长江三角洲北岸土体工程地质层组划分及其应用

鉴于研究区工程地质层划分标准尚不统一,不利于工程地质资料的交流和资料的社会化利用。为提高研究区工程地质资料的通用性,结合研究区11个1∶50000标准图幅工程地质调查工作,以100m以浅的土体为工程地质层组划分对象,考虑沉积时代、沉积环境、土体结构特征和物理力学参数,建立研究区基本地层结构层序和编码,将研究区土体划分为7个工程地质层组和21个工程地质层。通过工程地质层组的建立,找出了大区域工程钻孔土层的对应关系,使区域地层资料对比分析成为可能,有助于地质模型的概化和抓住工程地质问题。采用物理力学参数离散性、绘制的地质剖面和三维型分析了工程地质层组划分的合理性,证实了层组划分结果合理正确。通过工程地质层组划分,找出了2个软弱敏感层和4个优势持力层。本文的成果对长江北岸三角洲地区的工程勘察、城市规划和建设项目选址论证具有一定的指导意义。     

卷积神经网络及其在矿床找矿预测中的应用——以安徽省兆吉口铅锌矿床为例

大数据人工智能地质学刚刚起步,基于大数据智能算法的地质研究是非常有意义的探索性实验。利用大数据和机器学习解决矿产预测问题,有助于人们克服不能全面考虑地质变量的困难及评估当前模型在已有数据中的可靠性。元素地表分布特征量主要受原岩成分、成矿作用影响和地表过程的影响,它们携带某些指示矿体就位的信息,即矿体在地下空间就位时在地表的响应,且未在地表过程中消失。以往的地球化学勘查工作仅仅识别异常,但未能发现矿体在地表响应的成矿特征量。本文以安徽省兆吉口铅锌矿床为例,通过机器学习,利用卷积神经网络算法,不断挖掘元素Pb分布特征与矿体地下就位空间的耦合相关性。经过1000次训练后,可以得到准确率0. 93,损失率0. 28的卷积神经网络模型。这种神经网络模型就是矿体在地下就位时元素在地表分布的响应,可以用来进行矿产资源预测。应用该模型对未知区进行预测,结果显示第53号区域具有很大概率存在尚未发现的矿体。     

内蒙古克什克腾旗哲斯组沉积学和年代学研究及其古地理意义

为了重建西拉木伦河北部晚古生代沉积古地理格局并探讨古亚洲洋的闭合过程,本研究对克什克腾旗西部的哲斯组地层进行了详细的沉积环境与沉积相分析、锆石U-Pb年代学研究。来自哲斯组的凝灰质砂岩与安山岩样品的最年轻锆石组的谐和年龄为262±1Ma和258±17Ma,限定研究区内地层沉积下限为中晚二叠世。前人认为哲斯组地层为深海环境下的浊流沉积,本文研究表明哲斯组以近源快速堆积、快速沉降为特征,沉积环境为三角洲前缘相与前三角洲相的多次交替,同时受到火山活动的影响。通过详细的沉积岩野外剖面观察,区内地层并没有发现以单调的递变层理为特征的连续鲍马序列层序,而是出现从砂岩到粉砂岩或泥岩的不规则截然互层层序,并与巨厚层的滑塌堆积交替发育,表明研究区中二叠世古地理格局并不具备深海浊积扇的发育条件,也暗示中二叠世不存在向南发育的广阔大陆边缘;另一方面,结合古流向恢复和物源分析,哲斯组的沉积物应主要来自于南部的华北克拉通以及南造山带陆缘弧,说明研究区在中二叠世已与华北板块连通,并没有类似大洋俯冲带的阻隔作用。因此,研究区中二叠世古地理格局并不具备深海远源浊积扇的发育条件,而是处于近源快速堆积的伸展环境,中二叠世既不存在向南发育的广阔大陆边缘,也不存在代表板块边界的俯冲带。     

新疆西昆仑奥尔托喀讷什锰矿地质、地球化学及成因

近年来,西昆仑玛尔坎苏地区富锰矿找矿取得重大突破,新发现奥尔托喀纳什等大型锰矿床。该矿床层位稳定,厚度较大,Mn平均品位达35%以上,为中国最富的碳酸锰矿床,属于典型的海相沉积型锰矿床。锰矿体主要赋存于晚石炭世喀拉阿特河组地层中,该组岩性为一套浅海碳酸盐岩台地相沉积建造组合,可划分为台内浅滩、潮坪、开阔台地、局限台地等4个相类型。成矿分为三个期次,第一期为沉积成岩成矿期,矿石矿物由菱锰矿、锰方解石、硼锰矿组成;第二期为热液改造期,形成锰镁绿泥石、红锰矿、硫锰矿、锰方解石(脉)、重结晶菱锰矿、蔷薇辉石及滑石、石膏等;第三期为表生氧化期,发育少量软锰矿、水锰矿、硬锰矿等。锰矿石具有较低的Fe/Mn比值、V/(V+Ni)比值和强烈的Ce正异常,表明Mn是在氧化环境下,以氧化物或氢氧化物的形式沉积富集。含锰岩系顶、底板岩石中含较多成熟度较差的中酸性火山岩岩屑,以及具有较低Al/(Al+Fe+Mn)、Y/Ho、Co/Ni比值的锰矿石,说明其成矿物质来源于海底热水活动。奥尔托喀纳什锰矿具有"内源外生"的特点,锰矿石及菱锰矿具有负的δ~(13)C值(-23.3‰～-13.2‰),表明锰矿经历了先成锰氧化物或氢氧化物、再被还原转化成菱锰矿的过程。此外,有机质所导致的更为强烈的还原作用是本矿床富锰矿形成的重要机制。后期构造叠加致使矿体发生变形,矿体形态受褶皱控制。矿石受到强烈改造,形成锰镁绿泥石、红锰矿、蔷薇辉石等,晚期经历氧化淋滤作用形成软锰矿、水锰矿等。     

云南香格里拉市热林铅锌银矿特征及找矿标志

云南香格里拉市热林铅锌银矿,矿体呈似层状赋存于三叠系上统曲嘎寺组三段(T3q3)之大理岩与板岩的层间构造破碎带中,矿体明显受断层构造及特定的岩性控制.该矿床属断层破碎带控制的热液型铅锌银矿床,断层破碎带和地球化学异常是直接的找矿标志,沿断层带出现的硅化、碳酸盐化、黄铁矿化和角岩化是间接的找矿标志.     

蜀南地区下三叠统飞仙关组沉积相

蜀南地区飞仙关组的沉积相类型主要有浑水陆棚相与清水陆棚相,东部两者交替发育,中西部则主要为浑水陆棚相。浑水陆棚相以半局限海亚相为主,包括三个微相;清水陆棚相可分为颗粒滩、滩间、潮坪和内缘斜坡等四个亚相,包括若干微相。飞一段全区为清水陆棚相;飞二段全区均为浑水陆棚亚相的泥质半局限海微相;飞三段,西部为浑水陆棚相,东部为清水陆棚相,其间发育灰泥质半局限海微相;飞四段基本为泥质半局限海微相。飞一时,区内中部有一近南北向的台内凹槽,沿凹槽边缘鲕粒滩较发育。研究区以西的康滇古陆和龙门山岛链与华蓥山断裂带控制着区内沉积相展布,当康滇古陆和龙门山岛链构造活跃时全区发育浑水陆棚,华蓥山断裂带活动时东部发育清水陆棚;当康滇古陆、龙门山岛链与华蓥山断裂带构造的活动性都较低时,全区以清水陆棚为主。     

柴达木盆地第三系咸水湖相生油岩古沉积环境地球化学特征

在系统分析柴达木盆地西部第三系咸水湖相生油岩样品的微量元素、有机物、粘土矿物组成的基础上,揭示了研究区生油岩的无机元素和有机质组成特征;并结合沉积环境的无机与有机地球化学指标,重建了生油岩沉积时的古沉积环境。生油岩的B、Cl-及伽玛蜡烷含量表明,沉积时的水介质条件是半咸水—咸水环境,并伴随湖盆沉积中心北迁过程各层位地层的盐度发生规律性时空]化;生油岩的Fe2 /Fe3 、S2-、Pr/Ph比值、藿烷碳数分布模式及黄铁矿含量反映其沉积时的沉积环境为强还原性;粘土矿物组合特征及含量揭示这些生油岩的沉积相带介于中心咸湖相和边缘咸湖相之间。     

不同压实度下红层泥岩路堤沉降研究

压实系数是控制路堤填筑的关键性指标。本文通过分析不同压实系数下的强度特点以及通过离心试验和有限元对不同压实系数路堤沉降的模拟,得到粘聚力c值和摩擦角与不同压实系数的关系以及不同压实系数下路堤沉降的特点,为路堤填筑提供了重要的依据。     

Seismic images of a collision zone offshore NW Sabah/Borneo

Multichannel reflection seismic data from the southern South China Sea, refraction and gravity modelling were used to investigate the compressional sedimentary structures of the collision-prone continental margin off NW Borneo. An elongated imbricate deepwater fan, the toe Thrust Zone bounds the Northwest Borneo Trough to the southeast. The faults separating the individual imbricates cut through post-Early Miocene sediments and curve down to a carbonate platform at the top of the subsiding continental Dangerous Grounds platform that forms the major detachment surface. The age of deformation migrates outward toward the front of the wedge. We propose crustal shortening mechanisms as the main reason for the formation of the imbricate fan. At the location of the in the past defined Lower Tertiary Thrust Sheet tectonostratigraphic province a high velocity body was found but with a much smaller extend than the previously defined structure. The high velocity structure may be interpreted either as carbonates that limit the transfer of seismic energy into the sedimentary layers beneath or as Paleogene Crocker sediments dissected by remnants of a proto-South China Sea oceanic crust that were overthrust onto a southward migrating attenuated continental block of the Dangerous Grounds during plate convergence.     

Fundamental study for morphodynamic modelling: Sand mounds in oscillatory flows

Experiments on sand mounds in oscillatory flow, undertaken in controlled, large-scale laboratory conditions, have produced well-defined data sets for model comparison. Three bathymetries with different levels of submergence, including a surface-piercing case, were tested. The maximum slope was about 1:5.5. Sediment transport is due to bed load with ripple formation. The principal time-dependent bulk parameters are the vertical distance of the centre of gravity above the base and the volume of the mound. A semi-implicit finite-volume depth-averaged hydrodynamic model is used to drive morphodynamics, using van Rijn's sediment flux model generalized to take account of bed slope, and some justification is given for depth-averaged modeling in these conditions. Starting the model runs with the conditions at the end of the first cycle avoided initial atypical physical behaviour. In general good predictions were obtained with an angle of repose reduced from the standard value of about 30° for stationary beds to 15°. For these situations, morphodynamics was largely unaffected by a hydrodynamic roughness height in the range 2.5D₅₀ to 51D₅₀, with larger values accounting for ripple roughness. The reduced angle of repose may be physically expected with mobile beds but this specific value is only expected to be suited to this form of bed motion. In one case an exaggerated ripple formed near the top of the mound reducing agreement with experiment. For the submerged case with normal ripple structure excellent predictions were obtained. For the initially surface-piercing mound, the time of submergence was better predicted with a 30° angle of repose, presumably due to the prominent influence of the near stationary bed near the wet/dry interface, although long term predictions were better predicted with 15°. The occurrence of vortex shedding in the first cycle modeled was in agreement with experimental observation.