img

QQ群聊

img

官方微信

高级检索

黄金科学技术 ›› 2021, Vol. 29 ›› Issue (6): 771-780.doi: 10.11872/j.issn.1005-2518.2021.06.114

• 矿产勘查与资源评价 •    下一篇

基于大数据的深部找矿靶区定量成矿预测——以大桥地区金矿为例

王怀涛1,2,3(),王晓伟1,2,3,罗云之1,2,3,宋秉田1,2,3,罗建民1,2,3,徐磊1   

  1. 1.甘肃省地质调查院,甘肃 兰州 730000
    2.甘肃省地矿局地学大数据勘查工程技术创新中心,甘肃 兰州 730000
    3.甘肃省地学大数据工程研究中心,甘肃 兰州 730000
  • 收稿日期:2021-08-14 修回日期:2021-10-11 出版日期:2021-12-31 发布日期:2022-03-07
  • 作者简介:王怀涛(1985-),男,山东成武人,高级工程师,从事矿产勘查及地学大数据方法技术研究工作。359921417@qq.com
  • 基金资助:
    中国地质调查局地质调查项目“西秦岭成矿带典型矿区智能矿产地质调查评价试点示范”(WKZB1911BJM300369/018)

Quantitative Metallogenic Prediction of Deep Prospecting Target Based on Big Data:Taking Gold Deposit in Daqiao Area as an Example

Huaitao WANG1,2,3(),Xiaowei WANG1,2,3,Yunzhi LUO1,2,3,Bingtian SONG1,2,3,Jianmin LUO1,2,3,Lei XU1   

  1. 1.Geology Survey of Gansu Province, Lanzhou 730000, Gansu, China
    2.Geoscience Big Data Exploration Engineering Technology Innovation Center of Gansu Provincial Bureau of Geo-logy and Mineral Exploration and Development, Lanzhou 730000, Gansu, China
    3.Geoscience Big Data Engineering Research Center of Gansu Province, Lanzhou 730000, Gansu, China
  • Received:2021-08-14 Revised:2021-10-11 Online:2021-12-31 Published:2022-03-07

摘要:

常规的地球化学采样介质和遥感信息难以鉴别深部矿信息,地球物理信息可以反映深部矿的信息,但传统地球物理信息解释的多解性,影响了深部找矿靶区预测的精度和效果。为了消除物探信息的多解性,提高深部找矿靶区预测精准度,本文应用大数据思想和方法对甘肃省西秦岭大桥地区航磁数据进行深度挖掘,构建基于航磁信息的金找矿靶区定量预测系列模型,结合地质矿产信息,圈定并优选出金找矿靶区31处,其中Ⅰ级靶区6处(见矿率为16.9%),Ⅱ级矿靶区10处(见矿率为31.32%),Ⅲ级Au找矿靶区15处(见矿率为20%)。覆盖区的圈定找矿靶区,经验证发现了金工业矿体;靶区累积面积占研究区面积的2.4%,极大程度地缩小了找矿范围。研究认为基于航磁信息建立的找矿靶区定量预测系列模型,对研究区大桥式金矿找矿靶区的确定有着很高的准确性。

关键词: 航磁信息, 大数据, 定量预测模型, 深部找矿靶区, 大桥金矿, 西秦岭

Abstract:

Conventional geochemical sampling medium does not contain deep deposit information,and the remote sensing information is only the characteristics of surface images,which is difficult to identify deep deposit information.Geophysical information can well reflect the information of deep deposit and it is the best information choice for deep metallogenic prediction.However,the interpretation of geophysical information are multi-resolution,which has always seriously affected the accuracy and accuracy of deep prospecting target prediction.Big data is triggering a profound revolution in the field of Geoscience.New methods and technologies such as big data and artificial intelligence represented by statistical analysis methods and machine learning algorithms have been gradually applied to metallogenic prediction and achieved good prediction results.The Western Qinling area of Gansu Province is an important polymetallic metallogenic accumulation area in China,which has accumulated rich geological data.It is of great significance to carry out quantitative prediction of gold prospecting target by deep mining geological data with big data method for gold exploration and expansion of gold reserves in Western Qinling area of Gansu Province.In order to eliminate the multi solution of geophysical information and improve the prediction accuracy of deep prospecting target,big data ideas and methods were applied to deep mining of aeromagnetic data from Daqiao area in west Qinling of Gansu Province,and established the aeromagnetic database,the aeromagnetic information research unit database and the known ore unit database respectively.Through the discriminant analysis of aeromagnetic database and known ore unit database,a series of quantitative prediction models of prospecting target were constructed,and deep prospecting targets were delineated,combined with geological and mineral information to optimize the grade prospecting targets.A total of 31 gold prospecting targets have been delineated in the study area,including 6 Class Ⅰ targets (seeing ore rate 16.9%),10 Class Ⅱ ore targets (seeing mine rate 31.32%) and 15 Class Ⅲ targets (seeing ore rate 20%).Gold industrial ore bodies have been found in the prospecting target in the overburden area.The cumulative area of the target accounts for 2.4% of the area of the study area,which greatly reduces the scope of prospecting.The study believes that the series of quantitative prediction models for prospecting targets established based on aeromagnetic information have high accuracy in determining prospecting targets for Daqiao-type gold deposits in the study area.It provides a new idea and method for the metallogenic prediction of deep and concealed deposit.

Key words: aeromagnetic information, big data, quantitative prediction model, deep prospecting target, Daqiao gold deposit, Western Qinling

中图分类号: 

  • P618.51

图1

研究区大地构造位置图(a)和区域地质矿产图(b)1.第四系;2.新近纪甘肃群:砂砾岩建造;3.早白垩世鸡山组:砂砾岩建造;4.早白垩世周家湾组:长石石英砂岩建造;5.侏罗纪龙家沟组:炭质泥岩建造;6.中三叠世滑石关组:碳酸岩建造;7.中二叠世大观山组:生物碎屑—泥晶灰岩建造;8.晚石炭世岷河组:白云质灰岩建造;9.早石炭世伊娃沟组:碳酸盐建造;10.晚泥盆世双狼沟组:砂岩与板岩互层建造;11.中泥盆世黄家沟组:砂岩—粉砂岩—粉砂质泥岩建造;12.中泥盆世下吾那组:钙质粉砂岩建造;13.早泥盆世安家岔组:粉砂岩与板岩互层建造;14.中志留世卓乌阔组:长石石英砂岩建造;15.三叠纪石英闪长岩;16.地质界线;17.角度不整合接触界线;18.断层;19.矿床(点);Ⅲ-2-3-中祁连岩浆弧;Ⅲ-6-1-1-党川—利桥岩浆弧;Ⅲ-6-1-2-鸳鸯镇—关子镇蛇绿混杂岩;Ⅳ-3-1-松潘三叠纪前陆盆地;Ⅲ-5-2-鄂尔多斯西缘新元古代—早古生代裂陷带;Ⅳ-5-1-1-夏河—礼县陆缘沉积区;Ⅳ-5-2-1-泽库—武都裂陷沉积区;Ⅳ-5-2-2-白龙江隆起带;Ⅳ-5-2-3-阿尼玛卿裂陷沉积区;Ⅳ-5-2-4-三河口裂陷沉积区;Ⅳ-5-3-1-碧口古陆北缘"

表1

研究区大桥式金矿定量成矿预测系列模型"

模型

分类

参数

变化

系数

贡献

累计

贡献

有效性检验
模型1.1(常量)-1.437

R0=1.62

F0.01=2.32

Fp=30.18

有矿正判率:98.8%

无矿正判率:92%

ΔT11.6826.3726.37
ΔT20.0222.2548.61
ΔT3-0.0117.8666.47
ΔT41.1915.782.17
ΔT51.1214.6396.8
ΔT6-1.631.5898.38
ΔT7-0.010.9399.31
ΔT81.350.2199.91
ΔT9-0.780.09100
模型1.2(常量)-10.659

R0=0.66

F0.01=3.02

Fp=-14.54

有矿正判率:95.5%

无矿正判率:94.2%

ΔT910.4957.8157.81
ΔT10-5.4719.5677.36
ΔT115.3417.0494.4
ΔT1202.8897.28
ΔT61.582.72100
模型2.1(常量)21.295
ΔT13305.1744.7144.71
ΔT14-178.9526.270.91
ΔT15-57.817.878.71

R0=0.70

F0.01=2.32

Fp=-11.58

有矿正判率:97.1%

无矿正判率:80.5%

ΔT160.017.4986.21
ΔT17-47.626.2892.48
ΔT18-1.464.1396.61
ΔT19-24.351.297.81
ΔT40.811.198.91
ΔT2024.580.7199.62
ΔT213.680.38100
模型2.2(常量)-4.179

R0=0.69

F0.01=1.47

Fp=0.93

有矿正判率:95%

无矿正判率:76.6%

ΔT12-0.0531.3731.37
ΔT30.0426.657.98
ΔT70.144.8362.81
ΔT1156.234.4967.3
ΔT22-0.133.1770.47
ΔT18-0.672.3872.85
ΔT235 116.342.3575.2
ΔT24-42.071.8977.09
ΔT25-57.961.6978.78
ΔT26-0.041.5380.31

表2

定量成矿预测模型1.1预测结果统计"

预测单元分级预测有矿单元/个预测有矿单元占比/%有矿单元/个有矿单元占比/%有矿单元 /有矿单元总数较有矿单元平均值 提高倍数
单元总数/个48 157有矿单元总数/个106
Ⅰ级单元3820.794912.830.4658.34
Ⅱ级单元6181.28152.430.1411.04
Ⅲ级单元9992.08282.800.2612.73
预测总数1 9994.15924.600.8620.92

图2

定量成矿预测模型1.1预测结果柱状图"

表3

定量成矿预测模型1.2预测结果统计"

预测单元分级预测有矿单元/个预测有矿单元占比/%有矿单元/个有矿单元占比/%有矿单元 /有矿单元总数较有矿单位平均值提高倍数
单元总数/个48 157有矿单元总数/个106
Ⅰ级单元2470.51114.450.1020.23
Ⅱ级单元4000.8310.250.091.14
Ⅲ级单元6471.3491.390.856.32
预测总数1 2942.69211.620.197.38

图3

定量成矿预测模型1.2预测结果柱状图"

表4

定量成矿预测模型2.1预测结果统计"

预测单元分级预测有矿单元/个预测有矿单元占比/%有矿单元/个有矿单元占比/%有矿单元 /有矿单元总数较有矿单元平均值提高倍数
单元总数/个48 157有矿单元总数/个106
Ⅰ级单元1970.413417.220.3278.26
Ⅱ级单元3200.66237.200.2232.72
Ⅲ级单元5171.07112.130.109.67
预测总数1 0342.15686.580.6429.89

图4

定量成矿预测模型2.1预测结果柱状图"

表5

定量成矿预测模型2.2预测结果统计"

预测单元分级预测有矿单元/个预测有矿单元占比/%有矿单元/个有矿单元占比/%有矿单元 /有矿单元总数较有矿单元平均值 提高倍数
单元总数/个48 157有矿单元总数/个106
Ⅰ级单元1860.39136.9912.2631.77
Ⅱ级单元3010.63134.3212.2619.63
Ⅲ级单元4871.01122.4611.3211.20
预测总数9742.02383.9035.8517.73

图5

定量成矿预测模型2.2预测结果柱状图"

表6

研究区大桥式金矿预测找矿靶区信息"

靶区编码靶区面积靶区分级矿床点矿床规模判别值
Ⅰ-19.4722.04
Ⅰ-230.75西和县大桥金矿床超大型矿床19.54
Ⅰ-31.9719.38
Ⅰ-41.3116.61
Ⅰ-58.0116.24
Ⅰ-64.8016.21
Ⅱ-711.98小金厂C8金矿化点矿化点16.1
Ⅱ-88.43联合村金矿化点矿化点16.04
Ⅱ-93.9015.47
Ⅱ-1014.6615.35
Ⅱ-116.60西和县白崖沟金矿小型矿床14.98
Ⅱ-123.5414.75
Ⅱ-135.4314.64
Ⅱ-1419.7313.87
Ⅱ-157.6713.73
Ⅱ-163.6613.6
Ⅲ-179.2512.58
Ⅲ-186.7212.36
Ⅲ-191.15西和县红山金矿化点矿化点11.88
Ⅲ-202.87大坪金矿化点矿化点11.81
Ⅲ-211.8911.49
Ⅲ-229.4811.44
Ⅲ-2313.3611.14
Ⅲ-241.8011.08
Ⅲ-256.1410.74
Ⅲ-260.9010.44
Ⅲ-274.769.87
Ⅲ-280.719.73
Ⅲ-297.008.51
Ⅲ-304.667.51
Ⅲ-310.56安子山金矿点矿点7.49

图6

研究区大桥式金矿预测找矿靶区分布图1.第四系;2.新近纪甘肃群:砂砾岩建造;3.早白垩世鸡山组:砂砾岩建造;4.早白垩世周家湾组:长石石英砂岩建造;5.侏罗纪龙家沟组:炭质泥岩建造;6.中三叠世滑石关组:碳酸岩建造;7.中二叠世大观山组:生物碎屑—泥晶灰岩建造;8.晚石炭世岷河组:白云质灰岩建造;9.早石炭世伊娃沟组:碳酸盐建造;10.晚泥盆世双狼沟组:砂岩与板岩互层建造;11.中泥盆世黄家沟组:砂岩—粉砂岩—粉砂质泥岩建造;12.中泥盆世下吾那组:钙质粉砂岩建造;13.早泥盆世安家岔组:粉砂岩与板岩互层建造;14.中志留世卓乌阔组:长石石英砂岩建造;15.早志留世迭部组:碳质板岩建造;16.三叠纪石英闪长岩;17.地质界线;18.角度不整合接触界线;19.断层;20.矿床(点);21.验证钻孔位置;22.Ⅰ级靶区;23.Ⅱ级靶区;24.Ⅲ级靶区"

null Carranza E J M,2009.Controls on mineral deposit occurrence inferred from analysis of their spatital pattern and spatial association with geological features[J].Ore Geology Reviews,35(3):383-400.
null Carranza E J M,2011.Geocomputation of mineral exploration targets[J].Computers & Geosciences,37(12):1907-1916.
null Frankel F, Reid R,2008.Big data:Distilling meaning from data[J].Nature,455:30.
null Luo Jianmin, Zhang Qi, Song Bingtian, al et,2017.Application of integrated geophysical and geochemical data processing to metallogenic target zone quantitative prediction and optimization[J].Bulletin of Mineralogy,Petrology and Geochemistry,36(6):886-890.
null Luo Jianmin, Wang Xiaowei, Song Bingtian, al et,2018.Discussion on the method for quantitative classification of magmatic rocks:Taking it’s application in West Qinling of Gansu Province for example[J]. Acta Petrologica Sinica,34(2):326-332.
null Luo Jianmin, Zhang Qi,2019a.Big data pioneers new ways of geoscience research:Identifying relecant relationships to enhance research feasibility[J].Earth Science Frontiers,26(4):6-12.
null Luo Jianmin, Wang Xiaowei, Zhang Qi, al et,2019b.Application of geological big data to quantitative target area optimization for regional mineral prospeting in China[J].Earth Science Frontiers,26(4):76-83.
null Liu Yanpeng, Zhu Lixin, Zhou Yongzhang,2018.Application of convolutional neural network in prospecting prediction of ore deposits:Taking the Zhaojikou Pb-Zn ore deposit in Anhui Province as a case[J].Acta Petrologica Sinica,34(11):3217-3224.
null Liu Yanpeng, Zhu Lixin, Zhou Yongzhang,2020.Experimental research on big data mining and intelligent prediction of prospecting target area-application of convolutional neural network model[J].Geotectonica et Metallogenia,44(2):192-202.
null Ren Wenxiu, Luo Jianmin, Sun Bainian, al et,2018.Application of geochemical data in gold prospecting and target selecting:Taking the Yushishan area in Gansu Province as a case[J].Acta Petrologica Sinica,34(11) :3225-3234.
null Silver D, Schrittwieser J, Simonyan K, al et,2017.ing the game of go without human knowledge[J].Nature,550:354-359.
null Wigan M R, Clarke R,2013.Big data’s big unintended consequences[J].IEEE Computer,46:36-53.
null Wang Huaitao, Luo Jianmin, Wang Jinrong, al et,2018.Quantitative classification and metallogenic prognosis of basic-ultrabasic rocks based on big data:Taking the Beishan area as an example[J].Acta Petrologica Sinica,34(11):3195-3206.
null Wang Huaitao, Yang Jing, Du Jun, al et,2019.Comparison of the ultrabasic rocks associated with the Cu-Ni deposits in Beishan and Jinchuan,Gansu province and those associated within-plate environment:preliminary results of big data research[J].Earth Science Fromtiers,26(4):94-108.
null Wang Yuxi, Luo Jianmin, Wang Jinrong, al et,2018.The gold prospecting targets quantitative optimization based on information of the magmatic rocks (oxide) in Dunhuang block,Gansu Province[J].Acta Petrologica Sinica,34(2):319-325.
null Wang Yu, Zhou Yongzhang, Xiao Fan, al et,2020.Numerical metallogenic modelling and support vector machine methodsapplied to predict deep mineralization:A case study from the Fankou Pb-Zn ore deposit in northern Guangdong[J].Geotectonica et Metallogenia,44(2):222-230.
null Zuo R G, Xiong Y H, Wang J, al et,2019.Deep learning and its application in geochemical mapping[J].Earth-Science Reviews,192:1-14.
null Zhang Qi, Zhou Yongzhang,2018.Big data helps geology de-velop rapidly[J].Acta Petrologica Sinica,34(11):3167-3172.
null Zhao C B, Hobbs B E, Ord A,2009.Fundamentals of Computational Geoscience[J].Berlin:Springer.
null Zhou Yongzhang, Chen Shuo, Zhang Qi, al et,2018.Advances and prospects of big data and mathematical geoscience[J].Acta Petrologica Sinica,34(2) :255-263.
null Zhang Zhongping, Wu Yafei, Li Jianwei,2018.Characteristics and genesis of the silicified breccias in the Dqiao gold deposit,West Qinling orogeny[J].Geological Science and Technology Information,37(2):79-88.
null 罗建民,张琪,宋秉田,等,2017.物化探信息综合处理在找矿靶区定量优选中的应用[J].矿物岩石地球化学通报,36(6):886-890.
null 罗建民,王晓伟,宋秉田,等,2018.岩浆岩定量分类方法探讨—以甘肃省西秦岭地区为例[J].岩石学报,34(2):326-332.
null 罗建民,张旗,2019a.大数据开创地学研究新途径:查明相关关系,增强研究可行性[J].地学前缘,26(4):6-12.
null 罗建民,王晓伟,张琪,等,2019b.地质大数据方法在区域找矿靶区定量优选中的应用[J].地学前缘,26(4):76-83.
null 刘艳鹏,朱立新,周永章,2018.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[J].岩石学报,34(11):3217-3224.
null 刘艳鹏,朱立新,周永章,2020.大数据挖掘与智能预测找矿靶区实验研究——卷积神经网络模型的应用[J].大地构造与成矿学,44(2):192-202.
null 任文秀,罗建民,孙柏年,等,2018.化探数据挖掘在金矿找矿及靶区优选中的应用—以甘肃玉石山地区为例[J].岩石学报,34(11):3225-3234.
null 王怀涛,罗建民,王金荣,等,2018.基于大数据的基性—超基性岩定量分类及成矿预测研究——以北山地区为例[J].岩石学报,34(11):3195-3206.
null 王怀涛,杨婧,杜君,等,2019.甘肃北山和金川与铜镍矿床有关的超基性岩与全球板内环境超基性岩的对比:大数据研究的初步结果[J].地学前缘,26(4):94-108.
null 王玉玺,罗建民,王金荣,等,2018.甘肃敦煌地块基于岩浆岩(氧化物)信息的金找矿靶区定量优选[J].岩石学报,34(2):319-325.
null 王语,周永章,肖凡,等,2020.基于成矿条件数值模拟和支持向量机算法的深部成矿预测——以粤北凡口铅锌矿为例[J].大地构造与成矿学,44(2):222-230.
null 张旗,周永章,2018.大数据助地质腾飞:岩石学报2018第11期大数据专题“序”[J].岩石学报,34(11):3167-3172.
null 周永章,陈烁,张旗,等,2018.大数据与数学地球科学研究进展——大数据与数学地球科学专题代序[J].岩石学报,34(2):255-263.
null 张忠平,吴亚飞,李建威,2018.西秦岭地区大桥金矿床规划角砾岩的特征及成因[J].地质科技情报,37(2):79-88.
[1] 曹振梁,刘学龙,李守奎,刘思晗,李方兰,周博文. 滇西北燕山期含矿花岗岩与全球埃达克岩地球化学特征对比[J]. 黄金科学技术, 2023, 31(1): 15-25.
[2] 张斌,刘家军. 西秦岭寨上金矿床构造控矿特征与成矿规律[J]. 黄金科学技术, 2020, 28(6): 825-836.
[3] 陈彩华,孙斌,陈冀,陈力子 . 西秦岭中寨—十里铺地区水系沉积物测量找矿效果[J]. 黄金科学技术, 2017, 25(4): 1-9.
[4] 曹东宏,崔龙,辛小军,朝银银. 西秦岭马坞金矿床韧性剪切带构造控矿特征[J]. 黄金科学技术, 2016, 24(6): 24-31.
[5] 刘开君,樊小龙,余平辉,曾亮,成志雁,马锦龙. 西秦岭不同矿带金矿床成矿流体及其物质来源探讨[J]. 黄金科学技术, 2015, 23(6): 37-47.
[6] 郑卫军,刘新会,肖力,鹿青,姚薇,党若凡. 甘肃寨上—马坞地区矿产预测及找矿效果[J]. 黄金科学技术, 2015, 23(2): 8-16.
[7] 张帆,王根厚,王阳,周洁,吴春娇. 西秦岭阳山金矿床控矿构造特征探讨[J]. 黄金科学技术, 2014, 22(6): 39-45.
[8] 刘新会,陈彩华,赵天心,张科利,陈力子,刘爽. 西秦岭寨上金矿原生晕地球化学特征及成矿预测[J]. J4, 2012, 20(4): 7-15.
[9] 刘新会,冯明伸,刘家军,刘民武,张永文. 西秦岭寨上特大型金矿床碲化物与金矿物共生关系研究[J]. J4, 2011, 19(4): 6-12.
[10] 刘东海, 刘新会. 西秦岭寨上特大型金矿床黄铁矿特征及其含金性研究[J]. J4, 2010, 18(6): 8-12.
[11] 刘远华, 杨贵才, 张轮, 齐金忠, 李文良. 西秦岭阳山超大型金矿床花岗岩岩石地球化学特征[J]. J4, 2010, 18(6): 1-7.
[12] 刘新会,张永文,刘民武. 西秦岭寨上特大型金矿金的赋存状态研究[J]. J4, 2010, 18(5): 16-19.
[13] 刘新会,刘家军,陈彩华. 西秦岭寨上特大型金矿床硫盐矿物特征及其成因意义[J]. J4, 2010, 18(4): 6-11.
[14] 刘新会, 刘家军, 张争京, 郑卫军. 西秦岭寨上特大型金矿床成矿作用及成矿模式研究[J]. J4, 2010, 18(3): 20-27.
[15] 牛翠祎, 薛为民, 李绍儒, 卿敏. 西秦岭成矿带金矿资源综合信息预测评价[J]. J4, 2009, 17(2): 1-7.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!