基于RX多元勘查地球化学数据融合与异常识别

Developing techniques for identifying geochemical anomalies in complex geological settings is not only one of the hot topics in the field of mathematical geosciences and geochemical prospecting, but also is a significant way to alleviate the shortage of mineral resources and to improve the safety and security of national strategic resources. Based on the studies of mineral deposit geology and geodynamic background of Southwestern Fujian province, this proposal will explore the spatial distribution characteristics and enrichment patterns of geochemical multi-element, and investigate the spatial relationships between geochemical patterns and known mineral deposits, and then select the optimal elemental association related to mineralization. Meanwhile, this study will propose a hybrid method combining local spatial statistics, sliding window, and RX to process exploration geochemical data and to extract the weak geochemical anomalies. The identified geochemical anomalies will be evaluated in terms of geological environment for the formation of mineralization and the spatial correlations between geochemical anomalies and known mineral deposits. The high ranked geochemical anomalies will be further validated by field work. This study not only can enrich approaches for exploration geochemical data processing and geochemical anomalies identification in a complex geological background, but also can provide methods for selection of multi-elemental association related to mineralization and evaluation of geochemical anomalies. This study is expected to provide scientific clues for mineral prospecting in Southwest Fujian province.

研发复杂地质条件下弱缓地球化学异常识别和提取方法不仅是当前数学地球科学和勘查地球化学领域的研究热点，同时也是缓解矿产资源紧缺形势、提高国家资源安全保障的重要途径之一。该申请拟以闽西南铁多金属矿成矿区为研究区，进一步探究该区矽卡岩型铁多金属矿的成矿动力学背景和成矿规律，分析元素空间分布与富集规律及其与铁多金属矿的空间关系，据此选择矿致元素组合；研发基于局部RX方法进行多变量降维和弱缓地球化学异常识别技术，从异常与已发现矿床的空间相关关系和成矿地质环境2个方面进行异常评价和筛选，并进行野外查证。该研究不仅可进一步丰富和发展复杂地质条件下勘查地球化学数据处理与异常识别方法，而且还能为选择矿致元素组合和异常评价和筛选提供方法技术，有望为闽西南进一步找矿提供科学依据。

本项目围绕地球化学异常识别和评价这一关键科学问题，开展了基于大数据和机器学习算法的勘查地球化学空间模式识别和地球化学异常识别与评价的方法技术研究。本项目研究了闽西南地区的矿床地质和地球动力学背景，提出了一种基于空间分析方法选择矿致元素组合的新方法，以及一种以弱缓或负地球化学异常为重点的新的矿产勘查策略，开发了几种基于大数据分析和机器学习算法的地球化学空间模式识别和地球化学异常识别算法，以及评估与地球化学异常相关的不确定性的方法。该研究不仅丰富和发展了复杂地质条件下勘查地球化学数据处理与异常识别方法，而且还能为选择矿致元素组合和异常评价提供方法技术，为闽西南进一步找矿提供科学依据。上述成果在Earth-Science Reviews、Computers & Geosciences、Ore Geology Reviews等重要期刊发表12篇第一/通讯作者SCI 论文，授权1项国家发明专利和1项软件著作权，出版了一本学术专著。项目负责人在该项目的资助下获得了2项国家自然科学基金面上项目资助，当选为国际数学地球科学学会Council和国际应用地球化学家学会Council，入选了爱思唯尔2020中国高被引学者和伦敦地质学会Fellow，获得侯德封矿物岩石地球化学青年科学家奖，被邀请成为《Ore Geology Reviews》、《Computers & Geosciences》等期刊Associate editor，培养的研究生中6人获得国际数学地球科学学会研究生奖学金，2人获得研究生国家奖学金。