Geochronology,Geochemistry and Petrogenesis of the Bangbule Quartz Porphyry:Implications for Metallogenesis

The Bangbule skarn lead-zinc(Pb-Zn)deposit(>1 Mt Zn+Pb)is located in the western Nyainqentanglha polymetallic metallogenetic belt,central Tibet.Lenticular orebodies are all hosted in skarn and developed in the contact zone between the quartz porphyry and carbonate strata of the mid Paleozoic Middle to Upper Chaguoluoma Formation as well as in carbonate and sandstone beds of the Upper Paleozoic Laga Formation.As a newly discovered skarn deposit,the geological background and metallogenesis of this deposit remain poorly understood.Detailed petrological,geochemical and geochronological data of the ore-related quartz porphyry,helps constrain the mineralization age and contributes to discussion on the ore genesis of the Bangbule deposit.Both endoskarn and exoskarn are identified in the Bangbule deposit.From quartz porphyry to carbonate formation,the exoskarn is zoned from proximal garnet skarn to distal pyroxene skarn.Zircon U-Pb dating results show that the quartz porphyry formed at 73.9±0.8 Ma.Geochemical analysis results show that the quartz porphyry has high contents of SiO_(2)(71.40–74.94 wt%)and K_(2)O+Na_(2)O(3.76–8.46 wt%)with A/CNK values of 0.69 to 1.06.Besides,the quartz porphyry is enriched in large ion lithophile elements(LILEs)and light rare earth elements(LREEs)and have lowεNd(t)(from-8.25 to-8.19)and high initial(^(87)Sr/^(86)Sr)i values(0.713611–0.714478).Major,trace elements and whole-rock F concentration analysis results from the endoskarn samples show higher TFe_(2)O_(3),MgO,CaO,Pb+Zn,W,Sn,Mo and F etc.,and lower alkalis(K_(2)O,Na_(2)O,Sr and Ba)than those of fresh quartz porphyry,indicating that the early ore-forming fluids were an Ca-Fe-F-enriched fluid.Massive ore in the proximal skarn might be related to the high F content in the magma,which lowered the solidus temperature of the quartz porphyry magma and caused a lower temperature of the ore-forming fluids,as well as facilitating the precipitation of sphalerite and galena.Based on the geochemical characteristics presented in this study,we propose that the ore-related quartz porphyry was formed by partial melting of crust materials with some juvenile crustal component input.The partial melting of the middle-upper crust after the initial enrichment of lead and zinc elements are important for the formation of Pb-Zn deposits.The case study of the Bangbule deposit has proven that there is still a crust-derived magmatic source region in the western segment of the central Lhasa terrane.Therefore,there is still great potential for Pb-Zn mineralization and Pb-Zn exploration.

冈底斯构造带西段帮布勒铅锌铜矿床的剥露和保存:来自裂变径迹热年代学的证据

帮布勒铅锌铜矿床是近年在冈底斯构造带西段新发现的大型矽卡岩型矿床。采用低温热年代学方法,揭示帮布勒矿区的冷却和剥露历史,评估矿床保存情况,完善冈底斯构造带西段隆升剥露记录。结果显示,石英斑岩磷灰石裂变径迹年龄为24±2~32±3 Ma(1σ),记录了矿区在渐新世时的冷却和剥露时间。热演化历史模拟表明,帮布勒矿区存在3期冷却事件:Ⅰ始新世—渐新世晚期(48~24 Ma),冷却速率为4.56~10.64℃/Ma,与印度-欧亚大陆碰撞及碰撞后的地壳增厚有关;Ⅱ渐新世晚期—中新世中期(24~14 Ma),冷却速率为0.38~1.9℃/Ma,与印度俯冲板块折返、断离有关;Ⅲ中新世中期至今(14 Ma至今),冷却速率为2.28~9.88℃/Ma,与冈底斯构造带发育的南北向裂谷有关;其中5 Ma时,与冈底斯构造带普遍隆升有关。帮布勒矿区自始新世以来的剥露厚度为1.5~2.5 km,典中组火山岩的覆盖为矿床保存至今起到了保护作用。

基于LK-3DCNN网络特征的高光谱遥感图像分类算法

传统的高光谱图像分类方法均采用手工提取特征的方式,其表达能力有限,不能满足分类任务的高精度需求。因此,提出一种基于LK-3D卷积神经网络(Convolutional Neural Network,CNN)特征的高光谱遥感图像分类算法。该算法结合空洞卷积和三维卷积神经网络的双分支融合结构,对遥感图像进行深度特征提取,有利于提升小样本高光谱遥感图像数据分类的性能。在公开的3组高光谱遥感图像数据集上,与主成分分析(Principal Components Analysis,PCA)+3DCNN、PCA+3D-2DCNN算法进行对比,实验结果表明所提出的双分支融合网络提升了高光谱遥感图像的分类精度,具有较好的鲁棒性。

基于ML-3DCNN的高光谱遥感图像分类算法

高光谱遥感图像含有丰富的光谱和空间特征信息,在进行地物分类时这些特征的提取尤为重要。提出了一种基于融合拉普拉斯特征的三维卷积神经网络(Meld Laplace 3D Convolution Neural Network,ML-3DCNN)的高光谱遥感图像分类算法,该算法采用主成分分析(Principal Components Analysis,PCA)降维结合空间边缘细节特征提取组成的双分支网络结构对遥感图像进行特征提取,有利于提升分类性能。在公开的3组高光谱遥感图像数据集上,与PCA+3DCNN和PCA+3D-2DCNN算法进行对比,结果表明提出的双分支网络结构提升了高光谱遥感影像的分类精度。