松辽盆地大陆深部科学钻探2号井完井暨学术研讨现场会召开

正2018年5月26日,自然资源部中国地质调查局在黑龙江省安达市松科二井工程现场召开大陆科学钻探工程(松科二井)完井暨学术研讨现场会。中国地质学会常务副理事长兼秘书长、自然资源部中国地质调查局副局长、松科二井工程领导小组组长李金发宣布松辽盆地大陆深部科学钻探工     

基于岩石图像深度学习的岩性自动识别与分类方法

岩石岩性的识别与分类对于地质分析极为重要,采用机器学习的方法建立识别模型进行自动分类是一条新的途径。基于Inception-v3深度卷积神经网络模型,建立了岩石图像集分析的深度学习迁移模型,运用迁移学习方法实现了岩石岩性的自动识别与分类。采用此方法对所采集的173张花岗岩图像、152张千枚岩图像和246张角砾岩图像进行了学习和识别分类研究,通过训练学习建立岩石图像深度学习迁移模型,并分别采用训练集和测试集中的岩石图像对模型进行了检验分析。对于训练集中的岩石图像,每组岩石分别用3张图像测试,三种岩石的岩性分类均正确,且分类概率值均达到90%以上,显示了模型良好的鲁棒性;对于测试集中的岩石图像,每组岩石分别采用9张图像进行识别分析,三种岩石的岩性分类均正确,并且千枚岩组图像分类概率均高于90%,但是花岗岩组2张图像和角砾岩组的1张图像分类概率值不足70%,概率值较其他岩石图像低,推测其原因是训练集中相同模式的岩石图像较少,导致模型的泛化能力减小。为了提高识别精确度,对准确率较低的岩石图像进行截取,分别取其中的3张图像加入训练集进行再训练,增加与测试图像具有相同模式的训练样本;在新的模型中,对3张图像进行二次检验,测试概率值均达到85%以上,说明在数据足够的状况下模型具有良好的学习能力。与传统的机器学习方法相比,所提出的岩石图像深度学习方法具有以下优点:第一,模型通过搜索图像像素点提取物体特征,不需要手动提取待分类物体特征;第二,对于图像像素大小,成像距离及光照要求低;第三,采用适当的训练集可获得较好的识别分类效果,并具有良好鲁棒性和泛化能力。     

处置单元核素135Cs迁移数值模拟研究

根据高放废物地质处置库概念设计,以北山花岗岩为参考岩性,内蒙古高庙子膨润土为参考缓冲材料,使用数值模拟软件tough2对放射性核素¹³⁵Cs在处置单元及其围岩中的迁移行为进行了数值模拟研究,考虑放射性核素衰变释热对核素迁移的影响。研究表明:膨润土对核素具有较强的滞留作用,模拟200000年后核素在花岗岩中迁移了1 m左右,且峰值为10-6个数量级;前10 a模型的温度变化最剧烈且温差最大,10 a后模型整体温度开始降低,2 000 a后整体温度接近于初始温度;温度的变化会引起核素迁移速率的改变,但总体温度对核素迁移的影响不大。     

Geology and mineralization of the Sanshandao supergiant gold deposit (1200 t) in the Jiaodong Peninsula,China: A review

The Jiaodong Peninsula in Shandong Province, China is the world’s third-largest gold metallogenic area, with cumulative proven gold resources exceeding 5000 t. Over the past few years, breakthroughs have been made in deep prospecting at a depth of 500–2000 m, particularly in the Sanshandao area where a huge deep gold orebody was identified. Based on previous studies and the latest prospecting progress achieved by the project team of this study, the following results are summarized. (1) 3D geological modeling results based on deep drilling core data reveal that the Sanshandao gold orefield, which was previously considered to consist of several independent deposits, is a supergiant deposit with gold resources of more than 1200 t (including 470 t under the sea area). The length of the major orebody is nearly 8 km, with a greatest depth of 2312 m below sea level and a maximum length of more than 3 km along their dip direction. (2) Thick gold orebodies in the Sanshandao gold deposit mainly occur in the specific sections of the ore-controlling fault where the fault plane changes from steeply to gently inclined, forming a stepped metallogenic model from shallow to deep level. The reason for this strong structural control on mineralization forms is that when ore-forming fluids migrated along faults, the pressure of fluids greatly fluctuated in fault sections where the fault dip angle changed. Since the solubility of gold in the ore-forming fluid is sensitive to fluid pressure, these sections along the fault plane serve as the target areas for deep prospecting. (3) Thermal uplifting-extensional structures provide thermodynamic conditions, migration pathways, and deposition spaces for gold mineralization. Meanwhile, the changes in mantle properties induced the transformation of the geochemical properties of the lower crust and magmatic rocks. This further led to the reactivation of ore-forming elements, which provided rich materials for gold mineralization. (4) It can be concluded from previous research results that the gold mineralization in the Jiaodong gold deposits occurred at about 120 Ma, which was superimposed by nonferrous metals mineralization at 118–111 Ma. The fluids were dominated by primary mantle water or magmatic water. Metamorphic water occurred in the early stage of the gold mineralization, while the fluid composition was dominated by meteoric water in the late stage. The S, Pb, and Sr isotopic compositions of the ores are similar to those of ore-hosting rocks, indicating that the ore-forming materials mainly derive from crustal materials, with the minor addition of mantle-derived materials. The gold deposits in the Jiaodong Peninsula were formed in an extensional tectonic environment during the transformation of the physical and chemical properties of the lithospheric mantle, which is different from typical orogenic gold deposits. Thus, it is proposed that they are named “Jiaodong-type” gold deposits.©2021 China Geology Editorial Office.     

衡水地区地裂缝空间发育特征与地下水位降深关系

通过浅层地震法及高密度电阻率法物探对河北衡水地区最近发现的3条地裂缝调查,探明了地裂缝地下发育深度、宽度、倾向等,发现地裂缝在地表与地下发育特征有明显不同：地表以一条地裂缝存在,在地下深处则以裂缝带发育,裂缝带两侧裂缝埋深不同,地下裂缝形态表现为上窄下宽。对该地区深层钻探、水文地质和多年地下水开采等资料分析表明,该地区的地裂缝是地质历史时期地质构造活动产物,开采深层地下水等资源则是促使地裂缝发展的主要原因。通过对不同地点地裂缝发育深度与近36年以来地下水水位降深的关系分析,结果显示：不同地区地裂缝发育深度与相应深层地下水水位降深呈线性关系。研究成果对地裂缝机理研究及预防开采深层地下水对地质环境的影响具有指导作用。     

多种降阻材料搭配在变电站接地网改造中应用

在北海市三塘110KV变电站接地网系统改造工程实践中,通过不等长接地体技术及多种将阻材料搭配使用,克服了原接地网接地电阻值增大因素,并对变电站接地网改造和施工工艺进行设计。     

The connection between Labrador Sea buoyancy loss, deep western boundary current strength, and Gulf Stream path in an ocean circulation model

The sensitivity of the North Atlantic gyre circulation to high latitude buoyancy forcing is explored in a global, non-eddy resolving ocean general circulation model. Increased buoyancy forcing strengthens the deep western boundary current, the northern recirculation gyre, and the North Atlantic Current, which leads to a more realistic Gulf Stream path. High latitude density fluxes and surface water mass transformation are strongly dependent on the choice of sea ice and salinity restoring boundary conditions. Coupling the ocean model to a prognostic sea ice model results in much greater buoyancy loss in the Labrador Sea compared to simulations in which the ocean is forced by prescribed sea ice boundary conditions. A comparison of bulk flux forced hindcast simulations which differ only in their sea ice and salinity restoring forcings reveals the effects of a mixed thermohaline boundary condition transport feedback whereby small, positive temperature and salinity anomalies in subpolar regions are amplified when the gyre spins up as a result of increased buoyancy loss and convection. The primary buoyancy flux effects of the sea ice which cause the simulations to diverge are ice melt, which is less physical in the diagnostic sea ice model, and insulation of the ocean, which is less physical with the prognostic sea ice model. Increased salinity restoring ensures a more realistic net winter buoyancy loss in the Labrador Sea, but it is found that improvements in the Gulf Stream simulation can only be achieved with the excessive buoyancy loss associated with weak salinity restoring.     

Numerical modelling of deep sea air-lift

Deep sea air-lifting of solid particles from depth of 1600 m is simulated with a mathematical model of the three-phase flow in an upward pipe. The computations are carried out for an axisymmetric domain in a transient way. Phase distributions, pressure and velocity profiles together with flow rates for all phases are presented and analysed. The influence of the pipe diameter on the air-lift efficiency was studied for air-lift pipes of different lengths and found to be significant. The lifting efficiency increases with the increase of the pipe diameter due to the reduction of the wall friction influence on the flow. In addition, the efficiency also increases with the increase of the solid particles volume fraction at the inlet. The presented numerical model can be utilized during various stages of the design of the air-lift pumps to help answer fundamental questions on the process, and during their operation to select optimal process parameters and to address possible problems.     

Predicting suitable habitat for the cold-water coral Lophelia pertusa (Scleractinia)

Ecological-niche factor analysis (ENFA) was applied to the reef framework-forming cold-water coral Lophelia pertusa. The environmental tolerances of this species were assessed using readily available oceanographic data, including physical, chemical, and biological variables. L. pertusa was found at mean depths of 468 and 480 m on the regional and global scales and occupied a niche that included higher than average current speed and productivity, supporting the theory that their limited food supply is locally enhanced by currents. Most records occurred in areas with a salinity of 35, mean temperatures of 6.2–6.7  °C and dissolved oxygen levels of 6.0–6.2 ml l⁻¹. The majority of records were found in areas that were saturated with aragonite but had low concentration of nutrients (silicate, phosphate, and nitrate). Suitable habitat for L. pertusa was predicted using ENFA on a global and a regional scale that incorporated the north-east Atlantic Ocean. Regional prediction was reliable due to numerous presence points throughout the area, whereas global prediction was less reliable due to the paucity of presence data outside of the north-east Atlantic. However, the species niche was supported at each spatial scale. Predicted maps at the global scale reinforced the general consensus that the North Atlantic Ocean is a key region in the worldwide distribution of L. pertusa. Predictive modelling is an approach that can be applied to cold-water coral species to locate areas of suitable habitat for further study. It may also prove a useful tool to assist spatial planning of offshore marine protected areas. However, issues with eco-geographical datasets, including their coarse resolution and limited geographical coverage, currently restrict the scope of this approach.