujian Province.     

宁芜盆地钟姑矿田典型铁矿成矿岩体Hf-O同位素特征及地质意义

钟姑矿田位于宁芜盆地南部,受矿田内格状构造控制产出一系列早白垩世(129~132Ma)闪长岩及二长岩类侵入体,与铁矿床的形成密切相关,而区域内成矿岩体的研究尤其是岩浆岩的起源和演化有待进一步探讨。本次工作以矿田内的姑山辉石闪长玢岩、龙山辉石闪长岩、白象山闪长岩、钟九闪长岩和太平山二长岩等与铁成矿有关的岩体为研究对象,开展LA-ICP-MS原位锆石Hf同位素和SHRIMP原位O同位素研究。结果显示,钟姑矿田各成矿岩体具有较一致的锆石Hf-O同位素组成,~(176)Hf/~(177)Hf值为0.282425~0.282695,εHf(t)值为-12.3~-3.9,tDM2范围在1.26~1.86Ga;锆石δ18O集中于3.34‰~8.05‰,捕获锆石的176Hf/177Hf比值为0.282324~0.282487,εHf(t)=-15.8~-10.1,锆石tDM2年龄在1.72~20.8Ga之间,锆石δ~(18)O相对集中于5.90‰~6.51‰。钟姑矿田岩浆岩起源于富集的岩石圈地幔,同位素特征指示其携带板块俯冲和蚀变洋壳信息,钟姑矿田的成岩成矿作用应与伸展背景下古太平洋板块俯冲有关。与长江中下游其他矿集区岩浆岩对比研究表明,包括宁芜在内的长江中下游地区岩浆岩具有相同的地幔源区,以铁矿床为主的宁芜和庐枞盆地岩浆岩源区成分以交代岩石圈地幔为主,而以铜陵、繁昌为代表的铜多金属成矿区的岩浆岩受长江中下游新元古代基底的更多影响。     

西藏铁格隆南铜(金)矿床三维模型分析与深部预测

铁格隆南铜(金)矿床是近年来在班公湖-怒江成矿带西段多龙矿集区新发现的超大型Cu(Au-Ag)矿床。本文针对铁格隆南矿区深部找矿问题,以现代成矿地质理论和多元地学信息综合分析技术为支撑,以构建矿床找矿模型为指导,依托数据库技术、3S技术、三维建模与可视化技术及地质统计学理论与方法,开展基于矿产地质、地球物理、地球化学等成矿条件及找矿标志的三维地质实体建模与矿化异常三维空间重构,将铁格隆南矿床的预测评价研究拓展到三维空间,揭示了区内成矿地质特征、地球化学及地球物理异常表征,据此探讨了矿床的成因及矿体分布特征。并在此基础上,开展了矿区的地质-地球化学-地球物理综合信息分析与预测评价,以期减少单一信息多解性和成矿条件不确定性,为铁格隆南矿区深部找矿工作提供参考。研究结果表明:在地质找矿模型指导下,基于深部成矿空间三维结构重构基础上的三维地质、地球物理、地球化学异常信息提取与综合分析,可以有效的识别成矿地质体和矿致异常信息,实现深部矿产资源靶区空间定位预测,为深部找矿预测研究提供了新思路。综合分析结果显示铁格隆南矿床深部找矿潜力巨大。     

河南栾川西沟铅锌银金矿床三维地质建模与深部找矿预测评价

随着矿区浅部矿的日益减少,深部找矿越来越受到重视,三维地质建模技术在成矿预测、资源定量评价等方面得到了广泛的应用.本文利用三维地质建模平台GOCAD中的三维建模技术及地质统计学等方法,基于收集测试得到的地质图、钻孔和采样点的地层、岩性、构造、品位等数据,构建了西沟铅锌银金矿区的三维地质模型,包括断裂构造模型、矿体模型及...     

Fuzzy inference systems for prospectivity modeling of mineral systems and a case-study for prospectivity mapping of surficial Uranium in Yeelirrie Area,Western Australia

A Mamdani-type fuzzy inference system for prospectivity modeling of mineral systems is described. The system is a type of knowledge-driven symbolic artificial intelligence that is transparent, intuitive and is easy to construct by geologists because they are built in natural language and use linguistic values. No examples are used for training the system and expert-opinions are incorporated indirectly in terms of objective mathematical functions, which reduce the possibility of over-emphasizing the known deposits usually used as training data. The cognitive reasoning of the exploration geologist is captured in explicit if–then type of statements written in natural language using linguistic values. Conditional dependencies in the exploration data sets are managed through the use of fuzzy operators. A case study for surficial uranium prospectivity modeling in the Yeelirrie area, Western Australia, is used to demonstrate the approach. In the output prospectivity map, the SE-NW trending Yeelirrie and E-W trending Hinkler's Well palaeochannels show high prospectivity, while other channels show very low prospectivity ranges. The known surficial uranium deposits fall in high prospectivity areas, although minor showings and anomalies in the southern part of the study area fall in low prospectivity areas. A comparison of the prospectivity model with the radiometric image shows that several channels showing high surface uranium concentrations in the NW and NE quadrants may not be prospective.     

宁芜盆地深部地壳结构和岩浆侵入体形态特征及其对成矿的启示

在宁芜火山岩盆地部署了三条深反射地震剖面,通过精细的数据采集和处理,最终获得了高质量且可靠的深反射地震和初至波层析反演速度剖面。依据对反射结构和速度特征的分析,获得了以下认识:(1)利用层析反演速度成果刻画了岩浆侵入体在宁芜盆地深部空间的大体形态和分布范围;(2)依据深反射地震剖面反射同相轴和莫霍面结构特征,大体确定了深部岩浆侵入通道路径;(3)宁芜火山岩盆地下部地壳在6.0s上下位置发生解耦,在盆地正下方形成了地壳级别的鳄鱼嘴构造,控制着宁芜火山岩盆地深部的岩浆和构造活动;(4)阐明了宁芜火山岩盆地深部地壳结构和变形特征。这些成果的发现对探讨宁芜火山岩盆地深部成矿和动力学过程研究具有重要意义。     

宁芜盆地白象山铁矿床成矿作用过程数值模拟

白象山铁矿床是宁芜火山岩盆地钟姑矿田中典型的玢岩型铁矿床,主矿体赋存于闪长岩和黄马青组砂页岩接触带部位的内带-正带,呈似层状产出。本文采用数值模拟的方法研究探讨白象山铁矿床成矿过程的动力学机制以及汇流容矿空间的形成。在建立白象山矿床典型剖面以及三维实体模型的基础上,选取典型剖面,基于FLAC 3D系统,对白象山铁矿床的充填过程进行数值模拟。模拟结果显示,白象山铁矿床存在容矿汇流空间,其形成受力-热-流体的耦合作用制约;扩容空间的形成可为矿质的沉淀以及交代作用提供有利的成矿空间,并为流体的汇聚提供有利场所,也表明白象山铁矿床的成矿过程与力学作用密切相关。本文的模拟研究充分揭示了白象山铁矿床成矿过程中的物理过程,岩石不同的力学性质以及接触带的形态是控矿的重要因素,这为进一步的找矿工作提供了依据。     

宁芜和庐枞盆地含矿岩体地球化学特征对比

宁芜和庐枞盆地是长江中下游地区重要的铁矿产区,两盆地内含矿岩体的岩石学和地球化学特征既具有一定的相似性又表现出一定的差异性.文章对两个盆地内主要含矿岩体的岩相学、主量元素、微量元素、稀土元素、Sr-Nd同位素和锆石Hf同位素等进行对比研究,以期进一步探讨两盆地中成矿岩体的差异.镜下观察结果显示宁芜和庐枞盆地主要含矿岩体具有相同的矿物组合,主量元素特征表明这些岩体具有中硅、富钾和高镁的特点,均为中-基性火成岩.与宁芜盆地含矿岩体相比,庐枞盆地含矿岩体的∑REE、∑LREE和∑HREE含量均较高,明显富集Th、U等大离子亲石元素,具有明显的Eu负异常,说明庐枞盆地含矿岩体的地壳物质较多.两个盆地含矿岩体的Sr-Nd同位素组成指示岩体均来源于富集地幔,庐枞盆地含矿岩体较低的锆石Hf同位素组成说明在岩浆侵入和演化过程中可能有更多的地壳物质加入,导致岩浆的粘稠度降低,更有利于庐枞盆地铁矿的形成.     

Genesis of Early Cretaceous porphyrite-type iron deposits and related sub-volcanic rocks in the Ningwu Volcanic Basin,Middle-Lower Yangtze Metallogenic Belt,Southeast China

ABSTRACT

The Ningwu Volcanic Basin (NVB), located in the Lower Yangtze Metallogenic Belt, is characterized by the widespread occurrence of porphyritic iron-ore deposits. These deposits are clustered in three areas within the basin: the northern, central, and southern districts. Our study shows that the differences are existed in geochemistry of the rocks and the mineralization of the deposits among these three districts. Though results of Sr–Nd isotopes indicate that the parental magma of the sub-volcanic rocks in NVB were derived from EMI and EMII, however, the sub-volcanic rocks from the northern and southern districts have higher ε_Nd(t) values (t = 130 Ma) and lower initial ⁸⁷Sr/⁸⁶Sr ratios (t = 130 Ma) than those from middle district. The rocks from northern and southern districts are similar to EMI in geochemistry; in contrast, rocks from the middle district are close to EMII. Meanwhile, the deposits from different districts show some differences in mineralization. The magnetites from the northern ore deposits are rich in Cu, Mo, Zn, and Mg and poor in Ca, Al, and Ti, which probably indicates that the magnetites from the northern ore deposits closer to hydrothermal genesis than to magmatic genesis, but the magnetites from the centre are closer to both magmatic and hydrothermal geneses. Based on the fact that a part of the sub-volcanic rocks is closer to EMI and another part to EMII in geochemistry, we suggested that the incorporation of the lower and upper crustal materials into the source regions of the sub-volcanic rocks has taken place by different manner and processes. Integrated with analysis of the tectonic evolution of NVB, we come to the conclusion that the sub-volcanic rocks have recorded the geochemical imprints of the subduction of Yangtze block towards North China block during Indosinian Period and the delamination of the lower crust of Yangtze block during Yanshanian Period.     

Back-propagation neural network and support vector machines for gold mineral prospectivity mapping in the Hatu region,Xinjiang, China

Machine Learning technologies have the potential to deliver new nonlinear mineral prospectivity mapping (MPM) models. In this study, Back Propagation (BP) neural network Support Vector Machine (SVM) methods were applied to MPM in the Hatu region of Xinjiang, northwestern China. First, a conceptual model of mineral prospectivity for Au deposits was constructed by analysis of geological background. Evidential layers were selected and transformed into a binary data format. Then, the processes of selecting samples and parameters were described. For the BP model, the parameters of the network were 9–10???1; for the SVM model, a radial basis function was selected as the kernel function with best C?=?1 and γ = 0.25. MPM models using these parameters were constructed, and threshold values of prediction results were determined by the concentration-area (C-A) method. Finally, prediction results from the BP neural network and SVM model were compared with that of a conventional method that is the weight- of- evidence (W- of- E). The prospectivity efficacy was evaluated by traditional statistical analysis, prediction-area (P-A) plots, and the receiver operating characteristic (ROC) technique. Given the higher intersection position (74% of the known deposits were within 26% of the total area) and the larger AUC values (0.825), the result shows that the model built by the BP neural network algorithm has a relatively better prediction capability for MPM. The BP neural network algorithm applied in MPM can elucidate the next investigative steps in the study area.     

基于重磁联合反演的宁芜盆地钟姑矿田深部地质结构解析及成矿预测

宁芜盆地是长江中下游成矿带中重要的金属矿集区之一,以中生代火山岩和铁矿床广泛发育为特征。近年来,随着找矿深度的不断加深,深部找矿的难度不断加大,成本越来越高,急需利用地球物理方法推断深部地质构造,为深部找矿提供依据,以减少勘探风险。为了解宁芜盆地南段钟姑矿田的深部构造、岩体、地层的空间展布,本文基于区域重磁数据,并以地表地质、钻孔地质等信息为约束,开展深部地质结构的反演研究,厘定了地层及岩体的展布,并结合已知矿床的重磁组合方式和控矿地质要素,基于重磁反演结果在钟姑矿田内圈定了两处有利靶区,为下一步找矿工作提供参考。     

红格铁矿三维反演与地质建模

四川红格钒钛磁铁矿是我国重要的岩浆岩型铁矿石产出基地。在过去几十年中,对红格地区进行了不同程度的勘探,积累了大量地质、物化探资料。但是,这些工作比较分散,不同单位不同时期用多种方法开展了多种比例尺的地质、物化探工作,资料零散的分布于不同单位,没有进行过系统的梳理及研究。本文对红格矿区进行有地质约束的三维磁异常反演,建立了三维地质模型,并实现了三维可视化。分析所建三维地质模型,从宏观上推测了红格铁矿的成矿控矿规律及岩浆运移通道,微观上预测了红格矿区矿体空间位置、形态及控矿、容矿、含矿地质体的分布规律。另外,结合地质模型还进行了找矿潜力评价,分析了矿区外围和深部的找矿前景。     

Three-dimensional reservoir architecture modeling by geostatistical techniques in BD block,Jinhu depression,northern Jiangsu Basin,China

To reduce or avoid some of the ambiguities in a heterogeneous reservoir, a fine three-dimensional model would be inevitably established with the application of geostatistical techniques. BD oilfield is a low-permeability fractured reservoir with many fracture types, which composes a multi-azimuthal fracture system, causing strong anisotropy and heterogeneity. Moreover, fracture provides not only accumulation space for oil and gas but also channels for migration of hydrocarbons and water dashing and thus plays a leading role in controlling BD reservoir production. With the continuous exploitation of oil in reservoir for more than 30 years, BD oilfield has currently stepped into high water cut stage, exposing many problems such as low oil recovery, poor exploitation stability, rapid oil production decline, and one-way water intrusion. Therefore, a detailed reservoir characterization, constructed to describe reservoir behavior under strong water drive in a well-developed fractured reservoir, is urgently needed. In this study, combined with three-dimensional geostatistical techniques, an accurate and efficient reservoir parameter model that occurs in a strong heterogeneous fractured reservoir has been constructed through stratigraphic correlation and sedimentary facies analysis. Hence, all the study presented above will be available for well location, remaining oil potential-tapping and oil recovery improvement during the later study work.     

Exploration targeting for orogenic gold deposits in the Granites-Tanami Orogen: Mineral system analysis,targeting model and prospectivity analysis

A major challenge for mineral exploration geologists is the development of a transparent and reproducible approach to targeting exploration efforts, particularly at the regional to camp scales, in terranes under difficult cover where exploration and opportunity costs are high. In this study, a three-pronged approach is used for identifying the most prospective ground for orogenic gold deposits in the Paleoproterozoic Granite-Tanami Orogen (GTO) in Western Australia.A key input to the analyses is the recent development of a 4D model of the GTO architectural evolution that provides new insights on the spatio-temporal controls over orogenic gold occurrences in the area; in particular, on the role of pre-mineralization (pre-1795 Ma) D_GTOE–D_GTO1–D_GTO2 architecture in localization of gold deposits and the spatial distribution of rock types in 3D. This information is used to build up a model of orogenic gold minerals system in the area, which is then integrated into the three mutually independent but complementary mineral prospectivity maps namely, a concept-driven “manual” and “fuzzy” analysis; and a data-driven “automated” analysis.The manual analysis involved: (1) generation of a process-based gold mineral systems template to aid target selection; (2) manual delineation of targets; (3) manual estimation of the probability of occurrence of each critical mineralization process based on the available information; and (4) combining the above probabilities to derive the relative probability of occurrence of orogenic gold deposits in each of the targets. The knowledge-based Geological Information System (GIS) analysis attempts to replicate the expert knowledge used in the manual approach, but queried in a more systematic format to eliminate human heuristic bias. This involves representing the critical mineralization processes in the form of spatial predictor maps and systematically querying them through the use of a fuzzy logic model to integrate the predictor maps and to derive the western GTO orogenic gold prospectivity map. The data-driven ‘empirical’ GIS analysis uses no expert knowledge. Instead it employs statistical measures to evaluate the spatial associations between known deposits and predictor maps to establish weights for each predictor layer then combines these layers into a predictive map using a Weights of Evidence (WofE) approach.Application of a mineral systems approach in the manual analysis and the fuzzy analysis is critical: potential high value targets identified by these approaches in the western GTO lie largely under cover, whereas traditional manual targeting is biased to areas of outcrop or sub-crop amenable to direct detection technology such as exploration geochemistry, and therefore towards areas that are data rich.The results show the power of combining the three approaches to prioritize areas for exploration. While the manual analysis identifies and employs human intuition and can see through incomplete datasets, it is difficult to filter out human bias and to systematically apply to a large region. The fuzzy method is more systematic, and highlights areas that the manual analysis has undervalued, but lacks the intuitive power of the human mind that refines the target by seeing through incomplete datasets. The empirical WoE method highlights correlations with favorable host stratigraphy and highlights the control of an early set of structures potentially undervalued in the knowledge driven approaches, yet is biased due to the incomplete nature of exploration datasets and lack of abundant gold deposits due to the extensive cover.The results indicate that the most prospective areas for orogenic gold in western GTO are located in the central part of the study area, largely in areas blind to previous exploration efforts. According to our study, the procedure to follow should be to undertake the analyses in the following order: manual prospectivity analysis, followed by the conceptual fuzzy approach, followed by the empirical GIS-based method. Undertaking the manual analysis first is important to prevent explorationists from being biased by the automated GIS-based outputs. It is however emphasized that all of the prospectivity outputs from these three methods are possible, and they should not be treated as ‘treasure maps’, but instead, as decision-support aids. Therefore, a final manual prospectivity analysis redefined by the mutual consideration of output from all of the methods is required.The strategy employed in this study constitutes a new template for best-practice in terrane- to camp-scale exploration targeting that can be applied to different terranes and deposit types, particularly in terranes under cover, and provides a step forward in managing uncertainty in the exploration targeting process.     

宁芜盆地凹山铁矿床黄铁矿Re-Os同位素定年及其地质意义

宁芜盆地是长江中下游成矿带的7个矿集区之一,以中生代火山岩和玢岩铁矿的广泛发育为特征。凹山铁矿床位于盆地中部,是宁芜盆地玢岩型铁矿床的重要代表之一。本文首次利用成矿元素Fe的金属矿物对宁芜盆地凹山铁矿床的黄铁矿进行了Re-Os同位素测年,获得凹山铁矿床的黄铁矿Re-Os同位素等时线年龄为127.7±2.6Ma(n=5,MSWD=0.66),直接代表凹山铁矿床的成矿年龄。~(187)Os/~(188)Os初始比值为1.73±0.56,γ_(Os(t))值为527.18～535.93,介于地幔与地壳之间,结合已有研究成果,认为凹山铁矿床的成矿物质主要来源于富集地幔,少量来自地壳。     

Three-dimensional numerical modeling of land subsidence in Shanghai,China

Shanghai, in China, has experienced two periods of rapid land subsidence mainly caused by groundwater exploitation related to economic and population growth. The first period occurred during 1956–1965 and was characterized by an average land subsidence rate of 83 mm/yr, and the second period occurred during 1990–1998 with an average subsidence rate of 16 mm/yr. Owing to the establishment of monitoring networks for groundwater levels and land subsidence, a valuable dataset has been collected since the 1960s and used to develop regional land subsidence models applied to manage groundwater resources and mitigate land subsidence. The previous geomechanical modeling approaches to simulate land subsidence were based on one-dimensional (1D) vertical stress and deformation. In this study, a numerical model of land subsidence is developed to simulate explicitly coupled three-dimensional (3D) groundwater flow and 3D aquifer-system displacements in downtown Shanghai from 30 December 1979 to 30 December 1995. The model is calibrated using piezometric, geodetic-leveling, and borehole extensometer measurements made during the 16-year simulation period. The 3D model satisfactorily reproduces the measured piezometric and deformation observations. For the first time, the capability exists to provide some preliminary estimations on the horizontal displacement field associated with the well-known land subsidence in Shanghai and for which no measurements are available. The simulated horizontal displacements peak at 11 mm, i.e. less than 10 % of the simulated maximum land subsidence, and seems too small to seriously damage infrastructure such as the subways (metro lines) in the center area of Shanghai.     

阿尔金山喀腊大湾铁矿田地质特征与形成时代

阿尔金山喀腊大湾铁矿田位于北东向阿尔金走滑断裂与东西向阿尔金北缘断裂所夹持的地区。通过对已有铁矿床地质特征的研究,确认喀腊大湾地区铁矿床属于火山沉积型,因此,铁矿床具有成带状分布的特点,由八八铁矿床、八八西铁矿床、7914铁矿床、7915铁矿床、7918铁矿床和7910铁矿床共同构成喀腊大湾铁矿田。铁矿田范围内的地层以向北陡倾的单斜层为主,倾角75~88°;仅矿田西段八八西铁矿床、八八铁矿床一带沿走向发生直立陡倾伏(背斜)褶皱,可见地层和含矿岩系局部发生褶皱重复,形成铁矿带和含矿层位呈横卧"勾"状形态。铁矿床的产出严格受火山沉积岩系的特定层位和岩性控制,即铁矿床均产在卓阿布拉克组第三亚组第一岩性段(下部玄武岩段)的上部,其北侧为卓阿布拉克组第三亚组第二岩性段(大理岩段)标志层。与铁矿床关系密切的中基性火山岩锆石SHRIMP U-Pb年龄为(517±7)Ma,显示该套火山沉积岩系中基性火山岩的时代为早古生代晚寒武世,因此,铁矿床的形成(火山-沉积)时代为晚寒武世。铁矿床形成之后主要发生2期地质构造事件,其一是发生了构造变形,致使铁矿体及含铁岩系均呈陡立状态,并形成褶皱;其二是中酸性岩浆侵入活动,使已经形成的铁矿床受到中酸性侵入岩的吞食破坏,并发生接触变质矽卡岩化改造。矽卡岩化改造时代为早奥陶世末期(辉钼矿铼锇等时线年龄480 Ma)。     

Re-Os Dating of Pyrite From the Taocun Fe Deposit, Ningwu Basin, Lower Yangtze River Valley, Eastern China

Please refer to the attachment(s) for more details